Autecology of the Copperhead BY (2024)

O 1^ .1960

University of Kansas Publications „,,,.. , Museum of Natural History

Volume 13, No. 4, pp. 85-288, pis. 13-20, 26 figs, in text

November 30, 1960

Autecology of the Copperhead BY

HENRY S. FITCH

University of Kansas Lawrence 1960 UNIVERSITY OF KANSAS PUBLICATIONS MUSEUM OF NATURAL HISTORY

Institutional libraries interested in publications exchange may obtain this series by addressing the Exchange Librarian, University of Kansas Library, Lawrence, Kansas. Copies for individuals, persons working in a particular field of study, may be obtained by addressing instead the Museum of Natural History, University of Kansas, Lawrence, Kansas. There is no provision for sale of this series by the University Library, which meets institutional requests, or by the Museum of Natural History, which meets the requests of individuals. However, when individuals request copies from the Museum, 25 cents should be included, for each separate number that is 100 pages or more in length, for the purpose of defraying the costs of wrapping and mailing. * An asterisk designates those numbers of which the Museum's supply (not the Library's supply) is exhausted. Numbers published to date, in this series, are as follows:

Vol. 1. Nos. 1-26 and index. Pp. 1-638, 1946-1950. *Vol. 2. (Complete) Mammals of Washington. By Walter W. Dalquest. Pp. 1-444, 140 figmes in text. April 9, 1948. Vol. 3. *1. The avifauna of Micronesia, its origin, evolution, and distribution. By Rol- lin H. Baker. Pp. 1-359, 16 figures in text. June 12, 1951. *2. A quantitative study of the nocturnal migration of birds. By George H. Lowery, Jr. Pp. 361-472, 47 figures in text. June 29, 1951. 3. Phylogeny of the waxwings and allied birds. By M. Dale Arvey. Pp. 473- 530, 49 figures in text, 13 tables. October 10, 1951. 4. Birds from the state of Veracruz, Mexico. By George H. Lowery, Jr., and Walter W. Dalquest. Pp. 531-649, 7 figures in text, 2 tables. October 10, 1951. Index. Pp. 651-681. *Vol. 4. (Complete) American weasels. By E. Raymond Hall. Pp. 1-466, 41 plates, 31 figures in text. December 27, 1951. Vol. 5. Nos. 1-37 and index. Pp. 1-676, 1951-1953. *Vol. 6. (Complete) Mammals of Utah, taxonomy and distribution. By Stephen D. Durrant. Pp. 1-549, 91 figures in text, 30 tables. August 10. 1952. Vol. 7. *1. Mammals of Kansas. By E. Lendell co*cknmi. Pp. 1-303, 73 figures in text, 37 tables. August 25, 1952. 2. Ecology of the opossum on a natural area in northeastern Kansas. By Henry S. Fitch and Lewis L. Sandidge. Pp. 305-338, 5 figiu-es in text. August 24. 1953. 3. The silky pocket mice (Perognathus flavus) of Mexico. By Rollin H. Baker. Pp. 339-347, 1 figure in text. February 15, 1954. 4. North American jumping mice (Genus Zapus). By Phillip H. Krutzsch. Pp. 349-472, 47 figiu-es in text, 4 tables. April 21, 1954. 5. Mammals from Southeastern Alaska. By Rollin H. Baker and James S. Findley. Pp. 473-477. April 21, 1954. 6. Distribution of Some Nebraskan Mammals. By J. Knox Jones, Jr. Pp. 479- 487. AprU 21, 1954. 7. Subspeciation in the montane meadow mouse, Microtus montanus, in Wyo- ming and Colorado. By Sydney Anderson. Pp. 489-506, 2 figures in text. July 23, 1954. 8. A new subspecies of bat (Myotis velifer) from southeastern California and Arizona. By Terry A. Vaughan. Pp. 507-512. July 23, 1954. 9. Mammals of the San Gabriel mountains of California. By Terry A. Vaughan. Pp. 513-582, 1 figure in text, 12 tables. November 15, 1954. 10. A new bat ( Genus Pipistrellus ) from northeastern Mexico. By Rollin H. Baker. Pp. 583-586. November 15, 1954. 11. A new subspecies of pocket mouse from Kansas. By E. Raymond Hall. Pp. 587-590. November 15, 1954. 12. Geographic variation in the pocket gopher, Cratogeomys castanops, in Coa- huila, Mexico. By Robert J. Russell and Rollin H. Baker. Pp. 591-608. March 15, 1955. 13. A new cottontail (Sylvilagus floridanus) from northeastern Mexico. By Rollin H. Baker. Pp. 609-612. AprU 8, 1955. 14. Taxonomy and distribution of some American shrews. By James S. Findley. Pp. 613-618. June 10, 1955. 15. The pigmy woodrat, Neotoma goldmani, its distribution and systematic posi- tion. By Dennis G. Rainey and RoUin H. Baker. Pp. 619-624, 2 figures in text. June 10, 1955. Index. Pp. 625-651.

(Continued on inside of back cover) University of Kansas Publications Museum of Natural History

Volume 13, No. 4, pp. 85-288, pis. 13-20, 26 figs, in text

November 30, 1960

Autecology of the Copperhead BY

HENRY S. FITCH

University of Kansas Lawrence 1960 University of Kansas Publications, Museum of Natural Histort

Editors: E, Raymond Hall, Chairman, Henry S. Fitch, Robert W. Wilson

in test Volume 13, No. 4, pp. 85-288, pis. 13-20, 26 Bgs. Published November 30, 1960

Mils. COlVir. iLyOl

t

DEC 2 7 1960

.

University of Kansas Lawrence, Kansas

PRINTED IN THE STATE PRINTING PLANT TOPEKA. KANSAS 1960

28-4428 Autecology of the Copperhead BY

HENRY S. FITCH CONTENTS PAGE Introduction 89 Acknowledgments 92 Methods 93 Description Lepidosis 99 Color and Pattern 102 Size 103 Bodily Proportions 106 Dentition 108 Hemipenis 112 Relationships 113 Habitat 116 Range and Geographic Variation 121 Behavior Crawling 124 Coiling 126 Swimming 128 Climbing 128 Disposition 129 Combat Dance 131 Shedding 134 Hibernation and the Effect of Temperature 137 Movements 147 Reproduction Courtship and Mating 157 Fecundity of Females 162 Development of Ova and Embryos 164 Aggregating of Gravid Females 166 Time of Birth 168 Number of Young per Litter 171 Birth of Young 176 Behavior of Females 17S Defects and MortaUty at Birth 178 The Egg Tooth 179

(87) 88 University of Kansas Publs., Mus. Nat. Hist.

FACE Size at Birth 181 Appearance of Young 182 Growth and Development Utilization of Stored Yolk, and Early Growth 183 Later Growtli 183 Cessation of Growth 192 Food Habits Methods of Obtaining Prey 193 Luring of Prey by Young 196 Statements of Food Preferences 198 Composition of the Diet 199 Kinds of Prey 201 Amount of Food Consumed 211 Defense, Escape and Mortality Factors Defense and Escape 219 Natural Enemies and Predation 221 EflFects of CHmatic Extremes 226 Parasites, Diseases and Injuries 228 Composition of the Population 230 Numbers 236 Relation to Man Attitudes of tlie Pubhc 245 Survival Under Modem Conditions 249 Control 250 The Venom and Bite Adaptations Correlated with the Venom 253 Properties of the Venom 255 Quantity of Venom Produced 256 Toxicity of tlie Venom 256 Susceptibilit)' of Snakes 259 Circ*mstances and Outcomes of Bites 261 Hypersensitivity to Venom 262 Case History of a Bite 263 Treatment of the Bite 265 Summary 269 Literature Cited 277 INTRODUCTION

In 1948 when ecological studies were initiated on tlie newly created University of Kansas Natural History Reservation, the copperhead was one of the first species that attracted attention as meriting intensive investigation. As an abundant predator on small vertebrates, including both those that are primary consumers of the vegetation and those of higher trophic levels, it was recognized as a key animal in the local ecosystem. Despite persistent effort to study the copperhead, progress was slow, especially in the early stages of the investigation. Copper- heads were rarely seen engaged in their normal activities, and even when such individuals were found, observing them proved to be remarkably unrewarding. A copperhead found by chance usually lay motionless for long periods, either having "frozen" in the usual reaction to any alarm, or merely resting—the sluggish behavior that is characteristic of the species. Attempting to ob- serve such a snake severely tried the patience of the investigator. When the snake finally began to move, it might soon be irretrievably lost because of the perfection with which it blended with its back- ground, and the dense concealing vegetation and other cover in the situations frequented. In the summer of 1949 cylindrical wire funnel traps set for lizards at the edges of rock outcrops in the "Rat Ledge" area caught several copperheads, and many more of these snakes were trapped in similar situations in autumn of the same year. Thereafter, each autumn, trap lines were maintained in rock ledge habitat and cop- perheads were obtained in numbers at that time of year but not at other seasons. In 1957 trap lines were established in a variety of habitats not previously sampled, and tliis trapping was continued on a larger scale in 1958 and 1959. In tliese three years copper- heads were obtained in large numbers tluroughout the season of their activity. The present report is based primarily upon records obtained on the 590-acre University of Kansas Natural History Reservation, in the nortlieas tern corner of Douglas County, Kansas, and on the adjacent 160-acre Rockefeller Experimental Tract. Including a few miscellaneous records, such as tliose of snakes found dead on county roads, and of young bom dead in captivity, a total of 1532

(89) 90 University of Kansas Publs., Mus. Nat. Hist. individual copperheads obtained from the Reservation or imme- diately adjacent areas were recorded a total of 2018 times between July 1, 1948, and November 9, 1959. Supplemental information was obtained from numerous other copperheads collected or ob- served elsewhere in eastern Kansas, notably from large series taken near La Cygne by Vernon Mann, who kindly permitted me to examine the live snakes in his possession from time to time. Despite the rapid accumulation of data during the later years completion of the study was long delayed because of seeming in- adequacies or inconsistencies in the information obtained. The information gained from marked copperheads recaptured after sub- stantial intervals provides the core of this report and is the main basis for conclusions regarding movements, growth, longevity, and age distribution. However, such records of recaptured individuals were sparse in the early years of the study, and constituted a small minority even in my field work in 1959. Hence, the information obtained concerning some phases of the natural history is scanty and the conclusions drawn from it are tentative. Because the copperhead's range is in parts of the United States longest settled and most densely populated, the species figures prominently in folklore and much has been written concerning it, both in scientific and popular hteratiu-e. However, most published references to copperheads are brief and casual. Although 142 years have elapsed since the pubhcation of Rafinesque's (1819) "Natural History of ScytaJus cupreus, or the Copper-head Snake" no thor- oughgoing account of the species* natm-al history and ecology has been made heretofore. Ohver (1958) has published an excellent brief summary of the literature, with some new information. Es- pecially noteworthy contributions to knowledge of the copperhead are 1943 those of Gloyd and Conant ( ) concerning taxonomy, Gloyd (1934), Smith (1940) and Allen (1955) concerning reproduction, Uhler, Cottam and Clarke (1939), Clark (1949) and Hamilton and Pollack (1955) concerning food habits, and Minton (1951, 1954, 1956) concerning the venom. Also deserving of mention is Klau- ber's of the has ( 1956 ) monumental monograph rattlesnakes, which shed much light on the biology of the pit vipers, and on snakes in general, and has been a frequent reference source in the course of my work. In preparing the present report I have examined all available pubhcations pertaining to copperheads, and have drawn freely on AUTECOLOGY OF THE COPPERHEAD 91

Fig. 1. Map of area approximately six miles north-northeast of the University of Kansas campus at Lawrence, where field work on copperheads was concen- trated—the combined University of Kansas Natural History Reservation and Rockefeller Experimental Tract. Each large dot shows a location where at least one copperhead was captured. 92 University of Kansas Publs., Mus. Nat. Hist. these sources for supplementary or comparative material. For certain phases of the copperhead's biology that were marginal to my study or were already rather thoroughly investigated (for ex- ample, the venom and the bite) the material here presented is based mainly or entirely on pubHshed hterature. The present report v^'ill be of interest chiefly to specialists, but, it is hoped, will also have some practical value. The copperhead affects human affairs in various ways, as by destroying harmful rodents, by destroying certain beneficial animals including lizards and toads, and occasionally by injuring or even killing humans and domestic animals by its venomous bite. Its greatest importance is probably the devastating psychological effect that its appearance or suspected presence creates in many persons. It is to be hoped that a better understanding of its habits and limitations may even- tually help to dispel this unreasoning fear. Also, the information here assembled concerning movements, food habits, population turnover, and seasonal habits provides a background essential for the planning of control operations. Certainly wholesale control operations against the copperhead are neither practicable nor de- sirable, but locally, for example in suburban communities where high populations of copperheads in remaining blocks of woodland constitute a hazard to small children, control may be both necessary and feasible. ACKNOWLEDGMENTS

Financial assistance rendered by the National Science Foundation in 1957, 1958 and 1959 is gratefully acknowledged. Students who were employed under this NSF grant no. B-3444 as field and laboratory assistants include James W. Bee, William N. Berg, Donna M. Hardy, Robert M. Hedrick, Dale L. Hoyt, Robert M. Packard and A. Wayne Wiens. Special thanks are due to each of these persons for their sustained interest and energetic co-operation. In the early stages of the study Drs. Richard E. Freiburg, Richard B. Loomis, and Dennis G. Rainey, then graduate students engaged in field work on the Reservation contributed many specimens and records, as did John A. Knouse, Anthony N. McFarland, Kenneth E. Shain and Robert B. Wimmer, in the later stages. Mr. Vernon Mann, professional snake collector of La Cygne, Kansas, kindly co-operated in allowing me to examine the hve copperheads in his collection and to collect their scats and shed fangs. Mrs. Norma Rothman, Dr. Joseph P. Kennedy and Dr. Bayard Brattstrom generously contributed original field notes and Dr. Kennedy also contributed food habits material. Mr. and Mrs. Harold Hedges of Kansas City, Kansas, contributed a fine series of live copper- heads. Mr. Ian D. W. Sutherland of Tulane University contributed notes and photographs concerning the courtship of copperheads in captivity. I am AUTECX)LOGY OF THE COPPERHEAD 93

especially indebted to Dr. William Degenhardt who spent many hours in the field with me when I visited Big Bend National Park in July, 1957, taking me to several remote areas where copperheads had been collected in the Park, and drawing on his excellent knowledge of the region to clarify various matters regarding its physiography and ecology. Mrs. Eleanor E. Buckley of Wyeth Laboratories, Inc., kindly provided me with information concerning the im- proved antivenin serum and referred me to important recent literature con- cerning the treatment of snake-bite. Drs. Frederick H. Dale and William H. Stickel kindly checked the files on food-habits in the U. S. Fish and Wildlife Service on my behalf for records of predation on copperheads by birds and mammals. METHODS

Copperheads were obtained chiefly by means of live-traps of the same general style that I have already described (Fitch, 1951:77). These cylinders of galvanized hardware cloth wire, of quarter-inch mesh, were one foot to two feet in length (usually 15 or 18 inches), and most of them were approximately seven inches in diameter. At one end of the trap, or at both ends, an entrance fuimel was inserted. The large end of the funnel was approximately twice the diameter of the trap, and the opening at the small end was approximately an inch in diameter. These traps were modified and improved in various ways in the course of the study. The valve-type doors of transparent cellulose acetate, which were used in the early model were eventually abandoned; to discourage exit of the trapped snake, cut ends of the wire were shaped into a circle of inward-projecting prongs. A heavy wire pin nine inches long, with a sharp point at one end and with the other end bent into the shape of a hook, was used to lock the funnel in place on the end of the trap. The pin was curved in a bow shape. It was thrust through the trap and funnel ap- proximately an inch from the end of the trap, with the convexity outward. Before the terminal hook engaged the wire of the trap's side, the pin was ro- tated through 180 degrees, with the result that the funnel was drawn firmly into position against the trap. Some traps were reinforced by attaching a heavy wire ring to the edge of the hardware cloth at the trap's opening and attaching a ring of the same size to the funnel. When the trap was set, with the reinforcing rings of the funnel and the trap's end in contact, it was almost impregnable to predators. At times, especially in autumn, raccoons, skunks, and other predators tore open the traps not so reinforced, to eat the animals that had been caught in them. In most instances grasshoppers, beetles, frogs or mice that wandered into the traps attracted these predators. It is not known whether any trapped snakes were killed by the raiders. Once a young copperhead, partly eaten was found beside a trap that was broken open, but in this instance the snake had already been killed by freezing. Reliance was placed upon the strategic placement of the traps, and no bait was used. How- ever, various smaU animals that wandered into the traps may have served occasionally as attractants to snakes. The wire funnel traps were sometimes used with a funnel in each end, at other times were used with a funnel in one end and a plug in the other. The plugs consisted of wire cones or of metal disks (ends of "tin" cans) with wire loops attached, and were locked into place with pins in the same manner as the funnels. 94 University of Kansas Publs., Mus. Nat. Hist.

Usually funnels were used in each end in autumn when traps were set along hilltop rock outcrops. Sites for the sets were carefully selected, where vertical rock faces presented barriers, with the expectation that snakes would travel along the base of the outcrop either in contact with the rock or adjacent to it. In placing the trap, care was used to shape the outer edge of the funnel, molding it against the ground surface and the rock face so that no space sufficiently large to permit a snake to squeeze under or past the trap remained (Plate 13, fig. 2). In summer when the snakes had dispersed from the rock ledges where hibernation occurred, trapping could be carried on most effectively in fields, thickets and the edges of woodland. In the absence of natural barriers, drift fences were erected to direct snakes into the funnel entrances. Ten-inch boards set up on edge and supported by small wooden stakes, by rocks or by shrubby vegetation were utilized as drift fences. In earlier trapping, two such boards were set up to form a V converging into the funnel, at each end of the trap. Later it was found more effective to set up several boards end-to-end in a single straight fence, with a trap at each end. Used in this way each trap had only one entrance, and the opposite end was closed with a plug. The continuous drift fences were judged to be more effective than the double V arrangement because both sides of the drift fence functioned to steer snakes toward the traps. The traps were seldom checked oftener than once per day, or less often than once per week, but no regular schedule was maintained for checking them. Frequency of checking was influenced by the weather, the productive- ness of the trap lines from time to time, and the amount of time available for this purpose. Traps set at the hilltop rock outcrops were held in place by flat rocks, placed beside them and over them. The traps set in summer away from woodland were always placed in sites beneath a tree or bush, or some other shelter that provided shade for most of the day, especially the hotter part of it. As further protection piles of straw or other vegetation were placed on the traps. Ordinarily the snakes trapped were processed in the field and released without removal from the site. Techniques varied somewhat according to the circ*mstances. The plug or funnel was removed from one end of the trap and the snake was shaken into a cloth bag and weighed on spring scales of 500-gram or 250-gram capacity. Weights obtained in the field were accurate within a range of one or two grams. After weighing, the snake was emptied out of the bag and almost invariably it coiled on the defensive rather than attempting to escape. It was then held down, caught by hand, measured, and examined. Special technique was of course necessary to avoid the venomous bite of the snake while catching, handling and releasing it. In either catch- ing or releasing a copperhead I always held down its head with a stick or ruler to prevent it from striking while my hand was within its reach. In picking up the snake I grasped it firmly just behind the head with the forefinger of my left hand around its neck, and my thumb holding it firmly in place. This grip was maintained until the snake was released. While holding the snake in this manner it is important to keep the forefinger back away from the chin, as the snake may bite through its own lower jaw. In handling captive copperheads in the laboratory I often employed metal bottle tongs to grasp them. Although these tongs were only eight inches long, they AUTECOLOGY OF THE COPPERHEAD 95

provided sufficient reach to permit grasping of a copperhead of normal size, so long as a frontal approach was avoided. In the field I often carried longer metal tongs and used them to remove copperheads from traps without moving the latter from their position. For measuring, the snake was held suspended vertically and a steel tape, graduated in millimeters, was held against it with the "zero" mark on a level with the tip of the snake's snout, and the tape dangling parallel with the snake. Grasping the snake by the base of the tail with my right hand, I exerted steady down- ward pressure to stretch the body out full length. At first the snake usually would resist, but soon it would tire, momentarily at least, and its body would be extended full length against the tape, pennitting a reading of snout-vent- length to the nearest millimeter. Tail-length was then recorded by holding the tail against a plastic ruler with the tip on the zero mark. The snake's mouth was forced open with forceps or a small stick, and the fangs were examined. Loose fangs that were being shed were removed, wrapped in paper, and labelled. Sex was determined by probing the hemipenial invagin*tion in the base of the tail, with a loop of fine wire, or with the semi-rigid but soft ended shaft of a slender green stem of grass or other vegetation. Every hve copperhead that was examined was palpated for detection of food in the stomach. Objects detected were forced forward into the mouth, to be identified. To force a snake to disgorge I grasped it with my left hand just behind the head and my right hand at mid-body, exerting pressure forward and upward against any suspected food object. Usually such objects slipped forward easily through the gullet and could be examined without injuring the snake. Relatively few food records were obtained from stomach items, only 67 in 2018 examinations of snakes. That only approximately 3.3 per cent of the newly caught copperheads were found to have food in their stomachs can be attributed largely to the fact that a high proportion of them were taken in autumn along the rock ledges. Snakes preparing to hibernate are less inclined to feed than others. Even the individuals caught at other times are perhaps those least Ukely to have fed recently, since after feeding they are sluggish and tend to stay in sheltered places. The hungry snakes engaged in active prowling are most likely to be caught in traps or found in the open. Throughout the course of my field work feces were collected from time to time, from snakes that defecated while they were in traps, bags, or cages, or while they were being handled. In 1957 it was discovered that in snakes containing food too well digested to be palpated from the stomach, fecal material could almost always be palpated from the anus. Often the hind- quarters of a partly digested animal were palpated from the stomach, and parts of the same animal were found in the scat collected at the same time. In fact, if a snake had food anywhere in the digestive tract, a sample usually could be obtained and identified. A total of 315 fecal samples were collected on the Reservation. At the most, a scat was only a few cubic centimeters in bulk. Scats col- lected were wrapped in absorbent tissue paper or paper towel, labelled with the date, the location, and the scale formula of the individual snake. Each scat was soaked for a day or more in a detergent solution, then rinsed, dried on a paper towel, and transferred to a cellophane envelope. The contents of each envelope were examined under a dissecting microscope and compared 96 University of Kansas Publs., Mus. Nat. Hist.

with collections of reference material. Mammalian material was identified

chiefly from hairs. The length, diameter, shape in cross-section, taper, and coloration of hairs were in varying degrees diagnostic of the species. In some instances the cell pattern, under high powered magnification, provided useful characters. Because most of the scat material was from the Reserva- tion where the mammalian fauna was already well known, and the number of species of a size that could be swallowed by a copperhead was small, identification was much simphfied. Greatest difficulty was experienced in separating species of the same genus. The voles, Microtus ochrogaster and M. pinetOTum could be identified most readily by examining the hair without magnification. Two species of Peromyscus were recorded but most identifi- cations in this genus were made merely as "Peromyscus sp." Doubtless in most of these instances the animal eaten was the wood mouse (P. leucopus) since this is one of the most abundant mammals over most of the Reservation, and its habitat requirements correspond more closely with the copperhead's than do those of the deer mouse (P. maniculatus) which is relatively scarce and localized. No attempt was made to distinguish between the two species of harvest mice occurring on the area, but most or all occurrences in scats probably were of the common species, Reithrodontomys megalotis. Because identification was made from hairs there was no indication of the number of individuals of tlie same species represented in a scat. It is therefore assimied that each occurrence represents a single animal. How- ever, of the eleven copperheads found to contain prairie voles in their stom- achs, one had eaten four voles, a female and her three young. Probably other multiple feedings went undetected in the identification of prey from scats. The importance of voles, especially, in the diet thus tends to be minimized. Also, in some instances the hair of one species of mammal being more abun- dant or more conspicuous, probably masked that of anotlier kind in the same scat and caused it to be overlooked. Hair of harvest mouse or wood mouse associated with that of a vole, for instance, or hair of least shrew associated with that of short-tailed shrew would be difficult to recognize. Reptiles were identified chiefly from scale material. Size, shape, presence or absence of keels, pits, and terminal notches provided distinctive combina- tions of characters by which the local genera, at least, could be readily dis- tinguished. Bird material was represented only by feathers, and these were so matted and bedraggled that they provided Uttle indication of the kind of bird unless they were of distinctive coloration. The insects that were secondary food items were usually in fragmentary condition, so that identification to species was impractical, but some of the cicadas were more nearly intact than any other items found in the scats. The residual material in scats consisted of hard parts, chiefly integumen- tary structures such as hair, scales, feathers and fragments of chitin. Bones and even teeth were largely disintegrated by the digestive juices, but remains of them were often found and sometimes they were nearly intact when em- bedded in wads of fur or other material that partly protected them from dis- solution. The fangs and other teeth of the copperheads themselves were often found in the scats and are more resistant than the teeth of other animals. Occasionally an entire foot of a lizard or mouse was found nearly intact. The thoroughness of digestion seemed to be somewhat proportional inversely to the bulk of the meal. Chitin was found to be relatively resistant to diges- AUTECOLOGY OF THE COPPERHEAD 97

tion, with the result that the insects eaten were better represented than the vertebrates. Although certain large insects, chiefly cicadas and large caterpillars are eaten regularly, many of the insect remains found in scats were of small kinds which almost certainly did not represent primary food items. Small ants (Crematogaster and others) were found in 14 scats. Often they were intact and were represented by many individuals. The narrow-mouthed toad (Gas- trophryne olivacea) is abundant on the Reservation and is known to feed on ants of this type almost exclusively (Fitch, 1956:301). Among the 67 prey items found in stomachs the narrow-mouthed toad comprised six per cent, but amphibians are so completely digested that ordinarily no recognizable remains can be found in scats. The 14 occurrences of ants (in varying quan- tities) in scats were therefore all hsted as instances of predation on Gastro- phryne although in most instances no remains of the toad itself were distin- guishable. If the ratio of narrow-mouthed toads to other kinds of prey is representative for the items found in stomachs, some 31 occurrences should have been recorded in scats. Perhaps some were missed because they did not have suflBcient food in their stomachs to leave noticeable residue in scats. Other insect material was in much more fragmentary condition than the remains of cicada, caterpillar and ant, and most such occuiTences probably were secondary. Of 39 occurrences of insects seven were identified as beetle. Associations were nearly always with small insectivorous vertebrates: 7 with Cryptotis, 7 with Eumeces, 6 with Peromyscus, 5 with Blarina, 5 with Reithro- dontomys, 4 viath Microtus, 2 with Ophisaurus, and one each vdth Microtus, Sylvilagus, Terrapene, Coluber and "bird." The last three were all in the same scat. Also, 6 of the Eumeces-insect occurrences, 3 of Reithrodontomys-insect, 3 of Cryp^o^fs-insect and 2 of Microtus-insect were in association with other small vertebrates that are potential insect eaters. Six insect occurrences were not associated with vertebrate remains. In these instances, in the rabbit-insect as- sociation, and probably in some of the other occurrences, it seems most likely that the insects were in the digestive tract of an amphibian eaten by the copper- head and completely digested by it. These problematical occurrences of insects were tentatively assigned to the leopard frog (Rana pipiens) since this frog was found among the items identified from stomachs but presumably would have been completely digested and could not have been represented in scats except by the secondary prey items from its own digestive tract. In the later stages of the study many copperheads were tested for spenn. Samples of cloacal fluid were examined under a microscope for motile sperm as evidence of recent copulation in females or attainment of breeding condition in males. Before release, the copperhead was marked with quick-drying enamel paint of a bright color, red, orange, yellow or blue, to facilitate recording of molt. Permanent marks were made to render the individual recognizable by clipping of subcaudal scales. I clipped these scales with sharp scissors by holding the tail firmly in place between the middle finger and hand, meanwhile maintaining the original grip on the neck between the thumb and forefinger

(Plate 13, fig. 1). The subcaudals used for formulas were the second to the twentieth on the base of the tail. Each mark involved the excision of half a

subcaudal on the left side of the tail and half of one on the right. The scale and underlying skin were removed, laying bare the muscle layer beneath. The excision involved two cuts with the scissors. CUpping was begun with the 98 University of Kansas Publs., Mus. Nat. Hist.

scissor points at right angles to the tail; the skin was slit at the base of the scale to be marked. Then the scissor blades were laid against the tail with the half- scale, now loosened on one edge between them, and it was removed with another stroke. In the copperhead, subcaudals except those on the distal part of the tail are normally undivided. The first entire subcaudal behind the anus was not clipped on either side, as in its absence the "number two" scale might have been mistaken for the first of the series. Three hundred and sixty-one combinations were possible with the remaining positions that were used. After these combinations had been exhausted, a new series was initiated dupli- cating the first except that a ventral body scale, the second anterior to the anal plate, on the left side "G 2 L"—gastrostege nvunber 2 on the left side— was included in each formula. Subsequently other series, G2R, G4R, G5L and G6R were used in whole or in part. The G2R and G4R series were used ex- clusively for copperheads bom in captivity or those first captured when they were near the size at birth. Any recaptured snakes bearing these marks were therefore at once recognized as individuals whose records extended back to the time of birth or near it. Over periods of months the scale tissue always regenerated where the excisions were made, but the scar remained causing the area to differ slightly in color and surface texture from the intact scales nearby. Rarely the clipped scale was so well regenerated that only its narrowness, or an indentation on its posterior edge served for identification. The two or three excisions on an individual snake usually were not equally distinct at the time of recapture, and the factors affecting distinctness or obscurity after a period of years are not altogether clear. However if the anterior edge of the clipped scale remained, subsequent regeneration was much more complete. Many copperheads were obtained by turning flat rocks, but these were only a small proportion of the total number taken. On the Reservation four comprised the maximum catch in one day by this method, but on other areas where the population was higher, ten or more have occasionally been secured in a few hours of rock-turning. The snakes are to be found most concentrated in the spring before they have left the rock ledges where they have hibernated, and they tend to be diurnal while nights are still cool. Nevertheless, a thorough search involving turning of every loose rock that was not too heavy, might disclose only one or two copperheads or none at all along a stretch of ledge where dozens were known to be present from the data obtained by live-trapping in autumn. Occasionally copperheads might be found under rocks at any time in their season of activity, but in summer hunting them in this way was less productive than in spring, because the snakes had dispersed from the hilltop ledges, and because at the high prevailing temperature and humidity the strenuous activity of turning heavy rocks produced relatively rapid fatigue. A copperhead exposed by turning a rock usually lay motionless but alert for several seconds and then began moving slowly in search of shelter. Such snakes were usually caught without difficulty, but occasionally escaped when there were deep crevices readily available beneath or beside them. Copperheads also were caught actively prowling on roads, usually at dusk or after dark. Temperature of the air and of the snake's body was usually recorded on these occasions. Some of the copperheads were obtained by chance in the course of routine driving, but many evening drives were taken expressly for the purpose of collecting them. Chances of finding the snakes were best AUTECOLOGY OF THE COPPERHEAD 99 when air temperature was higher than 75" F. and when the soil and vegetation were wet from recent rain, with humidity high. Upon approach of an auto- mobile, a copperhead crossing a road usually retracted its head slightly and "froze" into immobility in a position from which it might strike an attacker or lunge for cover. Such individuals were more aggressive and irritable than those found under other circ*mstances; if touched or closely approached they would lash out wildly in self-defense, meanwhile thrashing and squirming in clumsy but animated attempts to reach shelter. However, such snakes found in open places usually could be held down with sticks and captured without diflBculty. Other types of data were obtained from the keeping of captives. An outdoor cage ten feet square, of quarter-inch wire in the shade of large elm trees near the Reservation headquarters was used to confine copperheads under condi- tions simulating their natural habitats. Natural vegetation grew in the pen but was kept trimmed to facilitate finding the snakes. Flat rocks and large boards provided shelter. A hibernation box was installed at a depth of three feet, with a removable insulation box between it and the surface permitting easy access to the snakes when they were dormant. A plastic tube with roughened inside surface (to permit traction as the snakes moved through it) provided a passageway from the hibernation box to the surface. General behavior, including feeding, breeding, activity, reactions to high and low temperature, and to sunshine, rain and other phases of the weather were ob- served in this cage. Another enclosure was constructed by installing a three- foot fence of quarter-inch wire extending in a semicircular arc 40 feet long, supported by metal stakes with each end against the outer wall of my residence. Much of the enclosed area could be seen from my bedroom wandow, which opened onto it, facilitating observations on natural activity. The pen contained natural vegetation and, as it had no top, birds, squirrels and other animals associated with copperheads in the wild moved freely in and out. Occasional opportunities to observe the mutual reactions of copperheads with such animals were afforded.

Still other copperheads were confined indoors, in cages in the laboratory or in my Uving-room. These provided types of information that rarely or never would have been obtained in the field such as frequency of shedding skin and fangs, rate of digestion, frequency of feeding, and various details of be- havior. However, under the admittedly unnatural conditions of captivity indoors, normal behavior and physiology may have been altered somewhat. DESCRIPTION

Lepidosis

The copperhead has lepidosis fairly typical of a generalized snake. Its cephalic scutes, for instance, correspond well in number and arrangement with those of most colubrids. In this respect the copperhead differs from crotalids of other genera, and even from some of the other species of Agkistrodon, in which there is a tendency for the scutes to be divided up into small, granular scales. In the copperhead the cephalic scutes vary but little in their arrangement, either individually or in geographic populations. The shape and relative size of each scute is characteristic, and distinctive of the species. In the 100 University of Kansas Publs., Mus. Nat. Hist. rostral plate, on the front of the snout, the width at the base exceeds the height but width at the top is slightly exceeded by its height. On top of the muzzle, bordering the rostral, are the paired intemasals. They are subtriangular, expanded posteriorly, and wider than long. The nostrils are on the side of the muzzle approximately one-third of the distance from the tip of the snout to the eye. Each nostril is between a prenasal and a post- nasal both of which contact the intemasals, above and the first supralabial, below. The paired prefrontals, on top of the head behind the intemasals, are wider than long, are rounded laterally, and are nearly as large as the frontal, supraoculars, and parietals, which are the largest cephalic scutes. On their posterior borders the prefrontals contact the frontal and supraoculars. The loreal pit has an aperture about half again as large as that of the nostril. It is situated at the level of the lower edge of the eye, between the eye and nostril, but a little nearer the eye. The pit is bordered above by a supraloreal, an oval scale somewhat more pointed dorsaUy, and is bordered below by the infraloreal, a small, inchned, rectangular scale. The pit is bordered anteriorly by the second supralabial, and posteriorly by the lower preocular. The frontal is a pentagonally shield-shaped scale on the middle of the forehead, almost straight across the anterior end, with an angle of usually slightly less than 90° at the antero-lateral comer, often approximately 135° at the posterolateral corner, and slightly less than 90° at the posterior comer. The paired parietals are half again as long as broad. They tend toward hexagonal or pentagonal shape but the posterior and postero-lateral margins are irregular and appear to be in process of breaking up into small scales. The anterior corner of each parietal forms an angle in the neighborhood of 90° (between the frontal and supraocular) and the two anterior sides are subequal. These two sides also approximate the length of contact of the two parietals with each other along the midline. The top of the head behind the parietals is covered with scales much smaller than those on any part of the body, arranged in irregular rows. A few of the more an- terior are smooth, the rest are weakly keeled. The supralabials are typically eight on each side. The first is low at the anterior end, and the length exceeds the maximum height. The second is higher than long and extends along the anterior margin of the facial pit to its upper edge. The third supralabial is much larger than the first or second, and subtriangular, low in front and high behind. The fourth is the largest, its anterior and posterior edges are nearly vertical; it is beneath the eye from which it is separated by two small suboculars. The fifth is similarly shaped but markedly smaller. The sixth is intermediate in size between the fourth and fifth, its rear edge inclined posteriorly. The seventh is higher than long and slanted posteriorly. The eighth is subtriangular, longer than high. There are two preoculars, horizontally divided, the upper approximately twice as large as the lower. The two suboculars are each about twice as long as high. There are three small postoculars (Plate 18, fig. 1). There are several rows of temporals. The lower row is in contact with the supralabials. In the lower row, the first is relatively small and contacts the fourth and fifth supralabials. Temporals of the upper rows become progres- sively smaller and grade into the small scales on top of the head posteriorly. AUTECOLOGY OF THE COPPERHEAD 101

Between the parietals and temporals, on top of the head, are four rows of moderately enlarged scales flattened and unkeeled and of somewhat irregular shape. On the chin there is a mental rounded anteriorly, and forming an angle of about 110° posteriorly, followed by the first pair of infralabials, which are broadly in contact on the midline and are pointed posteriorly. Behind them there is a pair of genials, which are more than twice as long as wide, are in contact medially, and are bluntly pointed posteriorly. Behind the genials and between the infralabials and anterior ventrals are the gulars in approximately six rows, but the rows are somewhat irregular. The gulars are mostly rec- tangular or vaguely hexagonal, from two to three times as long as broad. There are ten pairs of infralabials; those of the second pair are only about half the size of the third pair and a quarter the size of the fourth pair. Beyond the fourth pair size becomes progressively smaller. The last two are elongate, others are rhomboidal, slightly higher than long. On the neck the dorsal scales are markedly smaller than those elsewhere on the body (about one-fourth the dimensions of a typical body scale) and they resemble those on the posterior part of the head. There are slight ir- regularities in the scale rows of the neck and of the posterior part of the head, resulting from the merging in this region of rows having scales of different sizes and shapes—the gulars, the scales above the temporals, and the body scales. The body scales overlap slightly except when the skin is stretched. In the region of the throat, neck, and forebody the skin is especially loose and elastic. When the snake is swallowing prey, for instance, the skin may be stretched to the extent that two neighboring scales are separated by areas of skin much greater than their combined widths (Fig. 3). Farther posteriorly, especially past mid-body, the skin is much less extensible. The body scales are keeled (except those of the anterior part of the first row). A typical scale is ap- proximately twice as long as broad tending toward an oval shape with the posterior end the more pointed; scales are often faintly hexagonal. The dorsal scales near the midline are the narrowest; those farther down on the sides become progressively wider. Farther posteriorly on the body, and especially on the tail, the scales become smaller, and relatively wider, tending to a rhomboidal shape. There are 23 scale rows on the body for most of its length. In the neck region, however, counts of 25 rows or even more, can be obtained. The fifth row on each side drops out at a point averaging 11 per cent of the distance from snout toward the vent (in the neighborhood of the 17th ventral) leaving a total of 23 for most of the length. At a point averaging approximately 70 per cent of the distance from snout to vent ( in the neighborhood of the 102nd sixth on each side ventral ) the fifth remaining row ( originally ) drops out leaving a total of 21, and at a point approximately 87 per cent of the distance to vent the fifth of the remaining rows (originally seventh) drops out, leaving only 19 rows on approximately the posterior 13 per cent of the body. On the tail the remaining rows drop out in rapid succession. At the middle of the tail there are only ten rows and just ahead of the tail spine there are only three. The tail ends in a blunt spine, which is inclined downward slightly at the tip (Fig. 4). Cope stated (1900:1132) that the spine consisted of three scales, one ventral and two dorsal, ensheathing the last caudal vertebrae, an elongate, pointed splint. However, in the specimens that I have examined, 2—4428 102 University of Kansas Publs., Mus. Nat. Hist. the caudal spine consisted of a single scale, with a seamhke ridge along the mid-dorsal line but none ventrally, with another enlarged, platehke scale at the base of the spine on the dorsal surface of the tail. According to Gloyd and Conant (1943:168), ventrals averaged 148 in 820 specimens of A. c. mokeson, with no diflFerence in numbers between the sexes. In the same series, subcaudals averaged 46 in males and 44 in females. To judge from the relatively few counts made in the course of my study of the local population the average numbers of scales does not differ from the numbers recorded by Gloyd and Conant ( loc. cit. ) .

Color and Pattern

The color is predominantly brown, but with different shades, in a boldly contrasting pattern, from pale grayish brown, tan, or fulvous at one extreme to a deep chestnut, nearly black at the other. The range of shades is great when one takes into account individual variation, age variation (the yoimg are paler, with more vivid pattern and lack reddish suffusion), variation caused by the cycle of molt (colors become darker and duller as the time of shedding approaches), and sexual difference (the adult males are darker, with more reddish suffusion, as compared with most females). The head is reddish brown dorsally, having a color vaguely reminiscent of that of an old copper coin, hence the most common vernacular name of the species. In each parietal plate near its center but slightly displaced toward the midline of the head, there is a spot of dark chestnut, narrowly rimmed by fulvous. These parietal spots are conspicuous although they are usually less than one millimeter long, even in the largest specimens. The brownish hue of the head deepens in the temporal region, and is separated by a sharply defined line from the much paler cream-colored area of the supralabials. The line of separation passes from the eye posteriorly through the middle of the first temporal, through the lower part of the second temporal, and along the upper edges of the last two supralabials. The infralabials also are cream- colored, paler than any other part of the snake, but their ventral (medial) portions are darkened, with a sharp line of demarcation between the pale and dark portions, running continuously from the second to the last. This line on the infralabials joins or almost joins the posterior end of the line across the temporal region, the two forming a narrow loop around the comer of the mouth. The anterior part of the head is more grayish (less reddish) than the posterior part, and there are no markings on the rostral region. Over the entire body the scales are finely stippled with black dots. Typically there are from 20 to 40 per scale but the number cannot be counted accurately because some are in contact or partly fused with others. On the body there is a series of dark brown cross bands on the lighter brown ground color. The cross bands have been described as hourglass-shaped or dumbbell-shaped; they are constricted in the midline and widened laterally. At its medial con- striction a band of symmetrical shape is usually of a width equal to the combined lengths of three or four dorsal scales. On the side, at its widest extent, the band is usually eight to ten scales long—wider than the interspace which is usually three to six scales long at this level on the side. Many of the bands are not bilaterally symmetrical, but the half on one side of the midline is displaced either anteriorly or posteriorly with respect to its partner on the opposite side, with the result that the left and right halves have only a narrow AUTECOLOGY OF THE COPPERHEAD 103 connecting zone or are completely separated. Individuals having all their cross bands intact are in the minority. A band may be represented by only the left or right half, with no counterpart on the opposite side of the midline. Or on the left or right the band may be represented by a mere spot on the lower part of the side, not extending to the midline. The number of intact bands varied from seven to 16 in the copperheads examined, but 12 was the most frequent number. Livezey (1949:93) figured an abnormally patterned copperhead from Texas in which only tliree of the 14 bands on the body were intact. In individuals having the number of complete bands fewer, the number of half-bands, or blotches is correspondingly greater, and the proportion of the body covered by the chestnut markings seems to be remarkably constant (Figs. 5 and 6). Secondary sexual difFerences in the number and disposition of markings were not clearly indicated by the trend of the data. The trends for the left and right side showed no consistent differences either. The trans- verse dark markings of the body are continued onto the tail, but there they are not constricted middorsally, and the paler interspaces become progressively smaller until they are represented by only thin lines on the posterior part of the tail. On the average, there are approximately eight dark marks on the tail. The chin and throat are pale, cream-colored (except for the dark area on the infralabials, already mentioned). The ventral surface of the body has large, irregular, black marks that occupy the greater part of its surface. These markings are mainly on the ventral plates, but they invade or include some scales of the first row adjacent to the ventrals. The markings have sharply defined lateral edges, but elsewhere their edges are so difiFuse and ill-defined that no definite count can be made of the number present; rather the general effect is of marbling or heavy but uneven stippling over most of the ventral surface. The larger ventral markings are rounded and usually cover parts of three or four adjacent ventrals. Ordinarily there is one on each side beneath each dorsal cross band and one beneath each of the paler areas alter- nating with the cross bands. The markings are better defined on the anterior part of the ventrum than they are posteriorly; on the tail they are especially vague. The lining of the mouth is flesh-colored. The tongue is carmine, paling to white on the tips. The iris of the eye is pale gold vdth fine reticulations of dark pigment. As in all other pit vipers the pupil is vertically elliptical.

Size

Copperheads captvued on the Reservation ranged from 209 mm. to 936 mm. in snout-vent length (9.8 inches to 42.0 inches in over-all length). Many smaller than 209 mm. were bom in captivity, but probably most or all of these young were stunted by the unfavorable effects of confinement on the gravid females. In a sample of 1,678 records from the Reservation, 1949 to 1959, the average over-all length was 22.4 inches. Figure 7 shows the rela- tive numbers in each size class, of each sex. It is evident that maximum size is larger by one-fourth in males than in females. Most typical adult sizes are 28.5 inches for males and 26 inches for females, in over-all length. Almost nothing is known concerning geographic variation in size over the copperhead's extensive range, but there is some indication that western popu- lations do not grow so large as those in the eastern states. The largest copper- head ever recorded in the literature was 53 inches long, and was captured at 104 University of Kansas Publs., Mus. Nat. Hist.

White Plains, New York, in the northeastern part of the range (Ditmars, 1935:22). Others nearly as large have been recorded from this same general area. It is unlikely that this size is even approached by the largest individuals in Kansas. Wright and Wright (1957:904) recorded that a Mr. C. L. Love

Fig. 2. Bodily proportions, and relative sizes of scales in a copperhead. Column on left shove's two cross-sections of the head, five of the body and one of the tail. On the right are shown ventral scutes ( right halves only), and dorsal scales (shaded) in series of three, representing one from near mid-dorsal Bne (farthest left), one from halfway down the side (middle), and one from low on side, adjacent to ventral scute (right). Snake shown approximately X %, cross-section X %, scales X3. took a 52-inch specimen at Apopka in central Florida, but some mistake must be involved here as the locality is well outside the authenticated range of the copperhead. The same authors stated that adult size was 16 to 36 inches in the western A. c. laticinctus. Oliver (1958:40) stated that the Trans-Pecos AUTECOLOGY OF THE COPPERHEAD 103

!] 150 106 University of Kansas Publs., Mus. Nat. Hist. copperhead, A. c. pictigaster seems to attain a maxunum length of only two feet. However, only a few typical specimens of this subspecies are known, and all of them may be far short of maximum size.

Bodily Proportions

The form is that of a typical crotalid. The head is flattened and subtri- angular, much widened in the posterior temporal region, and abruptly taper- ing anteriorly to the muzzle, which is somewhat rounded. From the eye to the snout the top of the head is in a plane at right angles to its side, and a sharp edge, the canthus rostralis, is formed. The posterior part of the head is laterally elliptical in cross-section. The neck is constricted. The body is moderately robust, subtriangular in cross-section, and increasingly flattened posteriorly. The tail is round in cross-section, tapers abruptly, and is relatively short—usually from one-sixth to one-seventh of the snout- vent length, depending on the age and size of the individual. For 23 specimens, of both sexes and various sizes, that were measured when freshly killed and relaxed, the following proportions were obtained, expressed as ratios of snout-vent length: Length of head, 5.55 ± .11 per cent Width of head, 4.14 ± .08 per cent Circumference of neck, 6.81 ± .13 per cent Circumference at mid-body, 10.41 ± .16 per cent Circumference of tail base, 6.59 ± .12 per cent

None of these characters showed any significant differences between the sexes. However, relative head-length, and head-width was found to be greater in the smaller snakes, progressively decreasing as greater over-all

Table 1. Variation in Relative Tail-length According to Size and Sex IN A Population of Copperheads AUTECOLOGY OF THE COPPERHEAD 107 length is attained (Figs. 8 and 9). Klauber (1956:152) has discussed at length similar changes of proportions in the ratdesnakes. Like other kinds of snakes, the copperhead exhibits sexual dimorphism in the relative length and proportions of the tail, and in its ratio to body- length. However, in the copperhead the sexual dimorphism is relatively slight and tends to be obscured by ontogenetic changes. Table 1 shows the

10 108 UNivEBSiry OF Kansas Publs., Mus, Nat. Hist. difference is statistically significant. As the young snakes grow, the difference in proportions increases gradually. In young adults, males and females can be easily distinguished, and the sex is even more readily discernible in old adults. In both sexes the tail becomes relatively shorter as size increases. The trend is more pronounced in the females. However, growth continues longer in the males, and as a result, large adult males (900 mm. snout-vent) and large adult females (700 mm. snout-vent) resemble each other in ratio of tail to snout-vent length. AUTECOLOGY OF THE COPPERHEAD 109

The pterygoid bears 13 to 17 teeth, all strongly curved, and becoming smaller toward the posterior end of the series. The numbers of teeth mentioned above as characteristic for each dentigerous bone refer to the number of sockets, but at any one time some are empty, as teeth are shed frequently. The shed teeth pass through the digestive tract little altered, and a fecal mass that is examined microscopically usually is found to contain many of them. The maxillary is remarkably modified from the typical ophidian form of an elongate arcuate bone in the anterior supralabial region. Instead it has

Fig. 10. Right fang of a copper- head from (A) anterior view and from (B) lateral view, X 9. in the course of evolution become shortened chiefly by loss of its posterior part, but extended in a vertical plane, and it is considerably deeper than long. Its lateral face is excavated to accommodate the sensory facial pit. The maxillary articulates posteriorly with the prefrontal, palatine and ectopterygoid. Because of its shortened form, the maxillary can rotate through almost 90°, permitting its single greatly enlarged tooth, the solidly attached poison fang, to be either erected or folded against the roof of the mouth. Each maxillary bone bears twin sockets side by side, but, in the frequent replacement of fangs that occurs, there is alternation from one socket to the other. Both sockets bear fangs simultaneously only in a relatively brief interval when one fang is about to be shed and the other has just become anchored to the bone. The poison fang is similar in general appearance to the other teeth, but much enlarged. In its evolution it has been converted from a simple conical structure to a tube, functioning hke a hypodermic needle in the injection of venom. Presumably in remote ancestors of the pit vipers the fang first de- veloped a groove on its anterior surface, as a channel down which the venom might flow into the wound inflicted by the tooth. Eventually the groove be- came deeper, and its sides extended medially to contact each other, enclosing the venom canal as a tube. Even in the modem pit vipers the venom canal 110 University of Kansas Publs., Mus. Nat. Hist. is not entirely enclosed for the full length of the fang, but the lumen appears as an elongate slit on the anterior surface of the fang's distal one-third. The venom thus is injected to a depth somewhat less than the length of tlie fang. The tip of the fang is solid. Venom enters the fang through a notch near the base on its anterior face, enclosed deep in the sheath.

During its short functional life, the fang is rigidly attached to the bone, in its socket. After a new fang has assumed its position in the alternate socket, the first fang becomes weakened at its base, and breaks off by means of an irregular fracture when subjected to stress. Even after fracture occurs the fang may be retained for a day or more inside its fleshy sheath, loosely at- tached to the gum tissues or incompletely separated from its pedicel. Eventu- ally it becomes detached and is swallowed along with the food. The pedicel is then resorbed into the socket. At any one time a copperhead has several replacement fangs in various stages of development in the gum behind the socket bearing the functional fang. Each socket has its own series of replacement fangs. The tip of the fang is the part fonned first, and calcification gradually proceeds toward the base. The base of the fang is formed within its socket, and until calcifica- tion is completed in this region the fang is loose. In 1956, 1957, 1958 and 1959 I examined the fangs in 745 of the live copperheads collected on the Reservation. In 248 there was an accessory fang on either left or right side, and in 23 there was an accessory fang on both sides. Thus 294 fangs, or 19.7 per cent of the 1490 fangs in this entire group of snakes, were being replaced. These figures suggest that for about one-fifth of the time on the average, each fang is in process of replacement. A small adult female copperhead was kept active at room temperature through the winter of 1959-1960 and examined at irregular intervals of two or three days. On December 15 the right fang was being replaced and by December 19 the process was completed. On December 23 an accessory left fang was in evidence and on December 28 it was solidly attached in its socket beside the old left fang. On December 31 the old left fang was loose in its sheath and was removed with forceps. On January 15 an accessory right fang was again in evidence but was still loose and in a position behind the old fang. On January 17 there was little discernible change, but on January 20 new and old right fangs were side by side in their sockets. On January 24 the old right fang had been shed. On January 27 an accessory left fang was again in evidence still loose and behind the old fang. On January 30 the new fang was still sHghtly loose and behind the old fang. On February 2 one left fang, presumably the older, was loose and in a position behind the functional fang. In this snake, for the seven-weeks observation period, left and right sides alternated in replacement of the fang, with approximately a 33-day cycle on each side, and with known replacement periods extending over five, six and eight days, during which both new fang and old fang were in evidence on the same side. The new and old fangs were simultaneously functional during only a small part of the replacement period. Although the snake upon which these observations were based, was kept at slightly lower temperatures than AUTECOLOGY OF THE COPPERHEAD 111 those prevailing under natural conditions during most of the copperhead's season of activity, it seems reasonable to conclude that the normal cycle of replacement for each fang is slightly more than a month with a replacement period of approximately a week. Klauber (1956:726) estimated the normal functional lifetime of a fang to be six to ten weeks in an adult rattlesnake. Length of fangs is variable. In 52 shed fangs, measured from fractured edge of base to tip, in a straight hne, the average was .873 per cent of the snout-vent length. For 21 copperheads in the size range 500 to 600 mm.

( small adults ) the fang length averaged .88 ± .0144 per cent of the snout- vent length. At birth the young have relatively large heads and the fangs are longer in proportion to the snout-vent length than are the fangs of larger and older individuals. Although my samples were too small to show how these proportions change, the following figures are suggestive of the trend.

Table 2. Correlation of Relative Fang-length with Snout-vent Length

Snout-VENT Length (millimeters) 112 University of Kansas Publs., Mus. Nat. Hist.

.80 per cent to .89: Crotalus adamanteus, molossus, polystictus, and stejnegeri .90 C. exsul and ruber C. atrox, unicolor, pusillus, durisstis and horridus C. willardi, triseriatus, tortugensis, viridis and cerastes C. basiliscus and scutulatus Sistrurus catenatus and C. enyo Sistrurus miliarius and C. lepidus C. mitchelli and pricei C. tigris C. transversus C. intermedius

Hemipenis

As in other snakes, the paired copulatory organs are lodged in the base of the tail. There are distinctive differences between the hemipenes of different in kinds of snakes that are useful defining species and genera. Cope ( 1900, pi. 31, fig. 4) has illustrated the hemipenis of the copperhead, but his figure AUTECOLOGY OF THE COPPERHEAD 113

Klauber (1956:661-668) has well described the hemipenes of the many species of rattlesnakes and has illustrated several. In most respects these organs resemble that of the copperhead, but differ in various details. In the copperhead (Fig. 11) the hemipenis is deeply bifurcate (for about three- fourths of the total length). The lobes tend to be cylindrical and are only slightly tapered distally, but are enlarged at their bases, and are approximately three times as long as broad. There are approximately 35 large spines on the basal thirds of the two lobes. Most of these spines are straight but some are slightly hooked. The distal two-thirds of each lobe is covered with small flattened papillae, each ending in a spine. There are probably more than a thousand of these papillae on each hemipenis but they are not arranged in regular rows. The transition from the spiny basal portion of each hemipenis to the papillose distal portion is abrupt. There are no mesial spines in the crotch. The sulcus spermaticus is forked near its base.

RELATIONSHIPS

The copperhead is among the more primitive representatives of the CrotaHdae is of the ( pit vipers ) , which one most speciaHzed families of snakes. In lacking a rattle, in having enlarged head shields typical of the more generalized colubrid snakes, in having only those subcaudals on the proximal part of the tail undivided, and in having the poison fangs relatively short, the copperhead is less specialized than other crotalid genera, the rattlesnakes, {Crotalus and Sistrurus), fer-de-lance and its relatives (Bothrops), oriental pit vipers {Trimeresurus), or bushmaster (Lachesis). The genus Agkistrodon includes perhaps a dozen species (see Smith, 1943:494), but some of these are wide ranging and highly variable in their characters, and authorities differ as to the number of full species that should be recognized. Dr. Howard K. Gloyd is engaged in a taxonomic revision of the genus. Some characters of the genus are: pupil of eye vertical; head covered with symmetrical shields (or having the internasals and prefrontals broken up into small scales); facial pit between the preoculars and loreal; scales usually keeled (smooth in A. rhodo- stoma); anal plate single; subcaudals either undivided or divided in pairs—usually both conditions occur in the same individual on different parts of the tail; anterior genials large, posterior genials small or ill-defined. Though having these characters in common the species differ strikingly in size, color, pattern, numbers of scale rows, arrangement of cephalic shields, and also in habits and habitats. Obviously the genus is relatively old as compared with most other vertebrate genera, and its species are well differentiated. The cottonmouth (A. piscivorus) shares much of the copperhead's 114 University of Kansas Publs., Mus. Nat. Hist. range, and perhaps is its closest relative. The basic pattern is similar in the two species, and various behavioral traits are shared by both. The young are more nearly alike than the adults, and might easily be mistaken for each other. But in size, habits and habitats, and in various features of external morphology the two species are sharply differentiated. The cottonmouth is much larger; individuals six feet long are on record. It is partly aquatic and is usually found in or near streams, lakes or sloughs. It preys mostly on amphibious or aquatic vertebrates. The cantil (A. biline- atus) resembles the cottonmouth in habits but is perhaps less aquatic as on occasion it has been found far from water. The cantil's range is complementary to that of the cottonmouth, and is in Mexico, chiefly in the coastal lowlands. Its color pattern is readily derivable from that of the cottonmouth but is less like that of the copperhead. Of the Old World species, (Pope, 1935:386-403 and 1955:222-223; Smith, op. cit.; Schmidt and Inger, 1957:264-267) none seems to be closely related to the copperhead. Perhaps A. acutus of southern China resembles it most. The Old World species are confined to forests of southern (mainly southeastern) Asia except for A. halys which ranges from Japan westward through both forests and steppes to the Caspian Sea. Compared with the North American species the Asiatic species form a much less compact group and differ greatly among themselves in size, scalation and habits. A. acutus attains a length of five feet, but 34 inches is more typical for other species {rhodostoma, himalayanus), and the smallest species strauchi is only 20 inches long. Most of the species are viviparous but acutus and rhodostoma are oviparous. Scale rows number 21 in most of the species, but only 17 in hypnale and nepa. The scales are keeled in most species, but are only faintly keeled in hypnale and nepa and are smooth in rhodostoma. In acutus and to a lesser extent in hypnale and nepa the snout is drawn out into a pointed dermal appendage covered with small, irregular scales. A. hypnale is fur- ther peculiar in lacking hemipenial spines. A. rhodostoma and hyp- nale have been described as vicious while himalayanus and halys are inoffensive and will not bite even when handled. A. strauchi, mon- ticola and himalayanus are montane species of which the latter has been recorded at an extreme altitude of 16,000 feet. In all the species the colors run to grays and browns, with a series of blotches or rhomboidal or triangular or wavy markings on each side which seem to be hom*ologous with the bands of the copperhead. The wide geographic hiatus between the Old World species and AUTECOLOGY OF THE COPPERHEAD 115 those of the New World calls for comment. Schmidt (1946:150) mentioned Agkistrodon, along with various other reptilian genera (Natrix, Liopeltis, Elaphe, Leiolopisma [== Lygosoma], Eumeces, Ophisaurus, Alligator, Clemmys and Emys), which have represent- atives in both Eastern Nortli America and Eastern Asia and which seem to be remnants of a late Cretaceous or early Tertiary Holarctic fauna, which was forced southward and partly destroyed as north- em climates gradually became more severe in a trend that cul- minated in the Pleistocene glaciations. The genera mentioned are ancient and conservative and some of them are known to have changed but little throughout the Tertiary period when more pro- gressive groups of terrestrial vertebrates were undergoing rapid evo- lution.

The fossil record is too meager and too recent to shed any light on the evolution or relationships of the pit vipers. All known fossil pit vipers from North America are of Pleistocene age with the exception of those from Driftwood Creek, Hitchco*ck County, Nebraska, which have been tentatively assigned to the lower Plio- cene (Brattstrom, 1954:35). The fauna from these deposits includes

the and the . copperhead prairie rattlesnake ( Crotalus viridis ) The material to Brattstrom cit. consists copperhead according ( loc. ) of 31 vertebrae, which do not differ from those of Recent specimens. It is noteworthy that the locality is 160 miles farther west than the northern copperhead's present western range limits, and is also slightly farther north tlian the copperhead's range extends, except in the region of the Atlantic Coast. Evidence is provided that in earlier times the copperhead's range was perhaps more extensive and in any case extended well to the north and west of its present limits. Almost certainly, the deciduous forest habitat likewise ex- tended westward into this area which is now high plains. It is noteworthy also that fossils of copperheads and prairie rattlesnakes should be associated. At present these two species of crotalid snakes, whose combined ranges include most of the United States, overlap only slightly in Kansas, Oklahoma and Texas; their ranges are mainly complementary since their habitat preferences are en- tirely different. Presumably in the early Pliocene their habitats were less sharply segregated; also possibly the snakes themselves may have been less differentiated in their habitat preferences. Chaney and Ehas (1936:27) found that in the lower Pliocene the grasslands of the Great Plains were much less extensive than they are at present. Rainfall was estimated to be 15 inches higher than 116 Untversity of Kansas Publs., Mus. Nat. Hist. it is at present and supported a mesic deciduous forest some 180 miles farther west than the present limits of such forests. How- ever, the copperhead and all but one of its congeners are committed to a forest habitat, as are, predominantly, all the other genera of reptiles shared by North America and Eurasia. It seems that these genera, which Schmidt (1946:144) has termed "Old Northern," date back to a time in the early Tertiary when the North Pacific was bridged by land areas that were well forested. An early Ter- tiary flora dominated by deciduous forest is known from Alaska (Hollick, 1936:11), and a similar Miocene deciduous forest is known from the region of the Columbian Plateau (Axelrod, 1950: 230). Chaney (1947:147) described an Arcto-Tertiary deciduous forest widely distributed in the Northern Hemisphere, with striking uniformity throughout, as late as the Miocene. From the Miocene onward regional diversity developed, especially for the less cold- tolerant elements, for which climatic barriers developed. These Tertiary forests had many elements in common with the Recent Deciduous Forest Biome of southeastern North America. Probably this Recent deciduous forest was derived directly from the more northern Tertiary forests by gradual alteration, chiefly impoverish- ment, and retreat or shrinkage southeastward as a result of the trend toward cooler and drier climates. HABITAT

Usually associated with forests, the copperhead inhabits several types of deciduous forest climaxes and many of their serai phases. The species displays considerable versatility in adapting to varied habitats from swamp to desert and from sea level to high mountains over its wide range. Although geographic populations differ some- what ecologically, all have in common certain basic requirements. There is definite preference for ground that is shaded by a leaf canopy and blanketed with leaf litter from deciduous trees. Pref- erably this substrate should be wet or at least damp during the time that the snakes are active. However, copperheads may wander into brush, grassland, or weedy fields, and may prowl on a dry substrate.

The following briefly quoted statements from the literature and from field notes show the range of habitat preferences throughout the geographic range.

". . . chooses dark and shady places for its residence in general, though at times it is found in meadows of high grass." (Holbrook, 1838:71.) Northeastern states: ". . . rocky places, usually in the vicinity of mod- erately thick timber, marshy glades, or hollows." (Ditmars, 1907:422.) AUTECOLOGY OF THE COPPERHEAD 117

in the of thick timber. ". . . rocky ridges and ledges, usually vicinity fairly the often descend to the . . . During the height of summer they meadows and valleys." (Babco*ck, 1929:26.) "Ledgy, wooded hills with a base of wild, damp meadows, are the favorite prowling grounds of this snake, as it searches for small rodents, birds and frogs. During the summer it is often seen along old stone walls which might offer shelter and a congregating place for rodents." (Ditmars, 1935:24.)

Connecticut: ". . . meadows and low-lying groimd. . . . finds con- cealment in the rocky parts of the country, and still remains in the trap ridges of the Connecticut Valley. . . ." (Cope, 1900:1138.) ". . . ledgy hills, with base of wild, damp meadows or nearby heavy forest. During the smiimer it is often seen along old stone walls. . . ." (Lamson, 1935:26.)

". . . loose rock ledge. Cedars, pines, laurel and blueberry bushes are the predominant vegetation." (Finneran, 1948:124.)

New York: ". . . generally found in meadows, pastures, and the edge of woods." (Rafinesque, 1819:86.)

Pennsylvania: (Union County) ". . . mountains . . . [and] low- lands along streams." (Pawling, 1939:169. ) (Venango County) ". . . top of hills in August (many are killed cutting wheat) . . . back into the valley in September. In July there seem to be more along the weedy shore of the river than on the rocky hillside." (Swanson, 1952:176.)

Maryland: (Harford County) ". . . rocky mountain sides, ledges, and accumulations of talus. . . . During the summer . . . about stone walls, the edges of fields, old foundations, sawdust piles, and in wooded areas." (McCauley, 1945:131.)

Virginia: (Giles County) ". . . common in valleys throughout the county." (Hutchison, 1956:85.) (Stafford County) ". . . very abundant

. . . found on rocky hillsides bordering Aquia Creek and in the slab piles of abandoned saw-mill sites in the woods. A number were found, however, along road sides and in open fields." (Lynn, 1936:170.) (Princess Anne County) ". . . dense deciduous woods. . . ." (Werler and McCallion, 1951:251.)

North Carolina: ". . . old building sites, rock piles such as old stone fences and brier thickets in former clearings. It has been taken several times from under the fallen bark about the bases of dead chestnut trees. . . . not . . . above 2500 ". . . usually feet."^^ (King, 1939:577.) edge of a cypress swamp adjoining open fields." (Robertson and Tyson, 1950:143.) (Dismal Swamp region) On pine needles in trail, beside drainage ditch in woods of cypress and pine, with ferns and canebrake. Under log at edge of cornfield near wet woods and large water-filled ditch. Trail overgrown with honeysuckle, behind an abandoned shack. On dirt road a few hundred feet from extensive cypress swamp. In farmland, on road beside a drainage ditch bordered by thickets of honeysuckle and canebrake. In deep swamp, on road bordered on one side by deep drainage ditch and on the other by eight-foot- high impenetrable canebrake. Dirt road beside drainage ditch. Coiled on boards of bridge over drainage ditch, near sawdust pile overgrown with honey- suckle, cypress swamp beyond it. (Extracted from unpublished field notes of Barry Rothman and Norma Rothman. ) Ohio: "It frequents low swampy places in hilly regions." (Morse, 1904: 137.) ". . . in a variety of habitats . . . almost anywhere in un-

the . glaciated Ohio. ... in the valleys ( sometimes along streams ) . , hillsides to their summits and out over the farm land or woods on the flat- topped hills. Sometimes they were in cleared open country and sometimes in but more often were taken in heavy "woods, they scrubby second-growth or brush (Conant, 1938:110.)

Indiana: "hilly locations . . . common where there is timber with rock outcrops." (Minton, 1944:474.) "Dry, rocky wooded ridges are the preferred habitat, and the species may be quite numerovis locally." (Minton, 1951:318.)

3—4428 118 University of Kansas Publs., Mus. Nat. Hist.

Illinois: "The copperhead is most generally found in rocky, wooded areas, although it seems to like the proximity of water, perhaps because of the greater abundance of food." (Necker, 1939:36. ) ". . . common only in wooded, rock areas." (Smith, 1953:5.) Kentucky: (Mammotli Cave region) "Abundant, especially in damp woods." (Hibbard, 1936:281.)

Tennessee: ". . . high, dry and rocky regions. . . ." (Allyn, 1937: 220.) ". . . wooded uplands and hills. None have been found in low- lands or river bottoms." (Parker, 1948:28.) "It is most likely to be en- countered in woodlands since it has a preference for hilly or low mountainous country." (Gentry, 1956:248.)

Alabama: ". . . frequents rocky territory', often being fovmd by turning over boulders." (Haltom, 1931:94.)

Mississippi: (George County) ". . . swampy tj^e of country near Basin. . . ." (Allen, 1932:2.) (Jackson County) ". . . . specimens were taken in a deciduous woods of an unusually mesophytic nature for that region of pine meadows." ( Smith and List, 1955 : 123. )

Iowa: ". . . wooded, rocky bluffs on the Mississippi and lower Des

Moines rivers. . . ." (Bailey, 1941:1.)

Missouri: ". . . moderately common on rocky hillsides near streams.

. . . often taken about dwellings. . . ." (Boyer and Heinze, 1934: 198.)

Louisiana: (northeastern "Hill Parishes") ". . . most abundant in woody glades which lead back from swamps into the highlands. They are not uncommon, however, in swamps and marshes overgrown by trees. During the late summer they are frequently found in paths which border cultivated fields or wooded pastures. Habitat distribution was as follows: woody glades, 25; wooded swamps, 22; paths-field, 9; paths-pasture, 5." (Clark, 1949:258.) (Vernon Parish) "wooded bottomland" (Fitch. 1949:89.) (Lafayette Parish) "rare" in Highland Woods habitat, not recorded at all in the three other habitats listed, namely Swamp, Flood Woods, and Grassland. (Liner, 1955:41.)

Nebraska: (Richardson County) ". . . common along the heavily wooded Missouri River bluffs immediately south of the mouth of the Big Nemaha River." (Hudson, 1942:83.)

Kansas: (Riley County) ". . . among the rocks and vegetation be- side riffles of Wildcat Creek. They have also been found under flat, hillside rocks and in grassy, wooded bottom land." (Burt, 1927:8.) (Doniphan County) "among rocks on the bluff . . . under a pile of cottonwood slabs.' (Linsdale, 1927:81.) ". . . wooded areas, generally on hill- sides where rock is exposed. . . . Extremely heavy woods are not in- habited, for there is insufficient penetration of the sun between the trees to warm the snakes in spring and fall." (Smith, 1956:306.) (Osage County) Oak-walnut hillside forest, cultivated field, Buckbrush-sumac and Prairie habi- tats found to be used, in that order of preference. (Clarke, 1958:23.) "It is most frequently found in the vicinity of rocky ledges in oak-hickory woods.

. . ." (Clarke, 1959:7.)

Oklahoma: (Tulsa County) ". , . rocky, wooded bluffs and ridges." (Force, 1930:37.) (Marshall County) "... more common in the postoak-blackjack oak uplands than in the lowlands." (Bonn and McCarley, 1953:470.)

Texas: ". . . timber that borders our rivers and creeks; always selecting land that seldom or never overflows. They hide under logs, in de- cayed stumps, in holes dug by small animals." (Mitchell, 1903:27.) ". . . most common in rocky areas of mountainous country as well as in the wooded bottomlands. During the spring in some parts of the state they are found in numbers along streams and other moist areas where they spend most of the day hidden beneath decaying logs and other debris. . . ." (Werler, 1950:7.) (Terrell County) "Three specimens were taken from the mesquite- creosote association, three from the mesquite-sumac-condalia, two from the AUTECOLOGY OF THE COPPERHEAD 119 walnut-desert willow, 76 from the live oak, and five from the salt cedar associa- tion. One . . . was collected ... in the hackberry association. . . ." (Milstead, Mecham and McClintock, 1950:557.) (Grayson County)

". . . fairly common in the rocky brushland surrounding the lake [Texoma]. . . ." (Bonn and McCarley, loc. cit.) Dallas County ". . . wooded areas in hilly and lowland regions ... in spring it is found most frequently along the river under logs, pieces of tin, and boards." (Curtis, 1949:12.) In Trans-Pecos Texas the copperhead is represented by several small and disjunct relict populations in the Chisos and Davis Mountains, in canyons where there is mesic vegetation, and in live-oak groves along tributaries of the Pecos River. These far western populations exist partly by virtue of their increased tolerance for xeric conditions. In a typical locality in the Chisos Mountains, in Oak Canyon near its mouth, there is a grove approximately 100 yards long and 20 to 100 feet wide, of willow (Salix interior), oaks (Quercus robusta, Quercus sp. ), walnut (Juglans rupestris), hackberry (Celtis occidentalis), buckeye (Aesctilus sp.), persimmon (Diospyros texana), fragrant sumac (Rhus trilobata) and grape (Vitis arizonica). In places there are small accumulations of leaf litter; for the most part the ground is bare and rocky. Even within the grove there is xerophytic vegetation with such typical desert species as catsclaw {Acacia greggi) and prickly pear (Opuntia sp.). At the time of my visit in July, 1957, many of the larger trees were dead as a result of drought, and diversion of the limited water supply. Largest trees in the grove were oaks and willows approximately two feet in trunk diameter. Some of these canyons have endemic species and varieties of trees (especially oaks ) and other plants occurring as relicts, attesting to a long period of isolation since the climate has deteriorated and unfavorably xeric conditions have de- veloped in this general region. Some of the relict colonies of copperheads exist in situations where the habitat is so restricted as to support only a few individuals. In the Chisos Mountains suitable habitat probably totals less than one square mile; in the Davis Mountains there are more extensive scat- tered groves, mainly of live-oak (Quercus emoriji), and the available habitat doubtless totals several square miles.

On the Reservation in Douglas Count>', Kansas, the areas of rock ledge that are most frequented in fall by copperheads that are preparing to hiber- nate are also among the most likely spots to find these snakes at any time in summer (Plate 14, fig. 2 and Plate 15). It is obvious that some individuals remain in the vicinity throughout the summer, while others disperse for varying distances. Those individuals that have wandered far from the ledges return at diflFerent times. According to the figures obtained from my ten years of live-trapping, activity along the ledges attains a high level in the last week of September, reaches a peak in mid-October and tapers off abruptly in the last week of October and in early November. In autumn when the snakes are concentrated along the ledges there is a thick layer of new leaves on the ground, as throughout September there is some shedding of leaves, and this process continues at an accelerated rate in October until the latter part of the month when few leaves remain on the trees. The layer of loose leaves provides concealing cover which is effectively utilized by the snakes. Even though they are concentrated along the ledges in a density that might represent several hundred per acre, and are mainly diurnal in their activity at this time of year, they are rarely seen. In walking hundreds of miles along the ledges at the time of year when copperheads are most concentrated there, to check lines of live-traps, I have seen the snakes so rarely (except

for those actually in the traps ) that I would not have reahzed their abundance. Copperheads become diurnal in autumn when nocturnal temperatures are un- favorably low. Ledges that have southward exposure are optimum habitat. 120 University of Kansas Publs., Mus. Nat. Hist. but many that have predominately eastward or westward exposure are just as much frequented on the Reservation, and some that have partly northward exposure are frequented. Favorable characteristics of the ledge itself may outweigh the disadvantage of an exposure that is not optimum for receiving the maximum amount of sunshine. In summer, when copperheads have dispersed from the ledges, they oc- cupy almost every terrestrial habitat on the Reservation, but are unevenly distributed as some habitats are much preferred over others. Approximately half of the Reservation's area is woodland, and the other half is chiefly grass- land, much of which is in process of transition to brush or forest as plant succession progresses (Plate 14, fig. 1). Interspersion of habitats is so great that the home range of every individual copperhead encompasses a variety of habitat divisions. Effort to capture the snakes was most concentrated in those areas that had proven most productive previously. Numerical com- parisons of habitat preferences are not possible with the data on hand, but my impressions of the relative degree of use of several habitats best represented on the area are as follows.

1. Most preferred habitats, having concentrated populations of copperheads throughout the summer. with a. vicinity of intermittent streams, groves of elm ( Ulmus sp. ) , locust Cottonwood {Populus deltoides), (Gleditsia triacanthos) , various other trees, brush, including blackberry {Rubus argutus) and ground cover of various herbs and grasses. b. fence row lined with brush and saplings of elm, osage orange ( Mac- lura pomifera), locust, crab apple (Pyrus ioensis), and plum {Prunus americanus) . c. vicinity of pond, with grove of willow (Salix sp. ), and with dense ground cover of smartweed (Polygonum sp. ), day flower (Commelina communis) and rice cut-grass (Leersia onjzoides). d. upland thickets of ehn, locust, osage orange, crab apple, plum, sumac (Rhus glabra) and oak (Quercus prinoides) at edges of grassland dominated by brome (Bromus inermis) or blue-stem (Andropogon sp. ). 2. Less preferred habitats, having sparser populations. c. woods of oak-hickory (Quercus sp. and Carya ovata) and hack- berry (Celtis occidentalis). b. xeric thorny woodland of osage orange and honey locust with dense undergrowth. c. mesic woodland of elm, ash (Fraxinus americanus) coffee-tree (Glymnocladus dioica), redbud (Cercis canadensis.) d. weedy pasture, with brome, ironweed (Vernonia interior), vervain (Verbena stricta) and germander (Teucrium canadensis) . e. fallow fields dominated by weedy grasses, foxtail (Setaria sp.), and crabgrass (Digitaria sanguinalis). 3. Least preferred habitats, that are generally avoided, or are used only because they are adjacent to more favorable areas. a. mesic fallow field in an early stage of succession, with weedy vegeta- tion dominated by giant ragweed (Ambros-ia trifida) and sunflower (Helianthus annuus). b. more xeric fallow field, with common ragweed (Ambrosia artemisii- folia), three-awn grass (Aristida oligantJia), and lespedeza (Les- pedeza striata). c. cultivated field, either ahnost barren of vegetation or with com or milo. d. road.

The extent to which the less favorable areas were avoided was indicated by experience in collecting copperheads while driving over county roads at night, AUTECOLOGY OF THE COPPERHEAD 121 at times when weather conditions were most favorable for the snakes to be active. A ten-mile drive sometimes disclosed one or more of the snakes, but more often none was seen. It is estimated that in the course of a ten-mile drive approximately 25 acres of road were scrutinized, and if copperheads had been as numerous on the roads as they were estimated to be in nearby favorable habitats, more than 100 should have been secured on each such drive. RANGE AND GEOGRAPHIC VARIATION

The copperhead occurs throughout most of the southeastern one-fourth of Nortli America but is absent from peninsular Florida and enters that state only along its northern edge. The species is small confined chiefly to unglaciated regions, but has locally made scale penetrations into glaciated areas in Massachusetts, Cormecti- Mis- cut, southern New York, Pennsylvania, Ohio, Indiana, Illinois, souri, Nebraska and Kansas (Fig. 12). Most records from Pennsylvania are from the southern half of tlie state and the species is generally absent from the northern tier of counties. According to Smith (1945:70) the copperhead is absent from most of the glaciated areas of the state, but is begin- ning to penetrate them using the stream valleys as migration routes. He cited 14 records within glaciated areas, including those of Wis- consin Drift, Illinoian Drift and Pre-Illinoian Drift. In Ohio most records are from the unglaciated area of the Al- legheny Plateau or from near its borders, chiefly from vidthin the southeastern quarter of the state or from its southern edge. Re- corded occurrences in the glaciated areas are in or near the valleys of large streams (Conant, 1951:109 and 254). The copperhead is distributed over somewhat less than tiie southern half of Illinois, but with notable northward extensions in tlie main valleys of the in tliese main Mississippi, Illinois and Wabash rivers. Except of river valleys the species is limited to the part of the state south the Shelbyville Moraine. In Missouri the distribution has not been thoroughly investigated, but most records are in the southern two-thirds—near the Missouri

River or south of it. Along the Missouri River the range extends northward barely into the southeastern comer of Nebraska. In Kansas the species is limited to the eastern third of the state. It is abundant in several tiers of eastern counties but becomes increas- ingly scarce and localized farther west. The species reaches its northwestern Hmit in the Big Blue River drainage of Gage County, Nebraska. In Oklahoma the range is approximately the southeastern half, in oak-hickory woodlands. In eastern Texas the copperhead is generally distributed. It extends west across the central part of the 122 University of Kansas Publs., Mus. Nat. Hist.

state in oak woodland, and is common on tlie Edwards Plateau of west-central Texas. Farther west isolated populations occur in live-oak woods along streams in the Stockton Plateau, and in iso- lated deciduous forest relicts at higher altitudes in the Chisos and Davis mountains of Trans-Pecos Texas. In general, the copperhead's distribution corresponds to that of

100 Ti

-^..^

I 45 I "I

I

I

..(

1 *-k

^ ^-^J^' -36

._! cii V

Fig. 12. Range of the copperhead, showing marginal and near-marginal records, based upon a map shown by Gloyd and Conant (1943:153) but including additional records from Pennsylvania (Smith, 1945:70), Ohio (Conant, 1951:275), Illinois (Smith, 1953:2), Kansas (Smith, 1956:305), Oklahoma (R. G. Webb, unpublished thesis in the University of Oklahoma Library), Texas (Brown, 1950:212-213; Milstead, Mecham and McClintock, 1950:557), and Mississippi (Allen, 1932:12; Smith and List, 1955:123).

the Deciduous Forest Formation or Biome of the southeastern one- fourth of the North American Continent. Most of the Formation's associations, including the Mixed Mesophytic, Western Mesophytic, Oak-Chestnut, Oak-Pine, Oak-Hickory and Southeastern Evergreen (Braun, 1950), are mainly or entirely within the copperhead's range. Only the two most nortliem associations, the Beech-Maple AUTECOLOGY OF THE COPPERHEAD 123 and Maple-Basswood, are largely outside the range, as is the eco- tone of "White Pine-Northern Hardwoods" that is transitional to the northern Taiga. The copperhead's subspecies correspond roughly to the subdi- visions of the Deciduous Forest Biome. The southern subspecies, A. c. contortrix for the most part coincides with the Oak-Pine and Southeastern Evergreen associations. The northern subspecies A. c. mokeson coincides with the Oak-Chestnut, Mixed Mesophytic and Western Mesophytic associations in the East and with part of tlie Oak-Hickory Association in the West. The western, or broad- banded copperhead occurs chiefly within the southern part of the Oak-Hickory Association. The Trans-Pecos copperhead, A. c. pic- but tigaster occurs entirely outside the Deciduous Forest Biome, in small relict populations coinciding in distribution with isolated relicts of deciduous forest. The four recognized subspecies differ from each other, so far as that be known, chiefly in characters of color and pattern may adaptive to the different types of backgrounds in the several are types of forest climaxes where they occur. The differences of not striking as compared with those in some other species snakes, but are consistent and well defined. The northern cop- perhead, A, c. mokeson, is characterized as being reddish brown or chestnut, with relatively little contrast between ground color and the superimposed darker markings. The latter are hour- and have glass shaped—constricted mid-dorsally, widened laterally, rounded lateral edges. The ventral pattern is of more or less mottled distinct, subcircular blotches. The belly is usually dark, with gray or black. Gloyd and Conant (1943:150) mentioned in small or irregular spots between the crossbands some popu- Dutchess lations, and their photograph of a specimen from County, New York, has this type of pattern, which was not ob- served in the population that I studied. The snakes of the eastern United States seem to attain much larger size than those from any other region. It is notable that eastern and western popula- tions of mokeson are disjunct, separated by the subspecies contortrix where the latter extends north along the Mississippi River and its the northern tributary, the Illinois River as far as central Illinois, limit of the species' range in that region. The subspecies contortrix differs from mokeson chiefly in paler coloration, pale brown or tan, often with a pinkish tinge. The dorsal crossbands contrast strongly with the paler ground color, but they 124 University of Kansas Publs., Mus, Nat. Hist, shade into a paler hue in their central parts. They are strongly constricted middorsally. The belly is pale, not heavily marked. A. c. laticinctus is bright chestnut, or hazel brown, with strong contrast between the ground color and the darker crossbands. The latter differ from those of mokeson and contortrix in lacking mid- dorsal constrictions and in extending laterally to the ventrals (with no of their with the ventrolateral rounding edges ) , blending pattern of three more or less conspicuous spots to each crossband. This sub- species seems to be relatively small in average and maximum size.

A. c. pictigaster (Plate 16, fig. 1) resembles laticinctus in its dorsal pattern, but on the ventral surface there are bold and con- trasting dark markings continuing as extensions from the dorsal crossbands at both their anterior and posterior ends and with a U- shaped light area enclosing still another dark area on the ventral surface beneath the middle of each dorsal crossband. A. c. pic- tigaster seems to be a dwarfed race; it has one scale row slightly shortened, and on the average has several more subcaudals than have the other subspecies. There are indications of ecological differences between widely separated geographic populations, but the available information is inadequate to define these clearly, or to show whether they follow subspecific boundaries. A. c. contortrix of the southeastern states has often been found in a swampy habitat and prefers situations that are definitely more mesic than those frequented by the western and northeiTi subspecies. Both of the western subspecies, and especially pictigaster, are relatively tolerant of xeric conditions, although they are closely confined to woodland—of limited extent in the regions where they occur. Frogs seem to be far more im- in of contortrix in that of at portant the diet than mokeson ( least of its western representative). BEHAVIOR Crawhng

Klauber (1956:331-350) has described and explained in detail the modes of progression by crawling in snakes, particularly with reference to rattlesnakes. Four distinct types of locomotion are: horizontal undulatory, rectilinear, sidewinding, and concertina. Most snakes are capable of employing two or more of these types of progression. The mode of travel depends on the kind of snake, size of the individual, type of substrate, degree of excitation, and other factors. Horizontal undulatory locomotion is the most prev- AUTECOLOGY OF THE CoPPERHEAD 125

alent type in the majority of snakes, and is also exclusively used by limbless lizards such as the glass "snake" (Ophisaurus) , Travel- ling by this method, the snake's body is thrown into several lateral undulations or waves, conforming in its contours with irregularities in the ground surface upon which it rests. Pressure is exerted simultaneously on the outside and posterior surface of each curve, providing the force which drives the body forward on its course. On a smooth surface where suitable pressure points for pivots are lacking, the snake's lateral undulations are largely inefiPectual in causing it to move forward. In horizontal undulatory locomotion the distance gained by the snake is somewhat less than that actually travelled, because of the lateral motion. Each point along the snake's length tends to move along the same undulatory course, but actually the lateral movements are most pronounced in the anterior part of the body and least so in the head region. The mechanics of this type of crawling are complex. In the copperhead crawling is accomplished chiefly by means of horizontal undulatory progression. When there is cause for haste, the copperhead relies on this method exclusively. More leisurely locomotion may be partly or entiiely of the rectilinear type. This mode of progression is especially character- istic of large, heavy-bodied snakes such as boas, pythons, and large vipers. It depends upon the loose attachment of the skin to the body, with a powerful dermal musculature. In typical rectilinear locomotion the snake's body is extended in a straight line and seems to glide forward effortlessly. The weight is not evenly distributed over the ventral surface, but is supported on several well separated points. Along the intervening parts of the body, imperceptibly raised from the substrate, the loosely attached skin slides forward over the body. As each point upon which the weight is borne shifts posteriorly in a flowing fashion, the body is pulled forward within the sheath of skin. In crawling slowly copperheads often have the body extended almost straight, relying largely on the rectilinear type of locomotion. More often a combination of the rectilinear and horizontal undulatory types are employed. Sidewinding is a third type of progression; it is well developed only in the sidewinder rattlesnake (Crotalus cerastes) and in a few species of true vipers, all heavy-bodied snakes specially adapted for locomotion over a smooth surface of loose sand. In this peculiar type of locomotion most of the snake's body is held arched clear of the ground as it glides along with a rolling motion, with only two 126 University of Kansas Publs,, Mus. Nat. Hist. points in contact at any one moment. The track consists of a series of parallel lines. This type of locomotion is the nearest approach actually attained to that of the "hoop snake" of folklore. The hoop snake was supposed to take its tail in its mouth and roll down hill like a hoop. If a hoop is broken, and the free ends are pulled in opposite directions until the shape is stretched to a spiral with two loops, and if this spiral is rolled over a smooth surface, the motion resembles somewhat that of a sidewinder. The resemblance would be increased if the hoop were made of such flexible material that the loops sagged to an elliptical shape. Although few kinds of snakes use sidewinding regularly, others resort to it in emergen- cies, as when escaping over a smooth surface unsuitable for horizon- tal undulatory progression. These kinds include several species of rattlesnakes and even garter snakes, but in most of them side- winding efforts are crude. In the stubby and clumsy copperhead sidewinding locomotion is never well defined. However, indi- viduals startled as they are crossing roads or other open places with a relatively smooth substrate may make lunging movements, with part of the body off the ground, progressing, though inefficiently, in a manner that may be considered primitive sidewinding or its precursor. In the fourth method, concertina progression, the snake alter- nately anchors itself at the anterior end drawing the body forward in several sinuous curves, and then by straightening the body ex- tends itself out full-length anteriorly. In each cycle the snake ad- vances by the difference between the lengths in its straightened and waved positions. Concertina progression is not the regular method in any snake, but is used especially in a slow cautious ad- vance, as in the stalking of prey. I have never seen concertina progression used by the copperhead.

Coiling

The copperhead spends most of its time in a flattened pancake- like coil which is characteristic for the species. In this coil the tail is outermost and the body is compactly wound in from one to more than two complete cycles. Near the anterior end of the snake the direction of the coil is reversed, and the head and neck, near the center, assume a U- or S-shape. From this position the snake is able to strike, in a short jab, but ordinarily, upon the ap- proach of prey or an enemy, it would make preparatory movements including a slight raising of the forebody and a shifting of the coils to bring more of the length into the anterior loop, thus lengthening AUTECOLOGY OF THE CoPPERHEAD 127

the potential range of an eflFective stroke. Ordinarily the snake seeks a sheltered spot in which to coil, concealed under a rock or in leaf litter or in dense growing vegetation. Probably food is most often obtained by snakes that are lying in resting coils, wait- ing to ambush approaching prey. Before the prey has approached within reach of the snake, the latter is alerted by sight, scent, sub- strate vibrations, or the radiation receptors of the heat-sensitive

pits. The slight movements required to prepare for a stroke would not be readily noticed by the approaching victim. Copperheads kept in outdoor enclosures were remarkably ef- ficient at concealing themselves in their resting coils. In one pen which enclosed 100 square feet prolonged search was often neces- sary to find snakes that were not coiled beneath several wooden shelters provided for them. Usually they were found nestled amid screening vegetation. In their compact, flattened resting coils, copperheads presented a minimum surface to be seen, especially by a small animal approaching on the same horizontal plane. Since it was considered desirable to avoid unnecessary disturbance, and since hazard was involved in searching with hands or face near the ground, such snakes were often missing for periods of days. Those kept under observation were often coiled in just the same position over periods of hours, or even for several days. Those that were digesting a meal or were approaching the time of shedding were especially sedentary. From the behavior of these captive individuals under conditions simulating those in the wild it may be concluded that a copperhead often remains for 24 hours or more in the same spot. Movements within the enclosure were often motivated by changing conditions of sunshine and warmth within the daily cycle as the snakes sought to maintain body temperatures near their optimum level. On many occasions snakes kept in the enclosures were observed to return to the same spot to resume a resting coil after wandering about the cage. Any one spot might be used with some regularity for a week or more, but eventually would be deserted in favor of another. In the cages the number of potential resting places was limited. The nest or "form" shaped by a snake coiled in one place for a long period would naturally provide an ideal site for oc- cupancy on a later occasion. Under natural conditions, with free- dom of movement so that the snake wanders much farther in a period of foraging, and with abundant shelter on all sides it is doubtful whether an individual returns to the same spot with any regularity. On the few occasions when I have caught a copper- 128 University of Kansas Publs., Mus. Nat. Hist.

head twice in the same trap, the interval was short and it seemed probable the snake had blundered back into the trap after release, without having left the vicinity. Otherwise I have rarely found the same copperhead twice at any one spot. It has not been demon- strated whether the occasional aggregations of gravid females have any permanence but from the appearance of one shelter where an aggregation was found, I judge that it had been used for several days at least, and that the individuals involved may have left and returned again. Copperheads that are travelling move slowly. The motion is so gradual and so smooth-flowing that the snake remains extremely inconspicuous against its normal background, and might ambush prey almost as effectively as when it is in a resting coil. The rate is of course variable, but would be measured in yards per hour rather than in miles per hour. The crawling snake is seldom in motion for more than a few seconds without stopping for a longer or shorter period. Usually much more time is spent in the intervals of pausing than in motion. The route is usually circuitous. Swimming

Unlike its near relatives, the cantil and the cottonmouth, the copperhead has no special affinity for water. However, it does favor damp situations and many authors have mentioned its preference for mesic or riparian habitats. In captivity copperheads have been observed to coil in the water containers in their cages, in response to air temperatures that were either uncomfortably high or uncomfortably low. Like most snakes, the copperhead swims well and occasionally it enters the water voluntarily. In Penn- sylvania, Hudson (1954:72) recorded one found swimming across Unami Creek. Smith and Sanders (1952:214) noted one swimming across Lake Texoma, 500 yards from the Texas shoreline.

Climbing

Although it is obviously lacking in scansorial adaptations, the copperhead, like various species of rattlesnakes, has on occasion been recorded climbing in trees or bushes. Those I kept in an out- door enclosure sometimes climbed several feet off the ground in vines that were intertwined through the wire on the sides of the cage. Of the two Trans-Pecos copperheads that I collected at Independence Creek, Terrell County, Texas, in June, 1957, one was climbing two feet above ground among the roots of an up- rooted Uve-oak. Wright and Wright (1957:916) mentioned an AUTECOLOGY OF THE COPPERHEAD 129 individual of this subspecies found coiled in the fork of a live-oak four feet above the ground in the Davis Mountains. Vernon Mann told me that his brother was bitten when he placed his hand in the crotch of a tree six feet above the ground and unknowingly touched a copperhead that was inconspicuously coiled there. The Marais des Cygnes River of Kansas had flooded the woodland where this accident occurred, and probably the copperhead had climbed into the tree to escape the rising water. Johnson (1948:214) wrote of finding a half-grown copperhead coiled in the uppermost branches of a small tree on July 10, 1948, in McLennan County, Texas, and on other occasions found individuals four to five feet above ground level in piles of driftwood. Swanson (1952:176) mentioned find- ing several young copperheads climbing in laurel bushes a foot or more above the ground, in Venango County, Pennsylvania. Curtis (1949:12) wrote of finding several copperheads in Dallas County, Texas, climbing in low shrubs and trees after dark in mid- July, ostensibly to catch cicadas. Wm. Cutter told me of making similar observations in Marshall County, Oklahoma. When cicada nymphs are emerging from the ground in large numbers, and climb- ing stems and tree trunks, ready to metamorphose, copperheads may be stimulated by sight, scent or sound to climb after them, abandoning their usual tactics of waiting in ambush, in favor of active pursuit. The nymphs, and the newly metamorphosed adults of cicadas that are not yet completely dry, are so slow and clumsy that, unlike most prey, they could be readily overtaken and caught, even by the slow-moving copperhead. Judging from a few brief statements in the literature, some other species of Agkistrodon have the tendency to climb more strongly developed. Smith (1943:500) considered A. hypnale of India to be partly arboreal as it often climbs into low bushes. Koba (1938:247) studied an insular population of A. halys oflF the coast of southern Manchuria, and found that these snakes often climbed into trees or weeds, and fed mostly on birds.

Disposition

Widely diflFerent opinions have been expressed in the literature as to the copperhead's disposition; some writers have described the species as docile and inoffensive whereas others have considered

it the of : personification villainy. Atkinson ( 1901 152 ) described the copperhead as a ". . . sullen and treacherous snake, its disposi- tion is to remain concealed and it will not strike unless closely pressed or trod upon." Mitchell (1903:27) wrote of copperheads 130 University of Kansas Publs., Mus. Nat. Hist.

in Texas that tliey were rather lazy and sluggish until thoroughly aroused, but then became vicious. Branson (1904:421) wrote that in Kansas, "It strikes without warning and seems to be always on the lookout for something upon which to use its fangs." Morse (1904:137) also wrote, "It strikes, when approached, without warn-

ing. . . ." Surface (1906:186) wrote, "There is no creature more treacherous, despicable nor dangerous in this State [Penn- sylvania] than the Copperhead Snake. It lurks in bushes or grass or among stones, and strikes without warning and often without provocation." However, Brimley (1923:114) wrote that it is a

"gentle snake, much less aggressive or vicious . . . than most of our harmless snakes." Amaral (1927:70) stated that the copper- head is a rather vicious snake and strikes in any dnection without warning. Babco*ck (1929:26) described the copperhead as having

a ". . . shy and retiring nature, rarely becoming aggressive." Ditmars (1935:23) averred that it would seldom strike unless stepped upon, or otherwise attacked or annoyed. In Indiana, Min- ton (1944:474) noted marked individual differences in temperament and wTOte that individuals had been touched or even tiodden upon without sb'iking, whereas others would strike with but little provo- cation. McCauley (1945:132) wrote, "I have never seen a truly aggressive specimen." Oliver (1958:45) wrote tliat most copper- heads are quite mild and inoffensive. My own observations in general bear our Oliver's statement. In the many encounters with copperheads experienced by my asso- ciates and by myself on the Reservation, the snakes never behaved aggressively, but would attempt to defend themselves only when tliey were threatened or restrained. Even under these conditions the snakes sometimes did not strike or struggle when they were held do'ATi, grasped, and handled. When I discovered a copper- head in a funnel ti'ap, I would remove the end of the ti'ap and shake out the snake onto the ground. Usually the snake drew back into a coil and remained almost immobile until I was ready to handle it, but upon being grasped it would struggle to gain its free- dom by writhing and tlirashing, emitting jets of musk, voiding tlie contents of the cloaca, and making vigorous attempts to bite. An occasional copperhead found foraging at night away from shelter, showed more animated defensive behavior and even at- tempted to regain the nearest shelter by moving toward me with threatening lunges. Besides lack of the rattle, other sematic be- havior—hissing, inflating tlie body, assumption of menacing posture AUTECOLOGY OF THE CoPPERHEAD 131 with much of the body held clear of the ground, protrusion and slow, waving motion of tlie tongue—is much less developed than in ratdesnakes. Hence the accusation that copperheads strike with- out warning is not wholly unfounded. On one occasion, as I re- moved the dry grass covering a funnel trap, a small copperhead inside struck against the wire, and one fang scratched my knuckle. The only preliminaries to delivery of a bite may be a quick co*cking of the head toward the disturbing object, and perhaps a slight shifting of the coils. No effort was made to make pets of the copperheads that were kept in captivity from time to time, but such individuals even if bom in captivity, remained unpredictable in disposition, and any attempt to handle them without the usual precautions would have been fool- hardy. Although well adjusted to the presence of persons in the room where the cage was kept, the snakes were easily annoyed by any disturbance near their cage, and would prepare to strike or, on occasion, would strike against the screen or glass side of the cage at a nearby object. Both in confinement and under natural conditions there were fairly consistent differences in temperament between the sexes. Males, especially older individuals were more irritable and aggressive. Gravid females were much more docile than were other copper- heads. Combat Dance

The so-called combat dance has long been known in snakes, but until recently it was most often misinterpreted as mating behavior. Klauber (1956:671) expressed the opinion that the combat dance had only occasionally been mentioned in the literature because observers had nearly always assumed the intertwined snakes were actually mating, and may have avoided the subject, except in "the more Kinsey-like types of scientific publications." However, judg- ing from the frequency with which the sexual behavior of snakes has been described in the literature, the combat dance must be a relatively rare phenomenon. Combat dance has been described in many kinds of snakes, in- cluding colubrids, elapids and viperids, but seems to be best developed and most often observed in the crotalids, in which it has been well described by Shaw (1948:145), Gloyd (1947:3), and

Klauber loc. cit. . ( ) A typical combat dance occurs when two adult male crotaHds (nearly always of the same species) meet, and one, more aggressive, challenges the other, which accepts the challenge. 132 University of Kansas Publs,, Mus, Nat. Hist.

Facing each other the two snakes rear with their forebodies erect in a posture reminiscent of that of an angry cobra. As they come together their ventral surfaces are firmly adpressed, and the support thereby gained permits each snake to rear higher than it could other- wise. Swaying unsteadily in this position they intertwine their necks with slow, writhing movements, until one snake, momentarily gaining a favorable position, suddenly and violently contracts its body against the other in such a way that the opponent is thrown o£F balance and hurled forcibly against tlie ground. The snake that has thrown the other may follow up its advantage by moving onto the opponent and pressing him against the ground, thereby hindering and delaying his recovery, but the struggle, once joined, usually continues through many falls, with the same individual consistently playing the part of aggressor. In observed instances no perceptible damage to either opponent has resulted. Eventually one of the combatants may become discouraged and failing to respond to a renewed challenge, moves off whereupon the other does likewise without molesting him further. In captivity indi- viduals that have engaged each other in combat dance are likely to repeat the performance frequently over periods of weeks. The true significance of the combat dance, and its motivation are still poorly understood. In a few observed instances a female has been present when males were struggling, but most often, both in captivity and under natural conditions, no female was present and it seems unlikely that the combat is motivated primarily by sexual rivalry or to establish priority in mating. Sutherland (1958:23) related an instance of two adult males of the timber rattlesnake in captivity engaging in combat dance after both had grasped the same morsel, a dead blackbird, and each was obviously angered by the other's attempts to appropriate the meal. After one snake was vanquished, the other returned and ate the bird. On a later occasion Sutherland observed a smaller rattlesnake and a copper- head both interested in the same mouse. "They exhibited extreme agitation, inflating their bodies, emitting musk and v/eaving about with bodies elevated and necks arched." Unfortunately on this occasion one of the snakes was distracted by another mouse before there was opportunity for a typical combat dance. in lift. Somewhat similar behavior was described by Sutherland ( ) in a male copperhead whose courtship was interrupted. "While the male was courting, the observer gently stroked his body with a pair of long forceps. The male became agitated and moved AUTECOLOGY OF THE COPPEKHEAD 133 spasmodically. He inflated his body, emitted musk, and arched his neck. His movements became more violent with each stroke of the forceps, until the anterior part of his body was elevated six or eight inches above the floor of the cage. The same behavior pat- tern was demonstrated when another snake crawled over his body in the course of the courtship. Generally he returned immediately to the female when he was left undisturbed after these interrup- tions." Charles E. Shaw, in a letter to Sutherland, expressed the idea that the dance was an exhibition of, and a defense against, hom*osexual- ac- ity. Support for this idea is to be found in Shaw's (loc. cit.) count of the dance in rival male rattlesnakes (Crotalus ruber). The contest began after one male crawled onto the other, lying along him, facing in the same direction in a position similar to that assumed in mating. The lower male then raised the anterior third of his body and turned to face the opponent, who then also raised his forebody. cit. the observations of Gloyd ( loc. ) quoted Mr. Joseph Ackroyd made on two male copperheads near Winchester, Virginia, in late July, 1945, as follows: "The dance took place at 10:30 P. M. at the side of a farm lane bordered on one side by an uncleared fence row and wild blackberry patch and on the other by a long, wild meadow sloping dov/n to a small sti^eam across which is a woods of second-growth oak; elevation approximately 800 feet. . . . Pos- sibly two-thirds of the anterior portions of the snakes' bodies were entwined vertically with the exception of a portion of the neck. The heads were opposite each other and there was a slight swaying movement between them. About one turn of coil was wound and unwound, first in a clockwise and then in counter-clockwise direc- tion. At no time did the distance between the heads change during the rhythmic movements, and at no time did the snakes progress along the ground. It seemed as if the posterior ends were definitely 'anchored'. On three distinct occasions one of the snakes broke the rhythm of the dance by darting its head rapidly at the other." Shaw (loc. cit.) having observed the dance in captive copper- heads, stated, "The combat dance of Agkistrodon m. laticinctus is similar to that of Crotalus but differs quite markedly, insofar as our observations are concerned, in that one of the males acts as though he were afraid of being bitten on the head. ... al- ways ducks and dodges the head thrusts of the aggressor . . .

4—4428 134 University of Ka.nsas Fuels,, Mus. Nat. Hist.

although the aggressor never attempts to bite. . . . These snakes also seem to be more tensely alert during the dance, and the neck- twdning motions much more hurried, contrasting with the compar- atively leisurely motions observed in Crotalus" Mr. Delmer Ferguson of La Cygne told of seeing two large cop- perheads engaged in combat "dance" in a road near his home. He was not able to describe any details since the incident had occurred many years before. I have never seen a typical combat dance, either under natural conditions or in captivity. On September 26, 1957, the two males in a litter of copperheads bom in captivity only four days earlier, slowly approached each other, wdth their heads and necks elevated, and as they met, they reared until in each ap- proximately the anterior 2/5 extended up vertically from the sub- strate. Their ventral surfaces were pressed against each other, each one supporting the other's weight. The snakes remained bal- anced in this position for almost a minute. Although no evident hostility or combat was involved, the behavior seemed to be akin to that of the combat dance in adults.

The nocturnal habits of the copperhead perhaps explain in part, why the combat dance has so rarely been observed in such a com- mon species. In captivity normal behavior seems to be largely in- hibited in most individuals. The combat dance was not observed in the outdoor enclosure where several males were kept together and behaved more normally than those in closer confinement. In my opinion the combat dance is a rather rare phenomenon even in the wild, evoked only in certain individuals under special condi- tions, SHEDDING

The skin may be shed within a few days after birth but the interval is variable. Gloyd (1934:600) found that ordinarily all young of a brood shed about the same time, most often on the seventh or eighth day, but the range was from three to ten days. Chenoweth (1948:162) recorded a litter of five copperheads born on September 4 in which all shed on September 10. In two litters kept by Conant (1951:112) the young shed from five to ten days after birth. In litters kept by me shedding usually occurred within the time range indicated by Gloyd, Chenoweth, and Conant, but it was some- times delayed and under unusually dry conditions sometimes did not occur for several weeks. After such delayed shedding the slough did not come ofiF entire and there was a tendency for patches to remain. Some young had to be soaked for a day or more AUTECOLOGY OF THE COPPERHEAD 135 before they were able to shed. Especially in the stunted young bom from undernourished females, shedding was liable to be de- layed beyond the normal time. Young that did not shed promptly perhaps tended to outgrow their skins as they increased in length, living on the stored yolk. They became increasingly handicapped in their movements as they were stiffened by the dry layer of outer skin. Such young were unable to assume the characteristic compact resting coil, but usually lay with their bodies extended, either straight or with only minor flexures. In individuals less affected, and able to coil, the decreased flexibility of the skin was shown by bends or creases in the concavities of the coils. Probably feeding of young is delayed in most instances until shedding is completed. Even in their locomotion, and in striking to defend themselves, the young are handicapped by the stiffening effect of their unshed skins. After shedding, the young are far more alert and active; they may take food, and when quiescent they nearly always assume the typical resting coil. Copperheads that have recently shed have their patterns un- usually bright and vivid. Those that are preparing to shed are unusually dull and dark in appearance. However, the approach of molt is less evident than in some other kinds of snakes and the eyes do not assume the milky opaque appearance characteristic of many snakes. When the molt is approaching, the snake is unusually sluggish. In captivity, even those individuals that feed well at other times cannot be induced to take food when shedding is imminent. Immediately after shedding the snake shows renewed animation, moves about more than usual, and is eager to feed. Stabler (1939:228) presented data concerning the frequency of shedding in many species of common snakes that he kept in cap- tivity. One copperhead shed eight times in 24 months. Excluding a "rest period" or pseudo-hibernation that occurred in the poorly heated room where his snakes were kept from early October to April, Stabler obtained a figure of 1.8 months as the average interval between molts for this snake. Ahl (1930) recorded two molts in a copperhead kept for 12 months, and Carr (1926:150) recorded six molts in one that was kept 11 montlis and fasted throughout the entire period. Obviously the frequency of shedding varies and is influenced by many factors including the size and age of the individual, the amount of food that it consumes, and the temperature. Most of the copperheads captured in the course of my field study were marked 136 University of Kansas Publs., Mus. Nat. Hist.

with a daub of brightly colored enamel paint before they were re- leased, in an attempt to gain information concerning the frequency of shedding. Those retaining the paint at a later capture were known not to have shed in the interval. Nevertheless, the data ob- tained regarding shedding were remarkably meager. Most of the snakes recaptured were taken after intervals too long or too short to yield significant information regarding shedding. In warm, damp weather the paint tended to crack, peel, and wear away as the snakes crawled through ground litter and dense vegetation. After several weeks little or no paint might remain, even though no molt had occurred. Therefore, only the positive evidence provided by retention of the paint from one capture to the next was definitive. Most individuals caught after intervals of up to two weeks re- tained their paint. An adult male marked on September 27, 1958, retained paint when caught 37 days later, and a large adult female marked on June 20, 1958, likewise retained paint after 49 days. A small adult male marked on October 23, 1957, retained paint after seven months when recaptured on May 21, 1958, but tliis interval was mostly winter dormancy with probably not much more than a month of active existence. As opposed to these tliree positive records, copperheads tliat had lost their paint were recaptured after the following intervals, in days: 39, 40, 41, 45, 51, 54, 57, 61 and 61. Also many longer intervals were recorded. In rapidly growing young copperheads in captivity I recorded molt intervals of 31, 32, 34, 35, 41, 57 and 70 days, and, in an adult, 63, 85, 89, 94 and 96 days. The young that shed in 31 to 34 days were first-year individuals that were being fed maximal amounts and were making unusually rapid growth. The average interval of 85 days obtained for tlie captive adult may be typical of indi- viduals under natural conditions. The many gravid females that were kept tlirough September in order to obtain litters, mostly shed within three weeks after birth of theii- young. Whether shedding occurs regularly in the entire population at this season, or whether shedding is in part controlled by the physiology of the reproductive cycle was not determined. The sloughed skins are rarely found under natural conditions. They are often cast in burrows of rodents or deep rock crevices where they would usually be overlooked. Adults probably shed two or tliree times in the course of their season of activity, and juveniles probably shed three or four times in their first fuU growing season. AUTECOLOGY OF THE COPPERHEAD 137

HIBERNATION AND THE EFFECT OF TEMPERATURE

In the region of the Reservation copperheads spend at least half the annual cycle in hibernation, which normally extends from some time in late October or early November to some time in April. Earliest recorded dates of emergence in the 10-year study were: April 15, 1950; April 24, 1951; April 23, 1952; April 8, 1953; April 25, 1954; April 20, 1955; April 5, 1956; April 25, 1957; April 13, 1958; and April 6, 1959. Latest dates of record were: October 27, 1949; November 14, 1950; November 20, 1951; November 12, 1952; October 31, 1953; November 16, 1954; November 2, 1955; November and November 15, 1956; November 23, 1957; November 14, 1958; — 5, 1959. The average dates indicated by these figures April 17 for emergence and November 11 for retirement—represent the ex- tremes; most individuals emerge later and retire earlier. The figures for spring emergence are mostly based upon individuals found under rocks, and at this season the snakes spend much time basking under large rocks that are warmed by the sunshine on warm days but provide sufilcient insulation when the temperature is low. In autumn, the snakes rarely frequent such situations but tend to seek out deeper shelters that will serve as hibemacula. The autumn records are therefore based either upon individuals found in the open or those live-trapped at the hilltop ledges. For the latter, the date used is not necessarily the date on which the individual was found in the trap. Often the traps were checked on days too cold for the snakes to be active above ground. Snakes found in the traps on such days were known to have been caught in warmer weather, one or more days earlier. In April, early May, late Octo- ber and early November temperatures are often too low for the snakes to be active. Some individuals are dormant throughout these periods, while others are active intermittently when the tem- perature is sufficiently high. Copperheads are especially gregarious at the time of hibernation. Vernon Mann of La Cygne, Kansas, told of collecting several thou- sand copperheads, mostly at their hibernation sites, over a 30-year period. The dens were at tops of bluflFs in rocky situations that were hotter and drier than the surrounding habitats, usually where the exposure was mainly to the south. At times of emergence he found the snakes scattered along the ledges; rarely as many as 30 were found in the vicinity of one den entrance. On various occa- sions Mann had attempted to dig out dens but had never succeeded, as the dens were always deeper than anticipated, and were among 138 University of Kansas Publs,, Mus. Nat. Hist. rocks. On one occasion, he dug down through loose shalelike rock to a depth of more than four feet, whence the tunnel led beneath a massive boulder and could not be followed farther. Mann thought that the snakes sometimes travelled as much as two miles to and from the hibernation sites but that most travelled shorter distances. He thought that regular travelways existed be- tween denning areas and summer ranges. He mentioned so-called "snake rocks" along such routes which might shelter as many as six copperheads at one time. Presumably traihng by scent would account for such aggregations. In several different years I have first found copperheads in spring at the hilltop ledges under large flat rocks where there were deep crevices that probably led to hibemacula. On April 6, 1959, for example, after several hours' unsuccessful search, I turned a flat rock approximately 20 inches in diameter and three inches thick, and found an adult copperhead in damp soil beneath. A second was coiled in contact with the upper edge of the rock partly con- cealed beneath dry leaves. Raking through the heavy leaf litter that had accumulated on the uphill side of the rock, I uncovered two otiiers. Thorough search failed to reveal any more in the vicinity. A round hole one and a half inches in diameter extended downward almost vertically from the depression from which the rock was moved. As it seemed almost certain that the copperheads had emerged from this hole, I attempted to excavate it. The diameter of the tunnel enlarged to several inches, and the hole deviated slightly from the vertical, slanting back into the hillside until it con- tacted a vertical rock face, of the Toronto Limestone. At a depth of 16 inches the cavity divided into two almost horizontal branches running in opposite directions along the rock face. One branch was traced for approximately six feet and the other for four feet, but excavation had to be abandoned because of massive limestone slabs and boulders, wedged in crevices or too heavy to be moved. How- ever, it seemed that the cavity extended indefinitely in both direc- tions along the rock face, and that it was formed by pulling away of the loose soil, tending to slough downhill from the outcrop, with subsequent filling of the upper part of the crack by compacted dead leaves and other accumulated debris eventually forming a soil layer. Soil in the supposed den cavity was moist from seepage trickling over the face of the outcrop. At the time the snakes were found, in mid-afternoon, tempera- ture was 80° F—the highest of the season up to that time. Succes- AUTECOLOGY OF THE COPPERHEAD 139 sive maxima on preceding days had been 73% 74°, 55°, 79°, 70% and not 70' (on March 31). On still earlier dates maxima usually were above 60° (except for March 23 and 24 with maxima of 74°), and it is doubtful whether any copperheads had been active. Of the four copperheads found on this occasion, one was a small adult male, two were adult females, and one was a subadult female. The

Qen cr g 200 UJ cc ou.

£ 100 CD

SEP OCT. NOV.

Fig. 13. Numbers of copperheads caught along hilltop rock ledges in au- tumn, in area shown in Fig. 1, grouped in ten-day intervals. Eleven years' data are combined, but the trends dif- fered slightly from year to year. In general the trends reflect amount of activity of the snakes along the ledges, but trapping effort was somewhat less in early September than it was later in the autumn. In a typical year, activity along the ledges increases throughout September and until mid-October, then activity tapers off abruptly. male was in breeding condition and one of the adult females had sem*n in her cloaca, indicating that copulation had occurred recently, perhaps on the day that they were found. On the night following capture of the snakes there was a cold front with heavy snow, and freezing temperatures were frequent in the following week. Not until April 15 was temperature again high enough (70° F) for emergence. On April 16 also weather was mild, but another week of cool weather followed, and it was not until April 23 that mild weather returned. Probably many copperheads did 140 University of Kansas Publs., Mus. Nat. Hist.

not emerge at all until after this date. Changeable weather with temperatures often below the threshold for activity in copperheads, is typical of April in the area of my study, and much of the month is spent in dormancy in most years. Noble and Clausen (1936:314) excavated a hibernation den near Stony Point, New York, in January, 1933. The twenty-one copper- heads dug out of crevices in the rock ledges were encountered singly or in groups of two or three within an area twelve yards square and about four feet from the surface. Minton (1951:322) stated that in southern Indiana copperheads hibernate on wooded, rocky hillsides. McCauley (1945:132) mentioned the tendency to seek hilly rock-ledge areas for hibernation but cited an instance of 37 being found in early spring in a manure heap where they prob- ably had spent the winter. In Richmond County, Georgia, Neill (1948:112) found that not all copperheads frequent regular dens to hibernate. He sometimes found them hibernating beneath pine stumps or under the bark of stumps. A large den where many hibernated was located on the outskirts of Augusta, in a rocky outcropping bordering a golf course. Corrington (1929:58) found two copperheads hibernating in a log near Columbia, South Caro- lina. Similarly, Strecker (1935:26) found four hibernating in a hollow pin-oak log in McLennan County, Texas. Dundee and Burger (1948:1) found a denning area of copper- heads in Rogers County, Oklahoma, extending for about a mile and a half along a bluff of limestone and sandstone. Coachwhip snakes

(Masticophis f. flagellum) and western cottonmouths {Agkistrodon also piscivorns leiicostoma ) were numerous along the bluflF on April 5 and 6, 1947, when the observations were made. The three species were found together to some extent, but diflFerences in preferences of denning sites were evident. The copperheads were concentrated in drier areas, usually the east and southeast exposures of the bluff. The cottonmouths were concentrated on nortlieast facing slopes that were damp from seepage. The coachwhip snakes were found in more open areas than those occupied by the copperheads and moccasins. On the Reservation many species of snakes hibernated in dens formed by crevices and fissures in the hilltop rock outcrops, which totalled 9.1 miles on the area. Different species were found most concentrated at different stages of the season, but no segregation according to habitat was apparent. Most species spent the summer in other habitats, converging on the ledges in fall and dispersing again in spring. All species favored south- or southwest-facing AUTECOLOGY OF THE COPPERHEAD 141 exposures and tended to avoid exposures that faced north. The the yellow-bellied racer and red-sided garter snake, two of com- monest species, were often found in the same traps with copper- heads, and doubtless often hibernated coiled together with them in the same shelters. Less frequently the black rat snake, timber rat- tlesnake, blotched king snake, milk snake, and common water snake have likewise been trapped with copperheads. In all the instances of double or multiple captures with different species involved, there were no injiuries to either kind of snake. The association of the garter snake and racer with the copperhead in hibernation is of special interest because each of the three occasionally preys upon young of the others in summer. In autumn, when the snakes gather at the ledges there is no hostility, and in hibernating groups, re- gardless of the species involved, each individual may benefit from contact with the others in maintaining favorable conditions of temperatm-e and humidity. The timber rattlesnake, sharing most of the copperhead's geo- graphic range, and having similar habitat preferences, is a frequent associate of the copperhead in denning areas. On the Reservation the timber rattlesnake is relatively scarce, but most of those found are trapped in October along the same rock ledges where copper- heads are most abundant. In several instances a rattlesnake and a copperhead have been caught simultaneously in the same trap. Klauber (1956:567) quoted B. A. Eger of Buena Vista, Virginia, who had found copperheads and timber ratdesnakes coiled together in dens in cold weather, and Stephen H. Harwig of Pittsburgh, Pennsylvania, who had found numerous copperheads and some pilot black snakes at the same "den rock" with timber rattlers in summer, and beheved that all three denned together to some extent in winter. Swanson (1952:176) found three young timber rattle- snakes and ten copperheads together at a den near Mt. Alto, Frank- lin County, Pennsylvania, on September 28, 1924. Klauber men- tioned a newspaper account of a den that contained 193 adult timber rattlesnakes, 16 black snakes and a copperhead. The same author (op. cit. -.551) quoted John H. Stanley of Andrews, North Carolina, mentioning an instance of 30 adult timber rattlesnakes and copperheads killed at one time by dynamiting a den on the National Forest. loc. cit. Pisgah Noble and Clausen ( ) found eight timber rattlesnakes with the aggregation of 21 copperheads exca- vated from a den near racers Stony Point, New York. Two ( Colu- ber constrictor) were nearby. Hudson (1942:83) mentioned the finding of mixed dens of hi- 142 University of Kansas Publs., Mus. Nat, Hist. bemating timber rattlesnakes and copperheads in the course of quarrying operations in Richardson County, Nebraska. In January, 1954, a thermometer was installed 33 inches deep in a rock crevice at the base of a Hmestone outcrop near the top of a south-facing slope. Copperheads were known to hibernate vdthin a few yards of this spot, but depths of the hibernacula were, of course, not known. Even in bitterly cold weather of mid- winter, the temperature in the crevice never fell below 3° C; for most of the season of dormancy it ranged between 4° and 11°. Dur- ing the latter half of summer and early autumn the temperature remained near its maximum level, usually between 20° and 21°. From January through most of July there was a fairly consistent warming trend, and from mid-October through November, De- cember and January the reverse trend was noticeable. In the winter of 1956-57 copperheads were kept in a hibernation box one foot square and six inches deep, sunk 30 inches under- ground. A plastic tube two inches in diameter led from the box to the soil siu-face at an angle of approximately 30°. An insulation box filled with sawdust was on top of the hibernation box, which also contained a maximum-minimum thermometer. The interior of the hibernation box could be readily inspected by lifting out the insulation box and removing the lid. On November 18, 1956, when temperature was 8.8° C, two adult copperheads and two young of the year were placed in the box. None of the copperheads survived the winter in this artificial hibemaculum. The box was opened from time to time in the winter when air temperatures were well above freezing. On these occasions notes were taken, as follows.

November 30. Air 11.4", interior of box 9.5*, the minimum reading up to that time. Snakes sluggish but not dormant, and shifted position slightly as the box was opened. December 16. Air 6.7°, interior of box 5.7°. Copperheads all coiled to- gether; shifted coils slightly when lid was removed from box. One was re- moved from the box momentarily for observation. It could barely move but was too stiflF to attempt to bite and could be freely handled with impunity. December 26. Air 11.1°, interior of box 8.3°; snakes did not move when uncovered. January 2. Air 3.9°, interior of box 5.5°. The copperheads did not move until touched. When removed from the box, they seemed aware of their sur- roundings and responded to stimulation with slow squirming movements, but were imable to defend themselves. Even when touched on the head they usually did not attempt to bite. When so stimulated, one did open its mouth, but seemed unable to coordinate its movements for biting. They lay quies- cent except when placed near to cover; then, with slow, clumsy undulations they attempted to gain shelter. AUTECOLOGY OF THE COPPERHEAD 143

January 12. Air 4.4", after passage of a cold front with minimum of —9.7". Interior of box 6.7°, up from a minimiun of 5.0°. As before, the snakes were slow and sluggish but not completely dormant. January 19. Air 7.8° after a period of much colder weather. Interior of box 3.9°, the minimum reached up to that time. A copperhead that was near the top of the box drew back on the defensive as the lid was removed. January 21. Air 20.4° after arrival of a "warm front"; interior of box 6.1°, up from a minimum of 4.5°. February 8. Air 13.3°. Interior of box 4.4°. February 16. Air 10.0°. Interior of box 3.3°. February 27. Air 10.7°. Interior of box 4.4°. Two of the copperheads were dead. One of these, an adult that weighed 129.8 grams when placed in the box, had declined to 98.6 grams. Perhaps part of the weight loss had occurred after death. Approximately six inches of this snake's head and fore- body were in the exit tube, but the head was turned back toward the box. Dis- section revealed the snake's lungs to be inflamed. March 12. Remaining adult copperhead found dying on ground surface beside the entrance hole leading to tlie hibernation box. When stimulated, its tail twitched slightly, but it was incapable of any other movement. Ob- viously it had been caught out and frozen on the preceding night when tem- perature had been several degrees below freezing. The entrance hole of the plastic tube leading to the hibernation box was partly plugged, and the snake probably had been unable to gain entry. On March 11, weather was mild with a maximum of 20°, and the snake had probably emerged on this date. Just what stimulus impelled it to leave the hibemaculum is not clear, since tempera- ture within the box varied but little throughout the winter. The temperature range within the box was similar to that in the natural crevice where a ther- mometer was installed 33 inches deep at the base of a hilltop outcrop.

Copperheads released from live-traps at 10° C. were capable of vigorous movement. When dropped beside a rock crevice, such a snake would escape into it, moving briskly but rather stiffly. When handled, the snake would attempt to escape and to defend itself, in a slow-motion version of the usual behavior, writhing, thrashing, throwing the body into the characteristic kink, and attempting to bite. However, if released away from the immediate vicinity of shelter, the snake tended to He motionless indefinitely, without at- tempting to escape or defend itself unless it was actually touched. On April 21, 1959, 1 found a copperhead among dry leaves emerging from the entrance of what appeared to be a mammal burrow—an enlargement of a deep crevice in a hilltop rock ledge. Its oral temperature was 14.3°, but the cloacal temperature was only 10.2°, perhaps still near the temperature in the hibemaculum from which it had recently emerged. The foreparts had warmed more rapidly as the slowly emerging snake came in contact with tlie warmer sur- face air and, perhaps was warmed slighdy by insolation although 144 University of ICansas Publs., Mus. Nat. Hist. the sk>' had been mostly overcast. Probably 10° C. is near the threshold temperature at which torpid copperheads may rouse them- selves to activity. On April 23, 1952, a copperhead was found in the open when air temperature was 11.6°. Bodily temperature of the snake itself was higher as it was basking in sunshine, but presumably it had become active and left shelter at an air tempera- ture near 11°. On October 29, 1957, two copperheads were found in funnel traps where they had been caught on one of the two preceding days when temperatures reached maxima of 12.8° and 10.5, respectively. Even at such low temperatures snakes may be suflBciently active to respond to heat gradients and to move into the open if sunshine provides opportunity for them to bask. Extent of tolerance to low temperatures was demonstrated on January 4, 1959, when air temperature was down to —12°. Although an oil heater was burning in the laboratory building, the temperature in- side was below freezing. A group of copperheads being kept in the building, in screen cages and glass jars, were all immobile and lifeless in appearance and had temperatures somewhat below 0° C. Some that were nearest a window were frozen solid, but most were hmp. Although at first assumed to be dead, more than a dozen of these snakes survived after subsequent slow warming. About the same number, including all those that were frozen solid, failed to revive. On January 17 and 20 experiments were performed with the survivors by removing them from the artificially heated room to subfreezing outdoor temperatures. The snakes placed outside con- tinued to move about in their containers as the temperature dropped, and were still moving at 0° C. or even —.5° but they became increas- ingly slow and sti£F as temperature declined, and soon were com- pletely immobilized. Such a snake could be handled with impunity and was completely limp and lifeless in appearance and efiFectively anesthetized. One that was cooled to —1.5° C. and then warmed, showed the first signs of life at 6° and it recovered completely. Sev- eral others that were cooled to —1.5° and some of those that were cooled to —1° failed to revive. It seems that these temperatures are near the critical minimum. For reasons that are not evident individ- uals seem to differ notably in their capacity to withstand low tem- perature. At approximately —1.0° to —1.5° body fluids began to congeal, with release of heat, and even at air temperatures several degrees below freezing the decline in body temperature tended to pause at this level for periods up to a half hour. One juvenile that was kept for over an hour at a body temperature of —1° was somewhat stiffened by partial congeahng of body fluids, but it AUTECOLOGY OF THE COPPERHEAD 145 revived completely. While it was being revived frequent stimu- lation was applied to test its reactions. For many minutes it was completely limp and inert showing no responses to pinches

15 rrt I 10

A

20 q: UJ GQ

Z 15

10

D

n ' 1 i n J I ;

20° 30° 20° 30" 20" 30'^ TEMPERATURE, CENTIGRADE

Fig. 14. Records of temperatures (in degrees Centigrade) at captures of copperheads in eastern Kansas (Douglas Co., with a few from Anderson, JeflFerson, Linn and Miami counties). A. Temperatures of air at times and places where captures were made. The largest number of snakes was found at air temperatures between 24° and 26°. B. Bodily temperatures in gravid females found in the open. In the gestation period the females often bask and they tend to maintain relatively high temperatures. C. Bodily tempera- tures of non-gravid individuals (both sexes) found in the open. D. Bodily temperatures of copperheads found under natural conditions, including both those found in the open and those found under shelter. Favorite temperatures seem to be between 26° and 28°. E. Bodily temperatures of copperheads found under shelter. F. Bodily temperatures of copperheads found in the open; presumably these active snakes were regulating their temperatures to some extent by behavior, and their preferences for the range 26° to 28° is more evident than in those found under shelter. 146 University of Kansas Publs., Mus. Nat. Hist. or pricks until its temperature had risen to 13.9°. Then it moved sHghtly. First breathing occurred at 19.9°, and soon thereafter the snake had revived completely. Bodily temperatures were obtained from 112 copperheads found under natural conditions in the course of my study; 70 were in the open and 42 were under cover. Presumably most snakes of both groups were exercising some control over their bodily temperatures by their behavioral responses to environmental temperatures. How- ever, some of those found under rocks were cold and stiff and were obviously incapable of exerting any control over their temperatures, which were those of the surrounding soil and rock. The snakes that were in the open regulated their temperatures to a large extent by their movements between shade and sunshine, and by maintain- ing contact with objects in the environment having temperatures nearest their preferred levels. Therefore, these snakes that were found in the open fell within a slightly narrower temperature range than those under cover and they were more concentrated near the temperature that probably represents the optimum level (Fig. 14). The preferred temperature level seems to be between 26° and 28°, as previously concluded (Fitch, 1956:464). Temperatures were measured with a Schultheis thermometer inserted into the cloaca. In the elongate body of a snake different parts of the body probably differ notably in their temperatures. In one copperhead found basking beside a crevice with part of its body exposed to sunshine on May 5, 1958, at an air temperature of 19.5°, the oral temperature was 30.5° while the cloacal temperature was only 24.8°. Differences of similar magnitude probably occur frequently in individuals that are active. In those exposed to sun- shine the head and forebody tend to be warmer. The posterior end of the body and the tail sometimes remain in shade even when the snake seeks sunshine to warm itself, or at least their exposure to sunshine is delayed as the snake moves slowly into the open. All but one of the snakes having temperatures of less than 18° were found under rocks. Some of the lowest temperatures (17.0°, 16.4°, 15.5°, 14.0°, 13.7°, 12.4°) were recorded in early spring in copper- heads found under large flat rocks; these individuals probably had not yet ventured into the open but were still in the process of emerging from their hibernation shelters. Gravid females probably prefer temperatures somewhat higher than those preferred by other individuals; at least most of the copperheads found basking in sunshine in summer were females. AUTECOLOGY OF THE COPPERHEAD 147

In September, 1958, a gravid female, a yearling, and four adult males were kept together in an outdoor enclosure of 100 square feet. From time to time bodily temperatures of all snakes in this group were recorded within a few minutes. Eight sets of readings are shown in Table 3. In six instances the gravid female had the highest reading of the entire group and her average exceeded the aver- age of the group by 2.2° C. The yearling also tended to be warmer than the adult males; under the conditions of the experi- ment, with temperature of the air somewhat below optimum level, this small snake's ability to warm himself more rapidly by basking gave him a distinct advantage over the larger individuals.

Table 3. Bodily Temperatures, in Degrees Centigrade, of a GROxn» or Copperheads in an Outdoor Enclosure 148 University of Kansas Publs., Mus. Nat. Hist.

individual in the course of its routine activities. Almost all the movements are of this type. (2) Shifts that involve abandonment of an original home range and occupancy of a new one. These movements are relatively rare but ecologically significant. Some individuals may live their entire lives within the same home range. Others may shift several times. As compared with females, males seem to shift more frequently and farther. (3) Seasonal migration between a hibernaculum, usually at a hilltop rock ledge, and a home range that is occupied during the summer. Some individuals, especially first-year young and gravid females, tend to remain on rocky slopes throughout the summer, and hence have home ranges that encompass their hibernacula. Other individuals make trips of hundreds of yards, if not farther, between their home ranges and their hibernacula. The place used for hibernation and the route leading to it might be regarded as an extension of the home range; however the hibernaculum may be relatively remote from the sum- mer range and separated from it by areas of less favorable habitat, which the snake crosses in a fairly direct route. In an earlier discussion of home range in the copperhead (Fitch, 1958:125), based upon much less extensive data, especially for movements on the summer range, I failed to distinguish clearly between the three types of movements. In actual practice it is im- possible to distinguish between the three types in all instances. Data concerning size of home range were obtained chiefly in 1957, 1958 and 1959; few traps were operated in the summers of earlier years. For most individuals only two records are available, sepa- rated by a substantial interval. Therefore no home range can be plotted, but the distance between successive points of capture shows in a general way the size of the home range. Assuming that each snake wanders more or less at random within its home range as seem to do of ( copperheads ) , any two points capture for the same individual would be separated on the average by a distance equiva- lent to approximately half the diameter of the home range. Since home ranges tend toward a circular shape except when distorted by peculiarities of the terrain or habitat, the formula vr^ for the area of a circle should apply (Fitch, op. cit.:7S). Doubtless home ranges vary greatly in size and shape, but the measured move- ments of a sufficiently large number of individuals should indicate the size of an average home range. Table 4 shows a series of movements of 26 individuals that are thought to represent normal movements within a home range in each instance. In these sets of records elapsed time varies from two weeks to nearly five years. AUTECOLOGY OF THE COPPERHEAD 149

Table 4. Distances Between Successive Captures of Individual Copper- heads ON Their Summer Ranges

Intervals Between Captures 150 University of Kansas Publs., Mus. Nat. Hist.

have been excluded on the assumption that they probably represent shifts from one home range to another. These movements are shown in Table 5. Whether these exceptionally long movements differ merely in degree or differ in kind from the movements listed in Table 4 can not be proven with the information now available. Some individuals probably have home ranges that are much larger than the average, or are elongate and irregular in shape, or are composed of disjunct segments. It is doubtful whether the home range of a copperhead comprises a sharply defined entity. Insofar as is known, an indi- vidual snake has no central "home" to which it returns regularly for shelter, or any other purpose. Rather, it seems to wander about in a circuitous and irregular course. In view of the snake's sluggish

Table 5. Distances Between Points of Capture in Summer, of Copper- heads Each Thought to Have Shifted From One Home Range to Another Because of the Relatively Long Trips Involved

Intervals Between Captures AUTECOLOGY OF THE COPPERHEAD 151 after the several weeks that probably would elapse before the snake happened to return. Also, the snake does most of its travelling in darkness. Olfactory, auditory and tactile cues likewise seem in- adequate to keep the snake oriented within its home range, but per- haps all of these are involved. Certainly location memory must be well developed. Various authors, including Babco*ck (1929:26) in Massachusetts,

Table 6. Distances Between Sites of Axjtumn or Spring Captures at Hilltop Ledges Where Hibernation Occurs and Summer CAPTxmES Else- where, FOR Individual Snakes, Indicating Possible Extent of Seasonal Migration to and from Hibernacula

Snout-vent

Intervals Between Captures 152 University of Kansas Publs., Mus. Nat. Hist.

Vernon Mann, in conversation April 29, 1958, told me that he thought copperheads made much shorter seasonal movements than did timber rattlesnakes, but might make trips up to a mile or two. This was merely an impression, based on Mann's extensive experience with copperheads in Linn and Miami counties, Kansas. On the basis of my own experience I would consider a migration as long as a mile to be highly exceptional. In my study the marked snakes captured both at a hibernation ledge, and at a location away from a ledge in summer, had moved the distances shown in Table 6. These distances are assumed to represent trips between hiber- nacula and summer ranges although in some instances the snake may have been caught while still travelling toward or away from the summer range, and in other instances by the time of the second capture the snake may have permanently abandoned an original home range or hibernaculum. The circ*mstance of whether the capture in the home range happened to be in the part nearest the hibernaculum or most remote from it had considerable influence on the figures. For the 12 males in this group, distance averaged 2151 feet and for the 11 females 1333 feet, indicating that, on the average, males travel much the farther in their trips to and from hibernation shelters. Not included in this table are the records of 18 males and 18 females that made recorded movements of less than the com- of a feet for puted diameter home range ( 1162 males, 690 feet for females) and hence may have had their hibemacula within their summer ranges. The trend of available data suggests that nearly two-thirds of the snakes remain throughout the summer in a range that encompasses the rock ledge where the hibernaculum is situ- ated, and that for the remaining snakes migrations averaging a quarter mile for females and nearly half a mile for males are made to and from the hibemacula. Snakes that congregate in large numbers at communal hiber- nacula have been generally supposed to return to the same den year after year. Findings by Woodbury (1951:12) at a den in Tooele County, Utah, used by rattlesnakes (Crotalus viridis lu- tosus), bull snakes (Pituophis catenifer deserticola) and striped racers (Masticophis taeniatus taeniatus), bear out this supposition. Whether other snakes, which hibernate singly or in small groups, return consistently to the same locations has not been satisfac- torily demonstrated. However, the marked copperheads that I recaptured demonstrated notable fidelity to the stretches of ledge where they had been originally captured. AUTECOLOGY OF THE COPPERHEAD 153

A male, first caught when he was of four-year-old size on October 10, 1949, was recaptured along the same hilltop rock ledge on April four 26, 1952, October 1, 1955, and May 14, 1957; all captures within 210 feet of each other. His only other capture was made on the hill where May 18, 1952, at the head of a valley at the base of the ledge was located, 1000 feet from the nearest part of the ledge, and he presumably was either on his summer range or travelling toward it. Another male first caught on September 17, 1951, when he was of four-year-old size, was recaptured 205 feet farther along the same ledge on October 22, 1954, and on October 20, 1958, he was within 30 feet of the second location. A male of two-year-old size on November 12, 1954, was recaptured October 13, 1956, and October 31, 1957, within a 350-foot stretch of ledge. An old adult female was captured on the same ledge on October 15, 1957, No- vember 15, 1958, and October 17, 1959. The second capture was

355 feet from the first, but the third capture was made at the orig- inal location. Another female of three-year-old size when first captured on September 9, 1949, was recaptured 465 feet along the ledge on August 24, 1951, and on October 2, 1955, a third capture was made 110 feet from the second, in the direction of the first. A female of five-year-old size when fii'st captured at a ledge on Sep- tember 30, 1950, was recaptured at the same place on November 9, 1954; on August 12, 1952, she was found at another place on the ledge 110 feet from the site of the first and third captures. The snakes caught along the ledges in autumn and spring were not necessarily in the vicinity of their hibemacula, as some spend several weeks at a ledge in autumn before hibernating and may travel for varying distances along the outcrop in the interval. Hence it is not surprising that individuals were often found hun- dreds of feet along the ledge from a previous capture site. It is not definitely proven that an individual returns year after year to the same rock crevice to hibernate, but the fact that some indi- viduals had returned to the same spots on the ledge where they were captured in former years does suggest fidelity to a specific hibernaculum. For most individuals only two years' records of capture at the ledges are available. These records are shown in Table 7. Table 8 shows records of those individuals that were recaptured at diflFerent stretches of ledge, having moved many hun- dreds of feet from an original location, and perhaps shifted to an area altogether separate from the one originally occupied. The ratio of six males to two females in this series is noteworthy, and 154 University of Kansas Publs., Mus. Nat. Hist.

Table 7. Distances Between Capture Points of Copperheads Each of Which Was Caught at a Hibernation Ledge in Two Different Years

Intervals Between Captures AUTECOLOGY OF THE COPPERHEAD 155

Table 7. Distances Between Capture Points of Copperheads Each of Which Was Caught at a Hibernation Ledge in Two Different Yeaks— Concluded 156 University of Kansas Publs., Mus. Nat. Hist.

Table 9. Dispersal of Young Copperheads That Were Either Born in Captivity and Released at Capture Point of Mother or That Were CAPTtTRED Soon Before or After the First Hibernation and Were Recap- tured After Substantial Intervals. All Were in the Length Range 187 MM. TO 282 MM. When They Were Marked

Dates of Release and Recapture AUTECOLOGY OF THE COPPEBHEAD 157 several years; after periods as long as six years the snake might be found again in the same neighborhood. Table 7 shows the follow- ing intervals: 28 snakes were recaptured after one year, 14 after two years, four after three years, seven after four years and one each after five and six years. Many other copperheads were caught two or more times along a ledge in the course of an autumn's trapping, or were caught in autumn and again in spring—soon before retiring into hibernation and soon after emerging. For such snakes, the distances between successive captures tended to be shorter than in the snakes whose captures spanned longer intervals. For males the distances (in feet) involved were: 590, 270, 165, 160, 130, 100, 100, 90, 90, 90, 90, 70, 70, 70, 60, 60, 50, 30, 30, 20, 10, 10, and for females: 440, 300, 190, 175, 163, 140, 100, 80, 60, 10. Not included up to this point in the discussion or in the tables are the records of numerous young born in captivity or captured in autumn soon after birth, or in spring soon after emergence from their first hibernation, and recaptured after one or more seasons. The records of these individuals are assembled in Table 9. In general they are notable for the relatively long distances involved, as compared with the movements of other copperheads. Several exceed half a mile, probably indicating that young copperheads most often wander away from the immediate neighborhood of their birthplaces and become established in new areas. On the other hand some individuals were recaptured after periods of years in the neighborhoods of their birthplaces. In these snakes as in other groups of copperheads, the females are distinctly the more seden- tary. Of the 26 males, 35 per cent had travelled less than 600 feet; of the 15 females 47 per cent had travelled less than this distance. REPRODUCTION

Courtship and Mating

Sexual behavior has rarely been observed in the copperhead, probably because it normally occurs at night or under cover. Hay (1893:386) recorded a pair found mating on August 28. Knowing that young are born in late summer or early autumn, he erroneously concluded that the gestation period is nearly a year. Beyer (1898: 19) recorded a pair found mating on April 12, 1895, in Louisiana. He kept these snakes in captivity, and the female produced a litter of young on September 16. Guidry (1953:55) kept a female found 158 University of Kansas Publs., Mus. Nat. Hist.

in the act of copulation on May 3, 1952, in southeastern Texas, and she gave birth to a Htter on August 21. Finneran (1948:124) found a mating pair on April 26, 1941, in Connecticut. Vernon Mann told me that he had often seen mating pairs along rock ledges in spring when the snakes had recently emerged from hiber- nation, and he considered this time to be the breeding season. Gloyd (1934:588) did not observe actual mating, but he concluded that the breeding season in eastern Kansas was in April and early May, a five- or six-weeks period after emergence from hibernation. He frequently found males and females together under the same rock in this period. Twenty-one of 59 sexually mature females that Gloyd obtained in April and May were found to have active sperm in their genital tracts. On the Reservation, in the course of my study, pairs or groups including members of both sexes were found together on various occasions, in spring soon after emergence from hibernation. On April 6, 1959, a male and three females were found at a den en- trance, and probably were newly emerged on that date, as none had been found earlier, and v/eather had been unfavorably cool most of the time. The male and one of the females had active sperm in their cloacae, evidence of recent mating. Although copulation may be especially frequent at this time in spring when the snakes are concentrated along the ledges where they hibernate, there is no true breeding season. Gloyd (loc. cit.) examined vasa deferentia of freshly killed males in April, May, June, July, August and October, and all showed more or less active spermatozoa. A factor little understood until recent years is that of sperm storage in the female, and delayed fertilization, now known to occur in various a lizard chameleon reptiles, including ( ) , turtles, and snakes of several difiFerent families (Fox, 1956:521). Produc- tion of fertile clutches after intervals of isolation as long as six years in the night snake, Leptodeira anntdata polysticta (Haines, 1940: 116) and four years in the indigo snake, Drymarchon corais cou- peri (Carson, 1945:223) are known, although these records are exceptional. Allen (1955:228) presented conclusive evidence that the sperm may remain viable in the female copperhead for more than a year after copulation. An adult female from Texas (A. c. laticinctus) was received at the Highland Park Zoo in Pittsburgh in July, 1954. On August 24 of that year she gave birth to a litter of five young, and again on August 20, 1955, she gave birth to a Htter of five. "Since arrival at the Zoo she has shared her cage with no other snakes . . . had no contact with any males." AUTECOLOGY OF THE COPPERHEAD 159

of the involved in Fox (loc. cit.) studied the histology organs of the sperm storage of snakes, and found that the anterior part oviduct is thick and convoluted, with many highly modified alveolar of were glands which serve as seminal receptacles. Clumps sperm Carson loc. cit. often found tightly packed in these receptacles. ( ) speculated that in sluggish kinds of snakes, living in low population to assure fer- densities, powers of sexual search may be inadequate tihzation, so that prolonged sperm viability would have definite survival value. The copperhead would seem to be benefited by prolonged sperm viabihty, especially where it exists in sparse pop- ulations. In 1959 I tested many individuals of both sexes for active sperm. A few were sacrificed for dissection, but most were released un- harmed after collection of cloacal fluid. The cloaca was cleared of fecal material and uric acid by gentle massaging, and a small vial of Ringer's Solution was emptied into it. After further mas- saging (to express sperm from the vas deferens) the fluid was withdrawn into the vial. Nearly all adult males tested were found to have active sperm, and the sample included individuals col- lected in every month throughout the season of activity. Evi- dently adult males are continually in breeding condition. Relatively few of the females collected at any time of year were positive in these tests. Sperm probably persist for a relatively short time in the cloaca, but survive much longer in the upper end of the oviduct, in vascular tissues specialized as seminal receptacles. Adult females positive for active sperm in the cloaca, were recorded on the fol- lowing dates: April 6, April 27, May 19, May 20, June 1, October 20, October 25, and October 27. Others found to be negative for sperm were collected on April 6, April 21 (2), April 27 (3), June 12, July 12, July 23, July 24, July 29 (4), July 30, August 21, Octo- ber 17, October 21, October 25, October 27, November 4(3), No- vember 7, November 8(3). For rattlesnakes, Klauber (1956:692) has recorded matings in almost every month of the season of activity and some matings in captivity even in the winter. He concluded that in the southern United States rattlesnakes normally mate in spring, soon after they come out of their winter retreats but that farther north, where bi- ennial broods are the rule, the mating times may be more viddely dispersed and summer or fall matings may even predominate. In the copperhead it is my impression that mating may occur almost any time in the snakes' season of activity, especially when the snakes are concentrated along the hibernation ledges in spring and 160 University of Kansas Publs., Mus. Nat. Hist. fall, but that there is some tendency for concentrated sexual activity a few weeks after the spring emergence, at about the time when the females are ovulating—the latter half of May. On May 19, 1959, a pair was found in a trap. The female's cloaca contained sem*n swarming with active sperm—evidence of recent copulation. The female was kept isolated in confinement, and a second cloacal sample taken on June 24 was found to contain abundant motile sperm still. However, approximately three-fourths of the sperm in this sample were dead, and many others were no- tably slow in their movements. In middle and late Mav, 1957, several males were introduced at different times into the cage of a female which had been reared to small adult size in captivity. In each instance the male was in- troduced near dark or shortly afterward and observations were made from time to time subsequently over periods of hours. Court- ship, and probably copulation took place in each instance but the copperheads were remarkably sensitive to disturbance. When light was flashed onto the cage, activity was interrupted as the snakes "froze"; however their positions indicated that courtship had been underway. The male would be found follovdng the female or ex- tended along her, sometimes with his tail looped beneath hers. The first male introduced was a recaptured, marked young 32 months old. Although far short of adult size, he was sexually mature. At dusk, within a few minutes after being introduced into the cage, he evinced interest in the female, flicking her with liis tongue, and then following beside her or over her as she moved. The female's behavior indicated some stimulation by this courtship. Her move- ments became more animated and jerky than usual, with apparent nervousness. She tended to react to contact of the male by push- ing him away. Convexities of her coils were used as pressure points, and with sudden, perhaps involuntary spasmodic contrac- tions, she would "bump" the male forcing him aside a short distance, but she was not actively hostile. It might be expected that sexual attraction and inclination to mate would be reflected in the trapping records in "double" cap- tures involving an adult male and female in the same trap. Double or multiple captures were made from time to time throughout the season that traps were operated, especially in October when most copperheads were concentrated along the hibernation ledges, and seemed to trail each other to find suitable shelter. The combina- tions of individuals in traps usually indicated that sexual attraction AUTECOLOGY OF THE COPPERHEAD 161 was not the basis for the aggregations. However, in the latter half In of May pairs were caught more often than at other times. 1958, for example, eight of the 17 adults trapped from May 16 to 31 were else in pairs. Presumably they had entered the traps together or the males had trailed the females into the traps. In several instances I found copperheads in pairs under circum- stances suggestive of sexual activity in autumn. On the morning of September 26, 1951, a pair was found coiled together in dry leaves in thick woods of a southwest slope. For ten minutes that they were observed they remained motionless. The larger was then caught and was found to have his hemipenis partly everted. Pairs were caught in traps on September 25 and 27. Munro (1950:88) described what seemed to be courtship be- havior in a pair of copperheads said to be four years old and probably kept for a long time in captivity. When they were placed in the same cage they made bobbing head-movements each time they met in moving about the cage. The movements of the male were the more vigorous. After being kept together for several weeks they would still bob occasionally. During courtship (when the bobbing became "very agitated"), the tail and posterior parts of the body executed lateral wriggling and writhing movements not noticeably coordinated with the bobbing of the head.

Ian D. W. Sutherland has described (in litt.) courtship behavior in a pair of copperheads that were obtained in Kentucky in 1955 and kept in captivity for several years. The snakes were kept at a temperature averaging approximately 75° F. and did not hibernate. Coiu-tship occurred irregularly over an eleven-month period in 1957 and 1958. From August to January the male evinced consistent interest in the same female. From January to June only infrequent courting occurred, and then with reduced excitation. On December 19, 1957, the female copperhead moved from one comer of the cage toward the center, and the male in another part of the cage exhibited immediate excitation. He moved toward her and came in contact with the posterior part of her body. The female's move- ments ceased when the male made contact. The male then became moderately excited and began to rub his chin along her body in short spasmodic jerks. At first he moved toward the posterior end of her body but corrected his direction and progressed anteriorly, rapidly protruding his tongue, with the points widely spread. When he reached the anterior end of the female, he placed the posterior part of his body alongside and forced it under the female with a 162 University of Kansas Publs., Mus. Nat. Hist. rippling movement. When the female began to crawl, the male became extremely active and moved his entire body convulsively, causing the posterior part to disengage and thrash about wildly. The male's head was moved vigorously along the top of the fe- male's head. The posterior portion of his body Vv^as again brought alongside the female's body and his tail was twisted encircling hers. The posterior portion of the male's body contracted longitudinally, and forced the female's tail up and forward. The male's head movements increased and his body was in continual movement alongside the body of the female. The female remained impassive during the entire period of courtship.

Fecundity of Females

The proportion of adult females that breed each year cannot be readily determined. Gravid and non-gravid females differ somewhat in habits. As a result, samples are liable to be biased. Gravid females tend to stay nearer the hilltop rock ledges than do males, and spend a relatively large amount of time in basking. In a seven-year sample of summer records, 1950 through 1956, there were 54 gravid females and only 12 that were not gravid. All but six of the total of 66 were found on or near hilltop ledges. In the summers of 1957, 1958 and 1959, when extensive trap lines were maintained, chiefly in open fields, a much different ratio was ob- tained: 29 gravid females and 46 that were non-gravid. In the combined ten-year sample there were 76 from ledges, including 68 gravid and 8 non-gravid, and there were 57 from fields, includ- ing 12 gravid and 45 non-gravid. Probably the gravid females are represented by disproportion- ately high numbers in my sample because they are more easily found than those that are non-gravid being concentrated along the ledges and being more inclined to bask in the open. From the high proportion of females that are not gravid during the summer, it is obvious that a typical adult female does not produce a litter each year. It seems most plausible to suppose that litters are pro- duced in alternate years, and this supposition is borne out by the breeding records of all individuals that were caught as adults in more than one summer. A female, gravid when caught on June 11, 1957, weighed 168 grams. She was recaptured on July 22, 1958, and was then not gravid, weighing only 133 grams. Another fe- male caught on October 1, 1951, was emaciated (112 grams) hav- ing the appearance of recent parturition. Almost exactly three years later, on September 29, 1954, she was recaptured after the AUTECOLOGY OF THE COPPERHEAD 163 season's litters had been born; she had made substantial growth and was plump (222 grams) indicating that she had not produced a litter recently. Another female, gravid on July 5, 1954 (329 grams), was recaptured on September 7, 1957, and was not then gravid (247 grams) nor did she appear to have given birth to a htter recently. Five other female copperheads were captured when gravid and were recaptured after two-year intervals when again gravid. Two others that were captured when gravid were recap- tured again gravid after intervals of four years. In summary, of ten females originally recorded as gravid, three were not gravid when recaptured in odd numbered years, and seven were again gravid when recaptured in even numbered years; although this sample is small, it indicates rather convincingly that in the region of my study females ordinarily produce their litters only in alter- nate years. That this pattern may not hold throughout the range of the species and particularly in the southern part is indicated by Allen's (loc. cit.) record of a female from Texas that produced successive litters on August 24, 1954, and August 20, 1955. Since individuals grow at much diflFerent rates they may attain sexual maturity at different ages. At an age of three years most but not all are of small adult size and are sexually mature. Several of the young that were marked at birth were recaptured as young adults. One born on September 24, 1950, was recaptured on Octo- ber 12 and November 9, 1954, and on the latter date it was dead in the trap. Although it had attained a size typical of individuals in their sixth year, its ovaries and oviducts were small and obvi- ously it had not bred that year. Another female born on September 19, 1953, was recaptured on June 5, 1957, and then was of adult size (558 mm.) but seemingly not gravid (160 grams). A female born on October 24, 1954, was recaptured on July 8, 1958, and had to small adult grown size (544 mm.) but was definitely not gravid

130 . Another born on ( grams ) September 29, 1955, was recaptured on and August 9, 1959, was then definitely not gravid (122 grams). Another born in captivity on September 9, 1954, was recaptured late in her fifth year on June 30, 1959, and was tentatively recorded as on that gravid date. Another marked soon after birth on September 29, was when four 1955, recaptured nearly years old, on August 9, 1959, and she was then definitely not gravid. Still another female bom on September 11, 1954, was recaptured on June 25, 1958, and appeared to be gravid. This female was kept throughout the sum- but did not a litter and mer, produce by late August she no longer had the of appearance being gravid. Either the original diagnosis 164 University of Kansas Publs,, Mus. Nat. Hist. of pregnancy was erroneous or, possibly the ova were absorbed under the unfavorable conditions of captivity. Three individuals each v^'ere caught when approximately five years old; one bom on September 28, 1950, was recaptured on August 31, 1955, in late pregnancy (length 582 mm., weight 196 g.). Another born on September 6, 1952, was recaptured on November 1, 1957, and had the characteristic thin and wnrinkled appearance of those that have 622 106 recently given birth to a Htter ( length mm., weight g. ) A third bom on September 19, 1953, was recaptured on June 4, 1958, and appeared to be in early pregnancy then (length 536 mm., weight 110 g.). Thus none of the six four-year-olds in my records was breeding, whereas all of the four five-year-olds were pro- duced fitters. A female born on September 6, 1952, was recaptured on June 1, 1958; she did not seem to be gravid (length 606 mm., had birth to a litter the weight 143 g. ) and probably given pre- ceding September when she was five years old. The only deviate from the general trend of all these records was a female bom on August 28, 1952, recaptured as a typical two-year-old (476 mm.,

68 g. ) on October 5, 1954, and recaptured when seemingly gravid 1958. this ( 605 mm., 205 g. ) on July 5, Presumably pregnancy was her second, and probably she had produced her first htter at an age of four years. Although the foregoing records indicate a trend of 9 to 1 for litters bom to five-year-olds versus first fitters bom to four-year- olds, the sample is perhaps too small to be given much weight. Development of Ova and Embryos

At the time of emergence from hibernation females have only small whitish ova in their ovaries. In any one female these ova tend to fall into several size groups suggesting that they may be destined for successive broods, but there are always some of in- termediate sizes. Ova of the largest size group are ordinarily more numerous than are the young bom in a brood. In May the ova enlarge rapidly, and ovulation occurs in the latter part of the month, at approximately the time when copulation often occurs. By then the females' bodies are slightly distended posteriorly by the enlarged ova, causing them to have a gravid appearance. Females dissected at various times contained ovarian eggs, as indicated in Table 10. For the smaller eggs the measurements given are only rough approximations in most instances. Also, many of the more minute ova probably were overlooked. Because tlie smallest females recorded as producing litters or as definitely gravid AUTECOLOGY OF THE COPPERHEAD 165 were somewhat more than 500 mm. in snout-vent length, several of the females included in the table almost certainly were sexually immature. Some of the larger individuals, including both of those in were since taken June, not breeding ( females breed only in alter- nate years) and their ova would not have reached full size imtil

Table 10. Sizes akd 166 University of Kansas Publs., Mus. Nat, Hist.

probably most individuals ovulate at least three weeks earlier, in late May. After ovulation the eggs are elliptical, pale yellow, approximately twice as long as broad (average 34 x 17 mm. in eight clutches in

. and after several preserved specimens ) The embryos grow slowly, weeks have assumed the elongate serpentine form, coiled within the egg. At first the head is relatively large, and the protiniding eyes are conspicuous because of their dark pigment. Otherwise, the embryo is whitish and translucent. As development proceeds, the total volume of the egg increases; growth of the embryo and in- crease in the amount of amniotic fluid more than compensate for the yolk substance used up. It is estimated that the gestation period—time from ovulation and fertilization of the egg to parturition—is typically in the neighbor- hood of 105-110 days, including the last week of May, the first week of September, and the entire months of June, July and August. loc. cit. record of a 3 to Guidry's ( ) 108-day period ( May August 19) between copulation and parturition in soutlieastem Texas, probably represents a gestation period. As in other ectothermic ani- mals, the gestation period in the copperhead probably is variable, being influenced to some extent by external conditions, especially temperature. Certainly conception does not necessarily occur soon after copulation, as it does in most mammals. Precise measurement of a gestation period is impossible because the time of ovulation and conception cannot be determined without sacrificing the in- dividual. Aggregating of Gravid Females

A characteristic habit of pregnant female copperheads is that of gathering in small aggregations. An early account of this habit was published by Allen (1868:179) from the observations of a Mr. A. C.

Bennett in Massachusetts: "Of five specimens killed July 4th . . .

all were females. . . . They were all found in a heap. At another time, later in July, seven were killed, which like the others, were all found lying witliin the space of a square yard and all were females. Five of tliem were examined by Mr. Bennett, and found

to contain slightly developed embryos. ... In September . . . six specimens, all females, and all found in a heap, were killed, each of which either seven or nine stated had young." Gloyd ( 1934:592)

that ". . . on three occasions two or three gravid females were discovered in the same location. In 1927 two were in the crevice described [a pocketlike recess extending about a foot beneath a massive limestone slab, and having an opening about two inches AUTECOLOGY OF THE CoPPERHEAD 167 high]; in 1928 there were three; in 1927 two others were sheltered flat stone four feet across." Finneran 1948:124 and by a ( 1953:61) described a den location at Branford, Connecticut, where each year from 1940 to 1947, he found aggregations of gravid females in late summer. The den was at the summit of a hill, with a rock ledge and loose rocks, in woodland witli cedar, pine, laurel, and blue- berry. Each year groups of five to 11 females were found, either in actual contact or at least in a small area. On July 31, 1941, five were found and on August 1, four more. In 1942 groups of seven to nine were found. In 1946 tlie area of aggregation was shifted about ten feet to a partially shaded site, where on September 7, nine gravid females were found. Four were found in 1947. Newborn young were found at this same site in September of several years, and seemingly the place was occupied by copperheads through- out the season of their activity, Minton (1944:474) mentioned the gregarious tendencies of the copperhead, and stated "In addition to hibernating groups copper- heads are frequently found in pairs or threes during the summer/* Aggregations of gravid females have been found on the Reser- vation on only three occasions. This may be attributed to their relatively low population density on the Reservation as contrasted with the areas where some other workers have observed them, and also to the fact that suitable cover is extraordinarily abundant. On August 3, 1950, late in the afternoon, in a rocky area near the summit of a south-facing slope, I turned a flat rock approximately two feet in diameter and two to three inches thick. The rock's lower edge was deeply embedded in the soil and its surface was inchned at approximately a 30-degree angle, with the upper edge clear of the substrate. Two large adult female copperheads were coiled together beneath it in a nestlike depression. Slightly moist soil lining the cavity was packed and smooth probably indicating occupancy of the cavity, at least over a period of several days. When the rock was raised, the snakes uncoiled and began moving away; one temporarily found shelter beneath the edge of the rock and the other moved off into high grass. Several minutes were re- quired to capture and bag both the females. When this had been accomplished I heard the sound of a third snake gliding through the grass approaching the rock and the nest cavity. It was captured and was found to be another gravid female. On August 8, 1950, a second aggregation was found 100 yards from the flat first, under a rock five feet from the ledge itself, which was overgrown with a thicket of skunkbrush, hackberry, hazel, and 168 University of Kansas Publs., Mus. Nat. Hist.

dogwood. The rock was in an open place in low grass and weeds and was approximately 26 by 18 by 3 inches. Four gravid female copperheads were beneath the rock, all in resting coils, in contact with each other. When uncovered they lay still at first, and I retreated to find a stick with which to pin down and catch them. As I approached again all four snakes were becoming restive, look- ing about alertly and beginning to move. They were not inclined to leave the depression under the rock, but were running their snouts slowly over the ground, seemingly searching for concealment. I caught three but meanwhile the fourth escaped over the edge of the ledge into thick brush. On the following day, a gravid fe- male, probably the one that had escaped, was found coiled partly in sunshine on the sticks of an old wood rat house beneath a clump of skunkbrush at the ledge a few yards from the rock where the aggregation had been. The rock that sheltered this aggregation was exposed to sunshine from morning till late afternoon. On August 12, 1952, two gravid females were found coiled near to- gether beneath a board at the old quarry site. In contrast to these few instances when gravid females were found associated with each other on the Reservation, dozens of other gravid females were found or trapped alone. Even though there is some tendency to aggregate, the majority of individuals must remain solitary for most of the time. In the summers of 1957 and 1958 several gravid females and other copperheads were kept together in an outdoor enclosure of approximately 100 square feet. The females sometimes aggregated, but at other times rested sep- arately, and occasionally were associated with males or juveniles. Such social aggregations of gravid females seem to be almost unknown in other kinds of snakes. An indication of similar aggre- gations in another crotalid, the prairie rattlesnake, is provided by the field notes of A. M. Jackley, in South Dakota, quoted by Klauber

(1956:695). ". . . late in August they find suitable holes or cavities wherein they give birth to their young. These places I call rookeries, since it is common to find a dozen or more females quite close together. The distances of the rookeries from the dens vary a good deal, but I think the majority are about half a mile from the dens, and rarely are they closer than 600 feet."

Time of Birth

Actual birth dates for copperheads of 123 htters recorded in my study or gleaned from the literature range from August 3 to No- vember 6. But for any one locality and year the dates of normal parturition are far more concentrated than these dates indicate; AUTECOLOGY OF THE COPPERHEAD 169

probably most of them fall within a two-week interval. Most birth dates are based upon females kept in captivity for varying intervals, and for those kept for several weeks the birth of litters may be de- layed. Unnatural prolonging of gestation in captivity was strikingly dem- onstrated in 1950 and 1951, when I kept many females, captured in July and August in order to obtain their litters. These females were maintained in the laboratory singly or in twos and threes, in small cages, and frequently were offered mice as food. The mice were usually soon struck and killed, but most often were left un- eaten. Some of the snakes never fed and none ate normal amounts. As a result all were somewhat undernourished. Births of their litters were delayed, and averaged approximately a month later than the dates for other years. The delayed parturition of undernourished females is regarded as an adaptive response through which some or all of the fetuses may be saved from destruction. Under natural conditions, in times of food scarcity, rate of development of the embryos would be slowed. Improved food supply late in pregnancy would permit some, at least, of the retarded embryos to complete their develop- ment. None of the undernourished females gave birth to a normal litter. Sometimes all the young were bom dead, and stillbirths occurred in most of the litters. Even if born alive, the young were unusually small. The weight of some was as little as one-fourth of that of a normal newborn young. Despite their prolonged gesta- tion such young seemed to be premature, not only in their small size, but in fetal head shape, bulging abdomens still distended with yolk, and skin incompletely comified with underdeveloped scales. It is remarkable that several of these underdeveloped young bom in captivity, marked and released, survived under natural condi- tions and were recaptured years later as adults. Undernourished gravid females often passed small yolk masses with dead embryos from time to time as much as several weeks before actual birth of their litters. The females seem to have little capacity for the re- sorption of yolk material. Dolley (1939:170) dissected a gravid copperhead in which the left oviduct contained five embryos and the compacted yolk of an infertile or dead egg. Also, a fully de- veloped dead young of the previous year's brood was found at- tached ectopically to the intestinal mesentery. On October 7, 1957, I captured a gravid female on the Reservation, well past the time when most normal litters are bom. The female appeared undernoiu-ished. She died in late November, and still had not pro- 170 University of Kansas Publs,, Mus. Nat. Hist. duced young. Upon dissection it was found that a yolk mass of an infertile or dead egg, of rubbery consistency, was at the lower end of one oviduct, effectively preventing the passage of fetuses through the birth canal. Dr. Joseph P. Kennedy told me of dissect- ing a recently caught gravid female and finding all the fetuses dead, malodorous, and pardy decomposed.

Table 11. Birth Dates of Litters of Young Copperheads in Various Samples

Number of Place litters in sample

1. Kansas—Gloyd (1934), Barton (1948), La Cygne samples in 1958 and 1959

2. Kansas—Reservation in 1950 and 1951

3. Kansas—Reservation

in 1949 and 1952 through 1959. . .

Ohio—Conant (1938), New Jersey—Ditmara (1896), Pennsylvania—Barton (1948), Smith (1940), Swanson (1952) . . .

Texas—Allen (1955), Curtis (1949), Guidry (1953), Werler (1951), Oklahoma—Carpenter (1958), Chenoweth (1948)

Louisiana—Clark (1949)

Virginia—Hoffman (1945), North Carolina—Ditmars (1907), West Virginia—Reese (1926) — Connecticut Finneran— (1948), Massachusetts Babco*ck (1926) . . AUTECOLOGY OF THE COPPERHEAD 171

Certainly this is true of the Utters born on the Reservation in 1950 and 1951. Extreme effects of unfavorable factors in captivity are shown in the following five females captured many weeks before birth of their litters. The first date is the date of capture and the second is the date of parturition. In 1950: August 2, October 19; July 31, October 31; July 24, October 31; July 8, November 6. In 1951: August 3, October 20. In the same year other females, caught soon before the birth of their litters, were probably little affected by captivity. Summers of both 1950 and 1951 were characterized by subnormal temperatures with unusually heavy precipitation and many overcast days. Perhaps the time of breeding was delayed be- yond the normal time. In any event, there can be no doubt that in botli 1950 and 1951 gestation periods were extended beyond the usual time in females living under normal conditions. In these years many females were still gravid when captured at dates later than the latest recorded by Gloyd and Barton for parturition of Kansas females. The captures made by me in which parturition occurred later than usual were as follows (first date being that of captmre, and second date that of parturition). In 1950: September 17, Sep- tember 25; September 14, September 24; September 19, Septem- ber 29; September 9, October 7; September 19, October 13; in 1951 : September 17, October 2. An average parturition date of October 2 is indicated for this group, demonstrating that birth may be delayed for several weeks as a result of unsually cool weather. Making due allowance for delayed births because of low temperatures, and malnutrition in captivity, I regard mid-August to mid-September as the period when most natural births occur, with a trend toward the earlier dates in the more southern parts of tlie range, and toward the later dates at the northern edge of the range. Excluding the parturition dates that were obviously delayed, in 1950 and 1951, average dates each year for the eleven years of the study on the Reservation were: 1949—Sept. 18 (3); 1950—Oct. 2 (5); 1951—Sept. 26 (4); 1952—Sept. 2 (4); 1953—Sept. 9 (4); 1954—Sept. 11 (10); 1955—Sept. 9 (1); 195&—Sept. 13 (1); 1957—Sept. 17 (7); 1958—Sept. 15 (9); 1959— Sept. 10 (5). Number of Young per Litter

Because the copperhead is common in thickly populated regions, and because the gravid female is less secretive than other indi- viduals, much has been pubhshed concerning the numbers of young produced. Second-hand records and newspaper accounts of al- 172 University of Kansas Publs., Mus. Nat. Hist. leged litters numbering 42 to 80 young were discredited by Gloyd (1934:597), but even excluding such records, the literature reflects much difiFerence of opinion concerning the usual number of young per brood. Allen (1868:179), Hay (1892:533) and Stejneger (1895:405) stated that the young number seven to nine. Surface (1906:189) gave the number as six to ten. Ditmars (1910:338) stated that there are about a dozen young per brood, but later (1931:102) gave tlie number as six to nine and these figures were repeated by Lamson (1935:26). Other statements are: four to nine (Hurter, 1911:208; Stewart, 1929:11; Netting, 1939:132; Gow- anloch, 1943:47), three to ten (Necker, 1939:36), two to ten (Wright and Wright, 1957:912, quoting Conant and Bridges, 1939, in "What Snake Is That?", and Davis and Brimley, 1944, in "Poi- sonous Snakes of the Eastern United States"; Smith, 1956:307), and one to 17 (Wright and Wright, loc. cit.; Oliver, 1958:45). Schmidt and Inger (1957:266) stated that the young average no more than six, with ten a maximum. Minton stated that captive females in Indiana produced litters of four to eleven young, and Guidry (1953: 55) wrote tliat in southeastern Texas numerous broods bom in captivity averaged five. Specific records from pubhshed literature and from my own field study indicated a total of 1068 eggs or young from 203 females, with an average brood of 5.26 ± .147, From the histogram (Fig. 15) it is evident that broods of four, five and six, in that order, are by far the most frequent, followed by broods of seven and three with relatively few broods having fewer than three (down to just one) or seven to with one valid record of 17 . more than ( up 14, possibly ) The trends differ somewhat in different samples within the group. There is some indication of geographic variation in litter size, as shown by the following divisions: Subspecies mokeson, Kansas: 115 litters averaged 5.02 ± .385 Subspecies mokeson, eastern states: 55 litters averaged 6.16 ± .283 Subspecies laticinctus: 12 litters averaged 5.75 ± .945 Subspecies contortrix: 16 litters averaged 6.50 ± .56 The figures listed above are based upon the following litters mentioned in the literature: Allen (1955:228) 5, 5 (Texas); Anderson (1942:215) 5 (Mis- soxiri); Atkinson (1901:152) 6 (Pennsylvania); Babco*ck (1926:5) 6 (Massa- chusetts); Barbour (1950:106) 3, 6 (Kentucky); Barton (1948:198) 5, 9 (Pennsylvania), 6 (Kansas); Beyer (1898:19) 7 (Louisiana); Brimley (1923: 114) 8, 6, 4 (North Carolina); Burger and Smith (1950:432) 5 (Maryland); Carpenter (1958:115) 6, 4, 4, 4, 1 (Oklahoma); Chenoweth (1948:162) 5 (Oklahoma); Clark (1949:259) 10, 7, 7, 6, 6, 5, 5 (Louisiana); Conant (1938:112) 10, 10, 6 (Ohio); Curtis (1949:12) 7 (Texas); Dolley (1939:170) 5 (Mississippi); Dunn (1915:37) 7 (Virginia); Finneran (1948:124), 12, 10, AUTECX)LOGY OF THE COPPERHEAD 173

5 (Connecdcut); Gloyd (1934:596) 6, 6, 6, 6, 5, 5. 5, 5, 4, 4, 4, 4, 4, 4, 4, 4 3, 3, 2, 2, 2 (Kansas); Guidry (1953:55) 6, 5 (Texas); Hoffman, 1945:204)

1929 : 97 and ( 1945 : 134 ) 6, 5, 2 ( ( Virginia ) ; Lynn ( ) 7, McCauley 8, 7, 7, Mary- land); Moski (1954:67) 14 (Connecticut); Neill (1948:161) 11 (Georgia); Smith (1940:80) 10, 9, 7, 7, 5, 5, 5, 5, 5, 5, 5, 5, 4, 4, 3 (Pennsylvania); Stadelman (1928:67) 8 (Pennsylvania); Swanson (1952:176) 6 (Pennsyl- vania); Werler (1951:46) 4 (Texas); Wright and Wright (1957:906) 7, 6, 5, 4 (New Jersey, citing Hook).

25 174 University of Kansas Publs., Mus. Nat. Hist. slight. Several records may have been pubhshed because they were regarded as exceptional, for example a litter of 14 young men- tioned by Moski (1954:67), and one of 13 embryos mentioned by Hurter (1911:208). In my own experience, gravid females kept for periods of weeks in captivity often aborted an occasional embryo, perhaps as a result of injury sustained in capture, or unfavorable conditions in confinement. Possibly similar occurrences went imre- AUTECOLOGY OF THE COPPEKHDEAD 175 the idea that there was some loss in gestation, but it is not known how much of this loss was normal.

Embryos or late ovarian eggs: 36 litters averaged 6.19 ± .449. Young born: 162 litters averaged 4.91 ± .164. There may be local differences in size of litter even within one loc. cit. recorded part of the geographic range. Gloyd ( ) an average of 4.2 young in 20 litters from females collected in Franklin, Miami, Linn, Riley, Marshall and Bourbon counties, Kansas, He noted that on the average these litters contained fewer young than Utters re- ported in the literature from females collected in the eastern states. Many of the females were collected by Vernon Mann in the vicinity of La Cygne, Kansas. In 1958 and 1959 I obtained 16 htters from females collected by Mann near La Cygne. These difiFer signifi- cantly in number of young from the sample collected on the Reser- vation and immediately adjacent areas. Gloyd's sample and La Cygne 1958-59 sample: 36 litters averaged 4.03 ± .287. Reservation: 88 litters averaged 5.25 ± .206.

In Gloyd's sample, and in my owti sample from La Cygne, the fe- males averaged small, and I judge most of them to be three-year- olds or four-year-olds, which would produce relatively small Ut- ters. Whether these snakes are typical of breeding females in the population represented, or whether some sort of selection was in- volved in obtaining them is imcertain. Fig. 16 shows the number of young per Utter plotted against size of female. Correlation is evident but tlie number per litter for each size group of females is subject to wide variation. Females less than in snout-vent all 600 mm. length ( including nearly those bear- ing their first and second Utters) usuaUy have from three to six

Table 12. Average Nin^iBERS of Yoxwg for Females of Different Size Groups

Size of Female (millimeters, snout-to-vent) 176 UNivERsrry of Kansas Publs., Mus. Nat. Hist. young; those from 600 to 650 mm. usually have from four to seven young, and the largest females, those of more than 650 mm., usually have from six to thirteen young. Birth of Young Actual birth of young copperheads has been described by sev- eral observers. Gloyd (1934:593) noted that all but one of the 20 females kept by him underwent parturition at night. One that he kept under close observation during the process moved restlessly about the box, nervously twitching the posterior part of her body. The tail was elevated to an angle of about forty-five degrees and lowered at intervals. Soon a fetus appeared at the cloaca. The posterior third of the snake's body moved slowly from side to side, and a peristaltic wave pushed the fetus backward a few millimeters at a time. When about half extruded, the young snake straightened its neck and thrust its head through the membranes and a continuous wavelike motion of the female's body pushed it smoothly through the remaining distance. The female's only movements, besides those necessary for the expulsion of the fetus, were to flex the neck slightly, and she remained perfectly motionless for twenty minutes afterward. The extrusion of the young snake took slightly less than ten minutes. At birth the young snakes were folded two or three times witliin the membranes, with their heads toward the middle. The first part presented in the births observed was a bend of the neck. Unless the membranes were ruptured during parturi- tion, the young made no effort to break through it for about forty- five minutes. Gloyd noted that the duration of labor required for expulsion of each young was about ten minutes. Smith (1940:78) reported labor periods of 12, 11, 9, 5 and 6 minutes for the last five of a litter of ten. Intervals between emergences of these young were 16, 15, 19 and 12 minutes. In the course of my own observations, I found the length of time required for parturition and the time required for the young to rupture their natal membranes, both to be highly variable, depend- ing on temperature, on the condition of the female, and perhaps on other factors. Usually the process extended over many hours, and there was no distinct tendency for births to be concentrated at particular times of the day or night. On September 13, 1954, when air temperature was 23° C. at 8:30 a. m. a large female was found already to have given birth to four of her young. All were still enclosed in fetal membranes. At 8:35 birth of a fifth was completed. At approximately 9:35 the sixth of the litter was born and extrusion of a seventh was completed at 9:55. The eighth was bom at 10:40 AUTECXDLOGY OF THE COPPERHEAD 1T7

labor the contractions a. m., and the ninth (and last) at 11:55. In of the moved from side were slight. The posterior end body slowly to side several times, with noticeable contraction of the abdominal muscles. The fetal membranes appeared first, and usually within 20 seconds the fetus had been entirely extruded, although still resting in contact with the female's cloaca. in the on the A large Htter of ten young were born laboratory first was afternoon of September 3, 1952. At 1:15 p. m. the young noticed already about two-thirds emerged from the female's cloaca. and a ninth was By 5:30 p. m. eight more had been bom partly last of the htter was extruded. By 9:30 p. m. birth of the tenth and a. was found completed. On October 9, 1951, at 8:30 m. a female to be in labor with two young already bom, still enclosed in fetal membranes, beside her. At approximately 3:00 p.m. birth of the sixth and last young was completed. In most instances the young, extruded still enclosed in their fetal membranes, lay inert for varying lengths of time (Plate 20, fig. 1). In some, spasmodic twitching, especially of the head region, was noticed soon after birth, perhaps stimulated by the pressure of the female's muscular contractions in labor. Often many hours elapsed before the young showed signs of life, especially if air temperature happened to be substantially below the optimum level for activity. In many instances soon after birth of the young I ruptured the membranes artificially with a wire and prodded the snakes. When thus stimulated, they underwent violent muscular contractions, sometimes crawled clumsily a short distance, and immediately be- strik- came alert, showing awareness of their surroundings and even ing out with poorly directed strokes at any movement in the vicinity. In the young that were left undisturbed, activity was delayed and sometimes began with a lunging motion by which the head was thrust through the enclosing membrane. The young snake was then able to breathe and seemed to become aware of its surround- ings but it might remain coiled within the mptured membrane for several hours subsequently. Several young copperheads that were unusually small and feeble at birth, remained inactive within their natal membranes so long that, with the evaporation of fluids, they were eflFectively glued to the substrate and would not have been able to escape without aid. Lynn (1929:97) mentioned such an occurrence; the last of a brood of seven bom on September 24, 1928, was smaller than the others and was so slow to become active that it was imprisoned within the dried membrane for nine days, and broke loose only when it was moistened. 178 University of Kansas Publs., Mus. Nat. Hist.

Behavior of Females

Anderson (1942:215) expressed the opinion that females often remain with the young for several days after birth. Such associa- tion seems fairly plausible in view of the fact that the oviparous Agkistrodon rhodostoma remains with its eggs and guards them. Anderson reported finding numerous female copperheads under the same rocks with their young. In two instances ecdysis of the young had taken place and in other broods young were nearly ready to shed, suggesting that several or many days had elapsed since parturition. A female was observed in a rock crevice with three young on September 7, 1941, and the group was still together on the following day. On various occasions I have found litters of young still assembled but not accompanied by the female. In only one instance, on Sep- tember 24, 1958, have I found a female with a litter. In this in- stance the female was in the cavity beneath a rotten stump, and she contained a mouse which probably had been eaten since birth of the young. These young were scattered over an area of several square feet, some in the root cavities of the stump and others coiled on the surface, but partly concealed by sheltering vegetation. The young shed two days later. In this and other instances it seemed that dispersal of the family group was delayed not so much by their affinity for each other as by their extreme sluggishness, causing them to remain for long periods in the same spot, or to move such short distances that they remained in the vicinity of the same shelter and returned to it when its protection was required. Pregnant females were noticeably more docile than other copper- heads, but they underwent a noticeable change of disposition after the birth of their litters. They became irritable and would vibrate tlieir tails in response to any disturbance in the vicinity of their cages. When the recently bom young were disturbed or removed from the cages containing their mothers, the latter assumed a par- ticularly menacing demeanor, moving toward the disturbance wdth neck arched and tongue darting rapidly. Although their behavior clearly suggested defense of the young, they usually did not strike, perhaps failing to find a suitable target since the young were re- moved with wire hooks or metal tongs.

Defects and Mortality at Birth Among the copperheads born in captivity there were many still- births, and deformities were noted from time to time. One snake born while still far short of the usual size, was eyeless, and several AUTECOLOGY OF THE CoPPEKHEAD 179

loco- others had the spinal column kinked so severely that normal motion would have been impossible. The eflFects of captivity on the females in producing deformity and mortality in young cannot be evaluated, but much of the abnormahty probably is congenital and occurs under natural conditions. Klauber (op. cit. :G99) stated that female rattlesnakes, especially those long captive often produce in- fertile eggs and dead or defective young. He estimated that, on or the average, these defects would eliminate about three eggs young per litter. In the copperheads bom in captivity stillbirths were probably more frequent than they would have been under natural conditions. Occasional mortality probably resulted from the females lying on their newborn young and crushing them in the close confines of the cage, before the young had become active. Also, the handhng in- volved in capture and transfer, and the conditions of captivity, probably increased the number of deaths and defects in the unborn young. Unfavorably low temperature at the time of parturition may cause mortality in young that are otherwise normal. At 8:30 a. m. on October 9, 1951, after a night with temperature in the forties a female was found to be in labor, with two young already bom. Both were still enclosed in fetal sacs, and when the mem- branes were removed the young remained inert, and apparently lifeless. Later in the day when they had been warmed in the sun- shine, one of these young revived, and three of the four young bom subsequently also survived. Delayed activity in a young born at low temperature might permit the enclosing membrane to dry, suffocating the snake, or glueing it to the substrate, with fatal results. The Egg Tooth

Dunn (1915:37) and Gloyd (1934:595) mentioned the presence of an egg tooth in the newborn copperhead. According to Gloyd loc. cit. "It seems that in the this ( ) , probable ovoviviparous species structure, of such vital importance in the groups of snakes which produce tough-shelled eggs, is in the process of phylogenetic de- generacy." Although most crotalids are viviparous, it is remark- able that two members of the copperhead's genus, the Malayan pit viper (Agkistrodon rhodostoma) and the Chinese pit viper (A. acutus) are oviparous. Smith (1943:499) stated, concerning the Malayan pit viper: "Two females kept by me in Bangkok laid 13 and 30 eggs, respectively, on August 1st and September 1st, and guarded them until the young were bom, 42 and 47 days later. 180 University of Kansas Publs., Mus. Nat. Hist.

Development was already well advanced when the eggs were laid. They measured approximately 32 x 30 mm., and the young when bom were 150-160 mm. in length." Even in more specialized and strictly viviparous crotalids, the rattlesnakes, an egg tooth is re- tained. Klauber (1956:697-698) confirmed its presence in many species and quoted Trapido's (1939:230) observations on newborn timber rattlers which made series of upward thrusts of the head to ruptm-e the fetal membranes in a manner that may have brought the egg tooth into play. Klauber described the rattlesnake's egg tooth, which is situated medially in the front of the upper jaw just behind the recurved and indented edge of the rostral plate. It is so minute as to be scarcely discernible, and its position renders doubtful anv functional value.

Fig. 17. Palate of newborn copperhead, showing egg- tooth projecting horizontally from the moutli at lower edge of rostral plate, X 7.

In the copperhead, the egg tooth remains functional. In young preserved at birth, it can be easily felt as a thorny projection, when a finger is run lightly over the rostral region. The egg tooth is flattened, chisellike and the tip, which is slightly crenulate, is less than half . is the diameter of the base ( Fig. 17 ) The tooth rigidly attached to the palate near the midline just inside the mouth, behind AUTECOLOGY OF THE COPPERHEAD 181

in a horizontal the rostral plate, and is directed forward, lying plane. In some instances the edge of the rostral is slighdy wrinkled by the egg tooth pressing against its lower edge. In most observed in- stances the egg tooth was slightly to the right of the midline. Con- siderable variation was noted among twelve young or four Utters in the shape and position of the egg tooth. One individual had a symmetrical pair of egg teeth. This snake was abnormal in other side was respects; the eyes were lacldng and the facial pit on one open ventrally and continuous with the palate. Although relatively smaller than the egg teeth of some oviparous snakes, that of the copperhead is useful in permitting it more easily to puncture the enclosing membrane soon after birth, lessening the chance of suffo- cation. Occasionally the fetal sac enclosing the young copperhead was ruptured in parturition, but more frequently it remained intact. inside Typically, the newborn snake remained inert coiled the mem- brane for from several minutes, to several hours, if the temperatm-e was unusually low. The first sign of life consisted of feeble move- ments by which tlie snake oriented its forebody with the dorsal sur- face upward, and then slowly raised its head. The head, usually situated near the center of the snake's coils, was sometimes directed

almost straight upward. Over a period of perhaps half an hour the head would be gradually raised until the snout projected against the enclosing membrane as a distinct protuberance. Although no sudden or vigorous movement had been made, the pressure of the snout against the membrane, perhaps aided by the projecting egg tooth, eventually punctured the membrane and the edges collapsed about the side of the head leaving the nostrils exposed. Then, for several typically, tlie litde snake remained coiled motionless hours, with only its head or snout free of the membrane, permitting it to breathe. Size at Birth

Size at birth is subject to wide variation. Because many of the litters born in captivity were stimted, their sizes cannot be accepted as typical of those under natural conditions. Twenty-one young were collected in September of different years, and probably had grown but little since birth in most instances. They ranged from 247 to 209 mm. in snout-vent length and averaged 222.7 mm. (223.7 for 15 males and 219.2 for five females). Sex was determined for two hundred and thirty-eight young of 49 litters born in captivity. In 26 litters average length of young exceeded 210 mm., and these

7—4428 182 University of Kansas Publs., Mus. Nat. Hist. young were all considered to be normal. The remaining 23 litters, mostly born in 1950 and 1951, all averaged less than 210 mm. and were stunted in varying degrees.

Table 13. Snoxtt-vent Lengths (in millimeters) of Newborn Young

Female/ Number Male Female male in length length length sample ratio

Normal litters

born in captivity . . . 145 222.5(264-203) 219.0(256-203) 99.1%

Stunted litters

born in captivity . . . 102 194.1(214-170) 190.0(206-160) 97.9%

All captive litters 247 211.5(264-170) 206 (256-160) 97.3%

Young collected in September. 20 223.7(247-209) 219 (243-210) 97.8%

Forty litters bom in captivity contained young of both sexes and the males averaged larger in 29 Htters. The female/male length ratio varied from 105.9 per cent to 82.5 per cent in different litters but averaged 98.4 per cent. I conclude that, on the average, females are slightly the smaller at birth.

Appearance of Young

The newborn young are diminutive replicas of the adults in most respects. However, they differ in body proportions. In the young the head is relatively large, and as a result of differential growth in later development, the head of the adult comprises a smaller per- centage of its total length and total bulk. The same trend is, of course, equally true of other snakes, and of vertebrates in general. However, the different proportions have implications that apply specifically to the copperhead's way of life; in the juvenal copper- head the amount of venom, and the capacity to inject it deeply are greater than they would be if proportions of the body were those of adults. In young males and females the proportions of the tail and the ratio of its length to the total length are similar, and the differentiation that is characteristic of adults appears later in devel- opment. Coloration also differs from that of the adult in one detail. In newborn young the distal half of the tail is dull greenish yellow dorsally and bright yellow ventrally—a trait shared with various AUTECOLOGY OF THE COPPERHEAD 183 other crotalids. Also, the predominantly brown color of the body is tinged with gray. After shedding occurs, the grayish tinge is lost, but the young still lack the richness of tone characteristic of the adults. The chestnut hourglass markings in adults are delicately shaded, paler in the central areas and intensified along their edges, but those of the young are more uniform in shade. In adult males the coloration is more or less suffused with red or pink, but in the young, and usually also in the adult females, the ground color and markings are brown, with no trace of the reddish suffusion. GROWTH AND DEVELOPMENT

Utilization of Stored Yolk, and Early Growth

Newborn copperheads have varying amounts of residual yolk material enclosed with them in the fetal membranes. Also, each young snake retains a supply of yolk within its abdominal cavity. Gloyd (1934:600) noted the plump appearance of young during the first few days after birth. He dissected four newborn young and found that yolk within the abdominal cavity ranged from 13.8 to 29.2 per cent of the total weight of the snake. In each of two young dissected at an age of ten days the yolk was reduced to less than half a gram, and in one fifteen days old no yolk was pres- ent. While living on the stored yolk, the young snakes grow rapidly and also lose weight rapidly. In 1957, 1958 and 1959 I measured and weighed 39 young of eight litters just after birth and again after intervals averaging fifteen days. All snakes were kept during the interval in cages where water was available but there was no food. Each snake increased in length and lost weight. At birth the young averaged 216.0 millimeters in snout-vent length and weighed 11.9 grams. After the intervals averaging 15 days, they had gained on the average 12.5 millimeters (5.8 per cent), and lost 1.9 grams (16

cent . loss in per ) The weight these young, averaging more than one per cent per day, emphasizes their rather precarious situation. They must find food soon if they are to survive. In individuals which be- come emaciated and weakened, the chances of successful feeding are much reduced. Later Growth

In newborn copperheads there is wide variation in length and

weight ( from 264 millimeters and 14.9 grams to 160 millimeters and 7.1 in those in grams ) bom captivity, but ordinarily somewhat less in nature, since many of the captive females produced stunted 184 University of Kansas Publs., Mus. Nat. Hist.

young. Although ages of individuals cannot be determined with certainty, the trend of growth during the first year is indicated by the sizes of young at different stages of the season, as shown in Fig. 18. This figiu-e shows that in October and November many young are still of the usual size at birth or but httle larger although the average length has increased somewhat. In May and June there are still many young that are within the size range of the newborn and nearly all are within the length range 230 to 275 millimeters. For July and August the sample is especially meager, but during these months growth must be relatively rapid. By September there

450 AUTECOLOGY OF THE CoPPERHEAD 185 potential of the individual. By the time juvenal copperheads are one year old, some stimted individuals may be no larger than some of the next annual brood which happen to have been bom earlier than the usual time, and have made precocious growth. At the other end of the scale, the larger one-year-olds have surpassed the more retarded tsvo-year-olds in size, with wide overlapping between these annual groups. In fact, from the smallest one-year-olds up to adults, the young constitute an almost uniformly graded series, with no major breaks and with no concentrations in any particular size classes. Therefore age cannot be determined by sorting into annual size classes. The onl)' reliable information concerning growth was obtained from the marked individuals released and recaptmred. Of particular significance in this regard is a series of 44 copperheads bom in captivity, or else captured in fall or in early spring before there was much time to gi'ow beyond the size at birth. The actual or approximate birth dates of these young are therefore known. All of them were recaptured after substantial intervals of up to eight years. Thirteen of the young were undersized at birth and were judged to be stunted as a result of the unnatural effects of captivity on the gravid females. The stunted young probably were somewhat handicapped in their chances for survival by their small size and in other respects. It seems remarkable that so many sur- vived. Whether tlieir growth differed from that of other individuals that were normal at birth is uncertain, but at the time of recapture each appeared normal and healthy, evidently having overcome completely its early handicap. When recaptured, five of the thir- teen were still below average size for their ages, but tlie remaining eight were all above average size for their ages. In Table 14 showing growth in snout-vent length and in weight, the young are arranged (each sex separately) according to the time elapsed between original capture and recapture. Six were recaptured approximately one year after the time of birth, eleven after two years, six after three years, six after four years, three after five years, and one each after six, seven and eight years. In adult copperheads, the sexual diflFerential in length is much greater than it is in the newborn, and it increases with advancing age. In the one-year-old and two-year-old snakes hkewise the males presumably are the larger. This trend is not evident in my sample, probably because of the small niunbers of individuals involved. In both groups females averaged slightly larger; 364 to 338 mm. for the one-year-olds (with a sample of seven) and 480 to 476 mm. for the two-year-olds (in a sample of 20). The female-male length ratio gradually decreases in the older snakes until in those that are seven years old the females average between 84 and 85 per cent of the male's length (snout to vent). So great is the dispersion in size, that in my small sample the largest two-

is . year-old snake ( male ) as long as the smallest eight-year-old ( female ) 186 University of Kansas Publs., Mus. Nat. Hist.

Table 14. Gains in Length and Weight in Copperheads Marked Soon After Birth, and Recaptured in Subseqi^int Years

Original Record AUTECOLOGY OF THE COPPERHEAD 187

Table 15. Gains in Length and Weight in Copperheads Marked Be- fore They Had Completed Growth (and Recaptthved in Subsequent Years)

Original Record

^—V

Date 188 University of Kansas Publs., Mus. Nat. Hist.

Table 15. Gains in Length and Weight in Copperheads Marked Be- fore They Had Completed Growth— (and Recaptured in Subsequent Years ) Continued

Original Record Record of Recapture

ID bcg o3 *a^— T3 a Date Date > be c3 B J. a

'S3 a 03 CO AUTECOLOGY OF THE COPPERHEAD 189

Table 15. Gains in Length and Weight in Copperheads Marked Be- fore They Had Completed Growth— (and RECAPTxmED in Subsequent Years ) Concluded

Original Record Record of Recapture

^_^ a a bO 5 *3'^— bC -a Date *^ Date > bO j3 o- bC .i. CI en 190 University of Kansas Publs., Mus. Nat. Hist. able with certainty, data from the foregoing table reveals the trend in early growth. For instance an individual captured in autumn having a length of more than 300 millimeters and less than 400 millimeters is almost certainly a one-year-old, and any individual

oc LU

UJ

oI- z UJ .J AUTECOLOGY OF THE COPPERHEAD 191 up to average adult size can similarly be assigned a "probable age" and identified with one of the annual age classes. For such indi- viduals marked, and then captured again after substantial intervals, the original estimated age plus the time elapsed between captures provides an indication of the ages corresponding to various sizes of subadult and adult snakes. For those already adult when marked there is a greater element of uncertainty as to exact age, but even conservative estimates indicate ages up to 14 years for some of the largest copperheads recaptured.

400 —

300 —

< o

X 200 — O

100 —

23456789 10 II 12 13 14 AGE IN YEARS

Fig. 20. Weight in recaptured marked male copperheads, from area shown in Fig. 1, whose ages were known definitely or approximately. The general trend is for weight to increase rapidly from year to year, even in those indi- viduals that have attained sexual maturity. 192 University of Kansas Publs., Mus. Nat. Hist.

Cessation of Growth

In contrast to the majority of recaptured snakes, including all those listed above, a few individuals recaptured after substantial intervals failed to make any growth or grew but little. Some of these were old adults that were already near maximum size and had slowed their growth rates to a minimum level. Others, far short of adult size may have been handicapped by periods of adversity. In some instances the measurement obtained at recapture was ac- tually less than the original measurement. In measuring the elastic body of a live copperhead, the normal range of error was in the neighborhood of one per cent, but occasionally errors as great as three per cent were made. The trends of the combined length and weight records reflect the vicissitudes of the individual's career.

300 —

CO < oa: 200 —

XI- o UJ 5

100 —

12 3 4 5 6 7 AGE IN YEARS

Fig. 21. Weights in recaptured marked female copperheads of known ages. Circles enclosing dots represent averages for age-groups. Asterisks represent females that were obviously gravid, and these individuals are not included in the averages. AUTECOLOGY OF THE COPPEKEBEAD 193

Table 16. Records of Copperhe.\ds Captured and Marked as Adxh^ts, Which Had Made Little or No Growth When Recaptured in Subsequent Years

Original Record 194 University of Kansas Publs., Mus. Nat. Hist.

fully as noticeable to the snake as warm objects. The receptive field is an irregular cone extending in the horizontal plane about 10° across the midline in front and almost at right angles to the body laterally from the pit. Bullock and Fox (1957:231) have described in detail the anatomy of the pit, and have related it to function in sensing the presence and position of prospective prey. Receptor nerve endings average only two to five microns beneath the epider- mal surface inside the pit permitting prompt response, and detection of flickering and brief stimuli. The authors explain that because the pit's diameter is constricted at its mouth, radiant objects will not illuminate the whole sensory membrane but will cast shadows of the pit margin. This confers the possibility of deriving information about the direction of small objects or the edges of large objects. In captivity copperheads that were offered rodents struck and invariably snapped back into a coiled position immediately, releas- ing the prey. In such situations it seemed that the snake's response to the rodent's presence was in part defensive, and in the confines of a small cage the behavior toward prey may not be representative of that under natural conditions. Various authors have suggested tliat behavior of the copperhead differs in dealing with different kinds of prey and tliat some types are struck and released while others are held until the venom takes effect. Instantaneous release of the animal struck usually ensures that the prey will not deliver a retaliatory bite, which might entail serious injury or death to the snake; however, such release might often result in loss of the prey, which would wander too far to be found. Several authors have expressed the opinion that the copperhead characteristically retains its hold after striking a bird or frog, which might be irretrievably lost upon release, but that the snake v/ith- draws from biting a mammal. Conant (1938:112) concluded from observations at tlie Toledo Zoo that large and active prey is bitten and released, whereas smaller prey is retained in the jaws and help- less victims such as newborn mice are engulfed without being bitten, except that the fangs are employed to help work the food down the throat. McCauley (1945:135) reached essentially the same conclusions. Davis' (1938:183) record of a copperhead catch- ing a white-throated sparrow is perhaps the only detailed observa- tion of predation on a bird under natural conditions, and it bears out the supposition that the snake retains its hold awaiting death of the bird. I have never seen a copperhead catch prey under natural condi- tions. On September 8, 1948, a half-grown copperhead found AUTECOLOGY OF THE COPPERHEAD 195 coiled in leaf litter was ofiFered a dead shrew held in steel forceps. When the shrew had been pushed slowly toward the snake to within 1/2 inches, the snake suddenly turned its head toward the shrew and almost instantly struck at it, biting hard just behind the head and retaining its hold. For two minutes the snake did not shift this original grip, but rested motionless except for slight move- ments of its jaws as it sti'ove to embed its fangs more deeply, and injected venom. It then began swallowing the shrew without ever having released it. Swallov/ing was completed approximately 20 minutes from the time the bite was delivered. The behavior of this individual was probably representative of those ambushing prey under natural conditions. A shrew or mouse struck and pierced through the thoracic, abdominal or cranial cavities might succumb in a few seconds. Fast-moving small prey animals such as cicadas, slcinks and birds would usually escape if released and must ordinarily be held in the jaws after the original stroke. The innate caution and nervousness of the copperhead probably causes it to release any prey animal that struggles efiFectively and shows signs of retali- ating, while feebler prey is held down and immobilized until the venom subdues it. On June 22, 1960, when newly metamorphosed bullfrogs were numerous along the margins of the pond on the Reservation, my sons observed a large copperhead that was actively prowling at mid-morning. The sky was overcast and humidity was high. When first seen the snake was swimming near the shore line. For approxi- mately six minutes that it was watched it continued to swim or to crawl rapidly along the edge of the water. On four occasions it left the shore line to for head a bullfrog ( a different one each time ) resting on algae out in the water, and approached the frog with rapid, purposeful movements. Each time, the frog jumped while the snake was at least six inches away, and the snake struck in a futile attempt to secure it. The observers were watching from a boat, approximately 15 feet from shore. In the course of its wandering, the snake came up against the side of the boat and swam along it attempting to by-pass the obstruction but when one of the observers moved, the snake suddenly took alarm, swam rapidly to shore, and escaped into dense vegetation. The feeding of captive individuals may give erroneous impres- sions regarding the habits under natural conditions. Of the many individuals kept by me for varying lengths of time in captivity, few would eat regularly even though preferred natural foods such as voles, white-footed mice or ring-necked snakes were offered. 196 University of Kansas Publs., Mus. Nat. Hist.

Most of the food left in tlie cages was wasted, and when the food supply was limited, force-feeding was usually resorted to despite the fact that it involved some hazard to both the snake and the handler. Risk of injuries to the snakes were reduced by skinning the carcasses to be fed to them, thereby reducing friction as the food was forced into tlie mouth and down the gullet. Strips of raw beef were found to be much more easily swallowed by the snakes and the latter seemed to thrive as well on this diet as on natural foods. John A. Knouse informed me that several of a group of copperheads tliat he kept were induced to take raw ham- burger that had been warmed and offered on the end of a spatula. The acceptance of warmed hambm-ger is significant in connection with the copperhead's preference for warm-blooded prey. Beyer (1898:23) wrote that copperheads kept by him became

tame ". . . learning to take food, such as pieces of meat and fish from the fingers" and he stated that these snakes preferred fish over beef. Gloyd (1928:132) wrote that the copperheads he kept fed well on rats, mice, and sparrows but none showed the slightest interest in fish or frogs. In Ohio, Conant (1938:112) also found that captives readily ate mice and sparrows but none took frogs of the several species that were offered. Nevertheless in Indiana, Minton (1944:475) found that the young would eat small frogs. In Maryland, McCauley (1945:135) found that captives would eat frogs and salamanders as well as birds and mammals. In eastern Oklahoma, Chenoweth (1948:162) found that young copperheads ate small cricket frogs but refused small mice. In southeastern Texas, Guidry (1953:55) found that captives would feed fairly well on mice and small birds but would refuse frogs and toads.

Luring of Prey by Young

The yellow tips of the tails of juveniles may serve as lures to attract prey. Ditmars (1907:424) wrote "Quite frequently, when food is introduced into a cage containing small Copperheads, the tails of the little snakes wriggle and twist in a manner that instantly suggests their remarkable similarity to yellow grubs or maggots. When among dried leaves the colours of the young snakes blend so perfectly with their surroundings that it is almost impossible when a little distance away, to discover them with tlie exception of the bright yellow tail." Neill (1948:161) confirmed this idea with a litter of 11 young from a female caught near Augusta, Georgia. The young were placed in a wooden box with leaf litter. AUTECOLOGY OF THE COPPERHEAD 197 and several cricket frogs (Acris crepitans) were introduced. The box was left in one comer of a room, covered so that the interior was partly darkened. Later, peering into the shadowy interior, Neill saw "... a number of writhing, yellowish objects, for all the world like small worms or maggots . . . each httle copper- head was coiled up and was holding aloft its bright yellow tail, which was writliing slowly." When the frogs became alarmed, and leaped about the container, the snakes wriggled their tails with in- not interested in creased vigor. In confinement the frogs were feeding and none actually fell prey to the copperheads.

Fig. 22. Distal half of the tail of a newborn copperhead from dorsal and ventral views, X 5. The distal one-fourth of the tail is bright yellow ventrally, and dusky yellow with conspicuous markings dorsally.

I have kept several dozen broods of young copperheads, and have often introduced small vertebrates, such as frogs, lizards, snakes or slirews, into the containers with them as prospective food. On many occasions tliese animals have been struck and killed, and sometimes tliey have been eaten, but I have never succeeded in eliciting the tail-waving behavior described by Ditmars and Neill. Therefore I infer tliat the response is less strongly developed in the local population that I have studied than in populations in some other parts of the range. Possibly strength of the reaction is correlated with frog-eating habits, as suggested by Burger and Smith (1950:432). Frog-eating certainly is less prominent in the northwestern population of copperheads than in some others from elsewhere in the range. At any rate, there can be no doubt that the brighdy colored tail in the juvenal Agkistrodon is correlated with luring behavior to obtain prey, and tliat the reaction is more strongly developed in some other members of the genus than it is in the copperhead. Allen (1949:225) pubhshed a remarkable photograph illustrating tail-waving in juveniles of the Mexican cantil (A. bilineatus), and described in detail the behavior of these snakes. "The tail was

8—4428 198 University of Kansas Publs., Mus. Nat. Hist.

bright yellow for 3 cm. from the tip and on both sides, and the extreme end was grey, giving an unmistakable appearance of a

yellow caterpillar with a [grey] head. . . . The juvenile . . . carries its tail in a vertical position with the yellow tip in intermit- tent motion. It resembles nothing so much as a wriggling worm twisting about hungrily in search of food." The juvenal cantils kept by Allen successfully lured and struck various small animals placed in their container, including tree-frogs (Hyla) of different species, oak toads (Bufo quercicus) and an anole {Anolis carolinen- sis?) but usually tliey did not feed, and eventually starved. Even when the young cantils were at rest, with no prey in the vicinity, the tails were normally held elevated from 2 to 232 inches (one- fourth to one-fifth of the total length) and were set actively in mo- tion when prospective prey was near. Allen noted diat the tails were kept down for two days in individuals that had fed, and also were lowered at night. Henry (1924:257) made observations on a htter of the hump- in this nosed viper ( A. hypnale ) of India which showed that species also tail is as a lure. small skink the used effectively When a ( Lygo-

soma ) was introduced ". . . their tails, which were of a whitish colour, were protruded from the coils and caused to wriggle about in an extraordinary manner, looking for all the world like so many

very active earthworms . . . whenever a small lizard of any kind was put into the cage, the tail-wriggling immediately com-

menced. . . . On several occasions I saw small geckos actually seize a snake's wriggling tail and instantly receive a fatal wound from the venomous little creature."

Statements of Food Preferences

Many authors have made statements concerning the food habits of the copperhead. Some of these doubtless were based on original observations, but unfortunately cannot readily be separated from statements that merely reiterate the findings of earlier writers.

". . . their food [in New York] consists of birds, frogs, mice, and even squirrels, which they catch by surprise as they do not climb trees" (Rafinesque, 1819:86).

"Its usual food seems to be small birds and field mice. . . ." (Holbrook, 1838:71.)

"Slugs, birds and mice have been found in their stomachs. . . ." (Atkinson, 1901:152, in Allegheny County, Pennsylvania; the find- ing of slugs in the food has not been corroborated. ) AUTECOLOGY OF THE COPPERHEAD 199

". . . their food [in Texas] consists of small rodents, lizards, frogs. . . ." (MitcheU, 1903:27.) "The feeding habits are rather eccentric and seemingly relate to the possibihty of finding certain kinds of food during diflFerent phases of the season. . . . During the spring and fall it is very fond of frogs. . . . During the later spring, these snakes prefer young birds, showing in fact such a decided preference to this food that some snakes will fast unless provided with the feathered prey. During the summer months captive specimens will eat small rodents, such as mice and rats, or chipmunks." (Ditmars, 1907:424.) "The food consists of frogs, small birds and rodents." (Ditmars, 1910:338.) "Small rodents, birds and frogs." (Lamson, 1935:26.)

". . . known to eat mice, squirrels, shrews, birds, frogs, insects, salamanders and even opossums." (Gowanloch, 1943:47.)

". . . it feeds [in spring] mostly on mammals ... in the summer . . . large numbers of frogs and insects are consumed in addition to mammals." (Oliver, 1955:174.) "The catholic appetite is satisfied by small warm-blooded animals that as well as small cold-blooded ones even include insects" ( Pope, 1955:224). "The food consists chiefly of mice, with occasional birds, and even large insects and larvae of insects." (Smith, 1956:306.) "Its food consists mainly of small mammals, but when large in- sects, like some caterpillars, are available, they are regularly taken." (Schmidt and Inger, 1957:266.) "Small rodents, small birds, frogs" [subspecies contortrix]; "Small rodents, lizards, frogs, toads" [subspecies laticinctus]; "Small mam- mals, birds, insects, toads, salamanders, but mainly rodents and insects" [subspecies mokeson] (Wright and Wright, 1957:906-913).

Composition of the Diet

Of the 589 prey items identified in my study, 77 were from scats collected in the large cage at La Cygne, Kansas, where Vernon Mann had kept many copperheads and smaller numbers of several other kinds of snakes. Because these La Cygne scats could not be identified with individual snakes and a few of them may have been produced by snakes other than copperheads their records have been kept separate from those of the scats definitely known to have been produced by copperheads. Among the 512 items in the known copperhead scats and from 200 University of Kansas Publs., Mus. Nat. Hist.

stomachs of copperheads there were: 90 prairie voles {Microtus 80 cicadas ochrogaster) , {Tibicen pruinosa), 66 white-footed mice (mostly Peromyscus leucopus, though a few were definitely identi- fied as P. maniculatus and many were identified only to genus), 39 short-tailed shrews 35 {Blarina brevicauda) , ring-necked snakes (Diadophis punctatus), 33 little short-tailed shrews {Cryptotis parva), 30 five-lined skinks {Eumeces fasciatus), 29 caterpillars (Actios luna, Celerio ? and several otlier saturnids and sphingids not definitely identified), 24 pine voles (Microtus pinetorum), 18 harvest mice (Reithrodontomys megalotis, and possibly one or more R. montanus), 13 narrow-mouthed toads (Gastrophryne olivacea), 8 frogs (Rana pipiens and possibly others), 6 jumping mice (Zapus hudsonius), 6 slender glass lizards (Ophisaurus at- tenuatus), 6 cotton rats (Sigmodon hispidus), 4 each of worm snake (Carphophis amoenus), and house mouse (Mus musculus), 3 each of brown sldnk (Lygosoma laterale), and common garter snake (Thamnophis sirtalis), 2 each of racer (Coluber constrictor), east- ern wood rat (Neotoma floridana), eastern cottontail (Sylvilagus

Table 17. Estimated Percentages, by Weight, of Vahious Prey Species IN the Diet of the Copperhead at Two Localities in Eastern Kansas

Weight Percentage by weight of sample Kind of Pbey individual prey item in grams Reservation La Cygne

Prairie vole 30 38.6 1.4 White-footed mouse 18 16.3 6.0 Pine vole 30 10.5 4.3 Short-tailed shrew 12 6.7 2.3 Cotton rat 40 3.5 13.4 Ring-necked snake 6 3.1 .3 Five-lined skink 7 2.8 1.3 Little shortstailed shrew 6 2.5 .3 Harvest mouse 10 2.5 Cicada 2 2.1 .5 Jumping mouse 15 1.3 Eastern wood rat 40 1.2 1.9 Eastern cottontail 40 1.2 65.0

Luna and other moth larvae . 2 .6 .3 Narrow-mouthed toad 4 .7 .2 Glass lizard 10 .9 .5 House mouse 15 .6 .7 Bird 20 .3 1.0 Common garter snake 10 .3 .5 Leopard frog 10 2.9 Other 1.4 .1

100.0 100.0 AUTECOLOGY OF THE COPPERHEAD 201

floridanus), De Kay's snake (Storeria dekayi) and bird {Rich- mondena cardinalis?, Spinus tristis?) and one each of hatchling turtle ornata Great Plains skink ohso- box ( Terrapene ) , ( Eumeces six-lined black letus), racerunner {Cnemidophorus sexlineatus) , rat snake {Elaphe ohsoleta) and unidentified snake. Of the 77 items in the scats from La Cygne, 34 were cottontails, 7 were white-footed mice, 7 were cotton rats, 6 were cicadas, 4 were short-tailed shrews, 4 were five-lined skinks, 3 were pine voles, 3 were caterpillars, and there were one each of prairie vole, Httle short-tailed shrew, ring-necked snake, narrow-mouthed toad, glass lizard, common garter snake, wood rat, house mouse, and bird (Agelaius ?). Since the snakes had not been fed in captivity, these prey items represent natural feeding. In the copperhead, meals are relatively large and infrequent and the prey is invariably swallowed entire. Since the various prey species differ greatly in size, composition of the diet is best shown by calculating the percentage by bulk of each species in the total food consumed (Table 17).

Kinds of Prey

In this study the prairie vole proved to be by far the most im- portant species in the copperhead's diet, with more than twice the biomass of any other species. On the Reservation this vole is by far the most abundant small mammal (Martin, 1956:376; Fitch, 1957:131). By 1958 more than half the Reservation's area provided habitat favorable for the vole. The grassy and weedy fields where prairie voles occurred were of irregular shapes and were well dis- tributed over the Reservation. No point was more than 500 feet from such habitat, and probably almost every copperhead had voles within its home range. Spring dispersal of copperheads from the rock ledges in woodland where they hibernate to grassland habitat is perhaps motivated by the abundance of the voles in the grass- lands. At all ages and sizes the voles provide food for adult cop- perheads but the voles are in varying degrees unavailable to the younger snakes, which depend largely on other kinds of prey. The vole's habit of keeping to well-defined runways renders it easy prey for the copperhead, which often lies in or beside runways motionless but ready to ambush any small mammal that may come within reach. By day, when temperature is unfavorably high and humid- ity low the vole's burrows provide underground shelter for the copperheads in grassland habitat. The pine vole is relatively uncommon on the Reservation and at 202 University of Kansas Publs., Mus. Nat. Hist. times may be less than one per cent of the prairie vole's numbers (Fitch, 1958:80). In view of its relative scarcity, the pine vole is eaten with surprising frequency. In habitat and in over-all geo- graphic range it corresponds closely with the copperhead. Over the copperhead's range as a whole, it is possibly the one most im- portant food source. Surface (1906:189) found meadow voles (Microtus pennsylvanicus) in 13 stomachs of copperheads from Pennsylvania. In Vii-ginia, Uhler, Cottam and Clarke (1939:610) found microtines of four kinds in 20 stomachs; the meadow vole, pine vole, red-backed mouse (Clethrionomys gapperi) and south- ern bog lemming {Synaptomys cooperi). Hamilton and Pollack (1955:3) recorded one pine vole in a stomach from Georgia. Bush (1959:76) found pine vole (one individual?) to comprise one-sixth by bulk of the total sample from six Kentucky copperheads that contained food.

"Mice" collectively including two species of white-footed mice, harvest mice, jumping mouse and house mouse, were next in im- portance to voles. The white-footed mouse (Peromyscus leucopus) prefers the same woodland and edge habitat occupied by the cop- perhead, and is generally abundant over much of the copperhead's range. Therefore this mouse is probably a major food source. This mouse prefers the same sort of rock ledge situation which the cop- perhead chooses for hibernation, and as a result it figures in some of the earliest and latest seasonal records of the copperhead's feed- size of a ing. As it averages only a Httle more than half the vole, this mouse is available as food to copperheads over a wider range of size. The deer mouse was tentatively identified in 12 occurrences. It is limited to small areas of the Reservation, where vegetation is doubtless sparse. Most of the occurrences of "Peromyscus sp," per- tained to the commoner and more generally distributed white-footed mouse. The harvest mouse because of its relatively small size, is available as food even to young copperheads except those of the smallest size groups. Harvest mice were taken more often than mice of any other kinds except those of the genus Peromyscus, reflecting their abundance and extensive habitat on the Reservation. The favors interspersion of grassland and woodland habitats on this area but else- predation by the copperhead on this grass-living rodent, where harvest mice probably figure less importantly. The geo- mouse graphic ranges of the copperhead and the western harvest overlap but little. The range of the eastern harvest mouse (R. there are no humulis) is largely within that of the copperhead but AUTECOLOGY OF THE COPPERHEAD 203 definite records of predation on this species. The meadow jumping mouse occurred six times among the food items identified from the Reservation. The range of this mouse overlaps almost the northern half of the copperheads' range, and the habitats of the mouse and the snake are similar. The house mouse occurred four times among the recorded items, but probably comprises only a small part of the Clarke loc. cit. diet over the entire range. Uhler, Cottam and ( ) found white-footed mice (Peromyscus sp.) in seven digestive tracts and a meadow jumping mouse in one. McCauley (1945:135) re- corded a white-footed mouse, a house mouse and a meadow jumping mouse in stomachs of specimens from Maryland. Clark (1949:259) recorded 15 mice, species undetermined, among the stomach con- tents of 55 copperheads from Louisiana. Barbour (1950:106) found in one of two a jumping mouse ( Napaeozapus insignis ) copperheads examined from Big Black Mountain, Kentucky. Hamilton and Pol- cit. a white-footed mouse in lack ( loc. ) found ( Peromyscus sp. ) that one stomach and cotton rats in four. Bush ( 1959:76) recorded white-footed mice (two individuals?) comprised 58.3 per cent of the total sample in six copperheads from Kentucky that contained food. cit. in one Surface ( loc. ) found a white-footed mouse stomach, house mice in two others, unidentified mice in three, and unidenti- fied mammals in three. Rats made up only 1.6 per cent of the items recorded from the Reservation. The eastern wood rat was scarce on the area for most of the period of the study. The cotton rat was abundant in 1958 and 1959 but relatively scarce in some other years. Rat-sized ro- dents, when fully adult, are too large to be swallowed by any but the largest copperheads. However, during the summer, the bulk of the population consists of immature individuals. In the outdoor enclosure where copperheads were kept, partly grown cotton rats were eaten avidly whenever they were offered. There were no sciurids among the items identified in my study, but Uhler, Cottam and Clarke (loc. cit.) found chipmunks (Tamias striatus) in two loc. cit. also recorded and an unidentified squirrel in one. Surface ( ) an unidentified squirrel in one stomach. He also recorded opossums (Didelphis marsupialis) from three. These latter records are re- markable, since a young opossum at the time it first emerges from the mother's pouch, is already a large morsel for an adult copper- head. Rabbits make up a variable but sometimes important part of the diet. On the Reservation only one occurrence of the cottontail was 204 University of Kansas Publs., Mus. Nat. Hist.

recorded among the total of 512 items even though cottontails were common on the area. But in the samples from La Cygne 34 of the 77 items were cottontails. The snakes from which these scats were obtained were mostly gravid females which were collected in Au- gust, along woodland rock ledges. Perhaps cottontails were un- usually numerous at this particular time and place. Only young in the nest during their first week or two would be small enough to be eaten by a copperhead, and the snake would need to be fully adult. The eastern cottontail and other rabbits of the same genus occur throughout the copperhead's range and their young may con- stitute an appreciable percentage of the food. The shrews, Blarina brevicauda and Cryptotis parva, represented by 72 occurrences, constitute an important part of the diet, espe- cially for immature copperheads. The smaller kinds of shrews (Cryptotis and Sorex) are almost the only mammals within the range of the copperhead that are small enough to be eaten as adults by the youngest snakes. Both Blarina and Cryptotis coincide ap- proximately with the copperhead in their geographic ranges and Blarina has almost the same habitat preferences. The shrews are both diurnal and nocturnal in their activities, and occur in the same type of dense cover used by the copperheads. Surface recorded a short-tailed shrew in one and an unidentified shrew in another. Conant (1938:112) reported a half-grown hairy-tailed mole (Para- scalops hreweri) eaten by a copperhead in Licking County, Ohio. loc. cit. Uhler, Cottam and Clarke ( ) found shrews, including the short-tailed shrew, little short-tailed shrew and masked shrew Sorex cinereus in the ( ) stomachs of eleven of the copperheads from that examined. loc. cit. Virginia they McCauley ( ) recorded a short- tailed from a cit. shrew specimen from Maryland. Barbour ( loc. ) in found a shrew ( Sorex sp. ) one of the two examined from Harlan County, Kentucky. Small snakes were found 48 times in my samples, and proved to be important in the food of young copperheads. They were chiefly the ring-necked snake, which is by far the most abundant reptile of the Reservation, and is estimated to occur in population densities of ten or more per acre over extensive areas (Fitch, 1958:79). In captivity the ring-necked snake was almost the only prey taken vol- untarily by Juvenal copperheads. Ring-necked snakes seemed re- markably susceptible to the copperhead's venom; in less than a minute after being bitten they were incapable of normal locomo- tion, and would die after violent contortions, over a period of min- AUTEOOLOGY OF THE COPPERHEAD 205 utes. Other small snakes eaten include the worm snake, De Kay's snake, and young of the garter snake, racer and black rat snake. These and many other kinds are available throughout most of the copperhead's range. Probably any small snakes are taken more or loc. cit. a milk snake less indiscriminately. Surface ( ) reported (Lampropeltis doliata) in the stomach of one from Pennsylvania. loc. cit. a snake Tantilla Hamilton and Pollack ( ) reported crowned ( loc. cit. an instance of a coTonata ) in the food. Barton ( ) reported Juvenal copperhead bom in captivity which ate a small water snake (Natrix rhombifera). Six days later the copperhead died with the other snake's tail still protruding from its mouth. Of the lizards eaten, the five-lined skink, ground skink, Great Plains skink and glass lizard were taken in much different quanti- ties. The five-lined skink was sixth in number of all the species of prey taken. It is remarkably abundant in woodland and edge habitat; an estimate of 67 per acre was made for a 2/4-acre study area (Fitch, 1958:78). The relatively scarce ground skink, Great Plains skink and glass lizard were taken by copperheads on the Reservation in smaller numbers somewhat proportional to their abundance. Lizards are eaten chiefly by immature copperheads. Up to at least half-growTi size, five-lined skinks can be swallowed easily by newborn copperheads, and constitute an important source of food for them. Vernon Mann (in conversation April 29, 1958) mentioned finding a young copperhead in the act of swallowing a loc, five-Hned skink near La Cygne. Uhler, Cottam and Clarke ( cit. lizard undulatus in one of the 105 ) found a fence ( Sceloporus ) copperheads from Virginia examined by them. Minton (1944:475) recorded that a copperhead from Indiana disgorged a large fence cit. lizard. Hamilton and Pollack ( loc. ) reported undetermined lizards (Cnemidophorus, Sceloporus, Eumeces or Lygosoma) in two. Bush (loc. cit.) reported Lygosoma (two individuals?) com- prising 16.5 per cent of the total sample of food from six copper- heads from Kentucky. Robert G. Webb in an unpublished thesis in the University of Oklahoma Library, recorded that a copperhead from Comanche County, Oklahoma, contained a collared lizard in its stomach. Amphibians would seem to be one of the most available food sources. After summer rains copperheads are most active and at the same time dispersing frogs and toads, mostly juveniles, swarm over the fields and woodlands. Opportunities to feed upon them must occur frequently. It must be concluded that amphibians are 206 University of Ka.nsas Publs., Mus. Nat. Hist. low on the scale of preference since few were found in the snakes and most of these were the Great Plains narrow-mouthed toad (Gastrophryne oUvacea). Anderson (1942:216) recorded finding a juvenal copperhead in the act of swallowing one of these toads in Jackson County, Missouri. This small, terrestrial, and partly sub- terranean toad is numerous in the copperhead's habitat, but at times is much outnumbered by Btifo, Acris and Rana. Most identifications of narrow-mouthed toads were made from ants (Crematogaster in the as the di- sp. ) scats, toads themselves had been completely gested. One leopard frog ( Rana pipiens ) was found in a stomach. It is doubtful whether frogs would have left any remains that would have been recognizable in the scats. As already mentioned, re- mains of insects of kinds that probably would not have been eaten by the snakes were usually associated in the scats with remains of insectivorous vertebrates—mice, shrews, and lizards—in most in- stances. The seven instances in which they were not so associated were tentatively allocated as frogs, probably the leopard frog, but possibly including some of the other ranid, hylid, or pelobatid anurans occurring on the Reservation. Vernon Maim told me of finding a copperhead eating a small bullfrog {Rana catesbeiana) when he was attracted to the spot by the squalling of the frog. Among 55 items found in stomachs of copperheads in northern Louisiana, Clark (loc. cit.) found 30 frogs—22 Rana pipiens, seven R. clamitans, and one R. catesbeiana. Frogs seemed to be much higher on the copperhead's scale of preference in Louisiana than are in loc. cit. found salamanders diey Kansas. Surface ( ) two slimy (Plethodon glutinosus) in the stomachs of a series of 52 from Pennsylvania. Uliler, Cottam and Clarke {loc. cit.) found frogs in of ( Rana sp. ) the stomachs of two ( 105 ) from Virginia. These same authors found eight slimy salamanders, one red-backed sala- mander {Plethodon cinereus), and one red salamander {Fseudotri- ton ruber) in the same series. "Bird" was represented by three occurrences in my records, each of a different species. Because the remains of feathers were meager and in poor condition, definite specific determinations could not be made, but in each instance the color provided a clue. One scat from La Cygne, Kansas, in August, 1958, contained black feathers, which a red- blackbird or cow- probably were those of winged ( Agelaius ) bird {Molothrus). Contents of the digestive tract of a large male copperhead caught six miles east and one mile south of Arkansas City, Kansas, on June 1, 1954, contained yellow feathers that may AUTECOLOGY OF THE COPPERHEAD 207 have been those of a goldfinch {Spinus americanus) but possibly were from some land of warbler. The remaining scat, from a copperhead caught on the Reservation on October 22, 1949, con- tained red feathers, which almost certainly were those of a cardinal

(Richmondena cardinalis) . Opportunity to prey upon birds prob- ably comes when fledgings still unable to fly or climb efiPectively are wandering about on the ground. Several kinds that share the copperhead's woodland habitat on the Reservation and seem es- pecially vulnerable in this regard are the yellow-billed cuckoo (Coccyzus americanus), whip-poor-will (Caprimulgus vociferus), Carolina wren warbler (Thryothorus ludovicianus) , Kentucky (Oporornis formosus), summer tanager (Piranga rubra), cowbird Molothrtis ater towhee car- ( ) , red-eyed ( Pipilo erythrophthalmus ) ,

and field . dinal, sparrow ( Spizella pusilla ) Davis (1938:183) observed an instance of predation by a copper- head on a white-throated sparrow (Zonotrichia alhicollis), near Bastrop, Texas, on February 27, 1938. Thrashing movements among dead leaves drew the attention of the observer to the bird, still struggling, its head in the grip of the copperhead, which periodically clamped its jaws tighter as if to embed its fangs more deeply or inject more venom into the prey. The sparrow's struggles soon be- came feebler, and in three minutes it was limp and lifeless. The snake attempted to drag its prey back into a pile of litter, but re- leased it when disturbed. Wintering flocks of this and various other sparrows are probably subject to but little predation by copperheads because the snakes are normally hibernating at least throughout most of the birds' in their loc. cit. sojourn range. Surface ( ) recorded a fringilhd "sparrow" in the stomach of a copperhead from Penn- sylvania. Uhler, Cottam and Clarke ( loc. cit. ) recorded six occur- rences of birds, including unidentified passerines, a warbler (Den- droica a sp. ) and ruby-throated hummingbird (Archilochus colu- bris), in digestive tracts of 105 copperheads from Virginia examined by them. How a copperhead might secure such elusive prey as a hummingbird, or even a warbler, is a matter for speculation. Clark loc. cit. recorded ten birds in his ( ) ( species undetermined ) total of 55 food items from copperheads collected in northern Louisiana. In one exceptional instance a hatchling box turtle (Terrapene ornata) was found in the digestive tract of a large adult male cop- perhead, the same one, from near Arkansas City, Kansas, that had eaten a bird tentatively identified as a goldfinch. This snake also had in its stomach remains of two voles and a racer, more separate 208 University of Kansas Publs., Mus. Nat. Hist. items than ocxjurred in any other specimen. The irregular shape and protective armor of a turtle v^^ould render it difiBcult to swallow for any copperhead except an exceptionally large one. Eating of this

10 15 20 25 30 35 LENGTH OF SNAKE IN INCHES

Fig. 23. DifiFerences in food of copperheads of dif- ferent sizes. Voles and mice are eaten chiefly by adults. Cicadas are eaten by copperheads of all sizes, but most often by adults, whereas shrews (Blarina and Cryptotis), narrow-mouthed toads, and especially small snakes such as the ring-neck (Diadophis) are eaten chiefly by the first-year yoxmg. AUTECOLOGY OF THE COPPERHEAD 209 hatchling box turtle indicates a certain versatility on the part of the Pollack loc. cit. copperhead in its choice of prey. Hamilton and ( ) found three small musk turdes {Sternothaerus odoratus) in a large adult male copperhead from Georgia. Insects that were primary prey items were easily distinguished from those that were secondary (prey of the animals eaten by the snakes) in most instances. They were of larger lands and were more nearly intact. The common cicada (Tibicen pruinosa) was the favorite insect prey, with 80 recorded occurrences, and was second in frequency only to the prairie vole among the many kinds of prey eaten. Although many species of cicadas occur on the Res- ervation, only this one common species was definitely recorded in the food. Like other kinds of cicadas, Tibicen pruinosa has a long period of development; the nymphs remain underground for many years feeding on roots of trees. The nymphs emerge and metamor- phose in the latter half of the summer and the adults die with the advent of cold weather in late autumn. Ordinarily the nymphs are unavailable to copperheads throughout the period of their under- ground existence. The adults likewise are usually safe because of their wariness and their habit of perching several feet above ground. They are vulnerable mainly within a period of a few hours when the nymphs emerge to metamorphose. They crawl about slowly on the ground and then often climb onto a vertical surface such as a rock, stem, or tree trunk. The emerging imago is soft and helpless at first. Ten of the cicadas recorded as food of copperheads were nymphs, all full-sized and probably caught by the snakes soon after they had emerged to metamorphose. Probably most of the imagos eaten were caught soon after metamorphosis, before they had had time to dry thoroughly. Many cicadas of the same brood may emerge at about the same time within a few square yards, and their

: five total biomass is great. Atkinson ( 1901 152 ) found nymphs of cicadas in the stomach of a copperhead from Allegheny County, state Surface loc. cit. found stomachs Pennsylvania. In the same ( ) of six gorged with seventeen-year cicadas (Magicicada septen- decim). Gloyd (1928:132) wrote that a copperhead collected by him near Gould's Ford in Kansas had eaten several soft-bodied loc. cit. examined a cicadas just transformed. Conant ( ) copperhead in the Carnegie Museum from Washington County, Ohio, that had eaten a seventeen-year cicada that worked its way through the neck of the snake, causing the latter's death. In Dallas County, Texas, Curtis (1949:12) found several copperheads climbing in low trees and shrubs, and upon dissecting the snakes, found them to be 210 University of Kansas Publs., Mus, Nat. Hist. gorged with cicadas. McCauley {loc. cit.) reported "periodical" cicadas in the food of copperheads in Maryland. Gehlbach (1956: 370) reported that a copperhead found in Santa Elena Canyon, Brewster County, Texas, in mid-June, voided remains of both nymphal and adult cicadas. The cicadas were numerous in the tall grass among limestone slabs where the snake was discovered. Rob- ert G. Webb in "The Reptiles of Oklahoma," an unpubHshed manu- of script in the University of Oklahoma Library, recorded that three five copperheads caught about 10 p. m., had each eaten two cicada nymphs. Besides cicadas the only insects eaten regularly by copperheads are lepidopterous larvae of the famihes Sphingidae, Citheroniidae, Saturnidae, Ceratocampidae, and perhaps others. These accounted for 29 occurrences in my samples. On October 12, 1951, a larva, three inches long, of a sphinx moth (Celerio?) was found in the stomach of a small copperhead that had died in a trap. Three larvae of the luna moth (Actios luna) were found in the stomach of a copperhead collected in Polk County, Texas, on October 18, 1958. One of these larvae had begun to pupate and was partly enclosed in its cocoon, and probably the snake that ate it found it by scent. Other recorded occurrences of lepidopterans were aU from scats, and the remains were inadequate for specific or generic determina- tions. Adult moths have not been recorded in the natural food of copperheads. When a large hawk moth was released in a cage with two of the snakes, both showed unusual animation, alertly following the movements of the fluttering moth and lunging at it whenever it came within reach. One snake soon caught the moth and ate it. This three-year-old copperhead had been reared in captivity and had been sustained entirely by force-feeding, as it would not accept other kinds of prey that had been offered on various occasions. Sev- eral times subsequently hawk moths were offered to caged copper- heads, and were always avidly pursued and eaten. When smaller moths were introduced into the cages, the snakes watched them with seeming interest, turning their heads to follow the movements of the moths, but not attempting to catch them. the Surface {loc. cit.) found larvae of polyphemus moth (TeJca pohjphemus) in digestive tracts of two copperheads, of the io moth in of the {Atitomerus io) in two, of the oak worm (Anisota) two, the imperial moth (Eacles imperialis) and regal moth {Citheronia and Clarke found regalis) each in one. Uhler, Cottam (loc. cit.) 28 of the 105 ex- caterpillars of seven genera in copperheads they AUTECOLOGY OF THE COPPERHEAD 211 amined from Virginia. Orth (1939:54) found larvae of sphingid moths in the stomach of an adult copperhead from Harriman State Park, New York. Orth oflFered the larva of a polyphemus moth to a half-grown copperhead in captivity, and the snake soon ate it. Malnate (1944:731) found a nymphalid caterpillar in the stomach of a copperhead from South Carolina. Barbour {loc. cit.) found a in of of spliingid larva the stomach one ( two ) from Harlan County, loc. cit. Kentucky. Hamilton and Pollack ( ) reported a lepidopteran loc. cit. larva in one of 16 from Fort Benning, Georgia. McCauley ( ) reported caterpillars in the food of copperheads in Maryland. Bush (loc. cit.) also reported unidentified caterpillars, in a food sample from Kentucky. It is curious that insects so dissimilar as cicadas and larvae of large moths are highly preferred foods while other kinds of arthro- pods are rarely taken. Remains of a katydid were present in one scat, but they were in fragmentary condition and probably the katy- loc. cit. that did had first been eaten by a frog. Carpenter ( ) found a large male copperhead in Oklahoma contained a spider in its stomach, and McCauley (loc. cit.) reported spiders as part of the Pollack loc. cit. food in Maiyland. Hamilton and ( ) found a mantis (Stegomantis) and a locust (Scudderia) in the stomachs they ex- amined and both these large insects were considered to be primary food items since no other prey was associated with tliem. Neill and Allen (1956:172) questioned whether the mantis and locust recorded by Hamilton and Pollack were actually primary food items. The former authors cited instances of amphibians eaten by snakes being almost completely digested, while the insect prey ingested by the amphibians remained relatively intact. The spiders recorded in the food by McCauley and Carpenter might similarly be suspect as secondary items even though no remains of vertebrates were associated with them.

Amount of Food Consumed

Poikilothermal vertebrates in general and snakes in particular have low metabolism and their food requirements are correspond- ingly small. Doubtless there are important differences in quantita- tive food requirements between different types of snakes. Although no studies of this subject have been made, active and nervous snakes such as racers might be expected to require more food than sluggish kinds such as the copperhead. Certainly racers feed much oftener. Temperature affects the food requirements; in the locality of my 212 University of Kansas Publs., Mus. Nat. Hist.

study the snakes hibernate for more than half the year, fasting throughout this period and losing but Httle weight. Even in those parts of spring and fall that are included in the season of activity, food requirements are much reduced because of relatively low temperatures especially at night. Digestibility of the type of food taken also a£Fects the quantities required. When cicadas are eaten, the heavy exoskeletons are sometimes voided nearly intact. The chitin making up much of tlie biomass in such prey, is largely resistant to the digestive se- cretions of the snake, while some other kinds of prey, such as frogs, are so completely digested that no recognizable traces remain. The residue in dried feces from such prey is scanty and is of powdery consistency and dark greenish brown or nearly black. Crotalids, including the copperhead, are especially well adapted for fasting as compared with other snakes. The normal interval between successive meals is relatively long, the prey is large, and the snakes have the capacity to store quantities of fat in the ab- dominal cavity. This fat supply is drawn upon in times of enforced fasting, and the snake can fast for several weeks without deteriorat- ing noticeably in condition. A copperhead can survive for much longer periods of fasting but gradually becomes emaciated. Surface (1906:124) recorded one that lived for a year and three months in captivity without feeding. Carr (1926:104) wrote of one caught on July 7, 1924, that would not feed in captivity and was still fast- ing on June 17, 1925, although it had been active throughout the winter. Klauber (1956:650) mentioned fasts in rattlesnakes of several species at the San Diego Zoo, at a year-round temperature near 80° F., of: 23 months, 19 months, 16 months, 16 months, 16 months, 16 months and 15 months. Copperheads of similar sizes might survive as long. Since there is no need for food and but little loss of weight in winter hibernation, it is conceivable (but improbable) that an individual under natural conditions might live for three years or more witliout taking any food. Young in- dividuals certainly would starve to death much sooner. The normal food consumption is incompletely known, but cer- tainly the interval between meals is irregular, and the amount eaten at one time is highly variable. One basis for estimating the average food consumption is the rate of digestion in captive individuals and the proportion of those captured that have food in their stomachs or intestines. Another basis is provided by the amount of food AUTECOLOGY OF THE COPPERHEAD 213

in and consumed by captive copperheads. In one bom captivity reared to adult size, weight fluctiiation and amounts of food con- in a six-month that sumed (cliiefly through force-feeding) period corresponded to the maximum extent of a season of activity in this locaUty are shown in Table 18.

Table 18. Food and Weight of a Captive Copperhead in 1957

How Kind Weight Weight Date food was of of food of snake ingested food in grams in grama

257 April 15 fed mouse 14 254 April 25 fed beef 9 May 6 fed vole 28 280 May 24 ate 2 mice 34 June 8 ate mouse 17 June 23 fed beef 10 257 July 18 ate rat 26 254 July 30 fed beef 12

. fed beef 5 August 5. . September 6 fed glass-lizard 18 44 238 October 27 . fed beef and mouse

During the period covered by these records the snake gained in snout-vent length from 585 to 620 mm. Throughout, it was over- weight, and a weight of a htde more than half its average would have been typical for non-gravid individuals of the same length and sex. By the summer of 1958 this snake was fully adult and it made httle gain during the summer. It was recorded to be 630 mm. long on May 20 and 634 mm. on October 22. The food taken in this period is shown in Table 19.

Table 19. Food and Weight of a Captive Coppebhead in 1958

Date 214 University of ICansas Publs., Mus. Nat. Hist.

In tliis five-month period spanning the normal season of activity for copperheads locally, this individual consumed only a httle more than half the equivalent of its own body weight. Nevertheless it made a slight net gain. Klauber (1956:631) stated that mature rattlesnakes in captivity thrive on an adequate meal (presumably of about one-fourth the body weight) every 14 to 18 days, and that young need to feed somewhat more frequently to thrive. For them a weekly feeding was considered desirable. He speculated that in the wild, require- ments might be somewhat increased because of the more active life, with greater expenditure of energy. Klauber (op. cit. -.647) further stated that the annual food requirement in captive adult rattlesnakes, that were active throughout the year, amounted to approximately 4% times the snakes' body weights, and he implied that approximately half that amount might suflBce for those in the wild having a long period of hibernation annually. Copperheads, being closely related to rattlesnakes and somewhat Uke them in habits, probably have similar food requirements. Less than one-fourth of the snakes captured in my study were recorded to have anything in their digestive tracts. However, in the early years of the study the snakes were not thoroughly tested for food residues in their hind guts. Also the bulk of my records were ob- tained along the hibernation ledges in fall when the snakes were much less inclined to feed than they were in summer. In the periods June 1 to October 1 in 1958 and 1959 combined, 336 captures were recorded and in these snakes 186 prey items were obtained from scats and only 28 items were obtained from stomachs. In six instances the same kind of animal and presumably the same indi- vidual was recorded from both the stomach and the scat. In these instances the food palped from the stomach was pardy disintegrated by the digestive juices, especially the parts most posteriorly situated in the stomach. The ratio of 28 stomach items to 186 scat items might be interpreted as indicating that remains are retained in the intestines six times as long as they are held in the stomach, but such a conclusion does not agree with other types of evidence in- cluding those provided by snakes kept in captivity. Actually the relative numbers are probably much distorted by changed habits of the snakes that have recently fed. Captive copperheads that had ingested large food items were inclined to be unusually sluggish and spent most of their time coiled beneath any available shelter. Snakes living under natural conditions must have altered their be- AUTECOLOGY OF THE CoPPERHEAD 215 havior in similar fashion and as a result were less often caught in basis of with ti-aps or found in the open. On the experience captive snakes it is estimated that an average meal would be detectable in the stomach for from three to five days before being reduced completely or passed into the intestine. Digestion does not proceed uniformly in an object reposing in a copperhead's stomach; the a more posteriorly situated parts are digested most rapidly. In mouse swallowed head first the head and forequarters may be com- the feet pletely disintegrated after two or three days, while hind and tail are still intact. If the prey is large, the residues from the anterior portion (normally swallowed first by the snake) may be voided as a scat while the remainder is still being digested. If the prey is small, the scanty residues may be retained in the intestine until the remains of a second meal are added to it completing the formation of a scat. In only one instance did a copperhead have more than one prey item in its stomach. In this instance the prey consisted of a lactat- ing female vole and three small young of approximately the same size, probably her litter all eaten at the same meal. Of the 381 scats examined six contained nothing recognizable, 215 had only a single item, 52 had two items, seven had three items, and one had five items. Thus nearly 39 per cent of the items in scats were found to be associated with others. Doubtless the true percentage of such multiple occurrences was even greater, but the hair by which mammalian items were usually identified gave no clue as to the number of individuals. Therefore each occurrence of hair was re- corded as a single individual, although some such occurrences may have represented two or more animals. The much higher propor- tion of multiple occurrences in the scats seems to indicate that prey remains are retained considerably longer in the intestines than they are in the stomach. Ordinarily it was possible to count the number of individual animals in a scat only when they were of diflFerent kinds, but the chitin of cicadas and other insects was so resistant to digestion that the numbers could be determined readily. Often remains of two or three were associated in the same scat. For the summers of 1958 and 1959 the following figures were obtained: Total captures of copperheads, 336. Total prey items from copperheads, 208 (186 from scats and 28 from stomachs, of which 6 were the same). Total snakes with scats, 157. Total snakes with stomach items but no scats, 22. Total snakes with empty digestive tracts, 157. 216 University of Kansas Publs., Mus. Nat, Hist.

It is notewortliy that the numbers of snakes with empty digestive tracts and those with scats were exactly equal. However, those still digesting food in the intestines or recently finished probably were less active than those which had fasted and were caught in cor- respondingly smaller numbers. If this speculation is correct, the proportion of the population having food remains in the intestines at any one time may be considerably more than half. Certainly snakes with food in their stomachs were represented in less than their true proportions, partly because those that had recently fed were less active and less likely to enter traps, and pardy because snakes that were trapped usually spent a day or more in the traps before they were found and in many instances may have had time to complete digestion of a meal already in the stomach. Scats may have been lost also by being disintegrated and washed through the quarter-inch wire mesh of the traps in heavy summer rains. In the winter of 1959-60 two copperheads in the size range of small adults were kept indoors (diurnal maximum temperature 70° F. and nocturnal minimum 53° F. at the place where the cage was located) and were force-fed frequently. To facilitate feeding and avoid injury to the snake, the dead mice that were used as food were skinned back to the level of the hind legs. The skin was left attached to the body but turned inside out so that friction was re- duced as the carcass was pushed down the gullet. Digestion was hastened in the early stages by removal of the skin and hair. How- ever under natural conditions food is often digested at a tempera- ture sHghtly higher and the action is correspondingly more rapid. In the period from December 4 to February 4 one snake was fed five times and the other six times. For from three to five days after feeding the mouse stiU could be palpated in the stomach. Defeca- tion occurred eight times in each snake. The interval between in- gestion and first evacuation averaged 11.4 days (six to 18). There was a tendency for evacuation of the residue from one meal to occur soon after ingestion of a new meal. In four instances a meal was represented by tsvo separate evacuations, the second following two, three, five and nine days after the first. The interval required for complete digestion and evacuation of a meal varied from eight days to 19 days and averaged 13 days. Separate meals averaged 20 per cent of the snake's body weight. The smaller snake ate the equivalent of 120 per cent of its body weight, in six meals; the larger one ate the equivalent of 80 per cent of its body weight in five meals. Evidence from these feeding experiments indicates that on the average food is retained in the stomach for approximately one-third AUTECOLOGY OF THE COPPERHEAD 217 of the time from ingestion to final evacuation of remaining residues. Were it not for bias introduced in obtaining the snakes by trapping, my sample of 336 captures in summers of 1958 and 1959, yielding 157 snakes with food remains in their intestines, should have yielded nearly 80 snakes with food in their stomachs, but actually there were only 22. PubHshed reports based on samples of copperheads, obtained by methods other than trapping, all show higher ratios of individuals with food in their digestive tracts than does my own sample. The sample used by Uhler, Cottam and Clarke (1939:610) was collected by crews of workmen in Virginia engaged in such activities as con- struction of roads and trails. Presumably discovery of snakes by these crews was not dependent on the snakes* activity, but all or nearly all those within the limited areas being cleared or excavated were routed from their shelters, and immediately killed and pre- served. Because of these circ*mstances the collection should give a true ratio of the fed and empty snakes but no distinction was made as to the part of the digestive tract where the prey remains were found. In collections obtained by Surface (1906:189) in Penn- sylvania, Clark (1949:258) in Louisiana, and Hamilton and Pollack (1955:2) in Georgia, the techniques of collecting were not de- scribed. Conceivably some collecting techniques would yield sam- ples biased in favor of the snakes that were recently fed. Although recently fed copperheads are secretive, they are also sluggish, and once found would be less hkely to escape than would unfed in- dividuals. A feeding cycle averaging approximately 18 days is indicated, with food in the stomach for the first four days, residues in the intestine from the fifth through the thirteenth day, and the digestive tract empty from the 14th through the 18th day. At this rate of feeding, approximately seven meals would be consumed from May 1 to Sep-

Table 20. Ratios of Copperheads Containing Food Remains, in Various Samples

Sample 218 University of Kansas Publs., Mus. Nat. Hist.

tember 1. In the remaining weeks of activity during autumn I suspect that not more than one meal would be consumed, since the stomachs of most copperheads caught at the rock ledges at that season were empty. In 297 of the copperheads captured that had food residues in their digestive tracts, weights (exclusive of the food) ranged from 495 grams to eight grams and averaged 118. Weight of prey was calculated to average approximately 22 grams, 18.5 per cent of snake-weight, but individual prey items ranged from less than one per cent to more than 50 per cent of the snake's weight. Taking eight meals in its entire season of activity, an average copperhead

would consume 148 grams (approximately % lb. ), amounting to 13^ times its own body weight. If the food of such an average indi- vidual happened to coincide in its composition with that of the population as a whole, it might consist of two voles, two mice, two cicadas and one each of short-tailed shrew, little short-tailed shrew, skink, ring-necked snake, frog and young cotton rat. Actually such a distribution would involve several more meals than the snake probably would take. An individual of average size or above would concentrate on the larger kinds of prey and hence would require fewer separate meals. At a population density of five copperheads per acre—a conservative figure for the Reservation and nearby areas of similar habitat—it is estimated that the copperheads on a square- mile area would annually consume prey totalHng more than 1,000 pounds. The effect of this predation on prey populations is difficult to judge. The prairie vole being the favorite prey species, bears the brunt of the copperhead's effect. The annual toll of approximately eight adult voles per acre (or a correspondingly larger number of immature animals) would seem to be a substantial factor in the vole's ecology, but not a decisive one. Where the vole's population density attains a level of 50 per acre, or more, as it often does under favorable conditions, the copperhead's effect would be minor. But where the vole occurs in lower populations of ten to 20 per acre, the copperhead's levy would be felt more, even if the snake were partly diverted to alternate prey species. Other favored prey species in- cluding the several kinds of mice, the two kinds of shrews, the five- lined skink, ring-necked snake and the cicada, are all so numerous that the numbers taken annually by the copperhead would amount to only a small part of the annual increase. Rather than controlling their population trends, the copperhead merely exerts some stabiliz- ing influence. AUTECOLOGY OF THE COPPERHEAD 219

DEFENSE, ESCAPE, AND MORTALITY FACTORS Defense and Escape

The concealing pattern constitutes the first hne of defense against natural enemies; in time of danger a copperhead tends to lie quietly, resorting to defensive behavior only when actually attacked. One lying near a deep crevice may suddenly lunge for shelter when a person approaches, and within two or three seconds may slide down out of sight. In response to a less abrupt or less immediate dis- turbance, the snake begins to move hesitandy. It then crawls slowly (rate of perhaps two feet per minute) but directly to the nearest shelter. A copperhead that is in a resting coil and is not beside shelter reacts to the approach of a person by a sudden rota- tion of its head, which is turned to face the danger and co*cked up- ward at an angle of approximately 45 ". Without making any other movement than this inconspicuous flick of the head, the snake pre- pares to strike. A copperhead found in an exposed situation such as a road, sometimes "freezes," or sometimes makes clumsy but vig- orous attempts to gain shelter. With its head raised several inches above ground, it progresses by lunging with the anterior part of its body thrown into a loop, in a sidewinder-like type of locomotion. If closely approached, it may strike, lashing out wildly in the direc- tion of its tormentor even though he may be far out of reach. During this performance the snake does not hesitate to move directly toward its enemy, and the lunging movements with which it progresses are not distinct from the strokes with which it threatens or actually attacks. Vibrating of the tail is a response to severe alarm or disturbance; it was found to be characteristic of copperheads that are cornered, or those that have just been handled. The movement is a spasmodic twitching, resembling that of a typical colubrid snake, and much different from the rapid vibration of a rattlesnake's thick and mus- cular tail. A pattering or rattling or whirring sound is produced by the vibrating tail, depending on the type of material with which the tail comes in contact. A copperhead that vibrates its tail is thoroughly aroused and ready to strike. A copperhead that is held down with a stick may not resist or move at all, especially if it has been coiled inactive. If restrained on the posterior part of the body, or the tail, it may merely try to pull away but if the restraint is farther forward the snake may thrash with violent lateral movements, and with jaws widely open, turn its head about in an attempt to bite. It may bite the stick which is 220 University" of Kansas Publs., Mus. Nat. Hist.

used to hold it, or may bite its own body, if in the course of its struggles a coil comes within reach of the gaping jaws. Upon biting itself, the snake releases its grip almost immediately, but is not de- terred from repeating the bite as often as its body comes within reach. If grasped by the neck the snake throws its body in a cir- cular loop which is drawn up to a tight kink just behind the point where it is held, at the same time continuing to thrash and squirm vigorously. On one occasion an adult male grasped by the neck with metal tongs, thrashed and twisted so vigorously that vertebrae were dislocated. At intervals a struggling copperhead that is being held emits jets of musk in a fine spray, from the tail glands. Ordi- narily the musk is not emitted until the snake is grasped, or other- wise restrained. Then it is released in a jet of fine droplets like the spray from an atomizer. Several such jets may be released, from both sides, in the course of a few minutes while the snake is being handled. Besides the secretion actually sprayed, more oozes from

Fig. 24. Ventral view of tail-base of adult female copperhead from which skin and muscle layers have been dis- sected to expose musk glands. On right is posterior end of cloacal pouch, with ventral wall removed to show opening of duct from each gland.

the glands and accumulates at the margins of the anus. It seems that the sole function of the anal glands is defensive, and that their secretion does not serve a social function. Mashn (1950:460) de- scribed the females of A. halijs as having such enlarged postanal glands that their tail-bases were swollen resembling those of the males. In the copperhead, as in other snakes, it is the odor of the skin, not that of the anal glands, that is a stimulus in courtship and serves in the trailing of one individual by another. The musk is of a creamy appearance and consistency. The odor is distinctly dis- agreeable in high concentrations, but (to me at least) it is far less offensive than the scents of Thamnophis, Natrix, Elaphe, Coluber and other common snakes. The musk of copperhead has often been compared to the odor of cucumbers. When a copperhead is handled AUTECOLOGY OF THE COPPERHEAD 221

individual the vigor of the struggle varies greatly according to the and the circ*mstances; also adult males make a much more spirited defense than adult females. Gravid females are especially docile; when handled they thrash but little, or none at all, and they seldom Idnk the body in the manner described. six The bite is typically delivered as a short jab, often less than inches for an adult two to three feet long. Even at such close range the strike may be wildly inaccurate if the snake is highly excited or if the target is moving rapidly. The precision required in timing includes aiming, opening of the mouth, erecting of the fangs and hit the ejection of venom, and is such that the snake may often target with some part of its head without delivering an effective bite. Ordinarily withdrawal from the stroke is instantaneous, but a cop- perhead that is restrained and enraged may retain its grip for several seconds, straining to embed its fangs more deeply and inject more venom. Occasionally a copperhead that is cornered and is unable to defend itself by striking will react by coiHng with its head con- cealed and protected by part of the body. The hissing of chickadees and titmice disturbed while incubating, has been cited as warning behavior mimicking the hissing of a copperhead (Sibley, 1955:128). However, hissing is not a part of the normal defensive behavior in the copperhead. In fact I have never known a copperhead to hiss audibly under any circ*mstances, and the sematic behavior which is so characteristic of rattlesnakes is almost lacking in the copper- head. Natural Enemies and Predation

There are few published records of predation on copperheads. The United States Fish and Wildlife Service's food habits files, con- taining the records of analyses of contents from thousands of stom- achs of common predatory mammals and birds, included no record of a copperhead having been eaten by any animal. In general, potential predators seem to have an instinctive or learned aversion for these venomous snakes. First-year young are vulnerable to various predators that would not undertake attacking an adult. The mole (Scalopus aquaticus) is such a predator. I have often found tunnels of moles beneath flat rocks in the situations where small reptiles are likely to hide, and on the basis of circ*mstantial evidence, I concluded (Fitch, 1954:133) that, on the Reservation, at least, the mole is a frequent predator on nests of the five-lined skink. In the early summer of 1958 two moles were kept in a large terrarium for more than a 222 University of Kansas Publs,, Mus. Nat. Hist. month. The bottom of the container was kept covered to a depth of several inches v^dth damp soil. Meat scraps and insects that w^ere offered, were located by scent. The mole tunnelled upward beneath the morsel, and pulled it underground without exposing itself. Several times small reptiles were experimentally introduced into the terrarium. Usually after seeking shelter beneath a flat rock, they were attacked from below by a mole which dragged them underground and ate them. Twice, first-year copperheads were introduced and each time they were promptly attacked, dragged underground and eaten. It seemed that the snakes were unable to turn and strike in the narrow confines of the tunnel before being fatally bitten and immobilized by the mole. When a second- year copperhead was introduced, a mole soon was attracted by the odor or the sound and tunnelled up to it partly emerging. With- out touching the snake the mole appeared to sense its larger size and withdrew in panic. Neither mole attacked this snake, although it was left for several days in the terrarium. The opossum (Didelphis marsupialis) takes almost any animal food that is available, and occasionally preys on copperheads, and other such noxious animals that might be avoided by more skilled but warier predators. The opossum's thick, woolly pelage would provide partial protection, especially from small copperheads. However the clumsiness and lack of caution of the opossum might often cause it to be bitten. Such habits perhaps contribute to the short life expectancy of the opossum. Of 79 opossum scats ex- amined on the Reservation in the late summer and fall of 1951, one contained scales of a small copperhead (Fitch and Sandidge, 1953:323), and on at least one other occasion copperhead scales have been noticed in scats that were seen in the field but were not collected for detailed analysis. In November, 1957, partly eaten remains of an adult female copperhead were found, with hairs of an opossum adhering to them, at a rock ledge where many snakes hibernated. Schlenker (1942:60) described the behavior of two pet house cats that had often caught garter snakes, milk snakes, and other harmless kinds, when they were confronted with a four-foot cop- perhead, freshly killed and still twitching. One cat, when set on the ground nearby, leaped wildly and yowled in fright, but later re- gained his courage sufficiently to approach the snake several times and cuff it with a front paw. After each approach he would bound backward several feet to safety. The second cat became tense AUTECOLOGY OF THE COPPEBHEAD 223 and nervous as he approached the snake, and stopped short to examine it while still out of range, with his body extended forward to the maximum, ready to bound back at any sign of danger. Possibly the cats' wariness in this instance was occasioned by the relatively large size of this copperhead rather than by any recogni- tion of its venomous quahties, as the article seemed to imply. The conmion king snake {Lampropeltis getulus) is notorious for ophiphagy and is known to eat pit vipers as well as harmless snakes. Clark (1949:252) reported finding 17 copperheads, along with other prey, in the stomach contents of 301 king snakes (L. g. holbrooki) from northwestern Louisiana. Besides the copperheads, there were 27 other venomous snakes represented in the food sample. Minton (1951:322) mentioned finding a black king snake (L. g. niger) in a log pile in Floyd County, Indiana, which had a copperhead in its stomach. Six other copperheads were found in the same log pile.

Dr. Joseph P. Kennedy (in litt.) told of finding a juvenal copper- killed head in the stomach of a 47-inch king snake ( L. g. holbrooki ) on the road near Moss Hill, Liberty County, Texas, on May 1, 1958. The prey had been swallowed head first. Rattlesnakes have a characteristic and specific defensive behavior with which they respond to the presence of king snakes (Klauber, 1927:13; Cowles, 1938:13; Bogert, 1941:331). Olfactoiy cues are most important in detection of the ophiphagous enemy. The de- fense consists of raising the body in a vertical loop which is used to push or strike tlie enemy, while the rattlesnake presses its head against the ground. On July 6, 1959, a speckled king snake was introduced successively into several different containers where cop- perheads were kept, and the reactions of the copperheads were noted. The characteristic response described by Klauber, Cowles and Bogert in rattlesnakes was lacking. Nevertheless, the copper- heads showed some evidence of recognizing the king snake as an enemy. When the king snake was placed in a container with five young copperheads, the latter at once became alert and wary. They tended to avoid the king snake, and to strike at it whenever it moved within range. One of the young struck another, presumably excited by the sight and/or scent of the king snake. Within a few minutes all the young copperheads were gathered in one corner of the con- tainer, facing the king snake and ready to strike. Several times when the copperheads struck at the king snake, the latter jerked back so rapidly that it avoided the stroke, and none of the bites seemed effective. From time to time the king snake tilted its head 224 University oi^ Kansas Publs., Mus. Nat. Hist.

toward a nearby moving copperhead, as if about to seize it, and sometimes tested the other snake with its tongue, but did not actu- ally attack. Disturbed by recent handling, it vibrated its tail fre- quently and kept to one corner of the container, obviously on the defensive. The king snake involved in these observations was the only individual found on the Reservation in eleven years of field work. Because of its rarity on this area it cannot be considered an important natural enemy of the copperhead locally. Minton (1944:462-463) recorded that a milk snake (Lampropeltis doliata) overpowered and ate a young copperhead. He also re- corded that a captive prairie king snake (L. calligaster) ate a small dead copperhead that was oflFered. Another captive prairie king snake attacked a larger copperhead, but released it and backed away after it had been bitten on the neck. Keegan (1944:59) described the behavior of a captive indigo snake {Drymarchon corais couperi) which, when an adult copper-

head was introduced into its container, ". . . seized the prey by the head, and in fact seemed to avoid any other portion of the body. Before swallowing tlie copperhead, the indigo snake lacerated its head by 'chewing' with lateral movements of the jaws." Cope (1900:1138) mentioned an instance of a blacksnake {Colu- ber constrictor constrictor) caught near New Haven, Connecticut, which disgorged a well-grown copperhead. Branson (1904:412) recorded an instance of a racer (C c. flaviventris) disgorging a cop- perhead. Hurter (1911:171) wrote "On May 1, 1898, I caught a Blue-Racer just swallowing a copper-head about two feet long

. . . had about half disappeared." Vernon Mann told of finding, near La Cygne, a yellow-bellied racer that was eating a copperhead nearly as big around as itself. This is possibly the same incident referred to as in of by Gloyd ( 1932:403 ) occurring April 1929. The "racer had made its capture and was chewing the head and neck of its victim, which was thrashing about in violent efforts to free itself. He [Mann] observed the entire swallowing process, which lasted more than an hour." Mr. Delmer Ferguson of La Cygne also recalled an instance of a large racer found eating a small copper- head. adult red-sided sirtalis A large female garter snake ( Thamnophis parietalis) trapped on July 1, 1958, produced a scat in which the only recognizable materials were scales of a small copperhead. This garter snake only occasionally preys on other snakes and cer- tainly is not an important natural enemy of the copperhead. AUTECOLOGY OF THE COPPERHEAD 225

been found Pope (1937:99) wrote that no poisonous snakes had in the stomach contents of 3,693 hawks of kinds known to prey upon harmless snakes. However, Klauber (1956:1050-1052) has cited many instances of red-tailed hawks preying on rattiesnakes of several different species. Food habits of the red-tailed hawk were investigated by collect- the nests. Such col- ing pellets, chiefly from the ground beneath lections were made on the Reservation in 1952, 1955, 1958 and 1959. In the collections from each nest copperhead remains were represented, and the copperhead was the fourth most frequent kind of prey for the combined sample, with 40 occurrences in 224 sometimes were pellets or pellet fragments. Because the pellets trampled or broken in the nest before falling to tlie ground, or were of broken by striking branches in tlie fall, the actual number pellets was probably less than the number actually found. Also, the nest- each more lings, usually two or tliree in a nest, may have made than one meal from the same animal. The number of copperheads actually eaten was hence probably somewhat less than the number of recorded occurrences in pellets. Nevertheless it seems probable that each red-tailed hawk destroys several or many copperheads in the course of a summer, if these snakes are common on its terri-

tory. Although the hawk is diurnal and the copperhead is largely nocturnal, their periods of activity overlap after sunset and before sunrise; at these times of day the hawk is unusually active in search of prey. Just how the hawk secures a copperhead with impunity is unknown. Although the stroke of a pit viper is notable for its speed, the reflexes of a hawk are probably even faster. Aside from superior speed, a factor which favors the hawk is its relatively keen eyesight, and the near-sightedness of the snake. Swooping down upon the snake unperceived, the hawk may strike it a fatal blow or may secure a hold on its head or neck, rendering it helpless. The horned owl is the most abundant large raptor of the Reserva- tion and it might be expected to be an important predator on the copperhead, since the owl and snake are similar in time of activity and in habitat. Several hundred pellets of the homed owl from the area of the Reservation contained no remains of the copperhead. However most of the pellets were collected in the colder half of the year, when the snakes were not active. A homed owl reared in captivity had no instinctive aversion for copperheads or other snakes. On several occasions it was seen to fly down into the out- door enclosure (open on top) where several were kept, and once 226 University of Kansas Publs., Mus. Nat, Hist. lit on the ground and walked within a few inches of two of the snakes. The owl had its attention focused on a cotton rat that had been placed in the enclosure to feed the snakes and gave no indi- cation of noticing the snakes. The copperheads did not respond strongly to the presence of the owl either, but merely drew back their heads in readiness to strike. On a subsequent occasion when the owl was unusually hungry, it flew down into the enclosure and attacked a copperhead. The actual attack was not seen. The snake was carried for a short distance, and struck the owl one or more times high on the medial surface of the thigh. The owl uttered scolding sounds and dropped the snake. Almost immediately the owl showed signs of distress, and ceasing its usual activities perched hours quietly shifting its position from time to time. After several an oozing extravasation was noted, and a small pool of blood had collected where the owl was perched. Approximately eight hours after being bitten the owl suddenly collapsed and died.

EflFects of Climatic Extremes

Catastrophic effect of extreme weather conditions on a local pop- ulation of copperheads was illustrated by my observations in June, 1957, at Independence Creek, Terrell County, Texas. The herpeto- fauna and habitats of this area in the Stockton Plateau have been described by Milstead, Mecham and McChntock (1950:557). The University of Texas field party which collected in the area in June and July, 1949, found copperheads extremely abundant and ob- tained 89 during their three-weeks stay. More than twenty were taken in a single night. Nearly all the copperheads found by this field party were found in hve-oak groves in the immediate vicinity of Independence Creek (Little Canyon Creek) but a few of those taken overlapped into adjacent habitats. Encouraged by the account of copperheads in the publication by Milstead et at, and by conversation with Dr. Milstead, I had visited the area hoping to collect a large series of copperheads, but found them to be rare in June, 1957. I talked with many ranchers and other residents of the area. All were familiar with copperheads and agreed that in former years the snakes had been abundant, but that since 1954 they had been rare as the result of a devastating flood. On the night of June 29, 1954, as the aftermath of a hurricane that moved northwest from the Gulf of Mexico, a storm crossed the Stockton Plateau with torrential rain alleged to have totalled more than 20 inches by unoflScial observers at several places. In the re- AUTECOLOGY OF THE COPPERHEAD 227 suiting flash flood Independence Creek overflowed its banks and ex- tended across the valley, about a quarter of a mile wide. Water level rose as much as 20 feet. Many ranch buildings were swept away and several persons were drowned. Locally the hve-oak groves were mostly situated on low-lying ground adjacent to the creek, almost entirely within the flooded zone. Many of the trees were Plate 2 or were torn undermined by erosion and uprooted ( 16, fig ), out by the force of the current and transported debris. In situations exposed to the full force of the current almost every tree was up- rooted, including many of as much as two feet in trunk diameter. The uprooted trees had been swept downstream for varying dis- tances, and the interlacing tangles of roots on the upstream end of each such tree had collected huge piles of drift. Mr. Charles Chand- ler, a local rancher and long-time resident, told me that the live-oak groves had been reduced to less than one-third of their former ex- tent by the flood, and his estimate seemed reasonable on the basis of the evidence remaining in 1957. Presumably most of the copperheads living in the area in 1954 were swept away and drowned in the flood. Some may have sur- vived in the more protected areas by climbing into the live-oaks and keeping above the rising water level. A few may have been near the oak groves but in upland situations that were not flooded. Even though such survivors constituted potential breeding stock to re- populate the remaining oak groves, their habitat was mostly de- stroyed. The accumulated leaf litter, logs, and dead branches and even the soil had been swept away, leaving bare gravel. In 1957 there remained at least a dozen oak groves ranging up to a size of more than two acres, along several miles of the lower reaches of Independence Creek. In parts of these areas leaf litter had again accumulated, and habitat conditions appeared to be favorable for the snakes. The root tangles and great piles of debris where there are uprooted trees within the remaining groves or ad- jacent to them, provide abundant shelter. Of the two copperheads found by me on the night of June 27, 1957, one was climbing tsvo feet high on a pile of driftwood, the other was crawling over leaf litter beneath live-oaks. However, with the advent of improved roads into the area intensive use of the oak groves by humans has become a major factor. In summer, fishermen visit the creek in large numbers. Because shade is at a premium, they concentrate their activities in the groves. The remaining copperheads constitute some hazard to campers. Because of the limited extent of their 228 University of Kansas Publs., Mus. Nat. Hist.

remaining habitat and its intensive use by humans, it may be an- ticipated that the snakes will never again regain their former abundance, but will become even scarcer and eventually perhaps will be locally exterminated.

Parasites, Diseases and Injuries

The copperhead has various ectoparasites. Hyland (1950:494) first reported the common chigger {Tromhicula alfreddugesi) from copperheads; of six specimens collected in the Duke University Forest, four had chiggers, totalling 260. Most of the copperheads collected on the Reservation in the early years of my study were examined for ectoparasites. Loomis (1956) has reported upon the chiggers. Pie lists the copperhead as one of 16 important host species (including mammals, snakes, lizards and birds) of the com- mon chigger, locally. This chigger has been recorded from dozens of species of reptiles, mammals and birds, and in fact occurs on most of the terrestrial kinds that are abundant and share its habitat.

Because of this lack of specificity tlie chigger will attach even to humans. Unhke the natural hosts, man does not provide a suitable food source and the attached chigger dies without completing its development, but causes swelHng and irritation. Of 107 copper- heads examined, 80 carried common chiggers totalling 8,579; 5,898 in July, 1,340 in August, 1,204 in September and 137 in October. Even heavier infestations might have been found in late May or June, but Loomib obtained no samples from those months. A single copperhead may carry several hundred chiggers at one time. The chiggers burrow into the skin betsveen the scales, and often congre- gate in clusters. The larvae are usually on moist soil in sheltered situations, and they dirive in warm, humid weather. Among the 22 kinds of chiggers occurring on the Reservation, only three others, Trombicula llpovskyana, T. sylvilagi, and T. trisetica were found on copperheads. T. lipovskyana occurs chiefly in low, moist meadows having an abundant ground cover of grasses and weeds. Five copperheads from the Reservation had a total of ten of these chiggers, which also have been found on many species of birds, small mammals, snakes, hzards and even frogs and toads. T. tri- setica has been found chiefly in climax forest of oak-hickory, and has been taken mostly from hosts that are arboreal or semi-arboreal, the gray squirrel, wood rat, white-footed mouse, black rat snake and skinks {Eumeces laticeps and E. fasclattis). A single specimen was recovered from a copperhead. A single specimen of T. sylvilagi was AUTECOLOGY OF THE COPPERHEAD 229 recovered from a copperhead. Larvae of this species usually occur in well shaded places, often about decaying logs, and small mam- mals are the favorite hosts. Copperheads that were infested with chiggers showed no ill ef- fects and their infestations were relatively Hght as compared with those of some other local species, notably the common garter snake, yellow-beUied racer, and black rat snake. However, there is some vectors of diseases that possibility that occasionally chiggers are the afflict snakes. Various endoparasites inhabit the digestive tract, but insofar as known, none of these is pathogenic. Cloacal smears that were examined microscopically almost always contained large numbers of highly active ciliate protozoans that appeared to be mostly of one species but were not identified. Less frequently microscopic nematodes were found in cloacal smears, but these also were not identified. Crow (1913:123) reported a new species of fluke, Reni- material was fer kansensis, from the mouth of a copperhead. The from Kansas, but no definite locality was mentioned. Flukes of this group require intermediate hosts—a water snail which ingests the eggs and from which free-swimming larvae emerge, and a frog in which a later stage occurs. For the parasite to complete its de- velopment, the frog must be eaten by a snake. In the course of my study I examined mouths of several hundred copperheads without the finding any flukes in them, altliough flukes were abundant in yellow-bellied racers and garter snakes of the Reservation, espe- Lou Flukes of Tham- cially in early summer ( Peggy Stewart, "Lung nophis and Coluber in Kansas," an unpublished dissertation on de- posit in the library of the University of Kansas). In this connec- tion it is significant tliat the copperheads of the Reservation rarely so prey upon frogs, while the racers and garter snakes do frequently. Harwood (1933:66) examined 14 copperheads from the vicinity of Houston, Texas, and found these flukes {Renifer kansensis) in in rattlesnake Sistriirus two. The species was also found a pigmy ( miliarius) from the vicinity of Houston. Hughes, Baker and Daw- son (1941:39) hsted this same species (as Neorenifer kansensis) as a parasite of the copperhead, and also listed Renifer ancistrodon-

tis, which Harwood had considered a synonym of R. kansensis. Harwood found tlie diaphanocephalid nematode, Kalicephalus ag- kistrodontis, in stomachs of all of the 14 copperheads, also in the coral snake (Micrurus fulvius), hog-nosed snake {Heterodon pla- tyrhinos), bull snake {Pituophis catenifer), king snake (Lampropel-

10—4428 230 University of Kansas Publs., Mus. Nat. Hist.

tis getulus), water snakes {Natrix sipedon and N. rhombifera) and garter snake (Thamnophis proximus). He found the spirurid nematode, Physaloptera squamatae, in the stomach of one copper-

in the of skink . head, and also stomach a brown ( Lygosoma laterale ) Evidence of disease was noted in copperheads on the Reserva- tion from time to time, but especially in 1951. In the summer of 1951 precipitation was unusually high and temperature was low. Many of the copperheads trapped in autumn had necrotic patches on the ventrals and occasional blisterlike swellings on the dorsal scales. Such individuals often were emaciated, and snakes of other species were similarly affected. There may have been heavy mortal-

ity, as in 1952 and 1953, v^dth more traps and greater effort I was able to trap fewer copperheads per season than in 1949, 1950 and 1951. Otherwise most of the copperheads trapped appeared to be in good condition but occasional individuals showed evidence of in- jury or disease. Several adult males each had one hemipenis everted, dried and shrivelled. Injury to the tail involving the re- tractor muscles may have caused eversion in these instances, in which the organ was probably lost eventually. Copperheads rarely had scars of the type common in constricting snakes, that probably are bites inflicted by the struggling prey. COMPOSITION OF THE POPULATION

The true composition of the population, according to age groups and sex, is obscured because of differences in habits, which, in almost any sample cause certain groups to be represented by too few individuals or too many, in proportion to their true numbers in the natural population. The trends of the figures obtained de- pend upon the time and place of sampling. There is abundant evidence that in summer the males, especially the old adults, dis- perse far from the ledges where they are concentrated in autumn, and that the adult females, especially those that are gravid, tend to remain near the ledges. From year to year my samples varied accordingly, with bias toward one or the other group depending on the extent and location of trap lines. Some snakes living far from the ledges where they hibeiTiate, especially adult males, arrive relatively late in the autumn, and in September the population at the ledge is still biased in favor of the adult females. Compara- tively few data were obtained in spring, and dispersal begins promptly after emergence from hibernation. AUTECOLOGY OF THE COPPERHEAD 231

By October 1 the gravid females have nearly all produced their litters, and most copperheads have travelled from their summer ranges back to the rock ledges. Subsequently in all of October and usually the early part of November, the population is concentrated

HUJ Ul

X H o z 232 University of Kansas Publs., Mus. Nat. Hist.

it exclusively. In the eleven seasons of field work a total of 637 copperheads were recorded in October and November. Actually the number of individual snakes represented was somewhat fewer because the same individual might be recorded in two or more years, each time in a different age group. Actual age was known for relatively few of the snakes, namely for those that had been marked early in life. However, each was assigned to a probable age group. Since birth occurs in early autumn, this October-November sample consisted of discrete annual age-groups with no intermediates. A male of 620 mm. in snout- vent length, for example, was assigned to the four-year-old class on the basis of typical growth rates, although there was some chance that he might be an oversized three-year-old or an imder-

Table 21. Numbers of Copperheads of Various Size Groitps, Represent- ing Annual Age Groups, in an Autvmn Sample of 637 Records

Age AUTECOLOGY OF THE COPPERHEAD 233

the numbers of range arbitrarily assigned to each age group, and snakes of each sex in the total sample of 637 are shown in Table 21. differential Even these figures are biased in some respects by habits of the snakes, and do not represent the true composition of the population. The juvenal snakes are surely represented in less than their actual numbers. Newborn, and also one-year-olds, must outnumber two-year-olds but more of the last were obtained. Traps in which most of the snakes were obtained may somehow be se- lective, catching a higher proportion of the adults present than of the young. Small copperheads may avoid traps more easily by squeezing behind or beneath them, because they are able to pass through smaller openings. Or the small snakes may merely travel less. Their comparatively small bulk would permit them to utiHze relatively small fissures and interstices in the rock outcrops, whereas large snakes would less readily find hibernacula of sufficient size to accommodate them and would require longer search. Regardless of the availability of shelter, distance traveled may be in a general way, proportional to the size of the snake; for a foot-long individual the time and effort required to travel one foot might be approximately the equivalent of a three-foot movement in a snake three feet long. Of the 637 copperheads in the autumn sample, 288 were allocated as three-year-olds or as snakes of older groups, all sexually mature and past the period of most rapid growth. Of the 288 mature snakes, 106 were females. Presumably at least half of these sexually mature females had produced litters of young in the period of ten weeks preceding their captures; wdth the average litter 5.25 young, the 53 breeding females would have produced a total of 278 young. If females all breed for the first time in their third years and breed in alternate years thereafter, the breeding population would amount to more than half of the adults because three-year-olds are more abundant than any older age group. Of the 106 adult females actually recorded, 63 were tentatively classed as odd-year indi- viduals (3, 5, 7, 9, 11 or 13 years old), and if all these produced average litters of 5.25 young, the annual brood would amount to 330. However, some three-year-olds, those that lag in their develop- ment and remain undersized, fail to mature sexually and fail to ovulate, and it seems safer to assume that approximately half the adult females breed annually in this locality. Only 126 recently born young were actually captured in the sample, indicating that 234 University of Kansas Publs., Mus. Nat. Hist. more than half those that should have been caught were missing from the sample. The annual brood of 278 young calculated to have been produced by the 288 adult snakes might be expected to sustain losses in the subsequent three years suflBcient to reduce it to 85, the number of three-year-olds. If this reduction occurred at a constant rate, an annual loss of 33 per cent would be indicated, with reduction to 186 one-year-olds and 125 two-year-olds. The number of two-year- olds actually obtained was 141, suggesting that the two-year-olds are fully as well represented in the trap sample as are the older snakes. At a somewhat slower rate of loss in the adults, 29 per cent an- nually, four-, five-, six- and seven-year-olds would be represented by 69, 49, 35 and 25 individuals respectively. These figures correspond well with the numbers —85 remarkably actually caught ( three-year- olds), 63, 46, 34 and 26. If the same rate of loss were continued in subsequent years, the numbers would be reduced to one or two in the fourteen-year-olds, and obviously snakes of greater age would be rare. The oldest known copperhead detected in my study, for which a fairly definite age could be established, was a fourteen- year-old and several twelve- and thirteen-year-olds were also re- corded. Thus, an assumed annual loss of approximately 33 per cent in young up to an age of three years, and of 29 per cent subsequently fits best with the available data, although it might be expected that the rate of loss would change continually at different stages in the life cycle. For example, young of the year would appear to be much more vulnerable to predators than large adults. Of the 126 young of the year in my sample, 12 were not sexed and the remaining 114 included 84 males and 30 females. The 2.8 to 1 sex ratio in this sample approximates the ratio obtained from young born in captivity. In the older age groups, combined, females comprise approximately 39 per cent of the sample. Hence it seems that the heavy preponderance of males in the newborn snakes is in part compensated for by greater mortality in the males, espe- cially in theii- first year of hfe. It is not evident why the young males should be subject to heavier mortality than the young females. Recaptured, marked copperheads of known age are available in fairly substantial numbers to permit tracing of growth up to an age of seven years. Beyond this age the records are relatively few and their evidence is somewhat conflicting. For eight males known to be seven years old, length ranged from 709 to 791 mm. and aver- aged 744. Of 13 males known to be more than seven years old, AUTECX)LOGY OF THE COPPERHEAD 235

7*1

100

UJ

131

cf 1121 lltl IIOl I 91 81 7H

150 100 50 50 NUMBER OF INDIVIDUALS

Fig. 26. "Age-pyramids" for the copperhead on the University of Kansas Natural History Reservation and Rockefeller Experimental Tract. The upper figure shows distribution in the actual sample; the lower figure is hypothetical, showing the probable age- distribution in the natural population, but any sample collected is more or less biased because of diflFeren- tial habits in the sexes, and in young and adults. one had a length of 733, another 783, and the remaining eleven all exceeded 800. A length of approximately 775 mm. may therefore be established as the dividing line between those seven years old or less and those eight years old or more. Similarly, in females an upper limit of 650 mm. was established for seven-year-olds. Of 236 University of Kansas Publs., Mus. Nat. Hist. the 637 snakes in the fall sample (actually representing a popula- tion of 902, with the missing young), 34 were adults exceeding seven-year-old size, and these 34 were equally divided between the sexes, 17 males and 17 females. Assuming that in these large, old adults the same rate of loss continued (approximately 29 per cent annually) that had prevailed in each of the three preceding years, the original brood of 292 young would be reduced to a single sur- vivor in the fifteenth year. Perkins (1955:262), Conant and Hudson (1949:8) and Shaw (1959:337) have published many records of tlie longevity of captive snakes including copperheads. Individuals of two subspecies of copperheads have attained ages slightly exceeding eighteen years in zoos. Once adjusted to captivity, snakes in zoos have an excellent opportunity to live out their potential life span free to a large ex- tent from the usual hazards of predation, disease, parasitism and ex- tremes of weather which account for most of the mortality in natural populations. I suspect that in the vvdld, attainment of an age of 18 years is exceedingly rare. NUMBERS

For snakes in general, the published literature concerning popula- tion densities is meager, and impressions are liable to be erroneous. However reliable information concerning numbers is necessary for any appraisal of the species' economic or ecologic role. Because of the copperhead's secretive habits, no precise measurement of the population on any area was possible. It is certain that population densities differ greatly on neighboring areas, depending on their suitability as habitat, and also that numbers normally fluctuate somewhat from year to year on any area, although, of course, changes are much less rapid than in some other small vertebrates which have a higher reproductive potential. There are no definite published statements concerning population densities of the copperhead. Most authors have used rather vague terms, such as "common" or "scarce," but others have mentioned the number encountered in a season or in a single day. Excluding ag- gregations found at hibernation "dens," the greatest concentration has been reported by Guidry (1953:55) who captured 35 in a small area within a few hours, in southeastern Texas. Milstead, Mecham and McClintock (1950:557) found a comparable concentration near the Pecos River in western Texas and noted that because of their abundance the snakes constituted a hazard to campers and Hvestock. On a small island off the west coast of Manchuria, Koba (1938:247) AUTECOLOGY OF THE COPPERHEAD 237 found the Palearctic pit viper (Agkistrodon halys) to be extraordi- narily abundant, and he estimated that in the south part of the island the population density amounted to almost one snake per square meter. These snakes were feeding upon migratory birds. Doubtless the high concentration was made possible by peculiarities of the insular habitat, with an abundant food supply renewed from outside the ecosystem. Although the copperhead attains no such concentrations, this instance is of unusual interest in demonstrating that snakes may attain remarkably high densities under optimum habitat conditions. One clue to population density in the copperhead is the number actually collected on a sample area, at a time when the snakes are dispersed on their summer ranges. The area most intensively sam- pled by me was the small valley where the Reservation headquarters are located. Excluding surrounding wooded areas, cultivated fields to the west beyond the Reservation boundary, and a formerly cul- tivated field on the Reservation, this valley comprises a block of 25 acres, approximately 2,000 feet long, and 1,000 feet in greatest width near the middle, but narrowing at each end. In the summer of 1958, 66 copperheads were caught on this area. Although the area in question was completely surrounded by less favorable habi- tat, each copperhead caught on the 25 acres probably had a home range extending onto adjacent areas. Addition of a peripheral strip of the radius of a typical home range would greatly increase the sampled area and if it could be assumed that all snakes present were found, the population represented would be less than one per acre. Actually the 66 copperheads captured must represent only a small minority of those present on the area in 1958, because, up until the end of the summer, and in the following year, new snakes continued to be better represented than the old ones among those captured. Interpretation of the data bearing on population density must take into account the vagility of the snakes, and the diifferences in this regard between individuals of different ages, sexes, and stages of breeding cycle. Also it must take into account the rate of popu- lation turnover as indicated by known mortality and natality. Be- cause of the copperhead's secretive and elusive habits, a long time is required to collect a suflBcient number on any area for a census computation, and within such an interval there is sure to be some change in the population. Shifts in home range and movements within a range cannot be distinguished with certainty in the data now available. Until such distinction can be made, the rate of mixing of populations between a sampled area and adjacent areas 238 University of Kansas Pxjbls., Mus. Nat. Hist.

cannot be judged, and an unknown error is introduced into any cen- sus computation. Obviously, the shorter the time involved in sampling and the larger the area, the less important will be the error introduced by the mixing of populations. Also, in a census com- putation based upon the ratios of recaptured individuals to others, mortahty in the animals marked and their replacement by other unmarked individuals has to be taken into account. When mortality has occurred, an erroneously low ratio of marked individuals in the sample and an erroneously high population figure may result. However, in the copperheads the marking and handling entailed no appreciable mortahty, and it seems safe to assume that mortality rates were similar in the marked snakes and those that were un- marked, approaching 30 per cent annually in both groups. The young are all bom at approximately the same time of year and in their first year most are recognizable as a size group, hence are not included in the samples from which the ratios of marked in- dividuals to others are derived. The so-called "Lincoln Index" has been widely used in censusing of birds and mammals. Also, under a difiFerent name, it has been applied to populations of fish, and to a lesser extent has been used on populations of amphibians and reptiles. The technique of cen- susing involves two distinct periods of sampHng, which preferably should be short and close together. In the first period a substantial number of animals on the selected area are recognizably marked, and in the second period a sample is obtained that includes some of these same marked animals and demonstrates their relative abun- dance as compared with the remainder of the population. For example, if 100 individuals were marked in the first period, and in the second period another lot of 100 were obtained, including ten of the same individuals in the original lot, the ten to one ratio would indicate a total population of 1,000 in the area sampled. The as- sumptions are implicit, that: (1) sampling is random, covering the entire area uniformly and favoring neither the marked nor the new individuals, and (2) the population does not change within the periods of sampling nor between them. Actually these conditions are rarely satisfied, and census figures obtained by the method are usually more or less distorted. Whether the census yields fairly accurate information concerning the animal's abundance or creates highly erroneous impressions regarding it depends upon the quan- tity and quality of the data obtained and upon the judiciousness with which they are used. Various correction factors have been introduced into the formula of the Lincoln Index by different work- AUTECX)LOGY OF THE COPPERHEAD 239 ers, making it complex in some instances, but unfortunately the distorting factors usually cannot be measured readily. Sixteen different census computations, based on the Lincoln In- dex, have been made of the population of copperheads on parts of the Reservation. These figures vary over a wide range. To some extent they may reflect the differences in population density that occurred from time to time and from place to place. But it is be- lieved that the differences result chiefly from samples that are biased by various immeasurable sources of error, and are too small anyhow to yield highly accurate figures. Even though no one census figure can be considered accurate, the trends of the figures are con- sidered significant. Also, the trends provide some basis for judging the extent of error introduced by such variables as the time factor. Composite censuses based on several samples taken under similar conditions are deemed more reliable than a census from any one of the component sets of figures. The 1,532 copperheads obtained on the Reservation and adjacent areas in the course of my study were distributed over eleven seasons and represent several generations. Therefore the total indicates little concerning the numbers present at any one time. In 1958 and 1959, the last two seasons of field work when operations were most concentrated, a total of 616 was obtained. These were not all contemporaneous on the area, of course, as two successive annual broods of young were bom within the seventeen-month span of the two seasons' collecting. However, the young-of-the-year that were obtained comprised only a small minority of the total sample. The normal wandering of individuals within a seventeen-months period would result in some loss of the original population, with compen- satory gains from immigrants. However most individuals are be- Heved to retain small home ranges over periods of years, with regu- lar seasonal movements to and from hibernation shelters. Therefore, notwithstanding some replacement through mortality, reproduc- tion, and migration, a substantial majority of the 616 snakes must have been actually contemporaneous on the area. The area represented by the 616 copperheads caught in 1958 and 1959 cannot be definitely determined. Actually two populations, broadly overlapping, but not identical, were represented. Those caught along rock ledges in fall had gathered from various dis- tances and directions, some from within the areas trapped in sum- mer, and some from outside areas. Likewise, the snakes trapped in summer, in field or meadow habitat, had moved there from rock ledges at various distances and directions, some within the areas 240 University of Kansas Publs., Mus. Nat. Hist. of operation in the fall trapping, and some outside these areas. The area sampled was hence substantially larger than that over which the traps were actually dispersed, but was probably somewhat less tlian a square mile, A block of nearly 200 acres in the southern and eastern parts of the Reservation was not trapped. A minimum popu- lation density in the neighborhood of one copperhead per acre seems to be indicated by these data. The extent to which the actual population of the areas is repre- sented by the 616 snakes captured in tlie two seasons may best be judged by the ratio of new individuals to those previously captured in the final weeks of field work. In October and November, 1959, 97 individuals were caught, and only 18 of these were snakes caught previously in 1959, or in 1958. Even if young of the year are ex- cluded from the counts, the newly captured individuals exceed those four to one to 17 previously captured in a ratio of more than (73 ) , indicating a population of 2,222 copperheads on the area of the study. If it is assumed that by October, 1959, 30 per cent of the snakes captured and released in the preceding 15 months were al- ready eliminated through natural causes, the 17 individuals recap- tured would represent an original 243 individuals, and the Lincoln Index would indicate a total population of 1,725—a population density of 2.7 per acre, if it is assumed that the area represented is exactly a square mile. This figure is probably low since the figures apply primarily to adults and well-grown young. First-year and second-year young, which must be relatively abundant, are so poorly represented in all samples that it is estimated approximately 30 per cent of the population is overlooked. At the time of the annual maximum, in autumn, the figures obtained may represent a population of 2,450—3.6 per acre. An unbroken sequence of 11 consecutive seasons' records was obtained from trapping in autumn along the hilltop ledges. In general the stretches of ledge where traps were set, and the specific trap sites corresponded from year to year. However, operations were gradually expanded; more traps and larger traps were utilized. Also, stretches of ledge which were relatively unproductive in one year were often abandoned the following year. Such lack of cor- respondence between consecutive samples would tend to result that each snake returns in a too low ratio of recaptures ( assviming to its an census original hibernaculum ) and erroneously high figure. Further shortcomings are the year-long intervals between successive samples, and the concentration of traps along certain stretches AUTECOLOGY OF THE COPPERHEAD 241 of ledge, with extensive intervening unsampled areas. Obviously the samples are inadequately small, as in four different years no of recaptures from the previous season were made, while in each three other }'ears the census was based upon a single individual recaptured. Table 22 shows the samples obtained for each year, and the re- sulting census figures obtained by the Lincoln Index.

Table 22. Nxtmbers of Captures Each Avtitmn, and of Recaptures From THE Preceding Autumn, Made Along the Ledges Where the Copperheads Hibernate, Serving as a Basis for "Lincoln Index" Censuses of the Popu- lation

Yeak 242 University of Kansas Publs., Mus. Nat. Hist.

Table 23. Numbers of Captures in Autumn Along Ledges Where Cop- perheads Hibernate, and in Fields in the Following Summers, Serving AS A Basis for "Lincoln Indexes" of the Population

Year AXJTECOLOGY OF THE COPPERHEAD 243 year. In 1957, when both samples were small, no recaptures were made in fall, so census computations were possible only for 1958 and 1959, as shown in Table 24. Like the fall-to-summer census figures, the summer-to-fall figure needs to be adjusted by subtraction of perhaps 15 per cent, to com- pensate for normal loss and replacement of the marked snakes, and addition of 30 per cent to allow for unrepresented young. In this instance also, the population density indicated is approximately 13.6 per acre. The remaining censuses all are based on samples collected in summer in the valley and in several hilltop fields. These areas were sampled adequately only in 1957, 1958 and 1959; and in 1957 the samples were relatively small. Table 25 shows the two censuses derived from the consecutive annual samples, and the composite figure derived from them.

Table 25. Nxjmbers of Captures in Sxjmmers of Three Consecutive Years, in Fields, Serving as a Basis for a "Lincoln Index" Census 244 University of Kansas Publs., Mus. Nat. Hist. period. The census figures obtained for 1958 and 1959, and the composite from them are shown in Table 26 and apply to the same 25-acre area mentioned earlier, the valley on the west side of the Reservation, where the headquarters are located.

Table 26. NtrMBERS of Captures in Summer Samples of Three Different Years, in a 25-acre Valley. For the Purpose of Sampling, Each Sximmer IS Divided Into an Early Period, April-May-Jxwe, and a Late Period, July-August, the Numbers of Copperheads Caught in Both Early and Late Period Serving as a Basis for a "Lincoln Index" Census

Year AUTECOLOGY OF THE COPPERHEAD 245

between the figures are probably distorted by non-correspondence two samplings of each census giving an erroneously low return of marked snakes, and the lower figures are perhaps equally far off the mark because of overestimation of the several areas involved, or other sources of error. Aldiough no highly accurate census is fea- sible, the population density on the Reservation probably usually averages between five and seven per acre, in summer somewhat higher than this in the brushy fields, which are the snakes' preferred habitat, and somewhat less in woodlands. In autumn the popula- tion density is much higher tlian seven per acre in the hilltop edge areas where the snakes gather to hibernate. By a process of extrapolation, from the census data obtained from the Reservation, and from the relative abundance of copperheads there and elsewhere, as judged from the results of hunting them without the use of traps, I conclude that in Douglas County and adjoining counties a population of five per acre is fairly typical where favorable habitat exists, on rocl

Attitudes of the Public

Since the time of the early colonists the copperhead has been well known to Americans, and in the United States it is the one species of venomous snake most frequently encountered by the public. Wright (1950) and Wright and Wright (1957) have hsted the fol- lowing vernacular names applied locally to the copperhead: beech leaf snake, chunkhead, copper adder, copper-bell snake, copper belly snake, copperhead moccasin, copperhead viper, copper snake, copper viper, cottonmouth, deaf adder, deaf snake, death adder, dumb rattlesnake, dumb snake, harlequin snake, hazel-head, high- land moccasin, kupper schlange, lowland moccasin, moccasin, pilot snake, poplar leaf, rattlesnake mate, rattlesnake pilot, red adder, red eye, red oak snake, red snake, red viper, rusty moccasin, sand viper, thunder snake, upland moccasin, viper, white oak snake. Many of these names were originally listed by Rafinesque (1819:84) as used for the species in New York State. Some of the names listed above are based upon folklore prevalent in pioneer times, and perhaps adopted from earlier aboriginal ver- sions. A widespread superstition pertained to the "piloting" of the

11-4428 246 University of Kansas Publs., Mus. Nat. Hist.

rattlesnake by certain other snakes—most notably the copperhead and the "pilot" black snake (Elaphe ohsoleta). Klauber (1956: 1243) cited an early account by the Count de Crevecoeur in 1782 stating that the copperhead is called rattlesnake pilot because it comes out of hibernation a week earlier than the rattlesnake, and it in about. always precedes crawling Milling ( 1937:43) mentioned a belief, widely held in the southeastern states, that if a copperhead was killed it was necessary only to watch the body for a sufficient length of time, and a rattlesnake, following behind would appear and could be slain also. The copperhead was believed to be the fe- male of the rattlesnake. Strecker (1925:49) wrote of an old Texas bottomland myth to the effect that the copperhead leads the rattle- snake to its prey. Beck (1952:143) mentioned various beliefs of the backwoods people of the Blue Ridge region in the Appalachians, concerning snakes. The copperhead is one of the eight kinds of local snakes (some of which are legendary, Hke the hoop snake) known to these people, although many species of snakes actually occur in the Blue Ridge region. Elsewhere the copperhead is often confused with various harmless

snakes, especially the hog-nosed snake ( Heterodon platyrhirios ) and the milk snake (Lampropeltis doliata). The average person, espe- cially a suburbanite or city dweller, recognizes few kinds of snakes with certainty, and is inclined to regard all kinds as dangerous until he has definite evidence to the contrary. This attitude probably prevents some accidents, but unfortunately it results in the needless killing of many non-venomous and economically beneficial snakes. In extensive areas of the Midwest and Northeast, the presence of the copperhead is the basis for this uncertainty, since the easily recog- nized rattlesnake is the only other type of venomous snake found. An elderly farmer who owned land adjoining the Reservation, and had lived most of his life in the same neighborhood, told me, when asked, that he did not know whether there were any copperheads on his farm. He did not recognize them, or any other snakes except for rattlesnakes. His attitude was fairly typical of that of many local farmers, to whom "a snake is a snake" to be killed on sight, and most kinds of wildhfe are regarded as "varmints." Throughout most of the area it inhabits, the copperhead is, in varying degrees, hated and feared. Although, of course, attitudes toward snakes differ greatly among different persons in the same community, the copperhead is, in general, accepted rather casually. Compared with any of the several species of rattlesnakes that share AUTECOLOGY OF THE COPPERHEAD 247 its range, it occasions little alarm; its smaller size, more retiring habits, and lack of the rattle cause it to be less feared. Although it is usually killed on sight, as a matter of course, concerted efiForts rarely are made to reduce its numbers locally. In the one exceptional instance known to me in which copper- heads were purposely hunted and killed, at Trading Post, Kansas, in 1958 and 1959, the motivation seemed to be not so much the local extermination of the snakes as the competitive sport of killing them in large numbers where they were exceptionally abundant. The method of hunting consisted of driving at night over a certain stretch of road where the snakes were abundant. Those caught in the glare of the headlights usually "froze" to immobihty and could be clubbed without serious risk to the hunters. In localities where copperheads are scarce they are more feared. Persons in suburban communities, who lack first-hand familiarity with snakes, and know the copperhead only by its fearsome reputa- tion, are those most affected. Oliver (1958:46) stated: "It is no exaggeration to say that there are thousands of people around the New York area alone who are terrified by the possibility of an en- counter with a Copperhead. Last year I was consulted by two different persons who were considering selling their homes because of reports of Copperheads on their property. One lived in a section where no Copperheads had been found in twenty years, but a large milk snake was killed in her yard by a policeman who said it was a Copperhead." Occasionally landowners have found the presence of the copper- head an asset, and have been able to capitalize on the popular dread of snakes to prevent trespassing and vandalism. Oliver (1958: 41 mentions instances in which have 40, ) signs been posted reading "Beware of Copperheads," or "Do not feed or annoy the Copper- heads," which were effective in discouraging the public. Dread of the copperhead is not a major factor in the lives of the people who live in the areas where it abounds. In this respect, it contrasts with several species of the larger rattlesnakes, which fre- quently cause human deaths, and which are so greatly feared that their presence influences, to some degree, the habits and outlook of the people locally. Schmidt (1945:31) has v^itten of the social prestige gained in the community (formerly, at least) by the victim as the result of a rattlesnake bite, in the back-country of the Ed- wards Plateau in Texas. Furthermore, in certain parts of the coun- try, the killing of a rattlesnake is generally regarded as a feat estab- 248 University of Kansas Publs., Mus. Nat. Hist. lishing the valor and virility of the slayer. Such incidents are often the subjects of long and boastful accounts, in which the size, ag- gressiveness and proximity of the rattlesnake, are highly exagger- ated. In contrast, the killing of a copperhead seems to confer little, if any, prestige, and is not hkely to be talked of more than the killing of a non-poisonous snake. It is a curious fact that various harmless snakes, notably the racer {Coluber constrictor) are feared by rural people as much as the copperhead, and are widely credited with being venomous. The rapid, and sometimes aggressive, move- ments of a racer often cause panic, while the secretive and sluggish copperhead causes less excitement when it is encountered. The circ*mstances under which copperhead bites are sustained often illustrate complete lack of caution or failure to compre- hend and avoid the danger, on the part of the persons who are bitten. Bites are often inflicted on the bare feet or ankles, and often the victims are walking in the dark in places where the snakes might be expected to roam. I refer here to bona fide accidents happen- ing under natural conditions, to persons unaware of the presence of snakes until the bites were inflicted. However, many other bites are sustained by persons catching or handling copperheads. There are probably hundreds of persons in eastern Kansas and western Missouri who have handled live copperheads. Boy Scouts and high- school biology students for instance frequently hunt and catch them as a part of group activities out-of-doors. These inexperienced per- sons often grasp, handle, and release copperheads in such a manner that the snakes are able to bite. My work with copperheads on the Reservation generated wide- spread uneasiness and even hostility in nearby communities. The practice of marking and releasing snakes, especially, was looked upon with disfavor and was blamed for alleged alarming increases in the numbers of copperheads in the same county or those ad- joining. Farmers on land adjacent to the Reservation were urged by me to save any copperheads killed, so that these could be ex- amined, but none of the snakes was ever offered. Absurd rumors such as one that thousands of poisonous snakes had been brought from elsewhere and released on the Reservation, were seized upon by the local press for their sensational appeal, and as a result gained credence among many of the less well-informed country people. AUTECXDLOGY OF THE COPPERHEAD 249

Survival Under Modern Conditions

Early in this century Morse (1904:137) wrote that tlie copperhead in Ohio ". . . is not common as formerly and is undergoing cer- tain extermination." Although the copperhead and most other forms of wildlife have been eliminated from many areas owing to culti- vation or urbanization of the land, Morse's prediction is still far from fulfillment. Indeed the species has actually been favored by some of the changes brought about by man and its populations have increased in certain areas. Secretiveness, nocturnality, cryptic coloration, and a fairly wide choice of prey species are factors that have favored survival under altered conditions and in association with medium to dense populations of man. Also a relatively small individual home range probably is an important factor, as any snake that attempts to cross a thorouglifare with heavy motor traffic is usually doomed. Atkinson (1901:152) wrote that in Allegheny County, Pennsylvania, the copperhead remained fairly common in some localities for many years after both the timber rattlesnake and the massasauga had become extinct, surviving because of retiring disposition and eflFectively concealing pattern. Five years later Sur- face (1906:187) \vTOte that in the same state it had become extinct in most cultivated districts and that it was being gradually reduced in the wilder, mountainous parts. In the same year Stone (1906: 167) observed that the copperhead was becoming scarce in thickly settled districts, such as those of York and Fulton counties, Penn- sylvania. Ditmars (1935:23) wrote that the species was increasing in abundance along the Delaware River, and still occurred along the Palisades of the Hudson River, although the timber rattlesnake had been exterminated there nearly fifty years earlier. Strecker (1935:26) wrote that in McLennan County, Texas, the copperhead had become much less abundant than formerly, but that it was better able to withstand encroaching civilization than were some other kinds of snakes, because of its timid disposition and lurking habits. He explained tliat formerly when large rotting logs were abundant in the bottomlands of the Bosque and Brazos rivers, habitat was much more favorable for the copperhead there. In southeastern Oklahoma, Trowbridge (1937:298) noted that copperheads had decreased greatly in the last decade due to killing by man, but in northeastern Kentucky, Welter and Carr (1939:130) noted that cop- 250 University of Kansas Publs., Mus. Nat. Hist.

perheads had increased over a period of years, and they attributed this fact in part at least to the creation of the Cumberland National in abundant cover for the Forest, resulting snakes. Anderson ( 1942: 215) wrote that copperheads were still found in Swope Park in Kansas City, Missouri. Minton (1944:474) noted that in Indiana copperheads were difficult to eradicate, even in populous areas, and that a few survived even in thickly populated hills on the out- skirts of New Albany. Neill (1948:112) wrote that copperheads were found in large numbers in the northern outskirts of Augusta, Georgia, where a rock outcrop adjoining a golf course provided a favorable denning area. Conant (1952:14) wrote that the species persists across the Hudson River from New York City, in the Blue Hills near Boston, at Valley Forge near Philadelpliia, and within the city limits of Washington, D. C. Oliver (1958:43) recorded that the last copperhead found inside New York City was in the Bronx in 1936, and that in 1954 a single individual was killed in the Greenbrook Sanctuary in Alpine, New Jersey, within sight of New York's skyscrapers. Control

Techniques for controlling venomous snakes have not been satis- factorily developed. The bounty system has been most widely tested. It has been used against many kinds of snakes in various parts of the world, but in most instances no noticeable decrease in their numbers has resulted, and the financial outlay has been great in some instances. As usual with the bounty system, abuses have been common. The practices of keeping gravid females in captivity to obtain large numbers of young to submit for bounty, and of bringing in snakes found already killed by traffic on roads, or those killed in distant areas, to claim bounty from the granting agency of the state or county have often contributed to the breakdown of the system. South Dakota has long employed a rattlesnake-control officer. Mr. A. M. Jackley held this position for many years and became an expert in the mass extermination of rattlesnakes. Jackley's chief method consisted of clubbing or trapping the snakes as they emerged in large numbers from a hibernation den. Trapping was accomplished by partly closing the exit with cement, leaving only a narrow passage leading through a trap-door into a large wooden box where the snakes accumulated as they emerged. Although the technique was especially adapted to the conditions on the northern Great Plains, it is obvious that snakes which congregate in large AUTECOLOGY OF THE CoPPERHEAD 251 numbers to hibernate are thereby rendered more vulnerable to effective control operations. Mass slaughter of snakes (especially rattlesnakes) by clubbing, shooting and blasting in the vicinity of their hibernation dens has long been a common practice in various parts of the United States. However, in most instances the dens involved in these raids have been in areas remote from human habi- tations, where the presence of the snakes involved no special prob- lem. Rather, motivation for the killing has been the desire for sport or an ingrained disHke of snakes, or the fancied prestige gained by the recounting of exploits, perhaps substantiated in part by a display of numerous skins and rattles. To my knowledge the copperhead has never been subjected to systematic control operations, but in the course of my field work I received several inquiries as to how such control operations should be carried out. Details of the copperhead's natural history and ecology such as those set forth in this report provide a background essential in the planning of any control. Because of the copperhead's abundance and widespread range, control operations against the species as a whole are impractical. Also it is by no means certain that control operations resulting in complete suppression of the species would be desirable. Besides the harmful and obvious effects of its bite, the species affects man in various ways; the sum total of the beneficial effects may more than compensate for the occasional harm. Through its food habits, especially, the copperhead affects the ecosystem where it occurs and the prey consists chiefly of animals that are generally considered to be harmful. However, in local areas, such as remaining blocks of woodland in suburban communities, where the copperhead may be a distinct hazard especially to small children, if at all common, its control is desirable, and should prove to be feasible. An obvious method of control is by reducing the food supply. Wherever the species is abundant small rodents such as white-footed mice (Peromyscus sp. ) and/or voles (Microtus sp.) or perhaps other small vertebrates such as shrews, lizards or frogs, are sure to be common, providing the chief source of food. Control of the small rodents on a limited area ordinarily would be accomplished easily by use of poisoned baits. With the chief food supply removed the snakes would be starved out eventually. However, quick results could not be expected because the copperhead's capacity for fast- ing would permit individuals to survive throughout their entire sea- son of activity without any food. Even though the population of small mammals had been reduced to a low level by control opera- 252 University of Kansas Publs., Mus. Nat. Hist. tions, the high reproductive potential characteristic of most small mammals might permit building up to moderate or high numbers again soon enough to save the snakes from starvation unless the original operations were followed up at suitable intervals. More effective control could be apphed when copperheads were concentrated in their denning areas along rock ledges, in either spring or fall. Dynamiting of snake dens (usually those of rattle- snakes) has often been attempted, sometimes with spectacular re- sults. However, where shelter suitable for hibernation is abundant, the hibernating population might be much too well dispersed to be appreciably affected by dynamiting. On the 590 acres of the Res- ervation, for instance, the many hundred copperheads certainly hi- bernate in scores of different crevices and few could be killed with any one charge of dynamite. Klauber (1956:978) mentioned the possibility that blasting of dens (of rattlesnakes) might open up deeper or more extensive cavities favoring the survival and ultimate increase of individuals not killed in the blast.

Use of poison gas has often been tried as a means of killing snakes in their dens, but usually without much success (Uhler, 1944:8; Klauber, op. cit. :98S). The low rate of metabolism in snakes, espe- cially when they are dormant or partly so in their dens, renders them unusually resistant to the effects of poison gas. Flattery (1949:16) found nicotine sulphate to be highly toxic to snakes, and succeeded in killing large numbers of garter snakes out metal with half an ounce of ( Thamnophis sp. ) by putting trays nicotine sulphate dissolved in about 2/2 quarts of water where the snakes were so abundant as to be considered pests. The pans were covered with wire mesh to keep out other animals. The snakes were attracted to the water and were killed by the hundreds, but it seems doubtful whether the population in the general area was appreciably affected. Chemical sprays would be far more effective. Commonly used insecticides such as DDT, aldrin, dieldrin, toxaphene and hepta- chlorane are highly toxic to all kinds of reptiles. Experiments in which tracts of woodland or marsh were sprayed from planes with suflScient concentrations to eliminate insect pests have been shown to have devastating effects on the local populations of reptiles (Mills, 1952:289). Much wildlife of harmless and beneficial species is destroyed by the indiscriminate broadcasting of such poisons, and the cost per acre is high. Against copperheads most econom- ical and efficient control could be obtained by use of high concen- AUTECX)LOGY OF THE COPPERHEAD 253

in trations in a hand sprayer, with repeated heavy applications snakes concen- April and October along the rock ledges where the trate, with special attention to the holes, fissures, and crevices, which might serve as dens. In summer, applications of the same concentrated spray might be effective in killing snakes along rock walls, hedges, weedy fence rows and clumps of shrubbery which are the types of places where copperheads are most hkely to hide or travel in the infested areas. Smith (1953:1) and Minton (1951:322) have emphasized the importance of removing from one's premises potential shelter for venomous snakes such as old boards, shingles, wood slabs, card- board cartons, scraps of tarpaper, rock piles and vegetation such as weed patches, high grass, and clumps of shrubbery. Such shel- ter renders the habitat more attractive to the small animals that provide the copperhead's food, and to the copperhead itself. The snakes may be attracted from nearby areas by abundant cover and their chances of survival and successful reproduction may be increased. THE VENOM AND BITE

Adaptations Correlated With the Venom In venomous snakes the salivary secretions of the supramaxillary or parotid glands serve to subdue the prey. In the more primitive groups including many opisthoglyph and proteroglyph types, the development of toxic saliva and of teeth specialized as fangs to in- ject the venom has involved no evident modification in the snake's general habits or mode of life. But in the more specialized groups, notably the true vipers (Viperidae) and especially the pit vipers (Crotalidae) the development of a more advanced type of venom and injection apparatus is accompanied by other structural modifi- cations, and by increasing commitment to a mode of life different from those of the more primitive snakes. In both groups of vipers the body tends to be short, stocky and flattened. The head is triangular and widened posteriorly to accommodate the enlarged venom glands. The maxillary bone, bearing a single functional tooth, the speciahzed poison fang, has become much shortened, permitting it to rotate on a horizontal transverse axis, and the rigidly attached fang folds back along the roof of the mouth when not in use. By virtue of this device the true vipers and pit vipers have been able to develop poison fangs that are relatively much longer than the teeth of any other snakes. Using these speciahzed teeth in biting their prey, they are able to penetrate deeply into the 254 University of Kansas Publs., Mus. Nat. Hist. vital organs and inject the venom where its action is most rapid and effective. Typically, these snakes are nocturnal and feed upon relatively large prey animals, gorging themselves at each meal with long intervals of fasting. They are sluggish and hunt chiefly by lying in wait to ambush the prey rather than by active search. The copperhead is regarded as one of the most primitive of the pit vipers, yet it is fairly typical of the group as a whole. In retain- ing the large cephalic shields in a pattern similar to that of typical colubrine snakes, it differs from the more specialized crotalids (Crotalus, Bothrops, Trimeresurus) in which tlie shields tend to break up into small, granular scales—a tendency correlated with widening of the head and enlargement of the venom glands. The relatively short fangs and weak venom in the copperhead also seem to reflect its primitive position in relation to the other pit vipers. In general habits it is typical of the group, being slow and sluggish in most of its movements, somewhat secretive, and nocturnal. Like the majority of vipers it is terrestrial although partiy aquatic, fos- sorial and arboreal types have evolved among its more specialized relatives. Like the other crotaHds the copperhead preys mostly on vertebrates and especially mammals, but is unique in its hking for large caterpillars and cicadas, perhaps retained from a more generalized ancesti-al stock that preyed on invertebrates to a greater extent. The primary function of the venom is to cause rapid death in the small animals that are the usual prey, rather than to cause death or damage in humans or other natural enemies. From this view- point the copperhead's venom apparatus is as adequate and eflBcient as are those of other crotalids, although some of these ai-e far more dangerous to humans. In a typical bite the fangs penetrate the thoracic or peritoneal cavity of the prey, but rarely penetrate farther than the subcutaneous layer in a human victim. Forges (1953:50) in discussing the action of snake venom, stated: "Snakes cannot chew and mix the products of their salivary glands with the tissues of their prey. Instead they use a highly developed injection apparatus to apply digestive agents to their food." The action of enzymes plays an important part in the digestion of the prey. The sahva of poisonous snakes contains certain powerful enzymes that are not produced by other kinds of animals, notably ophio-oxinase. Forges stated that the tissues of a rat injected with venom are digested about twice as rapidly as those of an unin- jected rat. AUTECOLOGY OF THE COPPERHEAD 255

Properties of the Venom

The venom is a slightly viscous, watery hquid, yellowish but vary- ing from almost colorless to bright yellow or orange. Often it is cloudy, especially in large individuals; it is almost clear in some young. When copperheads are handled, they usually open their mouths in the course of their struggles and in some individuals the venom can be seen, mixed with other secretions in the buccal cavity. Struggling copperheads commonly erect their fangs, moving them in to bite. jerkily and opening and closing the mouth an attempt As the jaw muscles are contracted, venom often trickles from the for feet or tips of the fangs and can be forcibly projected two more in a spray of fine droplets. Githens (1931:82 and 1935:166) described the venom of pit vipers as a viscid yellow liquid containing from one-third to one-fourth solid matter, a complex mixture of mucous, fatty compounds, salts, act the epithelial debris, and several protein poisons which upon blood, walls of the blood vessels, the central nervous system, and other tissues. Snake venom has many components, each with its different dam- aging effect on the body of the victim, who may succumb from any of a number of different causes. The cause of death may differ according to the species of the victim, the site of the bite, the quantity of venom injected, and of course, the kind of snake deliver- ing the bite. The immediate cause of death may be paralysis of the central nervous system, or of the myoneural junctions, or of respiratory centers; it may be stoppage of the heart, or asphyxiation from massive intravenous clotting (Ghosh and Sarkar, 1956:191); or in more lingering cases it may result ultimately from the cumu- lative effect of more generalized tissue damage, including destruc- tion of the erythrocytes and the walls of the blood vessels. Even if the victim survives all these effects, he may succumb eventually to bacterial infections which thrive in the damaged tissues. The mouths and fangs of snakes are septic and some bites seed the wound with infectious microorganisms. Besides the toxic components in snake venom, there are relatively non-toxic substances which promote rapid spreading of the venom through the body of the victim. Buckley (1959:96) stated that per- meation of the victim's body is accomplished largely by enzymatic activity. Jacques (1956:291) attributed the rapid spreading of the venoms of snakes and other poisonous animals to their hyakuroni- dase content. 256 University of Kansas Publs., Mus. Nat. Hist.

Comparing symptoms in victims bitten by various crotalids, Hutchison (1929:50) noted that respiratory difBculty, an expression of the neurotoxic eflFect of the venom, is most pronounced in cases of poisoning by the copperhead, timber rattlesnake and western diamondback. In humans the bite of a copperhead does not cause extensive hemorrhage as do those of rattlesnakes or cottonmouths. Smith (1956:307) stated that serious secondary infections seldom occur after copperhead bites, although they almost always ac- company those from rattlesnakes. Copperhead venom has found medical uses, as a local coagulent, and in the treatment of epilepsy, neurasthenia, chorea, and shell- shock (Allen and Maier, 1941:249).

Quantity of Venom Produced

Amaral (1928:104) presented figures showing the amounts of venom secreted by copperheads, along with thirteen other species of "nearctic" pit vipers totalling several thousand specimens. Aver- age amounts of liquid and dried venom, respectively, for copper- heads of various categories were: young—.14 cc, 40 mg.; adults— .18 cc, 50 mg.; old adults—.21 cc, 60 mg.; exceptional individuals— .26 cc, 75 mg. The categories were not defined. The figures seem somewhat misleading in implying a rather narrow range of vari- ation between individuals of different size and age categories in the yield of venom. A large adult male copperhead may be as much as thirty times the bulk of a newborn young, and their yields of venom would presumably be somewhat proportional to their weights, although in the young the head and venom glands are relatively large. Githens (1935:167, Table 1) indicated that in 80 copperheads "milked" for venom, the yield of dried venom per snake averaged 56 mg. (maximum 90). Keegan (1956:414) gave 50 mg. as the average amount of dried venom obtained at each milking from a copperhead, and he estimated that in delivering an effective bite tlie snake may inject from 25 to 75 per cent of the contents of its venom glands.

Toxicity of the Venom

Githens (1935:171) performed extensive experiments to deter- mine the toxicities of various crotalid venoms. For the copperhead the average minimum lethal dose of dried venom for a 350-gram to .20 to pigeon intravenously injected was found be .12 mg. ( mg.

. have suflBcient .05 mg. ) Thus an average adult copperhead would venom to kill 470 such pigeons, each having many times the bulk of AUTECOLOGY OF THE COPPERHEAD 257 the snake's normal prey. Githens emphasized the fact that in his tests of venom the factor measm-ed was power to cause acute para- lytic death in the experimental animals. In the more primitive types of venomous snakes, including the elapids and the less speciaUzed crotahds, the neurotoxic components of the venom are relatively more prominent, wliile in the more speciahzed of the pit vipers, according to Githens, the hemolytic components of the venom are more developed. This hemolytic type of venom produces far more severe local symptoms, and more lasting effects—if the victim does not soon succumb. Describing the difference in effect according to the amount of venom, Githens (loc. cit.) wrote: "When given intravenously, espe- cially in excessive doses, the venom may kill within five to ten minutes by convulsions apparently asphyctic, and perhaps due to interruption of the circulation by intravascular clotting. After some- what smaller doses, paralysis is tlie usual manifestation. This begins in pigeons usually within fifteen minutes, as a weakness of the legs, the pigeon setthng to the floor of the cage. As paralysis advances, the neck becomes weak, the beak, and finally the entire head rest- ing on the floor. After this stage is reached, recovery is rare. Ac- cording to the dose, paralytic death may result in from fifteen minutes to twelve or eighteen hours." Effect of the venom is, of course, maximal when injections are intravenous. Using pigeons intravenously injected, as the experimental animals Githens tested the potency of 26 kinds of American crotalids, mostly rattlesnakes. If copperhead venom, with average minimum lethal dose of .20 mg. is used as a standard of comparison, with a rating of one unit, other kinds had the following relative potency

in of : rattle- ( numbers parentheses represent number assays ) red snake (Crotalus ruber) .2 (5), black-tailed rattlesnake (C. molos- sus) .4 (6), Florida diamondback (C. adamanteus) A (8), timber rattlesnake (C horridus) .6 (12), western diamondback (C. atrox) .9 (29), sidewinder (C. cerastes) 1.0 (4), cottonmouth (Agkistrodon piscivorus) 1.1 (16), cantil (A. bilineatus) 1.1 (2), Great Basin rat- tlesnake (C. viridis lutosus) 1.1 (11), South American rattlesnake (C. durissus) 1.1 (22), Pacific rattlesnake (C. viridis oreganus) 1.2 (11), massasauga (Sistrurus catenatus) 5.0 (6). Minton (1953:214) studied variation in venom samples from seven timber rattlesnakes and eight copperheads. He found that in the copperhead samples the strongest was three times as toxic as the weakest. Even greater variation in toxicity occurred in the seven samples from the timber rattlesnakes. Expressed in terms of 258 University of Kansas Publs., Mus. Nat. Hist.

'lethal the intraperitoneal dose 50" ( the dose capable of kilhng half it the experimental mice receiving injections of ) Minton found that of the rattlesnake venom 5.11 milligrams per kilo was required and of the copperhead venom, 6.36 milligrams per kilo. Thus the rattle- snake venom averaged approximately 1/4 times as potent as the cop- perhead venom. In discussing the validity of his findings w^ith re- gard to probable effect of the venoms on humans, Minton pointed out that the mouse is relatively more resistant to the venom than is a human, and that intraperitoneal injection is not normal in snake- bite when a human is the victim. However, he stated ". . . there is every reason to believe that the bite of a snake whose venom con- tains approximately ten thousand mouse lethal doses per milliliter would be considerably more serious to a human than the bite of a snake whose bite contains only about two thousand doses in the same volume. This limited study indicates that wide and appar- ently unpredictable individual variation in venom toxicity occurs among copperheads and timber rattlesnakes and probably among other species as well." However, in later pubHcations, Minton (1954:1079 and 1956:146) arrived at much different figures for the toxicity of these venoms on the basis of subcutaneous injections in mice. The 'lethal dose 50" per kilo of body weight of the victim of the copperhead was found to be 25.65 mg. Approximately the same potency was determined for the cottonmouth. Other crotalids tested were all more potent than the copperhead, in the following ratios (considering the cop- perhead's potency to be one unit): pigmy rattlesnake (Sistrurus miliarius), 1.06; red rattlesnake, 1.23; western diamondback, 1.33; cantil, 1.35; Florida diamondback, 1.76; timber rattlesnake, 2.8; side- winder, 4.7; massasauga, 4.9; southern Pacific rattlesnake, 7.2. Min- ton did not indicate the sizes of the samples on which these later figures were based, nor did he comment on the seeming discrepan- cies between these and his earlier figures. Criley (1956:378) used intravenous injections of venom on 18- gram mice to determine LD 50. Projecting his figures to the same units that Minton used, the LD 50 per kilo of body weight of victim was 6.95 mg. in the copperhead. The relative toxicities of other crotalids tested, considering the copperhead to be one unit, were: Mexican West Coast rattlesnake (Crotalus basiliscus), .6; cotton- mouth, 2.0; Pacific rattlesnake, 2.7; cantil, 4.3; Florida diamondback, 4.5; prairie rattlesnake, 6.3; South American rattlesnake, 55.0. Minton (1956:146) also compared the hemaglutinin, local necro- tizing action, hemolysin, and Paramecium lysin in several crotahd AUTECOLOGY OF THE COPPERHEAD 259 venoms. The three species of Agkistrodon were found by Minton to have similar venoms, of a type representing relatively primitive crotalid stock, with low lethal toxicity, relatively little necrotizing action, and well developed hemaglutinin and hemolysin. Minton noted that among the eleven species tested relationship between the intraperitoneal and subcutaneous lethal dose showed consider- able variation. "The greatest difference was observed with the venom of S. catenatus where the intraperitoneal LD 50 is approxi- mately one-twentieth the subcutaneous LD 50. By way of contrast, the intraperitoneal LD 50 for C. horridus venom is approximately two-thirds the subcutaneous LD 50." Obviously the copperhead made a relatively better showing when its partly neurotoxic venom was injected intraperitoneally or intravenously than when it was injected subcutaneously, and made a far better showing on the pigeon than it did on the mouse. There has been some difference of opinion concerning the ca- pability of young copperheads in producing venom and in delivering an effective bite. Reese (1926:357) related an instance of three newborn copperheads in captivity which repeatedly bit a rat placed in their cage without harming it. To test further the lack of potency in these young snakes Reese sacrificed them, ground up their venom glands, and injected an extract of the glands into young rats. The rats were not killed, but both earher (Atkinson, 1901:152) and later tests have failed to substantiate the idea that newborn young of copperheads and other pit vipers lack effective venom. Stadelman (1928:67) caused a newborn copperhead to bite his forearm for 60 seconds. Symptoms were relatively severe. Swell- ing steadily increased for 21 hours and subsequently did not lessen for the next twelve hours, subsiding gradually from the second to the seventh day after the bite. In further experimentation on the same subject, Stadelman (1929:81) resorted to use of a mouse as the experimental animal. He removed a captive-bom copperhead from its fetal membranes and forced it to bite the leg of the mouse. The animal died in 48 hours. Stadelman observed that the biting mechanism is not well co-ordinated in the first few hours after birth.

Susceptibility of Snakes There has been much controversy and misunderstanding regard- ing the susceptibilit}^ of venomous snakes to their own poison. In the course of my study many copperheads that were disturbed were observed to bite themselves or other individuals, and the results varied, from no perceptible effect at one extreme, to almost in- 260 University of Kansas Publs., Mus. Nat. Hist. stantaneous collapse, and death within a few minutes (Fitch, 1959: at the other. If a bite is into vital 21 ) venom from injected a organ, death may soon ensue, but ordinarily the recipient of the bite shows but little ill eflFect. A copperhead that is being handled will some- times bite through its own lower jaw by suddenly closing its mouth while the fangs are erect. Such bites sometimes cause severe tem- porary swelling of the chin and inflammation of the lining of the mouth. When a copperhead that is forcibly restrained bites its own body as a coil comes within reach in the course of its writhing, the damage is usually slight. Ring-necked snakes were often offered ahve to young copper- heads in captivity, and when bitten they usually died within a few minutes—as as rapidly any mammals. Keegan and Andrews ( 1942: 253) experimentally tested the venom of copperheads and rattle- snakes on many kinds of snakes, and found that all were susceptible, the ring-necked snake particularly so. Only one of 21 snakes tested by Keegan and Andrews wdth more than .233 mg. of venom per gram of body weight survived. However, it is noteworthy that this is than times the lethal dose more 650 ( injected intravenously ) for a pigeon of comparable size. Allyn (1937:222) experimented with a copperhead, a timber rattlesnake and a massasauga, causing each snake to be bitten at mid-body by the other two. In each instance there was slight swelling and the bitten snake was more sluggish than usual for several days, but recovered eventually. Swanson (1946:242-249) also experimented to determine and compare the effects of the venoms of North American crotalids on various snakes, including both harmless and venomous species. He found that snakes usually are able to survive normal or average doses of venom, although they are by no means immune to it. Copperheads curiously, proved to be more susceptible to venom of their own species than to other snake venom. Four minims killed a copperhead in 15 minutes; 2 minims in a half hour; 2 minims in 20 minutes; 1.5 minims in 5 hours; all snakes were of comparable sizes. Ten injections of copperhead venom (2.5 minims to 10 minims) into other kinds of snakes (water moccasin, massasauga, timber rattlesnake, common water snake, yellow-bellied racer, black rat snake, milk snake) all resulted in deatli in from 35 minutes to 29 hours and 45 minutes. PLATE 13

^^^^^^^^^K- PLATE 14

Fig. 1. Brushy field with dense herbaceous vegetation, a faNorable habitat for copperheads in summer, 100 vards east northeast of Reservation licadciuarters, July, 1959.

Fig. 2. Prominent hilltop limestone ledge with deep fissures, "Rattler Ledge" on the Rockefeller Experimental Tract, April, 1959. Numerous copperheads along with timber rattlesnakes and various harmless snakes hibernated in this vicinity. PLATE 15

'%'

'i *>,ii

Fig. 1. South-facing rock ledge at site of old ciuarry in May, 1959. The rocky and brushy habitat with woodland and grassland rendered this a favor- ite location for copperheads throughout the season of their activity.

Fig. 2. Nhissue Innestone slab at lower ledge on a southwestern exposure. Each autumn copperheads were trapped at the base of this slab. The cleft to the boy's right was believed to be the entrance to a hibernation den, November, 1958. PLATE 16

4. % 'J'*^^^

\ ^ -"V^^^

lujm^-

in live-oak at Fig. 1. rians-Pecos copperheacl ( cipproximately X '4) i^iost- Independence Creek, Terrell County, Texas. Sheep grazed in this vicinity and herbaceous vegetation was scanty; June 28, 1957.

t'lG. 2. Remnants ol live-oak grove devastated by flood, with piles ot drift and debris, at Independence Creek, Terrell County, Texas, June 28, 1957. The flood in June, 1954, destroyed much of the oak-grove habitat and drastically reduced the population of copperheads locally. PLATE 17

Fig. 1. Two-year-old female copperhead (above) reared in captivity; three the recaptured marked young ( below ) of same age but smaller bulk; approxi- mately X -A, May, 1957.

Fig. 2. Large adult male (left), large adult female (right) and newly matured two-year-old male showing differences in length and bulk; ap- proximately X 77> October, 1958. PLATE 18

Fig. 1. Head of live male 41-inch copperhead, X 2, showing physiognomy, scalation, and pattern; July 5, 1960.

n-

Fig. 2. Large adult male and one-year-old male, approximately X % show- ing difference in size, September 13, 1951. PLATE 19

Fig. 1. One-year-old copperhead (left), two-year-old (middle) and new- born young showing differences in size. ( Approximately X % )

three- Fig. 2. Tails and reproductive or,!j;.ui,s of female copperhcatLs: upper, year-old still not sexually mature; middle, adult that had not borne young recently (left ovary missing); lower, adult, recently parturient, with enlarged oviducts, October, 1959. (Approximately X%-)

Fi<;. 3. A group of copperheads from Buck Creek, seven miles southwest of Clinton, Douglas County, Kansas, showing irregularities in the dorsal cross- bands, that are typical of the population studied, April, 1959. (Approxi- mately X Va- ) PLATE 20

Fig. 1. Birth ot copperheads; to the right ot the female's tail a young just extruded lies enclosed in its fetal membranes. Two litter mates which pre- ceded it lie farther to the right, likewise still partly enclosed in membranes, September, 1959. (Approximately X%-)

of birtli in Fig. 2. First-year copperhead ( left ) kept from time captivity and well-fed even when it would have been hibernating under natural condi- tions, and recaptured litter mate of much smaller bulk, demonstrating effect of food upon growth. (Approximately X%-) AUTECOLOGY OF THE COPPERHEAD 261

Circ*mstances and Outcome of Bites

in United States In an early study of the snake-bite problem the data on 740 bites of venomous snakes Willson ( 1908:516) assembled of the bites were reported before 1908. Ninety-seven (17 per cent) those of copperheads and five of these copperhead bites resulted in death. Willson impHed that in each instance death might have been averted by better care, and that in the three adult men who died, the large amount of whiskey taken may have been the deciding factor. He expressed doubt that copperhead bite is ever fatal to adults in uncomplicated cases. Hutchison (1929:47; 1930:40) listed the seasonal distribution of copper- Antivenin Institute of head bites, for which reports had been received by the America for the years 1928 and 1929. For 1928 the distribution was: January, February and March, 1; April, 1; May, 8; June, 21; July, 50; August, 39; September, 36; October, 6; November, 5; December, none. For 1929: January, 1; February, none; March, 2; April, 10; May, 19; June, 22; July, 37; August, 28; September, 16; October, 2; November and December, none. None of the 303 copperhead bites recorded in these two years resulted in death, although there were many deaths from rattlesnake bite. For the 200 bites of poisonous snakes (mostly copperheads) reported in Virginia from 1941 through 1953 (Wood, 1954:938) the seasonal distribution was: April, 1.6 per cent; May, 9.5 per cent; June, 16.3 per cent; July, 31.6 per cent; August, 18.9 per cent; September, 19.5 per cent; October, 2.6 per cent. Minton (1951:320-321) tabulated the circ*mstances and results of copper- head bites in Indiana. The outcomes of nine bites were as follows: "Moder- ately severe symptoms. Recovery."; "Died. Said to have drunk a large inflicted this in quantity of whiskey." ( The snake that bite, Crawford County, Indiana, had been called a moccasin, but no cottonmouths are known from this area and Minton suspected a copperhead or rattlesnake was the real culprit.) "Recovery." "Uneventful recovery." "Severe serum reaction; recovery." "Recovery." "Recovery after rather prolonged illness." "Moder- ate symptoms with good recovery." In four instances the bite was on a finger, in three it was on a hand, and twice on a foot. Circ*mstances of the bites were summarized as follows: "Playing near a woodpile." "Working on construction gang near river." "Touched snake while climbing among rocks." "Measuring captive snake." "Picked up snake—mistaken for harmless species." "Transferring captive snake from cage to bag." "Moving a large rock in woods." "Reached into empty feed sack." "Playing in abandoned woodshed."

"Reached into com crib." Swartzwelder ( 1950:578) recorded copperhead bites in Louisiana while: "Trapping in marshes" (two instances); "playing around lumber," "stepping on snake," "in outbuilding," "picking moss," "hunting." Wood (1954:942) recorded the circ*mstances of a less typical accident in Virginia, as follows: "A camping party near Luray decided to sleep in a bam.

12—4428 262 University of Kansas Publs., Mus. Nat. Hist. and shortly after retiring one of the campers complained of a wasp sting on her right arm. About two hours later another member of the party was awakened by a stinging sensation on her thigh. A flashlight revealed the presence of a copperhead nearby, and examination of tlie wounds showed the deep twin punctures made by fangs of such a snake. . . . It is probable that the copperhead was foraging, detected the presence of a warm object with its Tieat-receptor' pits, and struck, injecting sufficient venom to immobilize its usual prey."

It is significant that three of the ten cases recorded by Minton involved handling of live copperheads. It is my impression that a similar ratio obtains in Kansas. Many of the copperhead bites that have come to my attention were sustained by persons handling the snakes, often motivated by curiosity or bravado. Githens (1935:165) in his tabulation of 2,342 snake-bites sus- tained in the period 1927 to 1934, indicated that 163 of the bites were received by persons who were intentionally handling poi- sonous snakes. Of the 72 persons who died from effects of bites of venomous snakes more than half were children less than 14 years old. Of the 2,342 total, from all parts of the country, 691 were those of copperheads, far more than were inflicted by any other species. The timber rattlesnake, with 411 bites, was second. Among 134 snake-bite cases in Virginia for which the species were known, the distribution was: copperhead, 119; timber rattlesnake, 12; cane- brake rattlesnake, 2; cottonmouth, 1 (Wood, 1954:937). In Loui- siana the cottonmouth is the chief offender and the copperhead in 11 of 161 bites figured only ( Swartzwelder, 1950:579). Swaroop and Grab (1956:441) estimated that there are from ten to 20 deaths annually in the United States from snake-bite, and tliey mentioned the diamondback [both species?], prairie rattlesnake and cotton- mouth as the chief killers. Certainly the copperhead is relatively unimportant as a cause of death, despite its prominence in snake- bite statistics. Hypersensitivity to Venom

Zozoya and Stadelman (1929:94) described an instance of hyper- sensitiveness to snake venom in a 22-year-old man who was em- ployed in the handling and collection of the venom. He had been bitten by a copperhead in June, 1923, and in later years received injections of the venoms of Crotalus, Agkistrodon, Bothrops and Naja. In August, 1928, he was again bitten by a copperhead. Late in 1928 the hypersensitiveness became evident, and increased in intensity over a period of weeks, with coryza and violent sneezing resulting whenever the dried venom was handled. Mendes (1952: AUTECOLOGY OF THE COPPEKEIEAD 263

a after 1328 ) described the case history of 29-year-old woman who, working at the Antivenin Institute at Sao Paulo, Brazil, for two years, developed acute asthma, rhinitis, and conjunctivitis upon proximity to the venom. These symptoms continued for eight years until the subject transferred to other employment; then they promptly disappeared. Cutaneous tests had shown strong reactions to venoms of Bothrops and Crotalus. No bites or injections of venom were mentioned in this account. Such hypersensitivity seems to be common in persons who handle snake venom habitually. Stanic (1956:181) stated that several members of the staff at the Central Institute of Hygiene, Zagreb, Yugoslavia, became hypersensitive in varying degrees to venom of vipers, and he described a desensitizing procedure which was partially successful for several months, but the hypersensitivity gradually returned. Parrish and Pollard (1959) studied the effects, on man, of repeated bites by poisonous snakes. Of 13 patients tested, four gave evidence of hypersensitivity to the venom, manifested by large wheals with pseudopods. None of the patients had experienced anaphylactic shock as a result of a second bite. However, the authors stated (op. cit. :284): "It seems entirely possible that occasional deaths from snakebites, in individuals who have been previously bitten, may result from snake venom allergy." The authors found that a bite by one kind of venomous snake might confer sensitivity to a closely related kind.

Case History of a Bite

A copperhead bite that I received in 1957 was perhaps fairly typical, and unusual opportunity to observe the effect of the venom was afforded because treatment was kept to a minimum, emotional shock occasioned by a horror of snakes was not involved, and symp- toms were set down in writing as they occurred. The bite was received at 8:15 p.m. on June 5. Driving on a county road near the Reservation, shortly after dark in a light rain, I saw the copperhead crossing in front of the car in the glare of the headlights, swerved to miss it, and stopped the automobile a few yards beyond. I ran back with a flashlight and located the snake, a large one, which was thrashing and lunging in an energetic at- tempt to gain the shelter of dense vegetation by the roadside. As I confronted the snake attempting to pin it with a foot ruler, an approaching automobile rounded a curve 200 yards away, and the glare of its headlights dazzled me. The snake did not deviate from its course, but sparred with me and lunged sideways partly avoid- ing the stroke as I pinned down its forebody. As a result, it was 264 University of Kansas Publs., Mus. Nat. Hist.

held down too far back behind the head and an instant later out- reaching me, it struck the middle finger of my right hand. Besides the slight lacerations made by the pterygoid teeth there were two distinct fang punctures l/s inches apart, one on the dorsal surface of the proximal knuckle joint, the other on the fleshy medial surface of the finger at approximately the middle of the basal segment. Immediately abandoning the snake, I concentrated on treatment of the bite, sucking hard and drawing small amounts of blood from the fang punctures. Spasmodic twitching of muscles at the site of the bite was soon noticeable. From the start there was a dull ache at the site of the bite. Over several hours it became progressively more severe.

No tourniquet was applied on the theory the bite certainly would not be fatal and that the venom could best be dealt with by absorb- ing it rather than allowing its effects to concentrate at the site. No physician was consulted. By 8:25 p. m. the site of the bite had become noticeably swollen and discolored. As a result of the swelling blood could no longer be sucked from the wound, and throbbing pain had become severe. At 8:30 p. m. to promote bleeding, tliree punctures were made with a 26-gauge hypodermic needle, and two aspirin were taken to alleviate pain. Bleeding from the needle punctures rapidly dimin- ished as the swelling increased. By this time it was becoming evi- dent that relatively little venom had been injected through the fang puncture on the knuckle joint. Probably the fang tip had struck bone near the surface preventing complete penetration of the slit end with the result that venom ejected had been partly spilled. Swelling was steadily progressing proximally. Site of the most severe pain had shifted from the lower fang puncture to an area about one inch in diameter in the palm adjacent to the base of the middle finger. The throbbing was accompanied by a sensation of numbness in the overlying skin. With a sterile razor blade a longi- tudinal incision half an inch in length was made through the fang puncture to a depth of approximately % inch. For several minutes thereafter blood flowed freely from the wound, but gradually it again became more meager as swelling increased, and soon little could be obtained by sucking. No further incisions were made. At 10:00 p.m. pulse was 58 (normal) and temperature was 99.0°. Swelling had progressed to a level about 2/2 inches above the wrist, with slight accompanying discoloration. By 10:15 p.m. swelling AUTECOLOGY OF THE COPPERHEAD 265 had .progressed to a level about four inches above the wrist. The throbbing pain in the palm was still severe and extended back along the lower side of tiie arai to the elbow. One-fourth grain of codeine and a glass of milk were taken. At 10:30 p. m. respira- antihistamine Chloro- tory congestion had become noticeable. An ( trymatron) was taken to counteract these symptoms; pulse 50. At 10:45 p.m. pain had reached its maximum, and was intense in the palm near the base of the middle finger, extending back as far as the elbow. Another Ji-grain of codeine was then taken. At 11:15 p. m. a feeling of nausea became prominent. At 11:30 vomit- ing occurred; pulse 55. At 12:45 a. m. no new symptoms had ap- peared; a sleeping capsule, nembutal grains VA, was taken. At 2 a. m. another IM grains of nembutal and /2 grain of codeine were taken. On the following morning systemic symptoms had largely disappeared. The bitten hand had swollen to nearly twice normal size and swelling extended shghtly above the elbow. Swelling and soreness in the afflicted hand and elbow subsided slowly, and it was nearly a montli before normal use of tlie hand was regained.

Treatment of the Bite

Thomas Say (1819:259) a pioneer American naturalist related from first-hand observation an instance of copperhead bite and a treatment that reflected belief in an old folk remedy. The bite was followed by rapid swelling and pain. The breast of a fowl was plucked and applied to the wound. ". . . in a few minutes the fowl died without having experienced any apparent violence from the hand of the applicant, the breast exhibiting a livid appearance." A second fowl was then laid open and placed upon the wound. The patient recovered. The "fowl treatment" is still widely believed in and sometimes practiced by rural people in various parts of the country, although it has no scientific basis. Other home remedies widely used in the past, but of no value in combating the action of the venom, and sometimes causing complications that prevent or delay recovery, are imbibing of hquor, and application of kerosene, or of potassium permanganate either powdered or in solution, to the wound. The controversial cryotherapy for snake-bite was first used by Crum (1906:1433) in Maryland on victims of copperhead bite. Technique consisted of spraying with ethyl chloride to freeze the tissues locally, supposedly slowing tlie action and spread of the venom. More recently cryotherapy has been championed by 266 University of Kansas Publs., Mus. Nat. Hist.

Stahnke (1953:35) but violently attacked by Shannon (1956:410) who maintains that freezing of the tissues or even prolonged ex- posure to water as warm as 55° F. produces serious and permanent damage and usually results in gangrene. Regardless of other measures taken, use of a tourniquet to delay spread of the venom, and incision and suction at the site of the bite to remove it have long been standard procedures. But re- cent experiments by Leopold, Huber and Kathan (1957:414) with rabbits have shown that both use of a tourniquet, and incision and suction shorten the time to death in tlie experimental animals. Rab- bits that were immobilized and injected with venom survived nearly four times as long as the controls, which were given the same amount of venom but were permitted to move about freely. Par- rish (1956:403) recommended early and extensive excision of tissue at the site of tlie bite combined with suction in serious cases after he had performed experiments in which two-thirds of the dogs in- jected with six MLD each were saved by such treatment, while others similarly injected, and treated with the customary incision and suction, all died. Antigenic serum effective against the venom of "Nearctic Crotal- idae," including the copperhead, was first manufactured in the late nineteen-twenties by the Antivenin Institute of America. Over periods of months horses were immunized by small but gradually increasing dosages of the venoms of eleven kinds of American crotalids. Although the serum has unquestionably saved many lives, its perfoiTnance proved to be somewhat less effective than had been generally anticipated when it was first made available. Githens (1935:172, table 3) presented statistics showing the incidence of mortality from bites of different kinds of North American crotalids, treated with antivenin and untreated. Of 152 bites for which no antivenin was administered, five resulted in death, but there were only two deaths from 539 bites treated with the serum. For severe bites massive dosages of many 10 cc. ampules were recommended. As most persons are more or less sensitive to horse serum, such large doses usually caused untoward reactions that were often of alarm- ing severity. In 1954 a new and much improved serum was made available (Criley, 1956:375) much more potent than the original product, yet more easily and cheaply produced. For this newer serum venoms of only four species (not including the copperhead) are used, yet it has been found to be effective against all crotalids. This serum is now manufactured and distributed exclusively by Wyeth Laboratories, Inc. AUTECOLOGY OF THE COPPERHEAD 267

In the eastern states and especially the Northeast, the copperhead is of increasing relative importance as the rattlesnake's numbers dwindle in the face of advancing urbanization, Antivenin serum is sold in many areas where the copperhead is the only venomous snake present, and perhaps this species figures more prominently than any other in the actual use of the serum or in its purchase and anticipated use. Since even the newer improved serum often has disagreeable and potentially dangerous effects, and since even untreated copperhead bites rarely result in death, the use of anti- venin in treatment has been seriously questioned. OHver and Goss (1952:270) stated: "Marked reaction to the horse serum may hter- ally make the 'cure worse than the bite' and terminate in sudden death, or produce less acute signs of distress." Wood (1954:940) discussing the circ*mstances and treatment of snake-bites (mostly of copperhead) in Virginia, stated that no fatalities were recorded in 168 cases. But in the 90 per cent of the victims that received serum, urticaria, pruritis, and angioneurotic edema were more pro- nounced than in those not so treated. Klauber (1956:920) stated that the bite of a copperhead is rarely serious enough to justify the use of antivenin. Shannon (1956:407-408) wrote that antivenin is "not an unmitigated blessing. . . . Preliminary skin testing may not reveal the presence of horse agglutinins, and serious de- layed reactions or even anaphylaxis may follow the use of large or small amounts." Shannon cited two instances of rattlesnake bite, seemingly not especially serious in themselves, which, when treated with antivenin resulted in violent and prolonged anaphylactic shock. One patient was comatose for a week, the other for six days. No information is available regarding the extent to which present meth- ods of refinement of the serum have eliminated untoward reactions. Minton (1954:1078) found that the serum, injected into mice that had had double the minimum lethal dose of copperhead venom, conferred no perceptible protection, as all the experimental animals died. The serum was found to be in varying degrees more effective in conferring some protection on mice that had been injected with the venoms of various species of rattlesnakes. In connection with the failure of the serum against Agkistrodon venom, Minton com- mented on the fact that this kind of venom was not used in the immunization process to which the horses producing the venom were subjected. However, the antivenin used in his experiments seems to have been the older type, made with eleven venoms in- cluding that of the copperhead, not the newer four-venom preparation. 268 University of Kansas Publs., Mus. Nat. Hist.

Criley (1956:374-376) explained that while the venoms of nine kinds of that of the fifteen rattlesnakes, plus cottonmouth ( per cent ) and copperhead (five per cent) were used in the manufacture of the original antivenin made for use against venoms of North Ameri- can crotalids, the improved serum more recently manufactured is based on only four species, the western diamondback, eastern diamondback. South American rattlesnake, and fer-de-lance. Be- cause there is strong antigenic relationship between the venoms of diflFerent crotalid genera, the use of venom from a large number of species in manufacturing serum is deemed unnecessary and unde- sirable. Tests have shown that antivenin made from the four species named a£Fords the broadest possible polyvalency, and is eflFective against not only North American crotalids but those of the Neotropi- cal and Oriental regions as well. Mrs. Eleanor E. Buckley of Wyeth

Laboratories stated in : "It is ( litt., January 12, 1960 ) the opinion of many physicians that the possible discomforts and risks in serum treatment do not equal those attendant on mechanical treatment, and recovery is certainly more rapid." loc. cit. Criley ( ) presented figures showing that the new antivenin prepared with the four-venom formula is more eflFective than the older preparation against all species of crotalids (17 kinds tested, including the copperhead). Mice weighing 18 grams were given the serum intravenously. The new type serum neutralized 22.4 lethal doses of copperhead venom per milliliter in these tests. EflFectiveness of the antivenin is inversely proportional to the potency of the kind of venom counteracted. In the more virulent species, Crotalus durissus terrificus (440 lethal doses per mg. of venom), Bothrops atrox (54 lethal doses per mg.), B. neuwiedi lethal doses and B. lethal (44 per mg. ) jararaca (44 doses per mg. ), the serum neutralized 187.0, 124.8, 90.2, and 74.8 lethal doses per milliliter, respectively. But in those kinds with weakest venom, B. jararacussu (3.2 lethal doses per mg.), C. basiliscus (5 lethal doses and contortrix 8 lethal doses per mg. ) , Agkistrodon ( per mg. ) the number of lethal doses neutralized per milliliter of serum is rela- tively small—7.2, 11.5, and 22.4, respectively. Similar trends were shown in figures published by Gingrich and Hohenadel (1956:382) based in part on the same lots of venom. The many conflicting statements in the literature, regarding the relative virulence of the copperhead's venom, and the eflFectiveness of serum or other remedies in combating its eflFects certainly empha- size the need for further investigation. However, it is obvious from the foregoing discussion that the copperhead has much less potent AUTECOLOGY OF THE COPPERHEAD 269 venom than most other New World crotalids, that the venom varies individuals greatly in both quantity and potency between different of the same population, and perhaps in the same individual at different times, and that different kinds of animals differ greatly in there- their susceptibihty to the venom. Measures of potency may fore show quite different results, depending on the type of damage inflicted, whether paralysis of the central nervous system, respira- of tory failure or massive histolysis, and on the species experimental animal utilized. SUMMARY

A field study of a local population of the copperhead was carried on from 1948 through 1959 on the 590-acre University of Kansas Natural History Reservation and the adjacent 160-acre Rockefeller Tract; 1,532 individuals were recorded a total of 2,018 times. Al- though the incidence of recaptures was low, even in the later stages of the study, the marked snakes recaptured yielded most significant data. Clipping of ventrals and subcaudals in different combinations provided formulas by which marked individuals could be recognized and the remarkably variable arrangements of the bands on the body provided a supplementary means of identification. Most of the cop- perheads recorded were caught in wire funnel traps. Rock ledges at hilltops where the snakes hibernated provided the most pro- ductive sites for traps. In the summers of 1957, 1958 and 1959, many funnel traps were placed in other habitats and supplemented with drift fences. In these seasons copperheads were caught in substantial numbers on their summer ranges. Records obtained on the Reservation were supplemented by records contributed by many co-operators, and by those available in published hterature. The copperhead is a medium-small snake; those from the Reser- 9.8 to 42.0 inches in vation averaged 22.4 inches snout-vent length ( over-all length). The coloration is reddish brown with seven to 16 chestnut cross-bands constricted middorsally to an hourglass shape. Maximum size is greater by one-fourth in males than in females. Most typical snout to vent lengths of adults on the Reservation are 28.5 inches for males and 26 inches for females. In the males the tails are, on the average, slightly longer than in females of the same size, but in both sexes relative tail length progressively decreases as larger size is attained. The poison fang, a relatively elongate hollow tooth that is borne on each maxillary bone, is only about half the relative length of a typical rattlesnake fang. The fangs are shed and replaced fre- 270 University of Kansas Publs., Mus. Nat. Hist. quently. Observations on a captive copperhead indicated that typically a fang is functional for a period of approximately a month, but the interval is variable. Nearest hving relatives of the copperhead are the partly aquatic cottonmouth of the southeastern states and the cantil of tropical Mexico. The genus also includes at least eight Asiatic species, most of which, like the copperhead, are forest dwellers. This zoogeo- graphic evidence indicates that the genus formerly was more north- ern in distribution and dates back to the early Tertiary, but the earliest known fossils of the genus and family are remains of the of rattlesnake copperhead ( found associated with those the prairie ) at Driftwood Creek in southwestern Nebraska in deposits that are probably of lower Phocene age. The geographic distribution of the copperhead corresponds closely with the extent of the Deciduous Forest Biome of Eastern North America, exclusive of its glaciated northern part. Small isolated populations exist far to the west of the Biome's present limits where relicts of deciduous forests remain in unusually mesic situations. The subspecies pictigaster of Trans- Pecos Texas consists of such relict populations. The other three subspecies correspond to major subdivisions of the Deciduous Forest Biome that are based upon the dominant genera of trees prevailing. A. c. laticinctus is confined to the Oak-Hickory Association which is the westernmost phase of the Biome. A. c. mokeson occurs mainly in the combined Oak-Chestnut, Mixed Mesophytic, and Western Mesophytic subdivisions, but a disjunct western segment occurs in the Oak-Hickory. A. c. contortrix occurs chiefly in the Southeastern Evergreen and Oak-Pine associations. A. c. pictigaster differs from the others most, as it has one scale row much shortened, and seem- ingly the snake itself is dwarfed. The other three subspecies differ from each other mainly in pattern. A. c. contortrix has the "hour- glass" markings and shading, which are characteristic of the copper- head, most highly developed. The copperhead lives chiefly in or near deciduous forest. Throughout most of its range it prefers rock ledges in hilly situations. In the southern states swamps and other lowland situations are frequented more than they are in other parts of the range. Where the copperhead occurs in arid regions at the western extreme of its range, the species inhabits the most mesic situations available. The copperhead spends most of its time in a characteristic flat resting coil, with the tail on the outside and the head, neck and fore- body in an S-shaped loop near the center. In this position the snake AUTECOLOGY OF THE COPPERHEAD 271 awaits approach of prey. Activity is chiefly nocturnal. In normal locomotion the snake crawls slowly, with frequent long pauses. Ordinarily a combination of the "horizontal undulatory" and "rec- tilinear" methods of crawling are employed. An individual moves about slowly under most conditions; in a 24-hour period it may shift only a few yards or may not move at all. There is a pronounced tendency to keep within a definite area, or home range. The areas of typical home ranges in summer were calculated to be approxi- mately 24.5 acres for males and 8.5 acres for females but records were inadequate to map any one range in detail. A home range may include a rock ledge situation where the snake hibernates, but more often is disjunct from the area of hibernation, and the snake's movements include annual migrations from hibernation den to home range in spring and back again in autumn. Most frequently the snake retvTrns to the same den each year. The average shift was found to be 1,715 feet for 21 males and 1,396 feet for ten females. Occasional shifts occur in hibernation dens and/or summer ranges. Skin shedding occurs at intervals that are variable for any indi- vidual. Most typically there are probably two sheddings, or at most, three, in a growing season of approximately six months in adults.

There is much diflFerence between individuals in their times of retirement into hibernation in autumn and emergence in spring, but on the Reservation the entire population ordinarily is dormant from the second week of November through mid-April. Hiber- nacula are in deep crevices in limestone outcrops along hilltops, where the exposure is to the south, east or west. Often several in- dividuals congregate in the same hibernaculum. In autumn, even after the advent of freezing temperatures at night, the snakes may continue to emerge to bask in afternoon sunshine. At times cop- perheads share their hibemacula with other species of snakes, in- cluding timber rattlesnakes, racers, rat snakes, water snakes, garter snakes, and king snakes. Copperheads are able to survive temperatures slightly below freezing but cannot survive having their body fluids completely congealed. Emergence from hibernation may occur at body tem- peratures in the neighborhood of 10° or 11° Centigrade. At tem- peratures only a few degrees lower, the snakes seem incapable of spontaneous movement and respond only to vigorous stimulation. At temperatures near freezing the snakes are completely dormant, appearing to be inert and Hfeless. Temperatures betwen 26° and 272 Untversity of Kansas Publs., Mus. Nat. Hist.

28° Centigrade (approximately 80° Fahrenheit) seem to be opti- mum. Activity is largely nocturnal and basking does not occur regularly in summer except in gravid females. Males become sexually mature in their second summer, usually many weeks before they have attained an age of two years. At sexual maturity they may be as small as 420 millimeters in snout- vent length—only half the length of a large adult and approximately one-tenth his bulk. Most females become sexually mature after their third hibernation and produce first litters when they are ap- proximately three years old, at a minimum snout-vent length of approximately 520 mm. Thereafter females normally produce Ut- ters in alternate years. The number of young is roughly correlated with size of the female. In primiparae, litters of three are common (sometimes only one or two young are born) and the largest fe- males frequently produce litters of eight or more (exceptionally 14). Courtship and copulation have rarely been observed, either in captivity or in nature. Presumably these activities are normally nocturnal. Cloacal smears from both sexes indicate that males almost always have motile sperm and that copulation may take place in any month throughout the snakes' season of activity. Per- haps copulation takes place most frequently in April soon after emergence from hibernation, and in the latter half of May, the season when ovulation occurs. The females carry their young throughout the summer. Gravid females sometimes gather in small groups. Births are concentrated in the first half of September but may occur in late August or early October. After insemination and fertilization of tlie matiure ova, sperm may remain viable in the oviducts for at least a year and may fertilize the eggs for a subsequent litter of young. At birth the sex ratio is remarkably unbalanced, with males outnumbering females by more than three to one. In gravid females that are undernourished, some of the embryos are aborted or resorbed; other embryos, although stunted, survive, and are bom as much undersized young after an abnor- mally prolonged gestation. In many instances such stunted young were known to have lived to become normal adults. Normal young are in the neighborhood of 220 mm. in length from snout to vent, and twelve grams at birth, and the males and fe- males are not noticeably different in size or proportions. The young differ from adults in the conspicuously yellow-tipped tail, a feature shared by the young of other members of the genus. Behavior suggesting the luring of prey within range, by squirming move- AUTECOLOGY OF THE COPPERHEAD 273 ments of the erect and conspicuous yellow tail, has been reported in young copperheads, but this behavior is much more strongly developed in their congeners the cantil, and the hump-nosed viper of India. Young copperheads grow most rapidly in their first two years and there is but little divergence between the sexes during this time. In the third year growth rate slows a little in the males and much more in the females. Typical snout-vent lengths for males one, two, three, four, five, six, and seven years old are, respectively, 354, 480, 560, 620, 668, 710 and 760 mm. Corresponding measure- ments for females are: 345, 450, 538, 578, 598, 626 and 643 mm. Copperheads that are fully adult may continue to grow for many years but the rate is variable and erratic. Unusually large indi- viduals of either sex are almost always more than seven years old, but the oldest individuals are not necessarily the largest. The food consists of small vertebrates and certain insects, notably cicadas and the larvae of large moths of several families. In the food of the adult snakes, voles (especially Microtus), mice (espe- and shrews are most im- cially Peromyscus) , (especially Blarina) portant, in that order of frequency. The young, still too small to prey freely on these animals, more often take small snakes, lizards, least shrews (Cryptotis), and narrow-mouthed toads (Gastro- phryne). Frogs and toads (other than Gastrophryne) are rarely eaten by copperheads of the local population studied; published reports indicate that in some other regions frogs are more important in the diet. Various birds are eaten occasionally but they comprise an insignificant percentage of the food. Exceptional items eaten turtles a and include young {Terrapene, Sternothaerus) , mantis, spiders. In summer approximately half of the copperheads caught had food remains in their digestive tracts, but only eight per cent had food in their stomachs. Because the traps were selective, catching chiefly the more active and hungry snakes, the figures are biased. Other workers, obtaining copperheads by techniques that did not include trapping, have found food remains in the digestive tracts of approximately 73 per cent of a combined sample of 249. A meal of average size (18.5 per cent of the snake's body weight) remains in the stomach three to five days before it passes into the intestine in a semi-liquefied state. Residues often remain in the intestine for as much as two weeks before evacuation. It is estimated that a typical copperhead consumes eight meals totalling approximately twice its own bodily weight in the course of a growing season. 274 University of Kansas Publs., Mus. Nat. Hist.

Presumably hundreds of copperheads die each year on the Reser- vation in the course of normal turnover of the population, but the causes of most of this mortality remain obscure. In three instances opossums were known to have fed upon copperheads, wliich the opossums may have killed. Experimental evidence bore out the suspicion that the mole may on occasion attack and kill juveniles. Other kinds of snakes including, among local species, the yellow- bellied racer and the common garter snake, sometimes prey upon young copperheads. Published records indicate that the common king snake is a natural enemy of major importance in some regions where it is abundant. Among the larger predators of the Reserva- tion, whose food habits have been investigated, the red-tailed hawk stands out as by far the most important predator on the copperhead. A total of 224 pellets of red-tailed hawks analyzed included 40 occurrences of the copperhead. An effect of a rare extreme of weather was observed at Independence Creek, Terrell County, Texas, where, in June, 1954, a hurricane with rainfall allegedly in excess of 20 inches resulted in flooding that destroyed most of the live-oak groves that were the copperhead's habitat locally. Four species of chiggers are common ectoparasites of the copperhead on the Reservation. Other known parasites include a fluke (Renifer kansensis), and nematodes (Kalicephalus agkistrodontis and Phy- saloptera squamatae). In the copperheads examined, evidence of disease was noted from time to time and especially in the summers of 1950 and 1951. In those years many individuals had necrotic patches of skin, and some were emaciated. The breeding population contains a tremendous excess of males because they are more numerous at birth and attain breeding maturity much sooner. Regardless of the method of collecting, the young up to t\vo years old are not represented in their true numbers. An annual loss close to 29 per cent in the population as a whole is indicated. In the young the mortality is a little higher than this and in the large adults it is a little less. Because of the long time required to collect a sample on any area and the impos- sibility of finding all individuals, any attempt at census must be based upon the ratio of marked individuals, and must make allow- ance for the rate of population turnover, with compensation for immigration, emigration, natality and mortality. Extent of normal movements and frequency of shifts are still inadequately known for a highly accurate census. Many different census computations were made, using different combinations of samples. The figures obtained varied over a wide range, perhaps reflecting actual changes AUTECOLOGY OF THE COPPERHEAD 275 in population density from time to time and from place to place, but probably reflecting, to a greater extent, inadequacies in the samples, resulting from small numbers or from sources of error in- herent in the method. Allowing for the probable errors produced by population turnover, it is conservatively estimated that the popu- lation on the Reservation in autumn after birth of the young slightly exceeds five per acre. No comparable data are available from any other area. Judging from the relative ease with which individuals are found by turning flat rocks in spring, the population density of the Reservation is fairly typical of other areas in northeastern Kansas having similarly favorable combinations of habitat features includ- ing woodland, meadow, brush, and south-facing exposures of fis- sured rock that crop out near the tops of liills. Certainly many areas within Douglas County and the counties adjoining it, have much higher population densities, probably exceeding ten per acre. The copperhead is to some extent hated and feared by the human population throughout its range. It is feared most where it is rare and is generally unfamiliar except as a mental image far more fear- some than the snake itself. Where the species is common, it is often accepted casually, and is feared less than some harmless snakes of more striking appearance and aggressive demeanor, which are be- lieved to be dangerous. In the region of this study, irresponsible journalism by local newspapers has done much in recent years to promote a dread of snakes by less well-informed segments of the public. Incidents involving venomous or supposedly venomous snakes are mentioned frequently by the press, almost always in a context of sensationalism, with gross exaggeration of size, venomous qualities and aggressiveness. Because of its small size, sluggish and secretive and nocturnal habits, and highly developed cryptic coloration, the copperhead has survived in areas densely populated with humans, even in the suburbs of large cities. Where it is abundant in such situations, it may constitute a hazard to small children, and should be controlled locally. Recommended control measures include removal or reduc- tion of the available food and shelter, and heavy spraying in spring

and fall, with concentrated solutions of insecticides at the crevices and fissures in rock outcrops where the snakes are known to hibernate. The sluggish habits and cryptic coloration are correlated with the development of venom glands and fangs for subduing the prey. In the United States the copperhead inflicts more bites on humans than does any other species of venomous snake, but the incidence of mor- 276 University of Kansas Publs., Mus. Nat. Hist. tality is low. Even in untreated cases, victims other than small children are almost certain to recover unless there are seriously aggravating circ*mstances. The venom of individual copperheads varies greatly and unpredictably both in quantity and in the potency of any given amount. Compared with the venoms of most rattlesnakes, of the cottonmouth, and of the Neotropical pit vipers, that of the copperhead is much less potent in its eflFect on humans, in a subcutaneous injection such as results from a typical bite, and is relatively strong in neurotoxic effects but causes less destruction of tissues. Grave sequelae, such as development of gangrene are comparatively rare. Most rattlesnakes have venoms two to six times as potent as that of the copperhead judging from the lethal doses required for mice; the venom of the South American rattle- snake is recorded to be 55 times as potent. Such comparisons are somewhat misleading and perhaps do not do justice to the copper- head. The venom kills in different ways according to the amount: by massive clotting and stoppage of circulation; by paralysis inter- fering with respiration; by gradual but cumulative damage to any one of several vital organs; or secondarily by septicemia, gas gan- grene, or other infections. The more recent improved antivenin serum manufactured and distributed by Wyeth Laboratories, Inc., is obtained from horses injected with a combination of Bothrops and Crotalus venoms of four species. Although no venom of Agkistro- don is used in its preparation, this new serum is found to be more effective against the bite of the copperhead than the original anti- venin which did rely in part on the venom of the copperhead. For- merly popular methods of treatment of snakebite—such as holding the raw flesh of a freshly killed fowl against the wound, treating the wound with kerosene, or with solution of potassium permanganate, cryotherapy, or imbibing of liquor—are all now in disrepute. Re- cently some authorities have expressed the opinion that use of anti- venin is not justified for treatment of the bites of copperheads, because of the risk of untoward reactions from the serum, and be- cause the venom is so weak that recovery is virtually assured without the aid of the serum. Other recent workers have presented evidence that the long-established treatments by use of a tourniquet to im- pede the spread of venom, combined with incision and suction to remove it, are more harmful than beneficial, and that the early use of antivenin is the only effective method of treatment. Compared with other vertebrates of similar size, snakes in general are more resistant to snake venoms. The copperhead is more than AUTECOLOGY OF THE COPPERHEAD 277 normally susceptible to the venom of its own species, and in cap- a bite. Such tivity one may occasionally be killed by well-placed occurrences are accidental, and so far as known the venom normally is used only to secure the prey or, secondarily, in defense against natural enemies. LITERATURE CITED Ahl, E. 1930. Reptilia (Kriechtiere). Tab. Biol., 6 (Suppl. 2):625-715. Allen, E. R. 1949. Observations of the feeding habits of juvenile cantils. Copeia, 1949:225-226.

Allen, E. R., and Maier, E. 1941. The extraction and processing of snake venom. Copeia, 1941: 248-252.

Allen, J. A. 1868. Catalogue of the reptiles and batrachians found in the vicinity of Springfield, Massachusetts, with notice of all other species known to inhabit the state. Proc. Boston Soc. Nat. Hist, 12:171-204.

Allen, M. J. 1932. A survey of the amphibians and reptiles of Harrison County, Mississippi. Amer. Mus. Novit., 542:1-20.

Allen, W. B., Jr. 1955. Some notes on reptiles. Herpetologica, 11:228. Allyn, W. p. 1937. Studies on poisonous snakes of Indiana. Proc. Indiana Acad. Sci., 46:220-224. Amaral, a. do 1927. Notes on nearctic poisonous snakes and treatment of their bites. Bull. Antivenin Inst. Amer., l(3):61-76. 1928. Studies on snake venoms. I. Amounts of venom secreted by nearctic pit vipers. Bull. Antivenin Inst. America, l(3):61-76. Anderson, P. 1942. Amphibians and reptiles of Jackson County, Missouri. Bull. Chi- cago Acad. Sci., 6(ll):203-220. Atkinson, D. A. 1901. The reptiles of Allegheny County, Pa. Aim. Carnegie Mus., 1:145- 157.

Axelrod, D. I. 1950. Studies in late Tertiary paleobotany. Carnegie Inst. Washington Publ. 590, 323 pp. Babco*ck, H. L. 1926. The copperhead. Bull. Boston Soc. Nat. Hist., 41:3-6. 1929. The snakes of New England. Boston Soc. Nat. Hist., Nat. Hist. Guides, No. 1, pp. 1-30. Bailey, R. M. 1941. The northern copperhead in Iowa. Iowa State Coll. Jour. Sci., 15(l):l-2. Barbour, R. W. 1950. The reptiles of Big Black Mountain, Harlan County, Kentucky. Copeia, 1950:100-107.

13—4428 278 University of Kansas Publs., Mus. Nat. Hist.

Barton, A. J. 1948. Snake litters. Herpetologica, 4:198. 1949. Ophiophagy by a juvenile copperhead. Copeia, 1949:232. Beck, H. P. 1952. Herpetological lore from the Blue Ridge. Midwest Folklore, 2(3): 141-150. Beyer, G. E. 1898. Contributions on the life histories of certain snakes. Amer. Nat., 32:17-24. BOGERT, C. M. 1941. Sensory cues used by rattlesnakes in their recognition of ophidian enemies. Ann. New York Acad. Sci., 41(5):331-343. Bonn, E. W., and McCarley, W. H, 1953. The amphibians and reptiles of the Lake Texoma area. Texas Jour. Sci., 1953(4) :465-471.

BoYER, D. A., and Heinze, A. A. 1934. An annotated list of the amphibians and reptiles of Jefferson County,

Missouri. Trans. Acad. Sci. St. Louis, 28(4) : 183-200. Branson, E. B.

1904. Snakes of Kansas. Kansas Univ. Sci. Bull., 12 ( 13) : 353-430. Brattstrom, B. H.

1954. The fossil pit-vipers ( Reptilia : CrotaUdae ) of North America. Trans. San Diego Soc. Nat. Hist., 12(3):31-46. Braun, E. L. 1950. Deciduous forests of eastern North America. The Blakiston Co., Philadelphia, xiv -f 596 pp. Brimley, C. M. 1923. The copperhead moccasin at Raleigh, N. C. Copeia, 124:113-116.

Brovitn, B. C. 1950. An annotated check list of the reptiles and amphibians of Texas. Baylor Univ. Studies, pp. xii -f 259. Buckley, E. E. 1959. Treatment of poisonous snake bites. Jour. Amer. Med. Assoc, 169(1 ):96. BtTLLOCK, T. H., and Cowles, R. B. 1952. Physiology of an infrared receptor: the facial pit of pit vipers. Sci- ence, 115:541-543. Bullock, T. H., and Fox, W. 1957. The anatomy of the infra-red sense organ in the facial pit of pit vipers. Quart. Jour. Micr. Sci., 98(2):219-234. Burger, W. L., and Smith, P. W. 1950. The coloration of the tail tip of young fer-de-lances : sexual dimor- phism rather than adaptive coloration. Science, 112(2911 ):431- 433. Burt, C. E. 1927. An annotated list of the amphibians and reptiles of Riley County, Kansas. Occ. Papers Mus. Zool. Univ. Michigan, 189:1-9. Bush, F. M. 1959. Foods of some Kentucky herptiles. Herpetologica, 15(2) :73-77. Carpenter, C. C. 1958. Reproduction, voung, eggs and food of Oklahoma snakes. Herpe- tologica, 14:113-115. AUTECOLOGY OF THE COPPERHEAD 279

Carb, W. H. 1926. The fasting of a copperhead. Copeia, 150:104. Carson, H. L. 1945. Delayed fertilization in the captive indigo snake with notes on feed- ing and shedding. Copeia, 1945(4) :222-225. Chaney, R. W. 1947. Tertiary centers and migration routes. Ecol. Monogr,, 17:139-148. Chaney, R. W., and Elias, M. K. 1936. Late Tertiary floras from the High Plains. Carnegie Inst. Washing- ton Publ., 476:1-72. Chenoweth, W. L. 1948. Birth and behavior of young copperheads. Herpetologica, 4:162. Clark, R. F. 1949. Snakes of the hill parishes of Louisiana. Jour. Tennessee Acad. Sci., 24(4):244-261. Clarke, R. F. 1958. An ecological study of reptiles and amphibians of Osage County, Kansas. Emporia State Research Studies, 7(l):l-52.

1959. Poisonous snakes of Kansas. The Kansas School Nat., 5(3) : 1-16. CONANT, R. 1938. The reptiles of Ohio. Amer. Midi. Nat, 20(l):l-200.

1951. The reptiles of Ohio. ( Reprinted with revisionary addenda. ) Amer. Midi. Nat, 20(1) :l-284. 1952. Reptiles and amphibians of the northeastern states. Published by the Zoological Society of Philadelphia, 41 pp.

CoNANT, R., and Hltdson, R. G. 1949. Longevity records for reptiles and amphibians in the Philadelphia Zoological Gardens. Herpetologica, 5:1-8. Cope, E. D. 1900. The crocodihans, lizards, and snakes of North America. Rept. U. S. Nat. Mus. for 1898, pp. 153-1270.

CORRINGTON, J. D. 1929. Herpetology of the Columbia, South Carolina, region. Copeia, 172: 58-83. COTT, H. B. 1940. Adaptive coloration in animals. Oxford Univ. Press, N. Y., xxxii -f 508 pp., 48 pis. Cowles, R. B. 1938. Unusual defense postures assumed by rattlesnakes. Copeia, 1938: 13-16. Criley, B. R. 1956. Development of a multivalent antivenin for the family Crotalidae. Pp. 373-380 in Venoms (ed. Buckley, E. E., and Porges, N., Amer. Assoc. Adv. Sci., Publ. No. 44). Crow, H. E. 1913. Some trematodes of Kansas snakes. Univ. Kansas Sci. Bull., 7:123- 134. Crum, C. W. R. 1906. Treatment of the bites of copperhead snakes. Jour. Amer. Med. Assoc, 46(19): 1433-1434.

C^TTTmS l_j 1949. The snakes of Dallas County, Texas. Field & Laboratory, 17:5-13. 280 University of Kansas Publs., Mus. Nat. Hist.

Davis, W. B. 1938. White-throated sparrow killed by copperhead. Condor, 40:183. DlTMARS, R. L. 1896. The snakes found within fifty miles of New York City. Abs. Proc. Linn. Soc. New York, 8:9-24. 1907. The reptile book. Doubleday, Page and Co., Garden City, New York, pp. xxxii -\- 472. 1910. Reptiles of the World. The MacMillan Co., New York, pp. xx -i-373. 1927. Occurrence and habits of our poisonous snakes. Bull. Antivenin Inst. Amer., l(l):3-5. 1931. Snakes of the World. The MacMillan Co., New York, pp. xi + 207, 84 pis. 1935. Serpents of the northeastern states. New edition, published by the New York Zoological Society, pp. 1-40.

DOLLEY, J. S. 1939. An anomalous pregnancy in the copperhead. Copeia, 1939:170. Dundee, H. A., and Burger, W. L. 1948. A denning aggregation of the western cottonmouth. Nat. Hist. Misc., Chicago Acad. Sci., 21:1-2. Dunn, E. R. 1915. Number of young produced by certain snakes. Copeia, 22:37. FiNNEBAN, L. C. 1948. Reptiles at Branford, Connecticut. Herpetologica, 4:123-126. 1953. Aggregation behavior of the female copperhead, Agkistrodon con- tortrix mokeson, during gestation. Copeia, 1953:61-62. FrrcH, H. S. 1949. Road counts of snakes in western Louisiana. Herpetologica, 5:87-90. 1951. A simplified type of funnel trap for reptiles. Herpetologica, 7:77- 80. 1954. Life history and ecology of the five-hned skink, Eumeces fasciatus. Univ. Kansas Publ. Mus. Nat. Hist., 8(1):1-156. 1956a. A field study of the Kansas ant-eating frog, Gastrophryne olivacea. Univ. Kansas Publ. Mus. Nat. Hist., 8(4):275-306. 1956b. Temperature responses in free-living amphibians and reptiles in northeastern Kansas. Univ. Kansas Publ. Mus. Nat. Hist., 8(7): 417-476. 1957. Aspects of reproduction and development in the prairie vole (Microtus ochrogaster ) . Univ. Kansas Publ. Mus. Nat. Hist., 10(4):129-161. 1958. Home ranges, territories, and seasonal movements in vertebrates of the Natural History Reservation. Univ. Kansas Publ. Mus. Nat. Hist, ll(3):63-326. 1959. A pattemless phase of the copperhead. Herpetologica, 15:21-24.

Fitch, H. S., and Sandidge, L. L. 1953. Ecology of the opossum on a natural area in northeastern Kansas. Univ. Kansas Publ. Mus. Nat. Hist., 7(2):305-338. Flattery, M. 1949. An eflFective way to control snakes. Pests and their control, 17(2) : 16-18. Force. E. R. 1930. The ampliibians and reptiles of Tulsa County, Oklahoma, and vicin- ity. Copeia, 1930:25-39. AUTECOLCXJY OF THE COPPERHEAD 281

Fox, W. 1956. Seminal receptacles in snakes. Anat. Rec, 124(3) :519-540. Gehlbach. F. R. 1956. Annotated records of southwestern amphibians and reptiles. Trans. Kansas Acad. Sci., 59(3):364-372. Gentry, G. 1956. An annotated list of the amphibians and reptiles of Tennessee. Jour. Tennessee Acad. Sci., 31(3):242-251. Ghosh, B. N., and Sarkar, N. K. 1956. Active principles of snake venoms. Pp. 189-196 in Venoms (ed. Buckley, E. E., and Forges, N., Amer. Assoc. Adv. Sci., Publ. No. 44).

Gingrich, W. C, and Hohenadel, J. C. 1956, Standardization of polyvalent antivenin. Pp. 381-385 in Venoms (ed. Buckley, E. E., and Porges, N., Amer. Assoc. Adv. Sci., Publ. No. 44). GiTHENS, T. S. 1931. Antivenin: its properties and standardization. Bull. Antivenin Inst. Amer., 4(4) :81-85. 1935. Snake bite in the United States. The Sci. Monthly, August, 1935: 163-167. 1935b. Studies on the venoms of North American pit vipers. Jour. Im- munology, 29:165-173. Gloyd, H. K. 1928. The amphibians and reptiles of Franklin County, Kansas. Trans. Kansas Acad. Sci., 31:115-141. 1932. The herpetological fauna of the Pigeon Lake region, Miami County, Kansas. Papers Michigan Acad. Sci. Arts and Letters, 15:389-409, 3 pis. 1934. Studies on the breeding habits and young of the copperhead, Agkis- trodon mokasen Beauvois. Papers Michigan Acad. Sci., Arts and Letters, 19:587-604. 1947. Notes on the courtship and mating of certain snakes. Nat. Hist. Misc., 12:1-4. Gloyd, H. K. and Conant, R. 1943. A synopsis of the American forms of Agkistrodon ( copperheads and moccasins). Bull. Chicago Acad. Sci., 7(2):147-170. Golan, H. G. 1941. Snake venom and allergy. Jour. Allergy, 12:11-23.

Goodwin, O. K., and Wood, J. T. 1956. Distribution of poisonous snakes on the York-James Peninsula: a zoogeographic mystery. Virginia Jour. Sci., 7 (new series): 17-21.

GOWANLOCH, J. N. 1943. Poisonous snakes of Louisiana. Louisiana Dept. Conserv., pp. 9-62, GUIDRY, E. V. 1953. Herpetological notes from southeastern Texas, Herpetologica, 9:49-56. Haines, T. P. 1940. Delayed fertilization in Leptodeira annulata polysticta. Copeia, 1940:116-118. Haltom, W. L. 1931. Alabama reptiles. Alabama Mus. Nat. Hist., Mus. Paper No. 11, vi + 145 pp. 282 University of Kansas Publs., Mus. Nat, Hist.

Hamilton, W. J., Jr., and Pollack, J. A. 1955. The food of some crotalid snakes from Fort Benning, Georgia. Nat. Hist. Misc., No. 140:1-4. Harwood, p. D. 1933. The helminths parasitic in the Amphibia and Reptilia of Houston, Texas, and vicinity. Proc. U. S. Nat. Mus., 81 ( 17) : 1-71. Hay, O. p. 1892. The batrachians and reptiles of the state of Indiana. Seventeenth Ann. Kept., Dept. Geol. and Nat. Res. Indiana, 1891, pp. 409-602, pis. 1-3. 1893. On the breeding habits, eggs, and young of certain snakes. Proc. U. S. Nat. Mus., 15:385-397. Henry, G. M. 1924. Notes on Ancistrodon hypnale, the hump-nosed viper. Spolia Zey- lonica, 13:257-258. HiBBARD, C. W. 1936. The amphibians and reptiles of Mammoth Cave National Park, Pro- posed. Trans. Kansas Acad. Sci., 39:277-281. Hoffman, R. L. 1945. Notes on the herpetological fauna of Alleghany County, Virginia. Herpetologica, 2:199-205.

HOLBROOK, J. E. 1838. North American herpetology, or a description of the reptiles in- habiting the United States. Vol, 2, J. Dobson, Philadelphia, 123 pp. HOLLICK, A. 1936. The Tertiary floras of Alaska. U. S. D. I. Professional Papers, 182, 171 pp. Hudson, G. E. 1942. The amphibians and reptiles of Nebraska. Nebraska Conserv. Bull., No. 24, 146 pp. HtJDSON, R. G. 1954. An annotated list of the reptiles and amphibians of the Unami Valley, Pennsylvania. Herpetologica, 10:67-72.

Hughes, R. C, Baker, J. R., and Davidson, C. B. 1941. The trematodes of reptiles. Part 2, host catalogue. Proc. Oklahoma Acad. Sci., 21:37-43.

HxmTER, J. H. 1911. Herpetology of Missouri. Trans. Acad. Sci. St. Louis, 20:59-274. HtrrcHisoN, R. H. 1929. On the incidence of snake-bite poisoning in the United States and the results of the newer methods of treatment. Bull, Antivenin Inst. Amer., 3(2):43-57. 1930. Further notes on the incidence of snakebite poisoning in the United States. Bull. Antivenin Inst. Amer., 4(2):40-43. HXJTCHISON, V. H. 1956. An annotated list of the amphibians and reptiles of Giles County, Viiginia. Virginia Jour. Sci., 7 (new series): 80-86. Hyland, K. E., Jr. 1950. The copperhead snake as a host for the chigger mite Trombicula

i Eutrombicula) alfreddugesi. Jour. Parasit., 36(5) :494. Jacques, R. 1956. The hyaluronidase content of animal venoms. Pp. 291-293 in Venoms (ed. Buckley, E. E., and Porges, N., Amer. Assoc. Adv. Sci., Publ. No. 44). AUTECOLOGY OF THE COPPERHEAD 283

Johnson, J. E., Jr. 1948. Copperhead in a tree. Herpetologica, 4:214. Keegan, H. L. 1944. Indigo snakes feeding upon poisonous snakes. Copeia, 1944:59. 1956. Antivenins available for treatment of envenomation by poisonous snakes, scorpions, and spiders. Pp. 413-418 in Venoms (ed. Buck- ley, E. E., and Porges, N., Amer. Assoc. Adv. Sci., Publ. No. 44).

Keegan, H. L., and Andrews, T. F. 1942. Effects of crotalid venom on North American snakes. Copeia, 1942:251-254. King. W. 1939. A survey of the herpetology of Great Smoky Mountain National Park. Amer. Midi. Nat., 21(3):531-582. Klaxjber, L. M. 1927. Some observations on the rattlesnakes of the extreme Southwest. Bull. Antivenin Inst. Amer., 1:7-21.

1956. Rattlesnakes. Univ. California Press, 2 vols., xxLx -\- 1476 pp. KOBA, K. 1938. Some notes on Agkistrodon halys (Pallas) from Syoryuzan Island, Kwangtung, South Manchuria. Zool. Mag., 50(5) :245-264. Lamson, G. H. 1935. The reptiles of Connecticut. State of Connecticut Gaol, and Nat. Hist. Surv., Bull. 54; 35 pp.

Leopold, R. S., Huber, G. S., and Kathan, R. H. 1957. An evaluation of the mechanical treatment of snake bite. Military Medicine, 120(6) :414-416. Liner, E. A. 1954. The herpetofauna of Lafayette, Terrebonne, and Vermilion parishes, Louisiana. Proc. Louisiana Acad. Sci., 17:65-85. 1955. A herpetological consideration of the Bayou Tortue region of Lafay- ette Parish, Louisiana. Proc. Louisiana Acad. Sci., 18:39-42.

Linsdale, J. M. 1927. Amphibians and reptiles of Doniphan County, Kansas. Copeia, 164:75-81. Livezey, R. L. 1949. An aberrant pattern of Agkistrodon mokeson austrinus. Herpeto- logica, 5:93. LooMis, R. B.

1956. The chigger mites of Kansas ( Acarina, Trombiculidae ) . Univ. Kan- sas Sci. Bull., 37, pt. 2(19): 1195-1443. Lynn, W. G. 1929. A case of delayed birth in Agkistrodon mokasen. Bull. Antivenin Inst. Amer., 2(4) :97. 1936. Reptile records from Stafford County, Virginia. Copeia, 1936:170. Malnate, E. V. 1944. Notes on South CaroUnian reptiles. Amer. Midi. Nat., 32:728-731. Martin, E. P. 1956. A population study of the prairie vole (Microtus ochrogaster) in northeastern Kansas. Univ. Kansas Publ. Mus. Nat. Hist, 8:361-416. Maslin, T. p. 1950. Snakes of the Kiukiang-Lushan area, Kiangsi, China. Proc. Cali- fornia Acad. Sci., 4th ser., 26(12):419-466. 284 University of Kansas Publs,, Mus. Nat. Hist.

McCauley, R. H. 1945. The reptiles of Maryland and the District of Columbia. Published by the author, Hagerstown, Maryland, 194 pp. Mendes, F. 1952. Snake venom allergy. Jour. Amer. Med. Assoc, 150 (13): 1328.

Milling, C. J. 1937. Is the serpent tale of Indian origin? So. Folklore Quart., 1(1): 43-55. Mills, H. R. 1952. Death in the Florida marshes. Audub. Mag., 54(5) :285-290.

Milstead, W. W., Mecham, J. S., and McClintock, H. 1950. The amphibians and reptiles of the Stockton Plateau in northern Terrell County, Texas. Texas Jour. Sci., 1950(4) :543-562.

MiNTON, S. A., Jr. 1944. Introduction to the study of the reptiles of Indiana, Amer. Midi. Nat., 32:438-477. 1951. Injuries by venomous animals in Indiana. Proc. Indiana Acad. Sci., 60:315-323.

1953. Variation in venom samples from copperheads ( Agkistrodon con- tortrix mokeson) and timber rattlesnakes {Crotalus horridus horri- dus). Copeia, 1953:212-215. 1954. Polyvalent antivenin in the treatment of experimental snake venom poisoning. Amer. Jour. Trop. Med. and Hyg., 3:1077-1082. 1956. Some properties of North American pit viper venoms and their correlation with phylogeny. Pp. 145-151 in Venoms (ed. Buckley, E. E., and Porges, N., Amer. Assoc. Adv. Sci., Publ. No. 44).

Mitchell, J. D. 1903. The poisonous snakes of Texas, with notes on their habits. Trans. Texas Acad. Sci., 5, pt. l(2):21-48. Morse, M. 1904. Batrachians and reptiles of Ohio. Proc. Ohio State Acad. Sci., 4, pt. 3, special paper No. 9:93-144. MosKi, H. 1954. A large litter of copperheads (Agkistrodon contortrix mokeson). Copeia, 1954:67. Mxn

Neill, W. T., and Allen, E. R. 1956. Secondarily ingested food items in snakes. Herpetologica, 12: 172-174. Netting, M. G. 1939. Hand list of the amphibians and reptiles of Pennsylvania. Bien- nial Rept. Pennsylvania Fish Comm. for 1936-1938:122-132.

Noble, G. K., and Clausen, H. J. 1936. The aggregation behavior of Storeria dekayi and other snakes with especial reference to the sense organs involved. Ecol. Monogr., 6:269-316. AUTECX)LOGY OF THE COPPERHEAD 285

Oliver, J. A. 1955. The natural history of North American amphibians and reptiles. D. Van Nostrand Co., Inc., New York, xi + 359 pp. 1958. The not-so-frightful copperhead. Animal Kingdom, 61(2);40-46.

Oliver, J. A., and Goss, L. J. 1952. Antivenin available for the treatment of snakebite. Copeia, 1952: 270-272.

Orth, J. C. 1939. Moth larvae in a copperhead stomach. Copeia, 1939:54-55. Parker, M. V, 1948. A contribution to the herpetology of western Tennessee. Jour. Tennessee Acad. Sci., 22(l}:20-30. Parrish, H. M. 1956. Early excision and suction of snakebite wounds in dogs. Pp. 399- 404 in Venoms (ed. Buckley, E. E., and Porges, N., Amer. Assoc. Adv. Sci., Publ. No. 44). Parrish, H. M., and Pollard, C. B. 1959. Effects of repeated poisonous snakebites in man. Amer. Jour. Med. Sci., 237(3), No. 1043:277-286. Pawling, R. O. 1939. The amphibians and reptiles of Union County, Pennsylvania. Her- petologica, 1:165-169. Perkins, C. B. 1955. Longevity of snakes in captivity in the United States as of January 1, 1955. Copeia, 1955:262. Pope, C. H. 1935. The reptiles of China. Amer. Mus. Nat. Hist., Natural Hist. Cent. Asia, vol. 10, lii -f 604 pp. 1937. Snakes ahve and how they live. Viking Press, New York. Pp. 238. 1955. The reptile world. Alfred A. Knopf, New York, xxv -f 325 pp. Forges, N. 1953. Snake venoms, their biochemistry and mode of action. Science, 117(3029):47-51. Ratinesque, C. S. 1819. Natural history of Scytalus Cupreus, or the Copper-head Snake. Amer. Jour. Sci., 1(18): 84-86. Reese, A. M. 1926. The venom of new-bom copperheads. Science, 53(1631 ):357. 1947. The hemipenes of copperhead embryos. Herpetologica, 3:206-208.

Robertson, W. B., and Tyson, E. L. 1950. Herpetological notes from eastern North Carolina. Jour. Elisha MitcheU Sci. Soc, 66(2): 130-147. Say, T. 1819. Notes on herpetology. Amer. Jour. Sci., l(2nd ed., part 10):256- 265.

Schlenker, J.

1942. [Letter re. cats and copperheads.] Nat. Hist. 49( 1 ) :60. Schmidt, K. P. 1945. The girl who had never been bitten by a rattlesnake. Chicago Nat., 8(2):30-3L 1946. On the zoogeography of the Holarctic region. Copeia, 1946:144- 152. 286 University of Kansas Publs,, Mus. Nat. Hist.

Schmidt, K. P., and Inger, R. F. 1957. Living reptiles of the world. Hanover House, Garden City, New York, 287 pp., many pis. Shannon. F. A. 1956. Comments on the treatment of reptile poisoning. Pp. 405-412 in Venoms (ed. Buckley, E. E., and Porges, N., Amer. Assoc. Adv. Sci., Publ. No. 44). Shaw, C. E. 1948. The male combat "dance" of some crotalid snakes. Herpetologica, 4:137-145.

1959. Longevity of snakes in the United States as of January 1, 1959. Copeia, 1959:336-337. Sibley, C. G. 1955. Behavioral mimicry in the titmice (Paridae) and certain other birds. Wilson Bull., 67(2): 128-132. Smith, A. G. 1940. Notes on the reproduction of the northern copperhead, Agkistrodon mokasen cupreus (Rafinesque), in Pennsylvania. Ann. Carnegie Mus., 28(5):77-82. 1945. Agkistrodon mokeson mokeson (Daudin) in Pennsylvania. Proc. Pennsylvania Acad. Sci., 19:69-79. Smith, H. M. 1956. Handbook of amphibians and reptiles of Kansas. Univ. Kansas Mus. Nat. Hist. Misc. Publ., 9:357 pp., 232 figs. Smith, H. M., and Sanders, O. 1952. Distributional data on Texan amphibians and reptiles. Texas Jour. Sci., 1952(2):204-219. Smith, M. A. 1943. Reptilia and Amphibia. Vol. IIL Serpentes in The fauna of British India including Ceylon and Burma. Taylor and Francis, London, xii + 583 pp. Smith, P. W. 1953. Some facts about Illinois snakes and their control. State of Illinois Dept. Registration and Education, Nat. Hist. Surv. Div., Biol. Notes No. 32:1-8.

Smith, P. W., and List, J. C. 1955. Notes on Mississippi amphibians and reptiles. Amer. Midi. Nat. 53 (No. 1):115-125. Stabler, R. M. 1939. Frequency of skin shedding in snakes. Copeia, 1939:227-229. Stadelman, R. E. 1928. The poisoning power of the new-bom copperhead with case report. Bull. Antivenin Inst. Amer., 2(3):67-69. 1929. Further notes on the venom of new-bom copperheads. Bull. Anti- venin Inst. Amer., 3(3) :31. Stahnke, H. L. 1953. The L-C treatment of venomous bites and stings. Arizona High- ways, 29:35. Stanic', M. 1956. Allergenic properties of venom hypersensitiveness in man and ani- mals. Pp. 181-188 in Venoms (ed. Buckley, E. E., and Porges, N., Amer. Assoc. Adv. Sci., Publ. No. 44 ) . AUTECOLOGY OF THE COPPERHEAD 287

SxEpjEGER, L. 1895. The poisonous snakes of North America. Rept. U. S. Nat. Mus., 1893(73):339-487. Stewaht, N. H. 1929. Reptiles of Pennsylvania. Bucknell Univ., Lewisburg, Pennsylvania, pp. 1-16. Stone, W. 1906. Notes on reptiles and batrachians of Pennsylvania, New Jersey and Delaware. Amer. Nat, 40(471): 159- 170.

Strecker, J. K. 1925. Reptile myths in northwestern Louisiana. Texas Folklore Soc, 4:44-52.

1935. Notes on the pit-vipers in McLennan County, Texas. Baylor Bull., 38(3):26-28. Surface, H. A. 1906. The serpents of Pennsylvania. Pennsylvania State Dept. Agri., Monthly BuU., Div. Zool. 4(4 and 5):114:208. Sutherland, I. D. W. 1958. The "combat dance" of the timber rattlesnake. Herpetologica, 14(l):23-24. SWANSON, P. L. 1946. Effects of snake venom on snakes. Copeia, 1946:242-249. 1952. The reptiles of Venango County, Pennsylvania. Amer. Midi. Nat., 47(1):161-182. SwAROop, S., and Grab, B. 1956. The snakebite mortality problem in the world. Pp. 439-466 in Venoms (ed. Buckley, E. E., and Porges, N., Amer. Assoc. Adv. Sci., Publ. No. 44).

SWARTZWELDER, J. C. 1950. Snake-bite accidents in Louisiana with data on 306 cases. Amer. Jour. Trop. Med., 30(4):575-589. Trapido, H. 1939. Partxrrition in the timber rattlesnake, Crotalus horridus horridus Linne. Copeia, 1939:230. Trowbridge, A. H. 1937. Ecological observations on amphibians and reptiles collected in southeastern Oklahoma during the summer of 1934. Amer. Midi. Nat., 18(2):285-303. Uhler, F. M. 1944. Facts about snakes. U. S. Dept. of Interior, Fish and Wildlife Serv., Wildlife Leafl. 257:1-10. Uhler, F. M., Cottam, C, and Clarke, T. E. 1939. Food of snakes of the George Washington National Forest, Virginia. Trans. 4th American Wildlife Conf., Amer. Wildlife Inst., Wash- ington, D. C, 1939:605-632. Welter, W. A., and Carr, K. 1939. Amphibians and reptiles of northeastern Kentucky. Copeia, 1939: 128-130.

Werler, J. E. 1950. The poisonous snakes of Texas and the first aid treatment of their bites. (Texas Game and Fish, February, 1950), Texas Game, Fish and Oyster Comm. Bull., 34:1-40. 1951. Miscellaneous notes on the eggs and yoimg of Texan and Mexican reptiles. Zoologica, 36(1): 37-48. 288 Untversity of Kansas Publs., Mus. Nat. Hist.

Werler, J. E., and McCallion, J. 1951. Notes on a collection of reptiles and amphibians from Princess Anne County, Virginia. Amer. Midi. Nat, 45(l):245-252. WiLLSON, P. 1908. Snake poisoning in the United States: A study based on an analysis of 740 cases. Arch. Int. Med., l(5):516-570. Wolfe, N. O., and Githens, T. S. 1939. Yield and toxicity of venom from snakes extracted over a period of two years. Copeia, 4:234.

Wood, J. T. 1954. A survey of 200 cases of snake-bite in Virginia. Amer. Tour. Trop. Med. and Hyg., 3(5):936-943. Woodbury, A. M. 1951. Symposium: A snake den in Tooele County, Utah. Introduction— A ten-year study. Herpetologica, 7(1):1-14. Wright, A. H. 1950. Common names of the snakes of the United States. Herpetologica, 6(6): 141-186. Wright, A. H., and Wright, A. A. 1957. Handbook of snakes of the United States and Canada. Comstock Publ. Associates, Cornell Univ. Press, vol. 2, pp. i-ix and 565-1106.

ZozoYA, J., and Stadelman, R. E. 1929. Hypersensitiveness to snake venom proteins. A case report. Bull. Antivenin Inst. Amer., 3(3):93-95.

Transmitted July 14, 1960.

D 28-4428

Autecology of the Copperhead BY (2024)
Top Articles
Latest Posts
Article information

Author: Domingo Moore

Last Updated:

Views: 5291

Rating: 4.2 / 5 (73 voted)

Reviews: 80% of readers found this page helpful

Author information

Name: Domingo Moore

Birthday: 1997-05-20

Address: 6485 Kohler Route, Antonioton, VT 77375-0299

Phone: +3213869077934

Job: Sales Analyst

Hobby: Kayaking, Roller skating, Cabaret, Rugby, Homebrewing, Creative writing, amateur radio

Introduction: My name is Domingo Moore, I am a attractive, gorgeous, funny, jolly, spotless, nice, fantastic person who loves writing and wants to share my knowledge and understanding with you.