REMOTE STORAGE PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE Aug 1 4 21113 !——— 2/17 203 Blue FORMS7DateDueForms_ZOI‘/.indd - 99.5 ABSTRACT POPULATION DYNAMICS AND ECOLOGY OF THE PAINTED TURTLE, CHRYSEMYS PICTA by J. Whitfield Gibbons A quantitative and descriptive analysis was made of the population dynamics and ecology of the painted turtle, Chrysemys picta, in southwestern Michigan. Individual growth rates, pOpulation structure, seasonal activity, and environmental conditions were determined for one population. Reproductive cycle and reproductive potential were deter- mined for another. Integration of these data resulted in a quantitative and descriptive life history of the species. One study p0pu1ation was in a marsh containing about 15 acres of open water. Survivorship and population size and structure were determined by a mark-release-recapture program over a two and one—half year period. More than 1000 Chrysemys were marked. The following information was re— I corded for each captured individual: date, locality in the marsh, sex, length, and age when possible. An indication of seasonal activity of turtles was Obtained by trap records and field observations. Growth rates were determined by age-length relationships and by recaptures. J. Whitfield Gibbons Reproductive cycles and reproductive potential were determined by dissection of 293 turtles from a population in a lake located two miles from the marsh. The two habi- tats are not connected by water. The male reproductive cycle is divisible into three phases. From late March to early May (the mating season) the testes are small but the sperm ducts are filled with sperm. Spermatogenesis occurs from July to October. This results in an increase in testis size. The winter phase is characterized by small testes. Most of the sperm are contained in the epididymis at this time. Transitional periods occur between each of these phases. Courtship be- havior occurs in spring and fall, but effective mating probably does not occur in the fall. Ovarian follicles begin to increase in size in September. By March the largest follicles are around 16mm in diameter. Immediately before ovulation these increase to about 18mm in diameter. Ovulation occurs around the middle of May and enlarged follicles, as well as the ovi- ducal eggs, are invariably present at this time. Presum- ably a second clutch is laid. The second ovulation probably occurs in middle June. Mean clutch size is about 6.5 eggs and the annual reproductive potential for an individual is around 13 eggs. Growth rates in the marsh population are similar throughout the juvenile years but decrease abruptly in both J. Whitfield Gibbons sexes when maturity is reached. Growth rate apparently de- creases to some degree as mature turtles grow older. The survivorship pattern, based on the study popu- lations, is a combination of a Type IV curve (egg-juvenile) and a Type III (adult). The former is characterized by high mortality in the egg stage (98 percent) followed by low mortality of the juveniles. Once maturity is reached by either sex, mortality is constant. Adults become active in early spring when water temperatures reach the minimum tolerance limits. Repro- duction is believed to be the incentive for this early activity. Distances traveled by mature males are greatest in the spring. By late May most Chrysemy§_become associated with areas of aquatic vegetation. Travel within the marsh is limited throughout the sUmmer. Juveniles and mature males are found in the same general area in successive summers whereas mature females have often traveled long distances. This is attributed to‘females leaving the water in one area for nesting purposes and returning to the water in another area. By September the turtles begin to leave summer feeding areas which are unsuitable for winter. Activity at low temperatures is reduced in the fall. 0, Copyright by James Whitfield Gibbons 1967 POPULATION DYNAMICS AND ECOLOGY OF THE PAINTED TURTLE, CHRYSEMYS PICTA BY Jg‘Whitfield Gibbons A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Zoology 1967 \J ACKNOWLEDGMENTS I wish to thank my guidance committee chairman, Dr. Marvin Max Hensley, and the other committee members, Dr. John A. King, Dr. T. Wayne Porter, and Dr. W. E. Wade, for offering suggestions before and during the period of, study and for critically reviewing the original manuscript. I thank Dr. T. Wayne Porter for suggesting Sherriff's Marsh as a study area and Dr. Walter F. Morofsky for assisting me in obtaining permission to conduct studies there. Mr. Fred Sherriff of Hickory Corners, Michigan, kindly consented to allow me free access to the area. Mr. Roswell D. Van Deusen permitted extensive collecting in Wintergreen Lake and other aquatic areas on the Kellogg Bird Sanctuary. Dr. William E. Cooper gave statistical advice concerning certain parts of the work. I am very grateful to Dr. Charles S. Thornton and' the Department of Zoology of Michigan State University and to Dr. George H. Lauff and the W. K. Kellogg Biological Station of Michigan State University for equipment, sup- plies, and laboratory space necessary for the study. A Sigma Xi Grant-in—Aid of Research provided field and photo- graphic equipment during part of the study. ii Special thanks go to Mrs. Bernadette Henderson for many services rendered during the period of study. Mr. Arthur Wiest was a valuable aid in locating and construct- ing field equipment.' His help is greatly appreciated. I am deeply indebted to many individuals for their assistance on field trips. Without their help a great pro- portion of the field data could not have been collected. I thank the following: R. L. Anderson, L. Babcock, D. R. Baker, Dr. D. A. Blome, M. Blome, R. G. Burbidge, H. T. Conley, R. L. Conley, Mrs. C. C. Gibbons, Miss L. E. Gib- bons, R. N. Glick, R. Hayworth,-Dr. M. M. Hensley, J. L. Hesse, W. C. Johnson, J. W. W. Johnson, Dr. J. A. King, Dr. A. W. Knight, R. M. Kocan, S. Mathews, D. L. McGregor, M. Myszewski, S. Nelson, D. R. Osborne, R. Pfohl, N. Shura- leff, B. A. Smith, Dr. D. W. Tinkle, D. Van Deusen, B. Van Deusen, Dr. W. E. Wade, and Dr. R. G. Wiegert. I am particularly grateful to Mr. Don L. McGregor for the many hours he spent assisting me in the field and in discussing aspects of the problem at the initiation and during the study. Finally, I thank my wife, Carolyn, for the essential role which she played in bringing this work to completion. My grandparents, Mr. and Mrs. J. Whitfield Moore, and my wife's parents, Mr. and Mrs. H. T. COnley, offered encour- agement throughout the study.- A debt of gratitude is due my mother, Mrs. Janie M. Gibbons, for guidance and under- standing which made it possible to undertake this work. iii TABLE OF CONTENTS INTRODUCTION MATERIALS AND METHODS The Study Species . . . Description of the Study Areas Sherriff's Marsh Wintergreen Lake Kalamazoo River Water Temperature and Depth Collecting Methods Hand or Dip Net Swim-in Traps Bait Traps Basking Traps Underwater Diving Terrestrial Traps Hatchling Traps Muddling . Seining Shooting . . Road Collecting Lures . Marking,Measuring and Aging . Head Counts . . . . . . . . Determination of Reproductive Cycles Growing Season . . . Nesting Experiments . Determination of Population Size Natality . . . Selection Against Marked Animals Growth During the Study > Mortality . . . . . . Migration . Behavior of Marked Turtles iv Page Random Redistribution . . . . . . . . . 26- Inherent Error . . . . . . . . . . . . . 27 RESULTS . . . . . . . . . . . . . . . . . . . . . 28 Physical Parameters . . . . . . .1. . . . . .' 28 Reproduction . . . . . . . . . . . . . . . . 31 Size and Age at Maturity . . . . . . . . 31 Seasonal Changes in Males . . . . . . . 34 Seasonal Changes in Females . . . . . . 37 Mating . . . . . . . . . . . .‘ 41 Reproductive Potential .4. . . .’. . . .' 41 Growth Rates . . . . . . . . . . . . . . . . 42 Population Dynamics . . . . . . . . . . . . . 48 Population Size and Density . . . . . . 48 Sex Ratio . . . .~. . . 52 Age Structure and Survivorship . . . . . 52 Mortality . . . . . . . . . . . . . . . 57 Seasonal Activity . . .-. . . . . . . . . . . 60 Aquatic . . . . . . . . . . . . . . . . 60 Terrestrial . . . . . . . . . . .'. . . 72 Local Travel in Sherriff's Marsh . . . . . . 74 DISCUSSION . . . . .1. . . . .1. . . . . . . . . . 79 Reproduction . . . . . . . . . . . . . . . . 79 Size and Age at Maturity . . . . . . . . 79 Seasonal Changes in Males . . . . . . .‘ 81 Seasonal Changes in Females . . . . . . 82 Mating . . . . . . . . . . . 84 Reproductive Potential . . . . . . . . . 85 Growth Rates . . . ._. . . . . . . . . .'. . 87 P0pu1ation Dynamics . . . . . . . . . . . . . 89 Sex Ratio . . . . . . .' 89 Age Structure and Survivorship . . . . . 89 Mortality . . . . . . . . . . . . . . . 93 Seasonal Activity . . . . . . . . . . . . . . 94 Aquatic . . . . . . . . . . . . . . . . 94 Terrestrial . . . . . . . . . . . . . . 97 Local Travel in Sherriff's Marsh . . . . . . 100 SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . 103 LITERATURE CITED . . . . . ._. . . . . . . . . . . 107 Table 1. 10. 11.‘ 12. 13. LIST OF TABLES Relationship between maturity and plastron lengths of Chr sem s picta males from Sher- 5 1n I9§6 . . . . . .‘. riff's Mars . . . . . Seasonal changes in average number and size of large follicles of Chrysemys picta from Wintergreen Lake. . . . . . . . . . Results of.£ tests for pairs of observations in Table 2. . . . .‘. . . . . Results of nesting experiments during summer of 1966 . . . Growth of Chrysemys picta in Sherriff's Marsh . . . . . . . . Recapture ratios of Chrysemys picta from Sherriff's Marsh. . . . . . . . . . . Population size of Chrysemys picta in Sher- riff's Marsh. . . . . . . . . . . . . . Number of juvenile Chrysemys picta in Sher- riff's Marsh. . . . Population size structure of Chrysemys picta in Sherriff's Marsh . . . . . . . . Age structure of Chrysemys picta population at Sherriff's Marsh . . . . . . . . Chrysemys picta found dead in Sherriff's Marsh . . . . . . . . . . . . Percentage of Chrysemys icta captured in swim- in traps in Sherrif arsh during 1966. o o o o ' Chrysemys picta collected in swim-in trap located in the outlet of Wintergreen Lake during spring of 1965 vi Page 32 38 39 43 48 50 51 53 55 58 59 61 63 Table 14. 15. 16. 17. 18. 19. 20. 21. 22. Chrysemys pictacollected in the swim- in trap in'ZOne 5 of Sherriff's MarSh during 1965. . . . . .g. . . . . . . . . . . . . Chrysemys p1cta captured in three swim- in traps in entrances to Zone 5 in Sherriff's Marsh during 1966 . . . . . . . . . . Chr sem s icta captured in baited traps at Sherr1f¥'s Marsh during 1965. . . . . . . . Head count observations of Chr sem s icta from three areas in Sherriff's Marsh Euring 1965. O O O O O O O O O O O O O O O O . O O IO Head counts of Chrysemys picta in four Zones in Sherriff‘s Marsh on July 25, 1966. Records of Chr sem s icta2 found on land from May 2, 1965, to une 1966.. . . Chr sem s icta which changed zones in Sherriff's Marsh and traveled more than 100 meters. . ‘ .fi. . . . . . . . . Chrysemys picta from Sherriff's Marsh which were recapturEH in the same or an adjacent zone each time (0) compared with those which were recaptured at least once in a non-adjacent zone (N). Chrysemys picta from Sherriff's Marsh re- cordéd as traveling more than 400 meters vii Page 64 65 67 68 69 73 75 78 Figure 1. 10. 11. 12. LIST OF FIGURES Sherriff's Marsh . Zoning system at Sherriff's Marsh . . . . . Wintergreen Lake . . . . . . . . . . . . . Relationship between Wintergreen Lake and Sherriff's Marsh . . . . . . . . . . . . Water level changes at Sherriff's Marsh . Water temperature at Sherriff's Marsh Relationship between plastron length and maturity of female Chrysemys picta . . Relationship between plastron length and weight of the testes in Chrysemys picta Age-length relationships of Chrysemys picta from Sherriff's Marsh, not including mature males . . . . . . . . . . . . . . . . . . Relationship between age and plastron length of juvenile Chrysemys picta from Sherriff's Marsh . . ... .7. . . . . . . . . Relationship between age and plastron length of mature males from Sherriff's Marsh . Age-length.relationship of Chrysemys picta from Sherriff's Marsh based on an estimated growth rate of l mm/year . . . . . . .~. . viii Page 12 13 29 30 35 36 45 46 47 56 INTRODUCTION Cagle (1953) points out that most ecological studies on reptile populations have not met the “critical standards" necessary for formulating ecological principles and advanc- ing ideas. He further indicates that the numerous facts con- cerned with reptile ecology are of such a diverse nature that they can seldom be satisfactorily integrated. The same is true today, particularly in regard to freshwater turtles. Thorough natural history studies have been made on a few species of North American turtles, namely Sternothaerus odoratus (Risley 1933), Gopherus agassizi (Woodbury and Hardy 1948), Pseudemys scripta (Cagle 1950; Webb 1961), Ter- rapene ornata (Legler 1960), and Graptemys pseudogeographica (Webb 1961). Several authors (Cahn 1937; P0pe 1939; Breckenridge 1944; Conant 1951; Carr 1952; Smith 1961) have included brief accounts of the natural history of Chrysemys picta in general treatises of tdrtles. These accounts provide little. quantitative information regarding the ecology and life his- tory of the species.‘ Because of its widespread distribution and abundance, however, Chrysemys picta has been the subject of a number of studies involving particular aspects of its natural history. Pearse, Lepkovsky, and Hintze (1925), Raney and Lachner (1942), Marchand (1942), and Lagler (1943) investi- gated the food habits of Chrysemys. The general consensus was that the species is omnivorous. Pearse (op. cit.) and his cohorts also reared laboratory individuals on various diets and concluded that growth was influenced by the vita- min content of the fobds eaten. Studies on home range and migration in turtles have been particularly pOpular. Such studies on Chrysemys (Pearse 1923a; Cagle 1944a; Williams 1952; Sexton 1959a) have formulated a general picture that this species con- fines its activity to'a restricted area if conditions are suitable but will travel great distances if such conditions do not exist. Sexton (1959a) investigated the environmental factors influencing movement and determined the seasonal activity cycle of the species. The presence of surface vegetation has been hypothesized as being the important factor in determining the suitability of an aquatic area (Meseth and Sexton 1963). Ortleb and Sexton (1964) experi- mented with several factors which might be important in Chrysemys orientation. Limited information has been collected concerning growth rates by Pearse (1923b) and Cagle (1954) who point out that juveniles grow much faster than adults. The grow- ing season of-Chrysemys in southern Michigan was defined by Sexton (1965) as being June 1 to August 31. Cagle (1954) gives an account of the life cycle of Chrysemys by consolidating the-previous literature and sup- plementing this with personal observations. He states that there is a need for more information concerning natality, mortality, attainment of maturity, and population densities of turtles. The remainder of the information on the natural his- tory of Chrysemys picta is to be found in short reports of casual observations. A systematic investigation of the life history and ecology of the species has not been previously accomplished. In reCent years, increased interest in population ecology has placed an emphasis upon studies of natural p0pu- lations of animals. Numerous animal species have been in- vestigated in attempts to analyze their population dynamics and to relate these changes to the ecology of the species. A successful analysis of p0pulation dynamics, however, is dependent upon the qualities of the species studied. Emi- gration and immigration often change the complexion of the p0pulation. Ages of individuals are often difficult to determine. Dead animals are seldom_found, so that few clues are given as to the causes of mortality in most species. Also, the investigator himself often becomes a major en- vironmental factor. Chrysemys picta has many attributes which make it a favorable study species in investigations of the dynamics of natural populations. Chrysemys are usually found in dis- tinct, well defined pOpulations. Aging is possible to a much greater extent than is true of most other animals and size can be accurately defined by measuring plastron lengths. Information on mortality can be gained because of the dura- bility of turtle shells after death of the individua1.~ Chrysemys can be permanently marked for individual identi- fication without injuring the animals or distrubing the population. The sexes can be distinguished externally. The two objectives of the present study were (1) to investigate the natural history of the painted turtle, Qh_y— semys picta, and (2) to analyze the population dynamics of a natural population of animals. The study quantitatively analyzes the life cycle and population dynamics of Chrysemys picta in southwestern Michigan, and provides descriptive information concerning the ecology of the species. Reproductive cycle and repro- ductive potential were determined for the turtles from one population. Individual growth rates, population structure, seasonal activity, and environmental conditions were deter- mined for another p0pulation. Integration of data from these two sources results in a quantitative and descriptive life history of Chrysemys picta. MATERIALS AND METHODS The Study Species Chrysemys is a genus of New World, aquatic, emydine turtles generally considered to contain a single species, Chrysemys picta. The genera Pseudemys and Trachemys are regarded as being most closely related to Chrysemys (Carr 1952), and have been included as subgenera of Chrysemys (McDowell 1964). Chrysemys picta is composed of four rec- ognized subspecies. The present study was conducted within the range of g. p. marginata Agassiz. Description of the Study Areas The major portion of the study was conducted at two localities, Sherriff's Marsh and Wintergreen Lake, in Kalamazoo County, Michigan. In addition, observations were made on a few turtles from the Kalamazoo River in Allegan County. Sherrifffis Marsh Sherriff's Marsh (Figure l) is located in the north- west corner of Kalamazoo County, Michigan, in Sections 2, 3, and 11, T15, R9W. The marsh is comprised of more than 75 acres of a grass-sedge association and approximately 14 acres of open w O ater 5 Figure l. Sherriff's Marsh “coco cocoa ROAD/1""000300 ° Fishing Lake 6 '-————| 0 I00 m E] Open water Iii. Marsh Ea Cultivated lan —-+- -f--L— —+—-—+— -+- E] Trees The primary water source is Hamilton Lake, a hards water lake located one mile to the east. A fast-flowing stream from Hamilton Lake enters the southeast corner of the marsh, whereupon it widens and becomes a winding, slow- moving channel. The channel narrows again at the northwest corner of the marsh where it connects with Augusta Creek. A small dam was constructed across the channel in the north- west corner in the early 1930's.' This raised the water level several centimeters throughout the marsh and resulted in the death of a stand of tamarack along the eastern border. The vast majority of the marsh consists of grasses and sedges growing in soggy, often flooded areas away from the main channel. The roots of the plants form a dense mat. The Open water in the marsh consists of approximately a mile of winding channel and its extensions, a 5-1/2 acre lake, a one acre "pond," and the entrance of a small stream from the north. All of these bodies of water are interconnected. The Open water areas were divided into 15 zones to facilitate reference to particular locations in the marsh (Figure 2). With the exception of Zones 3, 4, and 6 in which the places for division were arbitrarily picked, all zones represent habitats which are distinct from the ad- joining zones. The area north of the marsh was also desig- nated as a zone (15). The main channel varies from 10 to 20 meters in width. The bottom is silt and not solid in most areas. Figure 2.--ZOning system at Sherriff's Marsh O ’6 O I. “ 0 4 O. J O \ . . O ‘ 3 o. O '0 l a .' . . . . , g I o o 5 I I ' ' a .‘ 2 - - - » 0 O O o o I O . a o 0 ‘ o- O I . .. O - .’ I o .0 I " O .O I 0‘ . O’. O o . O - 17 o O . 'e a O O I . i f 14 The water-depth was measured from the top of the silt layer to the surface of the water and ranged from one to two meters in depth in the main channel through the marsh. At two places the channel widens into larger bodies of water (Zones 8 and 11) which are 80 and 100 meters at their widest diameters. Zones 2 and 9 are side extensions of the main channel and are similar to it in most respects except for being narrower (6-8 meters) and shallower (1/2 to 1 meter). Surface vegetation in the form of Lemna and Spirodela is much more prevalent in these side channels than in the main channel. Aquatic vegetation is similar in all parts of the main channel. Coontail (Ceratophyllum demersum) is the dominant submergent. This species reaches such a density that by midsummer it occupies most of the aquatic habitats in the marsh. Clumps of Chara sp. occur occasionally, but Ceratophyllum demersum is by far the most abundant aquatic plant. Duckweeds (Spirodela pg1yrhiza and Lemna minor) form a surface mat in many places, particularly along the shore. Water lilies (Nymphaea odorata and Nuphar advena) are abundant. The most common emergents are cattails (Typha.latifolia), swamp loosestrife (DecodongvflrtiCillatUS), arrow arum (Peltandra virginica), bulrushes (Scirpus validus), and arrowhead (Sagittaria latifolia). Several species of sedges (Carex) and grasses (Gramineae), the most common being Calamagrostis canadensis, occur along the margins. 10 The pond (Zone 5), although connected directly with the main channel, is vegetatively distinctive. Marginal vegetation is similar to that in the rest of the marsh. Ceratgphyllum demersum is the dominant submergent, but does not reach the density of that in the channel. The apparent reason for this is that Spirodela and water meal (Wolffia punctata) cover almost 100 percent of the water surface by July, thus preventing underwater plants from receiving ade- quate light. Presumably the pond's unbroken layer of Spirodela and Wolffia results from the wind protection af- forded by clumps of trees to the north and east and by a row of bushes along the small stream to the west. The length of the pond is 110 meters; the width varies from 15 to 50 meters. The depth is less than one meter. The bot- tom, in contrast to the rest of the marsh, is relatively solid. The stream entrance (Zone 7) is only a small area of the marsh_but should be mentioned because of its phy- sical and vegetational differences from the main channel. Near the road the stream is less than 30 centimeters deep and one meter wide. The area where it enters the channel is still water, 10-30 centimeters in depth, and often a few degrees colder than the main channel in the summer. Spirodela covers up to 50 percent of the surface in the summer.. A large Chara bed occupies the northwest corner. Ceratophyllum is present in insignificant amounts. The ll stream entrance is margined by two parallel ridges of dirt, each about one meter high. Red osier dogwoods (Cornus stolonifera) and swamp roses (Rosa palustris) are common on these ridges. Fishing Lake (Zone 14) is notably different from the remainder of the marsh. Its larger size and greater depth (up to four meters) precludes the density of vegeta- tion characteristic of the main channel. Most of the aquatic vegetation, primarily CeratOphyllum and bladder- worts (Utricularia vulgaris), is restricted to the margi- na115 meters. The middle portion of the lake is devoid of plants. The short channel connecting Fishing Lake and the main channel is completely filled with lily pads during the summer. Wintergreen Lake Wintergreen Lake (Figure 3), on the W. K. Kellogg Biological Station of MichiganState University, is located in Section 8, TlS, R9W, Kalamazoo County, Michigan. The lake is two miles southwest of the western edge of Sher- riff's Marsh. The two habitats are not connected by water (Figure 4). Wintergreen Lake is highly eutrophic with organic bottom sediments.’ Nymphaea odorata and Nuphar advena are common around the margins. CeratOphyllum demersum and pondweeds (Potamogeton pectinatus) are the dominant sub- mergents but do not attain the density that is found in the 12 Figure 3.--Wintergreen Lake outlet SWALES Lagoon O IOOM 13 Figure 4.--Map showing relationship between Wintergreen Lake and Sherriff's Marsh in northeastern corner of Kalamazoo County, Michigan. Duck Lake \‘ Wintergreeo ake Gull \ Lake Scale 1 cm = 640 m *————b 1 kilometer AUGUSTA CREEK Lawrence oLake O Sherriff's Marsh Stony Lake 14 marsh. The lake prOper is approximately 20 acres in extent with a maximum depth of slightly less than seven meters. A narrow channel On the western edge connects the lake with several elongate swales which were cleared in 1961 to provide nesting areas for geese. These swales, though containing water throughout the year, are seldom more than a meter in depth. On the southeastern edge of the lake and connected directly with it by two narrow channels is an 80 x 30 meter "lagoon." The depth is 1-2 meters. Although fences in the channels and on the land around the lagoon isolate it from the rest of the lake, limited interchange between the lake and lagoon turtles occurs overland through broken areas in the fence. Kalamazoo River A few of the Chrysemys used in the study were taken from a polluted stretch of the Kalamazoo River in Section 21, TIN, R12W, Allegan County, Michigan. Most of the tur- tles were collected from a shallow mud flat in an area of sparse aquatic vegetation. The river in this area is pol- luted by industrial and domestic wastes from the city of Kalamazoo only 15 miles upstream. 15 Water Temperature and Depth Water temperature at the marsh was taken with a hand thermometer at the start of each collecting trip. The temperature was always taken at the same spot in the chan- nel at a depth of 30 centimeters. On occasion readings were made at other locatiOns in the marsh and at several depths for comparative purposes. Water depth was deter- mined on each collecting day by reading a meter stick placed in the water near the dock area. CollectingMethods Twelve techniques were used in capturing turtles from Sherriff's Marsh and Wintergreen Lake. The relative merits for each of these will be discussed separately. Hand or Dip Net The most successful collecting method throughout the study was simply looking for turtles in the water and then catching them either by hand or by the use of a dip net. A wooden-handled net with 1/2 inch mesh near the rim grading into 1/4 inch mesh at the tip was found to be suit- able for capturing turtles of all sizes. At Wintergreen Lake most of the turtles collected by this method were in the water along the edge of a walkway. In the marsh, the majority of the collecting was done from a rowboat. More than 1000 of the turtles collected durinnghe“5tudy were captured by this technique. 16 Swim-in Traps Traps constructed of three-inch wire mesh also were successful in collecting turtles. The body of the trap consisted of a three-foot diameter, five-foot high cylinder standing on end. The bottom was of wire mesh but the tOp was Open. A one-foot'quare was cut out of the cylinder about six inches from the bottom. The square was bordered by wire mesh extending six inches into the trap. Traps were placed along shore in shallow water. Accompanying each trap was a wire drift fence six feet long and two feet high. One end was placed at the side of the door away from shore, and the fence extended aWay at a 45° angle to the shore. The result was a funnel formed by the shore and the fence.‘ Turtles traveling parallel to shore entered the trap through the open door. Thirteen of these traps were used in the marsh. Most were never moved from their original locations. The traps have the advantage of not requiring any maintenance, such as baiting, yet they are always operative. Since the tOp is Open, there is no chance of a turtle drowning should the trap not be checked at regular intervals. The major disadvantage of these traps is that they are restricted to use in shallow water near shore, and also, the use of three— inch mesh undoubtedly allows smaller turtles to escape. 17 Bait Traps Several bait traps, constructed of 1/2 inch mesh galvanized hardware cloth, were used at the marsh. The traps were two feet long and one foot square with funnels at each end.' Sash cord Was tied at the top sothat the f trap could belowered to any depth and the cord tied to some object along shore. Dead fish or chicken necks, placed in a hardware cloth bait container, served as bait. Basking Traps A trap of this type was used to a limited extent at Wintergreen Lake. The trap consisted of a three—foot square wire mesh basket with attached floats. A short wooden plank was connected to each side in such a way that it extended about a foot into the water on the outside of the basket. The other end of each plank was elevated slightly over the inside of the basket. Turtles crawling onto the plank to bask would often enter the water off the elevated end and become trapped in the basket. Underwater Diving This technique used at the marsh was extremely suc- cessful under certain conditions. A rubber dry suit, fins, an underwater face mask, and a snorkel (or SCUBA) were used during certain times of the year for capturing turtles. By this method the investigator could swim along until a turtle 18 was seen. The fins enabled one to swim faster than most Chrysemys. This technique was used to best advantage in the spring and fall when the water was clear and vegetation at a minimum. It was also found to be an excellent method of approaching basking turtles and catching them before they entered the water. Terrestrial Traps In an attempt to capture females during the egg lay- ing period a terrestrial drift fence was used. A funnel~ composed of two eight-inch-high fences of hardware cloth each about 30 feet long extended from the water's edge onto land at a 45° angle. At the apex a space of 6-8 inches be- tween the two fences led into a circular, fenced-in area about three feet in diameter. The opening was bordered by hardware cloth extending six inches into the enclosed area. Female turtles leaving the water to lay eggs were funneled into the circular area. These traps proved to be valuable in providing information concerning nesting. HatchlingTraps A drift fence was used to determine when hatchlings leave the nest in the spring. Fences were 100 feet long and constructed of eight-inch-high hardware cloth placed parallel to the shoreline. Two-pound coffee cans with the Open end level with the surface were embedded in the ground 19 beneath the fences at 20-foot intervals. Hatchlings encoun- tering a fence would follow it until they fell into one of the cans. This technique proved successful for the purpose for which it was designed. Muddling This technique, employed by most turtle investiga- tors at one time or another, consists of feeling by hand through mucky areas until a turtle is found. In the pre— sent study, the greatest success was attained in November of 1965 in Wintergreen Lake. Attempts at muddling in the marsh were seldom rewarding. Seining A 20-foot seine was used on one occasion at Winter- green Lake. Since only three turtles were caught in several hauls, and since operation of a seine requires two people, this technique was not employed during the remain- der of the study. Shooting A pellet gun was used on one occasion in an attempt- to shoot basking turtles at Wintergreen Lake. Usually, when one got close enough to shoot, he was also close enough to run into the shallow water and catch the turtle with a dip net. There also is the disadvantage that specimens might be damaged when hit with the pellet. 20 Road Collecting Many turtles were picked up during the spring and fall on highways in the area or on the roads bordering Win- tergreen Lake. Although many specimens were obtained in this manner, the chances of finding turtles on any given day made planned terrestrial collecting unprofitable. Lures Near the completion of the study it was discovered that some Chrysemys would follow a fishing lure being pulled over the top of thick aquatic vegetation. Several individuals actually followed the lure up to the boat and’ were captured. Whether this technique could be used for extensive collecting was not determined. Marking,_Measuring, and Aging Turtles in Sherriff's Marsh were marked for indivi- dual identification and usually released a few minutes after capture. Notching the marginal scutes with a small rectangular file was found to be the most efficient method ofmarking adult Chrysemys.. Cutting v-shaped notches with fingernail clippers was most effective on the flexible shell of juveniles. Though marked marginals occasionally bled, no permanent damage was known to occur as a result of marking. The notches on turtles marked in 1964 and recap- tured in 1966 were not noticeably different from notches on 21 turtles that had been marked for less than a month. It was not necessary, therefore, to remark individuals and it is assumed that all previously marked turtles were recognized as such when they were recaptured. As an aid to determining the distance traveled by marked individuals without making a capture each time, let- ters of the alphabet were painted down the center of the carapace with white enamel. ‘In this way the letters on basking or swimming individuals could be read without hav- ing to capture the turtle. This method was very effective in early spring. With the onset of dense vegetation, little open water was present in Which to see swimming turtles and those basking often had carapacial algae which obscured the letters. This method of marking was not permanent, since Chrysemys shed their scutes annually. Plastron length was used as a measure of size and was taken to the nearest millimeter with a plastic ruler. Age was determined by the number of annuli present on the left pectoral plate. Estimates of age in older individuals- using the method given by Sexton (1959b) were not made, be- cause variability in the length of annuli was too great. All ages given in this studyyjtherefore, are based ongthe actual number of annuli. 22 « :‘V‘£-; Head Counts -, ‘h.~. ' I The number of Chrysemys heads at the surface were counted in selected areas of Sherriff's Marsh during 1965. The boat was stopped 30-40 meters from the area to be ob- served, and the heads counted by using binoculars. Each area was 54 square meters. A record was kept of the sur- face vegetation present in the area at the time of each observation. Determination of Reproductive Cycles The study specimens were usually dissected within 24 hours after their capture. The testes were removed, blotted-with paper towels, and weighed to the nearest one hundreth gram on an Ohaus Triple Beam Balance. Follicles were measured individually at the widest diameter with Vernier calipers. The average diameter of follicles in the largest set was determined only if a distinct size class was apparent. The anatomical criterion for maturity in males was the presence of testes weighing more than 0.01 g. The cri- terion utilized for ascertaining female maturity was the presence of follicles greater than 10mm in diameter (Cagle 1954). Long foreclaws is a good indicator of maturity in male Pseudemys scripta (Cagle 1948). In the present study, 86 Chrysemys males were examined internally and all mature 23 individuals possessed long claws (third claw greater than 10mm in length). With one exception, all immature males had short claws (third claw less than 9mm in length). The presence of elongated front claws was therefore considered to be dependable enough to be used as a criterion for dee termining maturity in males from Sherriff's Marsh. Growing Season Sexton (1965) considered that most of the growth in Chrysemys from southeastern Michigan-occurred from June through August. Mark-recapture results during the present study have shown that at least some growth occurs during May and September. Individuals from the study populations are considered as having one growing season (May through September) per year. NestinggExperiments Twelve mature females, collected in the spring of 1966, were used in controlled nesting experiments. Seven of these were kept in outdoor pens which were placed in an area with soil type similar to that in which local Chrysemys often dig their nests. The five remaining females were kept indoors in large aquaria cOntaining sand to a depth of seven centimeters. 24 Determination of POpulation Size The Lincoln Index (Allee et al., 1949) was used to estimate the number of individuals in the different age, sex, and size classes. The following formula was used. Number of individuals X Number of individuals which were already (marked and unmarked) marked at the beginning which were captured of the collecting period during the collecting period Number of marked individuals captured during the collecting period Each size-sex category was estimated separately. The final estimate for each category was derived by taking the average of the estimates for six collecting periods. Estimates ob- tained by this method may be subject to error for a number of reasons. The possible sources of error in estimating the size of the present population are listed and evaluated as follows. Natality New turtles were added to the population in the spring of 1965 and 1966. Hatchlings of 1965 were treated as part of the juvenile population in deriving an estimate. Hatchlings of 1966 were not included in making an estimate of the juvenile population. The reason for this is that all estimates prior to the summer of 1966 were based only on the turtles which were in the population at the start of the summer of 1965. Including the 1966 hatchlings which 25 appeared in the final collecting period would have led to an underestimation of the total juvenile population in 1966. Selection Against Marked Animals No marked turtles in the present study were known to be harmed or disabled by being marked. The methods of marking, therefore,are assumed to have had no influence on an individual's chances of remaining in the population. Growth During the Study A few turtles passed from immature to adult cate- gories during the census period. Such individuals were classed in their original category for purposes of esti- mating population size. This caused a small degree of overestimation in the juvenile category and underestimation in the adult categories. The effect on final estimates was negligible. Mortality Death of turtles during the study is assumed to have had little effect on estimation of numbers. Death is not a factor of error in estimating numbers for any given collecting period, since marked and unmarked turtles have equal Opportunities to die. 26 Migration Migration of turtles to or away from Sherriff's Marsh was probably minimal and had little effect on esti- mation of pOpulation size. Some interchange undoubtedly. occurred between turtles in the study area and those in Zones l6, l7, and 18 in whiCh no collecting was done. 1. Movement of turtles between these zones and the study area would result in an overestimation of the population size. Behavior of Marked Turtles A possible source of bias was that marked animals may be less likely to be captured than unmarked animals. Evidence that some Graptemys and Pseudemys become more wary after being captured and released is presented by Tinkle (1958). Sexton (1959a) found that unmarked Chrysemys often allowed the collector to get closer in a boat than did pre- viously marked individuals. Behavior of this type was not observed in the present study, but if it did occur, an over- estimation of-pOpulation size would result. Random Redistribution One of the assumptions in using the Lincoln Index is that a previously marked individual has as much chance of being collected as any other individual in the popula- tion. This assumption cannot be met in a natural population of animals. Although turtles were collected throughout the 27 prescribed study area, certain areas were collected more extensively than others. Hence, a greater proportion of turtles were marked in some zones than in others. However, during each of the six collecting periods on which the estimation was based, some amount of collecting was done in each part of the marsh. The error caused by non-random redistribution was therefore minimized. Inherent Error The Lincoln Index is based on chance. Even if it were possible to sample completely at random, the estima- tion of population size would probably still be aberrant from the true population size. Each estimate therefore must be considered as a general assessment of population size, not as a calculated number of individuals. Neverthe- less, a knowledge of the approximate size of the population is useful and imperative when dealing with the dynamics of an animal community. RESULTS Physical Parameters Fluctuations in water level at Sherriff's Marsh from March 1965 to November 1966 are shown in Figure 5. A rise in water level began in August 1965 and continued into March 1966. The lowest level for each year was reached in July but was 33 cm higher in 1966. Water temperatures varied between 1965 and 1966 (Figure 6). Thawing was earlier in 1966, occurring in the middle of March, whereas in 1965 the ice did not begin melting until the first week in April. Temperatures re- mained several degrees higher in 1966 than in 1965 until the first week in April when a cold spell sent water tem- peratures near the freezing mark. During the last half of April, temperatures were similar for both years. Water temperatures taken in the first half of May were consistently higher in 1965 than in 1966. The incongruity of the dates on which readings were taken in late May and early June makes comparison of the two years difficult. However, temperatures averaged in the low 20's during this period in both years. Water temperatures were relatively constant during June, July, and August but were generally 3-4° C higher in 28 29 12 ec> PU) P< “*3 -w "'2 -< '-z -—m "h “'9 L-z r-o “U3 -<: “r5 “r5 ‘2 -< ~22 .coma monso>oz Op mama zone: Eouw nmhmz m.wwfippocm um momcmco Ho>oH houmz mo mommpo>m xazucoz --.m opsmfim (°mo ut) Tenet 1919M 30 4— .gmsez n.mmHnnngm pa tom m ocON mo Hoccmnu came :H Eu om mo canoe m we manumuomsou young--.o ousmflm (3°) ainieiadmai 1919M 31 1966. Temperatures dropped considerably in late September of 1965. Readings were taken on too few dates in the fall of 1966 to compare with 1965, though presumably water tame peraturespfollpwed a_similar trend.' Most of the marsh was frozen by late December of both years. Reproduction Size and Ageat Maturity Eighty-two Chrysemys picta males from Wintergreen. Lake and four from the Kalamazoo River were dissected-and examined for purposes of determining the annual reproduc- tive cycle and the size or age at which maturity is attained. All Chrysemys males with a plastron length exceeding 81 mm were mature (weight of testes > 0.01g). Two specimens mea- suring 75 mm and one with a plastron length of 79 mm provedv to be immature. Since mature males can be distinguished by the pos- session of long claws, this secondary sex characteristic, rather than dissection, can be used to indicate the size at which maturity is reached in a pOpulation. Table 1 gives the percentage of turtles between 70 and 85 mm recorded as mature males in the marsh population. Since, assuming a 1:1 sex ratio, half of the animals would be expected to be males, the paucity of those below 80 mm recorded as males is presumably a result of there being few mature individuals. 32 Table l.--Re1ationship between maturity (based on long fore- claws) and plastron lengths of Chrysemys picta males from Sherriff's Marsh in 1966. Sample Plastron % Recorded as Males Size Length on basis of foreclaw (mm) length 5 70 0 7 71 29 9 72 0 2 73 O 8 74 0 3 75 0 5 76 0 10 77 20 7 78 14 3 79 0 7 80 29 8 81 63 3 82 67 7 83 57 9 84 56 8 85 88 33 Regression equations were determined for the growth of immature individuals from the lake and marsh populations. The 95 percent confidence interval for the regression co- efficient of the lake population is 3.55 1 0.69; that for the marsh pOpulation is 1.96 i 0.15. The growth rate is ap- proximately 18 to 26 mm per year for the lake turtles and 9 to 10 mm per year for those from the marsh. Since hatch- lings are usually close to 25 mm in plastron length, males from the lake attain a length of 80 mm in their third to fifth year, whereas males from the marsh are in their Sixth or seventh year. A total of 207 female Chrysemys picta from Winter- green Lake was dissected to determine the reproductive cycle and the size at which maturity is attained. As with the males a comparison was made between females from Winter- green Lake and those from Sherriff's Marsh to establish if reproductive maturity was dependent upon size and not age. Four females 118 to 125 mm in plastron length from the marsh had oviducal eggs present as determined by palpa- tion during June 1-7, 1966. Seven females which were not carrying eggs were 90 to 100 mm in length and seven to 12 years old. Although no large samples of Chrysemys were re- moved from the marsh pOpulation, two females were dissected in August 1966. These were 103 and 106 mm in plastron length and nine and 11 years old. Both had immature ovaries. Six mature females from Wintergreen Lake ranging in size from 116 to 125 mm were seven to 10 years old. 34 All dissected females with plastron lengths of at least 117 mm were mature (follicles > 10 mm in diameter) or showed signs of follicular enlargement (Figure-7). All specimens be- low 112 mm in plastron length showed no development of the ovaries. Four individuals (44 percent) of nine ranging from 111 to 115 mm and three (15 percent) of 20 from 116 to 120 mm were considered immature. Of 26 females with plastrons measur- ing 121 to 125 mm, only one individual (4 percent) was not mature. Seasonal Changes in Males Testis size and development are dependent upon size of the turtle and season of the year (Figure 8). At least two .seasonal phases of testicular development were Evident. In the spring phase (March-June) testes were reduced in size though the sperm ducts were often enlarged, thickened, and heavily packed with sperm. In the summer phase (July-September), testes were greatly enlarged but sperm ducts were small and lacked sperm. Three specimens collected in October tended to have testes similar in size to those of the spring phase, whereas testes of three specimens examined during the last week in June approached more closely the summer phase. Three individuals dissected during the first week in January had testes with‘ weights corresponding to the spring phase. The sperm at this time were contained primarily in the epididymis. Enlargement of the sperm ducts was most prevalent _ in April. Fifteen, of 16 males dissected, had the vasa deferentia filled with sperm. During the first half of May 35 .ucoswumaco mamsvfi>fivaw ensue: mu Hmasuwaaom oEom use mOAHm>o vomoHo>op xaamfiupmm nu“: mOHSHmEEH mu moahm>o pedoHo>opas :ufiz mOHSHmEEH mm AEEU nausea couummam CNH cad ocH mm _ -_- e _ 8.2 E i omH P H -2 «yuan meomanu onEom mo xuflHSOmE can :HMCOH dehumaam coozumn magmaofiumaom--.n ousmfim sIenptAtput go laqmnN Weight of testes (g) 36 Figure 8.-—Re1ationship between plastron length and weight 1. 1. of the testes of Chrysemys icta from Wintergreen Lake. Regression equation for July-September is y = -l85.2 + 2.55X. Regression equation for March-June is y = r48.2 + .69X. :1_- O JUIy-September . 0 June (last week) 0— Q October . O1 .9__ A January 0 March-June . .8— 9 .7- P . . 9 A5— 9 A .5— e 0 o 4 '0 e . -4 O O o 9 0A 0 .3_4 C) OC>C> O O o Q 8 “Bo z§é% .2- 3 O O O o o .o a ' <9 0 .1— e 0 oo ' C. .0 O 0 I I I I j 70 80 90 100 110 120 Plastron length (mm) 37 only three out of nine animals had enlarged ducts. All in- dividuals had small vasa deferentia during the latter half of May and June and testes were still small. All Specimens examined from July to October possessed small ducts. Seasonal Changes in Females Mature females were collected in every month from March to October. No specimens were available from Novem- ber through February. A distinct set of large follicles was present in 103 specimens so that determination of average number and size of follicles was possible during different seasons (Table 2). The t test was used to determine if sig- nificant differences existed between appropriate pairs of observations (Table 3). The 5 percent level of significance was used. The number of largest follicles when oviducal eggs were present did not vary significantly in either year from the number present after egg laying. The average follicle size, however, was significantly smaller after egg laying. In 1965 the number of large follicles in females prior to ovulation did not differ significantly from the number in females containing oviducal eggs. The follicles did increase in size. Both number and size of large folli- cles increased significantly in females collected in late September and October compared to those collected in July through early September. 38 any RN .nuo N.eH w.R w -HN .eem may flog mm Assn w .eem menses H.MH H.NH m.~ e.m e SH -HH sash .-m.sfisw was ennw< AOL Amy ea mesa e .cmh Hammond o.RH o.RH m.e m.e A OH .4 mesa -eN Re: ammo Heusen>o Amy mm as: w.e w.e m .ea As: Afiuudeoumom .HOHQENHQ GNOZ adv—Z .oxmq :ooumuopcfiz Eonw muuwm m>Eomanu mo moHOfiHHom omhmfi mo ONHm use Hones: ommuo>m ca momamnu Hmeommom--.~ manna 39 Table 3.--Results of t tests for pairs of observations in Table 2. The 5 percent level of significance was used. Significant Significant Pair of Difference Difference Observations Between Average Between Average Tested Number of Follicles Diameter of Follicles 1965A 1965B No Yes 1965B 1965C No Yes 1965C 1965D Yes Yes 1966A 1966B No Yes 1966B 1966C Yes No 1966A 1966C Yes Yes 1966C 1966D No Yes 4O Follicle diameters increased significantly in the week before ovulation began in 1966. There was no signi- ficant change in the number of follicles. Conversely, in the following two weeks, when ovulation had occurred, the number of large follicles decreased significantly but there was no change in follicle size. Two female Chrysemys collected on May 14, 1965, had ovulated, as compared to only one of three collected May 23, 1966. No additional specimens were examined until May 26 in 1965, but following that date all mature females contained oviducal eggs or corpora lutea. Seven females dissected be- tween May 16 and 21 apparently had not ovulated in 1966. Egg deposition in 1965 (eight females observed) commenced on June 5 and continued until June 23 while in 1966 the observed nesting period lasted from June 10 to July 3 on the basis of five individuals recorded. These egg lay- ing dates may be extended to May 28 and July 5 in 1966, since a specimen collected at the earlier date had corpora lutea whereas one at the later date still contained oviducal eggs. Corpora lutea were present in all of 13 mature fe- males examined during June, 1966, but only one of four had corpora lutea in July. When oviducal eggs were present, the number corresponded to that of the large corpora lutea in all but one instance. Usually the corpora lutea were 4 to 5 mm in diameter. In five turtles a second set was pre- sent. These were 1 to 2 mm in diameter. 41 Mating Instances of sexual behavior were recorded as they were encountered in the field.) Sixteen observations were made; one in March, six in April, five in May, one in August, and three in October. In most instances (14), the activity involved titillation between males and females of larger size; however, on two occasions (March and April) males were observed chasing other males and attempting titillation. Two observations (May) involved attempts by males to restrain the female by biting her plastron. Reproductive Potential The average number of oviducal eggs was 6.6 in 1965 and 6.1 in 1966. There was no significant difference be- tween the two years. More than 60 percent of the females contained either six or seven oviducal eggs. Only two of 41 individuals had less than five eggs and only two had more than eight. The relationship between size of the female and number of oviducal eggs was tested in 41 Chrysemys. The 23 turtles less than 130 mm in plastron length (mean length of those in sample) averaged 6.2 oviducal eggs. Those 130 mm or above had an average of 6.7 eggs per individual. A t test was run and there was no significant difference between thw two size classes at the 5 percent level (t - 1.2 with 39 d.f.). 42 During the egg laying season of 1966 nesting experi- ments were conducted with captured animals in an attempt to determine if individuals lay more than one clutch per-year. Abnormal egg deposition occurred in all of the turtles, presumably as a result of the experimental conditions. None of the animals dug nests, but, instead, eggs were deposited on the surface. In most instances eggs were laid sporadi- cally, usually no more than one or two per day. A record was kept of the number of eggs laid and on July 30, all of the turtles were dissected. Oviducal eggs and follicles were counted and follicle diameters were determined in 10 individuals. The results are shown in Table 4. The average number of eggs laid was 5.7. The total number (eggs laid plus those in the oviducts) averaged 8.7. This was significantly higher than the average clutch size (6.6) of wild females from 1965. Large follicles ranging from 12 to 17 mm and averaging about 14 mm in diameter were present in nine of ten individuals. When the number of large follicles was added to the number of eggs, a total of 13 to 15 was reached. The individual without enlarged fol- licles had_a total of 11 eggs. Growth Rates Determination of juvenile growth rates in Sherriff's Marsh was particularly effective because of the large number 43 Table 4.--Resu1ts of nesting experiments during summer of 1966. Numbers in parentheses are averages per turtle in cages having more than one individual. Number of Total Number Number Oviducal Number Total of Eggs of Eggs Eggs on of Large Number Plus - Turtle Laid July 30 Follicles_ of Eggs Follicles l 7 0 6 7 l3 2 6 5 0 11 ll 3 9 0 5 9 l4 4 9 0 — 9 - 5 (4.6) 4 8 (8.3) (15.0) 6 (4.6) 3 6 (8.3) (15.0) 7 (4.6) 4 6 (8.3) (15.0) 8 (4) 4 4 (9.0) (13.7) 9 (4) 7 3 (9.0) (13.7) 10 (4) 4 7 (9.0) (13.7) 11 (5.5) 3 4 (8.0) - 12 (5.5) 2 - (8.0) - f Averages 5.7 4.0 C N 5.4 8.7 13.7 44 which could be aged. Growth rate could also be determined by measuring the increase in size of individuals recaptured during the period of study. The mean plastron length, range, and standard de- viation for each age class from one to 11 are given in Figure 9. Males over 80 mm in plastron length are not_in- cluded. Plastron length increased steadily through the 10th year. The 12 turtles in their 11th to 13th years were only slightly larger than those in their 10th year. Ex— cluding turtles in their 11th to 13th years, the increase in mean size in successive years varied from 13.0 mm be- tween the first and second years to 4.3 mm between the fifth and sixth. The average increase in mean size in successive years from one to 10 was 7.7 mm/year. Regression equations were determined for the 187 juveniles one to four years old and for 31 mature males (Figures 10 and 11). The 95 percent confidence interval for the regression coefficient is 1.96 1 .15 for the juve- niles and .203 1 .07 for the males. Average growth rates determined by this method are 9 to 10 mm per year for the young turtles and about 1.1 mm per year for the mature males. On the basis of recaptures, females 110 mm or more averaged 1.1 mm per year, mature males 1.2 (Table 5). Com- bining recapture data for both sexes, the average growth rate was 1.1 mm/year. The trend in both sexes was for the Plastron length (mm) 45 Figure 9.--Age-length relationships of Chrysemys picta from Sher- riff's Marsh, not including mature males. Vertical lines represent range; horizontal lines represent mean; 120 110 100 90 ‘80 70 60 50 40 30 bars represent standard devi ation. Numbers in paren- theses are sample sizes. F” +— _. T74) (3) __, (5) __ 1" (9) (9) ‘7 (11) -1 5) -1 (28) .LJZZ) _T-HA F- LJCSZ) (35) .- ”l (47) F (59) .14 (29) 1 2 3 4 5 6 7 8 9 16 £1 12 13 Age in years 46 Figure 10.--Relationship between age and plastron length of juvenile.Chr sem s icta from Sherriff's Marsh. Regression equatlon 15 y = 33.2 + 1.96X. 80 -— 30008 00 oo \‘ OA 0 70 00 Plastron Length (mm) Growing season (May-September) Plastron length (mm) 47 Figure 11.--Re1ationship between age and plastron length of 80 mature males from Sherriff's Marsh. Regression equation is y = 82.1 + .203X. ITIIIIQIIIIIII1FIFI 5 6 7 Growing season (May-September) 48 H.H N.Hw Nm he 00 HMHOH mOHmEom can monz N.H N.Hm mm oN mN Hmuoh e. w.mH o m NH HHH-ooH m. m.m m m m mm-om m.N ¢.NH wN NH NH mm-ow mOHmz H.H o.om mm 5N mm Hmuoy o.H w.mN mN NH AH mmH-oNH H.H «.mH NH m NH mHH-mHH N.H w.OH mH A w eHH-OHH monEmm monsummoom HEEV nuzohu auzouo peanummuom omcmm ONHm memo» n commom cam mOHSHQmu o>HumHsenu mcHzonm honesz wcHzouo\Eev HmchHho coozuom Honesz num:OH.:H. mcommom mcHSOHu omMOHucH ommpo>< mo Honesz O>HumHzesu .commom MOHSOHM map mchzw mxme mu cmflu OHoE Houmm OOHSHQmOOH mHmst>chH do women swam: m.meHHoam cH muuwm mNEomxmcu mo :HZOHw--.m OHnmh 49 larger adult size classes to show less growth than smaller ones over comparable lengths of time. Population Dynamics Population Size and Density From July 1964 to October 1966, 1010 Chrysemys were marked at Sherriff's Marsh and 408 recaptures were made on 258 marked individuals. The percentage of recaptures gener- ally increased during successive collecting periods from early 1965 to late 1966 (Table 6). An estimate was made of the number of Chrysemys in each size-sex category (Table 7). More than 1400 immature turtles, composing approximately 59 percent of the total population, were estimated (Lincoln Index) to be present in Sherriff's Marsh. The estimated number of mature turtles was 925. Actual captures were 521 (52 percent) immature and 480 (48 percent) mature. Chrysemys in Sherriff's Marsh were restricted to the 14 acres of open water. The minimum density, derived from the estimate of 2328 individuals, would therefore be 166 turtles per acre. However, approximately four acres of Zone 14 were not utilized by turtles and should not be considered in determining density. Excluding these four acres, this results in a density estimate of 233 Chrysemys per acre. 50 Table 6.—- Recapture ratios of Chrysemys picta from Sher- riff's Marsh. Total Number Recapture Size-sex Category Collecting period Captured Percentage Juveniles less Apr-May 1965 34 .00 than 80 mm in Jun—Jul 71 .06 plastron length Aug-Oct 56 .09 Mar-May 1966 21 .17 Jun-Jul 108 .28 Aug-Oct 85 .30 Immature females Apr-May 1965 22 .05 80-114 mm in Jun-Jul 44 .23 plastron length Aug-Oct 32 .15 Mar-May 1966 45 .13 Jun-Jul ' 59 .21 Aug-Oct 39 .23 Mature Males Apr-May 1965 46 .04 Jun-Jul 54 .17- Aug-Oct 43 .22 Mar-May 1966 86 .22 Jun-Jul 67 .33 Aug-Oct 34 .29. Mature Females Apr-May 1965 54 .07 Jun-Jul 41 .22 Aug-Oct 36 .36 Mar—May 1966 36 .22 Jun-Jul 57 .23 Aug-Oct 29 .55' 51 Table 7.--Population size of Chrysemy picta in Sherriff's Marsh. Percentage Percentage of All Number of A11 Number Estimated Actually Captured Size-sex Category Estimated Individuals Captured Individuals Juveniles less than 80 mm in 862 .37 305 .31 plastron length Immature females 80-114 mm in 541 “.23 216 .22 plastron length Mature males 490 .21 265 .27 Mature females 435 .19 215 .22 Total 2328 1001 52 The density observed by head counts in small, heavily vegetated areas was the equivalent of more than 1000 turtles per acre on many occasions. The highest recorded density was 60 turtles in 54 square meters in Zone 8 on June 9, 1965. This was the equivalent of more than 4000 Chrysemys per acre. Sex Ratio .+ The sex ratio of mature individuals (males more than 1 80 mm and females more than 115 mm in plastron length) cap- tured in the marsh was 1.2 males:1.0 females. The ratio estimated by the Lincoln Index was 1.1 males:l.0 females (Table 7). There is no significant difference between the sex ratios determined by the two methods (x2 = 0.61). The sex ratio is not significantly different at the 5-percent level from a 1:1 ratio when based upon the number of estimated individuals (x2 = 3.3), but is significantly different when based upon actual captures (x2 = 5.2). Age Structure and Survivorship An estimate was made of the number of individuals in each of the first four age classes by using 1965 as a precensus period and 1966 as a census period (Table 8). Assuming a 1:1 sex ratio, the 1962 through 1965 cohorts were estimated to have the following numbers of turtles of each sex in 1966: 124, 69, 35, and 78. 53 Table 8.--Number of juvenile Chrysemys picta in Sherriff's Marsh. Number Marked Number Number and Re- Originally Number Total Estimated Age leased Captured Recaptured Captured to be in Age Class in 1965 in 1965 in 1966 ., in 1966 in 1966 Population 1965 l 20 41 6 47 157 1964 2 l7 l9 6 25 70 1963 3 38 39 15 54 137 1962 4 33 30 6 36 198 54 Due to difficulties in aging and complications brought about by the changes in growth rates as males neared 80 mm a satisfactory estimate of the number of in- dividuals in the 1961 age class was not made. Age could not be determined by counting annuli in most immature females over 80 mm in length. A general age-length relationship is indicated in Figure 9. Most females between 80 and 89 mm were in their seventh year, although this size range also included some individuals in older and younger age classes. The 90-99 mm size range was composed mostly of eight and nine-year-old animals. The 100-109 mm size range primarily included individuals nine to eleven years old. Essentially, then, the immature females which were 80 to 109 mm in length included almost all females eight or nine years old, more than half of those 10 or 11 years old, about two-thirds of those that were seven years old, and about one-fifth of those five or six years old. The 80 to 109 mm size range was thereby composed of immature females from the equivalent of approximately four age classes. Since 358 females were estimated to be 80 to 109 mm in length, (Table 9), this size range would average about 90 individuals per age class. The age-length relationship of adult Chrysemys from the marsh is shown in Figure 12 and is based upon a growth 55 Table 9.--Population size structure of Chrysemys picta in Sher- riff's Marsh Size range (plastron length in mm) Mature males 80- 85- 90- 95- 100- 105- 110-116 Total 84 89 94 99 104 109 Number captured 38 59 54 46 35 28 5 265 Percentage of total 14.3 22.2 20.4 17.4 13.2 10.6 1.9 Estimated number based on percentage of 490 70 109 100 85 65 52 9 490 Immature females Total 80- 90- 100- 110- 89 99 109 114 Number captured 42 42 59 73 216 Percentage of total 19.5 19.5 27.3 33.8 Estimated number based on percentage of 541 105 105 148 183 541 Mature females 115-119 120-124 125-129 130-134 135-140 Total Number captured 83 69 43 15 4 215 Percentage of total 38.6 32.1 20.2 7.0 1.9 Estimated number based on percentage of 435 168 140 88 30 8 435 M 00 Plastron length (mm) of 56 Figure 12.--Age-length relationship of Chrysemys picta from Sherriff's Marsh based on an estimated growth rate of 1 mm/year. 120.. "140 0+ 01- 110“ ‘H O ._.130 I; 5 '5 100 - 3:" d) —-120 H I: O H ‘5’. 90" m H Q.‘ —110 80 .- I I I F I I I I 5 10 15 20 25 30 35 40 45 50 Age in years 57 rate of approximately one mm per year. Consequently, each 5 mm size range should include five age classes. The age structure of the adult pOpulation can be estimated using Figure 12 in conjunction with the number of individuals in each 5 mm size range (Table 9). The age structure of the Chrysemys picta pOpulation in Sherriff's Marsh is shown in Table 10. Mortality Twenty-seven Chrysemys were found dead in the marsh during the study (Table 11). The animals found in 1964 were scattered around the periphery of Zone 5. The heads were missing on all individuals, and a few were also missing limbs. Slight scratches were present on the shells. The grass was flattened down for a half meter or so around each dead turtle. Some specimens appeared to have been dead less than a day, whereas, others had been deceased for several days. All indications were of a predator such as a mink or raccoon which operated over a period of several days. The cause of death was uncertain in the turtles from 1965 and 1966, with the exception of the ones from July 15, 1965, and August 2, 1966. Each of these carcasses was missing its head, presumably as a result of predation. The eight dead Chrysemys found on May 15, 1965, were heaped together along shore. Only the shells remained and no clue was given as to the cause of death. 58 Table 10. --Age structure of Chrysemys picta population at Sherriff's Marsh. Estimates fOr years 1 to 4 are based on a Lincoln Index of turtles in which age was known. A 1:1 sex ratio is assumed. The collective estimate for immature females 6-11 years old was derived from the age-length rela- tionships (Figure 9) and the estimated number in each size class (Table 9). Estimates of mature turtles were based on a growth rate of 1.0mm/ year for both sexes and on the number estimated to be in each size class. Males ' Females Plastron Estimated Plastron Estimated Length . Age Number/ ‘Lengthr Age Number/ in mm in 1965 Age Class in mm in 1965 Age Class 1 79 l 79 2 35 2 35 3 69 3 69 4 99 4 99 80-84 6-10 14.0 6-11 90 85-89 11-15 21.8 110-114 12-16 36.6 90-94 l6-20 20.0 115-119 17-21 34.6 95-99 21-25 15.0 120-124 22-26 27.0 100-104 26-30 13.0 125-129 27-31 17.6 105-109 31-35 10.4 130-134 32-36 6.0 110-116 36-41 2.5 135-140 37-42 1.3 59 Table ll.--Chrysemys picta found dead in Sherriff's Marsh. Number of Previously Dead Turtles Date Found Length Sex Zone Captured (mm) 1964 6 July 3 75,83,107, -- 5 No 109,115,129 3 July 3 -- -- 5 No 1 August 3 90 -- 5 No ‘1965 1 May 8 39 -- 6 No 8 May 15 70,84,94, -- 16 No 106,106, 109,112,133 1 July 15 125 F 5 No 1 October 12 116 F 8 Yes 1966 1 April 5 130 F 3 No 1 June 10 68 -- 8 No 1 July 4 58 -- 11 No 1 August 2 115 F 4 Yes 1 August 19 119 F 14 Yes 1 August 29 109 M 5 No 60 Predators were a major source of mortality on eggs in the nest.. The fate of seven of nine nests, whose con- struction was observed in 1965, was predation by mammals. Four of these were destroyed the night following egg lay- ing; the other three were not disturbed until at least two weeks later. Numerous eXamples of predation on nests which had not been previously located were noted throughout each summer. In 1966 four neSts were covered with hardware cloth immediately after the female left the nest site. -All eggs hatched successfully in two of these. In one of the other nests eight eggs did not hatch and the single hatch- ling died. Two eggs hatched successfully and two hatchlings died in the remaining nest. Seasonal Activity Aquatic Swim-in traps: The swim-in traps in Sherriff's Marsh served as indicators of the seasonal activity of Ch y- semys, based on the assumption that increased movement and activity in the population was reflected in an increase in trap captures. The ratio of the number of Chrysemys captured in swim-in’traps to the number of these traps checked was de- termined for selected periods in 1966 (Table 12). The acti- vity in March was followed by a period of inactivity in early April. The capture ratio during the second half of April was 61 Table 12.--Percentage of Chrysemys picta captured in swim- in traps in SherriffTs Marsh during 1966. Number of Traps Capture Dates Males Females Total Checked Percentage March 14-22 4 4 8 36 21 April 5-11 0 0 0 l7 0 April 15-30 13 2 15 65 23 May 3-14 8 4 12 22 55 May 18-23 0 0 0 30 0 May 25-June 10 ll 16 27 46 59 June 20-27 5 8 14* 34 41 July 4-28 7 14 21 59 36 August 2-29 7 11 18 61 30 September 7 0 0 0 10 0 October 17 0 l l 7 14 *Includes one juvenile. 62 similar to that of March. In the first two weeks of May the ratio more than doubled that of the preceding two weeks only to be followed by a week with no trap captures. Late May and early June had the highest capture ratio of the summer. After June 20 the ratio declined steadily. A single trap in the outlet channel of Wintergreen Lake in 1966 showed a trend similar to that of the marsh traps. A brief period of activity from March 17-22 was followed by three weeks in which no turtles were captured (Table 13). Another period of increased activity began on April 15. No turtles were caught in the Wintergreen Lake trap after April 22. Only one swim-in trap, placed near the eastern en- trance to Zone 5, was used extensively in the marsh during 1965. This trap had a smaller, more funnel-shaped entrance than other traps, and an enclosed top. Few, if any, turtles escaped once they had entered the trap. At least one Ch_y- sgmys was captured in this trap on each day it was checked from July 15 to August 31 (Table 14). Only three individuals were caught during 13 collecting days in September and October. Three swim-in traps were in the entrances to Zone 5 from March to October 1966. Although a few individuals were captured sporadically after the middle of March, con- sistent captures in these traps did not result until June (Table 15). During the summer months the traps in Zone 5 63 Table l3.--Chrysemys picta collected in swim-in trap located in the outlet of Wintergreen Lake during spring of 1965. Date Males Females Total March 17 0 2 2 18 0 ’ 1 1 21 l l 2 22 2 2 4 23,24, 0 0 0 25,26, 28,31 April 2,4, 0 0 0 7,11 15 1 l 3* 16 3 5 8 20 4 2 6 22 0 1 0 25,28,30 0 0 0 *Includes one juvenile 64 Table l4.--Chrysemys icta collected in the swim-in trap in Zone 5T6f erriff's Marsh during 1965. Date Males Females Total July 15 5 5 11* 16 1 0 1 l7 0 3 3 18 1 l 2 23 2 5 7. 26 0 3 3 28 2 3 5 Aug. 4 0 4 4 11 3 3 6 l4 4 3 7 21 3 3 6 23 2 4 6 24 0 1 l 28 3 4 7 31 0 0 0 Sept. 2,5, 11 12 0 0 0 21 0 2 2 23,27 0 0 0 Oct. 5 0 l 1 12,19, 23,27, 28 0 0 0 *Includes one juvenile 65 Table 15.--Chrysemys picta captured in three swim-in traps in entrances to Zone 5 in Sherriff's Marsh during 1966. Date Males Females Total March 14 0 0 0 l7 0 0 0 21 1 0 l 22 l l 2 April 5,11,15 0 0 0 17 l l 2 22 1 0 l 25,28,30 0 0 0 May 3 0 0 0 7 0 l 1 l4 0 l l 18,20,21, 23 0 0 0 25 l 1 2 June 1 l 4 5 2 1 0 l 7 1 2 3 10 l 4 5 20 0 2 2 22 0 0 0 26 l 2 3 27 0 0 0 July 4 l 3 4 6 l 4 5 15 0 0 0 16 l 1 2 18 0 O 0 25 1 0 l 28 0 l 1 Aug. 2 0 0 0 9 l 0 l 13 0 2 2 19 l 0 1 24 0 l l 29 1 1 2 Sept. 7 0 0 0 Oct. 17 0 l l 66 yielded 39 Chrysemys in 21 collecting trips. The traps were not checked extensively after August. Baited traps: During the spring of 1965, 33 Ch y- sgmys were collected in baited traps at Sherriff's Marsh (Table 16). All but two of the turtles were caught between April 29 and May 11, though many traps were set before and after these dates. Head counts: One indication of activity was the number of heads visible at the water's surface. Head counts for three equal areas of water in the marsh during 1965 are given in Table 17. The two areas in Zone 8 were within 50 meters of each other and differed primarily in that one had surface vegetation and the other did not. The area in Zone 5 lacked surface vegetation. Both the area in Zone 5 and the area with surface vegetation in Zone 8 averaged more than 15 heads per observation in June and July. The other area in Zone 8 averaged less than one head per observation in June. Head counts began to decrease during August and no turtles were observed after September 23 in any of the areas. Two heads were seen at the surface during a two hour period at the marsh on March 17, 1966, but none were observed on March 21, April 5, or April 11. Approximately a dozen heads were seen in a two hour period on April 15. Most of these animals were in masses of algae floating at the surface. Comparative head counts were made in four Zones on July 25, 1966 (Table 18). The most heads 67 Table 16.--Chrysemys picta captured in baited traps at Sher- riff's Marsh during 1965. Number Number of Total Captured Traps Trap Number of Captures Per Trap Date Set Hours Males Females Total Hour March 17 l 1 0 0 24 l l 0 0 27-28 2 42 0 0 April 24 l 29 0 0 25 l 19 0 0 29 6 35 3 2 5 .14 May 1-2 6 130 4 9 13 .07 4 l 5 0 l 1 .20 8-9 10 360 2 8 10 .03 ll 11 64 0 2 2 .03 15-16 8 168 0 0 21 2 l4 0 1 l .07 22 - - 0 l 0 - 25 l - 0 0 31 3 - 0 0 June 9 4 36 0 0 68 Table l7.--Head count observations of Chrysemys icta from three areas in Sherriff's Mar§h during 5. Zone 5 Zone 8 (Surface (Surface Zone 8 (No Vegetation Vegetation Surface Date 'Present) Present) Vegetation June 9 16 60 0 12 8 18 0 l3 8 15 0 20 18 11 0 22 A 2 4 1 26 9 15 2 29 24 .. - 30 6 3 0 July 4 21 17 - 8 23 32 - 15 40 - _ l6 8 - - Aug. 4 6 - - 24 0 - - 28 0 4 1 Sept. 2 6 - - 5 0 5 1 12 2 3 0 21 0 3 1 23 3 0 0 27 0 0 0 Oct. 5,12,13,19, 23,27,28,3l 0 o o 69 1"?" Table l8.--Head counts of Chrysemys picta in four Zones in Sherriff's Marsh on July 25, 1966. (At this time Zones 2 and 5 were completely covered with duckweeds. Zones 3 and 4 had very little sur- face vegetation.) Number of Zone Heads Counted 2 31 3 l3 4 3 5 171 70 were seen in the two zones with abundant surface vegetation. Head counts were made for all of Zone 5 on August l3, l9, and 29. These were respectively 106, less than 25, and less than 15. This shows a steady decline from the 171 of July 25. Basking: Although basking activity was not observed in a quantitative fashion, field notes taken at Sherriff's Marsh indicate certain trends. Considerably more Chrysemys were seen basking in April than in any other month. Bask- ing occurred throughout the summer and early fall but in most instances the number of turtles basking was but a small proportion of those seen in the water. Few or no turtles were seen basking in the marsh on windy days, regardless of cloud cover or air temperature. Observations of juveniles: Since smaller individ- uals were usually not collected in any of the traps, an indication of juvenile seasonal activity must necessarily be based on either observation or hand collecting. One trend which was prevalent in 1965 and 1966 was that indi- viduals in the third and fourth growing seasons were first collected in early May, two-year-olds were first collected in late May. Young of the year were neither seen nor col- lected in Sherriff's Marsh until middle June in 1965 and late July in 1966. Individuals in each age group were col- lected regularly after the dates of their first appearance until middle September of both years. 71 Observations at low temperatures: On two occasions during the study period large numbers of Chrysemys were ob- served on the bottom of Wintergreen Lake. More than 100 were seen in 10 minutes on April 6, 1965, in an area of about 600 square meters. Most of the lake was frozen at this time and turtles were sought only in the area of open water. Many disappeared into the bottom silt when the boat approached, others swam slowly away. A sample was taken of which 20 were mature males, 13 mature females, and 2 immature females. Only a single Chrysemys was seen in Sherriff's Marsh in the four-hour period preceding the observation at Wintergreen Lake. The water temperature was 4° C at both localities. A similar phenomenon to the above was observed on December 15, 1966, beneath newly-formed ice at Wintergreen Lake. The turtles behaved as they had in April. They were in the same place as before, however, the remainder of the lake was not investigated. It was not determined, therefore, if the activity was localized in that part of the lake. A collection of 21 contained 5 mature males, 14 ' mature females, and 2 immature females. Other observations of Chrysemys at water temperatures of 4° C were of single individuals on January 3 and November 29, 1965, and January 8, 1966. The bottom in three areas in the marsh was searched for turtles on October 17, 1966. An underwater suit was ,72 used and the collector probed at least 50 cm into the slit and mud. Many parts of the marsh still had thick clumps of CeratOphyllum and Chara,’and could not be‘investigated by this method. Turtles were neither seen nor collected. The following areas were examined: approximately 500 in Zone 2 5, about 300 m2 in Zone 9; and about 200 m in Zone 13. Terrestrial A record was kept of all Chrysemys seen on land in an attempt to quantitatively evaluate the periods of ter- restrial activity. Early spring travel occurred during both years (Table 19). The 10 specimens from late April and early May 1966 were represented by mature males and females and by immature individuals. After the first week in May of each year, there was a two-week interval in which no turtles were sighted on land. A few individuals were again seen during late May of both years. A total of 29 were found on land during June and July of 1965 and 1966. All of these were mature females. No Chrysemys were observed on land in late July or August. Fall travels were not recorded in 1966, but three Chrysemys were seen crossing roads in middle September 1965. Terrestrial drift fences for hatchlings were set up in early March of 1966 in five areas around Wintergreen Lake. Three of the fences were 15 meters long and two were 30. No Chrysemys hatchlings were trapped between March 13 73 Table l9.--Records of Chrysemys icta found on land from May 2, 1965, to June , 966. Mature Mature Immature Other Date Females Males Females Juveniles Total 1965 May 2-8 2 2 4 May 25-31 2 1 l 4 June l-July 9 14 14 Sept. 17-21 1 l l 3 1966 April 20-May 6 2 2 l 2 10* May 21-26 2 ' 2 June 4-20 15 15 *Includes three individuals over 80 mm in which sex and state of maturity were not determined. 74 and April 19. The first hatchling was collected on April 20, and traps were checked every day thereafter until May 7. Baby turtles were captured on April 21 (1), April 24 (3), May 5 (l), and May 6 (1). Traps were checked on seven days between May 7 and May 24, but no more turtles were collected. Local Travel in Sherriff's Marsh Less than 15 percent of the Chrysemys recaptured in Sherriff's Marsh changed zones and traveled more than 100 meters during a single summer (Table 20). There were no significant differences between the distances traveled by mature males, mature females, and immature individuals. Between successive summers, mature females changed zones and traveled distances of more than 100 meters sig- nificantly more than mature males (X2 = 7.5) or immature individuals (X2 = 6.0). Travel by mature males between successive summers was not significantly different from that by immature turtles (X2 = 1.7). Mature females also traveled interzonally more often than males or immature individuals, based on all recaptures throughout the study (Table 21). Over 30 percent of all re- captured females had traveled to a non-adjacent zone, whereas this was true for less than 20 percent of all other turtles. Thirteen of the 16 individuals recorded as traveling 400 meters or more were females, nine of them mature (Table 75 Table 20.--Chrysemys picta which changed zones in Sher- riff's Marsh and traveled more than 100 meters. Percentage Changing Number Zones and Traveling Recaptured 100 Meters During single summer Immaturee 58 12 Mature Males 19 5 Mature Females 23 13 Between successive summers Immature 70 17 Mature Males 19 5 Mature Females 21 43 76 Table 21.--Chrysemys picta from Sherriff's Marsh which were recaptured in the same or an adjacent zone each time (0) compared with those which were recap- tured at least once in a non-adjacent zone (N). Immature Mature Males Mature Females Total 0 N O O O N O N Number 95 19 66 15 44 20 205 54 Percentage 83 17 81 19 69 31 79 21 77 22). Four of the trips were made in less than three weeks time, one immature female traveling 600 meters in a maximum of eight days. The longeSt distance traveled by any turtle during the study was that of a mature female captured cross- ing the road north of the marsh on July 4, 1965, and recap- tured on August 19, 1966, a straight-line distance of more than one kilometer. Twenty males were captured two or more times in a single spring (March to early May); thirteen were taken in the same zone each time they were captured while seven had changed zones and moved 150-500 meters. One of these ani- mals, captured four times in the spring of 1965, was in a different zone each time.‘ Interzonal travel by males in the spring-time was significantly higher at the 5 percent level than travel in the summer (X2 = 5.29). III. I III III! {I ll 3' Ill.llllll I'l' Il.3.i I. ill-I ‘l i I 78 Table 22.--Chrysemys picta from Sherriff's Marsh recorded as traveling more than 400 meters. Distance Zone of. Number of Traveled Original Zone of Days Between Sex (m) Capture Recapture Captures F 1100 15 ' 14 411 J 800 3 11 327 F 760 5 14 100 F (imm) 750 5 14 307 F (imm) 620 3 9 20 F (imm) 600 2 8 8 F 500 5 2 382 M 500 9 13 13 F 470 3 8 303 F 470 5 2 743 F 460 5 2 392 F 460 5 2 710 F 460 5 2 322 F (imm) 440 5 2 409 F 400 9 13 19 M 400 7 9 49 DISCUSSION Reproduction Size and Age at Maturity Reproductive maturity in turtles is correlated with either age (Risley 1938) or with length (Hildebrand 1932; Cagle 1948, 1954). Cagle (1948) commented that male Pseud- emys scripta troosti from two Illinois populations reached maturity at about the same size although their growth rates, and hence age, were different. Tinkle (1961), on the other hand, found that Sternothaerus odoratus males from different geographical localities reached maturity at different lengths but at the same age. Whether Tinkle's findings are a result of geographic variation or a difference between sternothaerine and emydine turtles is not known. Age-length maturity relationships in Chrysemys picta from Wintergreen Lake and Sherriff's Marsh indicated that sexual maturity is correlated with length, not age, at these localities. Males from both the lake and the marsh reach maturity at,a plastron length of approximately 80 mm. The average growing seasons at which these lengths are reached differ in the two pOpulations, being the fourth year of age in turtles from the lake and the sixth year in those from the marsh. 79 80 Cagle (1954) stated that males from northern Michi- gan reach maturity at about 90 mm and those from Illinois at about 70 mm. Since the populations of the present study are geographically intermediate between Cagle's two popula- tions, the size at which maturity is attained in males may decrease southward. This possible geographical trend is strengthened by Cagle's mention of two Louisiana males in around 60 mm which were mature. However, the appearance of a geographical trend must be considered with caution since Cagle's presentation of data in scatter diagrams shows no males between 75 and 85 mm from his Michigan population and only two below 75 mm from his Illinois population. A statement cannot be made with certainty, on the basis of the data shown, that most males from these populations do not reach maturity between 75 and 80 mm, the same as Ch_yf sgmys males from Wintergreen Lake and Sherriff's Marsh. Some Sherriff's Marsh females as old as 12 years were immature whereas certain seven year old individuals from Wintergreen Lake were mature. Maturity in most in- stances coincides with plastron length within a narrow size range at both localities. The transition zone from the immature to the mature state appears to be between 110 and 120 mm. As a working definition, a plastron length of 115 mm will be considered as the size at which Chrysemys picta from this region reach maturity. This varies from results obtained by Cagle (1954) 81 who stated that Chrysemys females reached maturity at a plastron length of 120 to 130 mm. Other than possible geo- graphical differences, no explanation for this disparity is readily available. Seasonal Changes in Males Annual changes in the testes of turtles have been investigated quantitatively in Sternothaerus odoratus (Ris- ley 1938) and Terrapene carolina (Altland 1951). These in- vestigators also determined the spermatogenetic cycle. Changes in the testes of Chrysemys picta, based on the pre- sent study, are similar to those of Sternothaerus and Terrapene. Testicular changes in Chrysemys may be conveniently classed in two or possibly three seasonal categories. The testes are reduced in size from March through early June. The sperm ducts are enlarged and filled with sperm during part of this spring phase, primarily in April. By late May and June the ducts are small, presumably because mating has occurred and the sperm released. Late June is a transitional period in which the testes begin to enlarge. In Sternothaerus, Terrapene, and presumably in Chrysemys, this is a result of the beginning of Spermatogenesis. Spermatogenesis reaches a peak in the summer phase, which lasts from July through September in Chrysemys. Apparently all spermatozoa remain in the testes, it . {I 3 ll. . l III 111! 1" l I '1 .IJII I'll: 82 resulting in the extremely large size of testes and small size of the ducts during this period. October is the transitional period following the summer phase. At this time the testes are beginning to diminish in size, presumably because of the cessation of Spermatogenesis and the passage of spermatozoa into the I k .- epididymis as is true of Sternothaerus. A winter phase is suggested on the basis of three Wintergreen Lake Chrysemys which were dissected in January. This phase is apparently one of small testes and ducts with most of the sperm being contained in the epididymis. Ris- ley (1938) found that the epididymis in Sternothaerus was largest during the winter. 'Presumably there is a third transitional period when the sperm pass from the epididymis to the vas deferens in preparation for the breeding season. Seasonal Changes in Females The female reproductive cycle basically consists of enlargement of the follicles, ovulation, and egg laying. The follicles are smallest in the period from July to Septem- ber. The largest follicles at this time are about 12 to 13 mm in diameter and undoubtedly represent the potential eggs for the following May. The follicles change very little in size during the summer but in September they begin to in- crease in diameter. No specimens were available from November through February but specimens in March contained follicles 15 to 83 16 mm in diameter. The follicles remain this size until May. Eight specimens in 1966 were examined the week pre- ceding ovulation and the largest follicles were found to be almost 18 mm in diameter, greater than for any other time. Presumably, then, follicles undergo a rapid growth immediately before ovulation. During the time when oviducal eggs were present the follicles averaged 17 mm in diameter both years. These ob- servations were made in late May and early June. These en- larged follicles may represent second cluthes which would have been laid in late June or July. The fluctuation in average number of follicles which was apparent throughout both sampling years was pro- bably artificial in some instances. Determining which follicles in July and August actually belong to the largest set is often difficult due to their overall small size. Hence, certain follicles which appear much smaller than those in the largest set might increase enough in size to be ovulated the following spring. Follicular atresion has been observed in box tur- tles by Altland (1951). There is no direct evidence of atresion in Chrysemys of the present study; however, in both 1965 and 1966 the average number of oviducal eggs was smaller than the average number of enlarged follicles in the spring. The lower number of eggs is possibly the result of some follicles not ovulating properly. I91: .. ||lll~n.lr 1 |.| _ E ‘ IE I ll. '1 .. 84 Ovulation occurred as early as the second week in May. Most females had ovulated by June. An idea of how long the eggs are retained after ovulation is afforded by comparing known egg laying dates with the times when ovi- ducal eggs were present. In 1966 ovulation was not observed before May 23 but females were seen laying eggs on June 10. This indicates that the oviducal eggs were retained less :1 than three weeks. On the Other hand, females were still laying eggs in the first week of July. This indicates egg retention of more than six weeks, unless more than one clutch is laid. The time of the first ovulation appeared to be about a week earlier in 1965 than in 1966. No turtles were dis- sected between May 14 and May 26, 1965, so whether ovulation was in effect throughout this period is not known. However, the two Chrysemys collected_May 14, 1965, had oviducal eggs whereas none of seven collected from May 16 through 21, 1966, had ovulated. If ovulation actually occurred earlier in 1965 the most obvious causal factor would be temperature, since warm temperatures prevailed in May, 1965. Whether such a hasten- ing of ovulation could result in an extra clutch being laid is unknown. Mating Mating apparently begins as early as March and lasts .through May. Based on the appearance of the sperm ducts, 85 sperm transfer could be effected during April or early May. Courtship behavior is negligible during the summer, but in the fall courtship again occurs though apparently not at the intensity observed in the spring. Reports of other workers on Chrysemys indicate both spring and fall courtship. Taylor (1933) reported titilla- tion by several laboratory males on October 16. No instance of copulation was observed. Finneran (1948) noted copula- tion in a pair of Chrysemys on May 2. Cagle (1955) observed hatchling Graptemys and Pseud- emys titillating in the laboratory. He pointed out that this occurred in immature individuals incapable of "attain- ing the assumed goal." He suggested that sex hormones are probably the usual cause of such behavior. Assuming this to be true, such male sex hormones are possibly activated in the fall when the mature male turtle is undergoing testi- cular changes, so that, although courtship behavior and even copulation may occur, there is no sperm transfer. Reproductive Potential The annual reproductive potential may be considered as being the number of eggs a female is capable of laying in a year. In Chrysemys picta of the present study mature females averaged a minimum of 6 or 7 eggs per season. This figure can probably be doubled, since several lines of evi- dence indicate that there are two clutches per year: 86 (l) Ovulation occurs during the third or fourth weeks in May and egg laying begins by the first week in June and lasts until the first week in July. It is unlikely that oviducal eggs would be retained less than three weeks in some indivi- duals but more than six in others. Instead, eggs are probably retained 2 to 3 weeks after ovula- tion and then deposited. A few days later ovula- tion might occur again and these eggs are also deposited in 2 to 3 weeks. Such a schedule coincides with nesting observations. (2) Two size groups of corpora lutea were often present in females in late June and July. Since corpora lutea are no longer visible in August and September, the small group would not be a holdover from the previous year. Also, two groups were not observed in early June before a second ovulation could have occurred. (3) In females with oviducal eggs in early June there was always a set of follicles of a size large enough for ovulation. Since follicles this large are not present in females after the egg laying period, they are presumably ovulated. (4) The experimental animals used in controlled nesting studies produced a significantly greater number of eggs per individual than normal (Table 87 4). Also, Upon dissection in late July many of these females had follicles which appeared large enough for ovulation. When the number of these follicles was added to the number of eggs produced by an individual, the resulting total was approxi- mately double that of the average number of ovi- r,— ducal eggs normally found. I The evidence is greatly in favor of two clutches per year though no female under natural conditions was ob- served laying eggs twice. The size of the second clutch is apparently similar to the first. Two clutches seem to be the maximum for one year. Egg laying in alternate years, as is true of some reptiles (Tinkle 1962; Wharton 1966), is not characteristic of Chrysemys. Thus, reproductive poten- tial for female Chrysemys p1cta from southwestern Michigan is apparently 13 or 14 eggs per year. Cagle (1944b) reported older Pseudemys with senile ovaries. A difference between the ovaries of old and young mature female Chrysemys was not observed in the present study, indicating that reproductive potential is constant with size and age of the female. Growth Rates Immature Chrysemys picta grow at a relatively steady rate which decreases markedly when maturity is reached. {'1 [- ‘ [I '1 l ‘l I I'll I all... i I II 88 Growth rates are similar for each of the immature age classes, but the fastest growth rates were observed between hatching and the end of the second year. After the second year, immature turtles at Sherriff's Marsh increase in plas- tron length at an average rate of 7 to 8 mm per year. A striking reduction in growth rate occurs when fe- males reach about 110 mm and males about 80 mm in length. Growth rates in mature turtles were approximately 1 mm per year. This estimate was obtained both from the regression line of age-length relationships in a series of males (Fig- ure 11) and by means of the recapture method (Table 5). Females in different size categories were found to have similar growth rates with a slight reduction in the larger ones. The estimates for the different size classes of males, however, were less homogeneous. The smaller males had growth rates of more than 2 mm per year, whereas, the large ones grew less than .5 mm per year. There is a slight discrepancy between the growth rate of smaller males esti- mated from the regression line (about 1.1 mm/year) and that estimated by the recapture method. The former figure was based on a much greater sample size and hence is a more accurate method. Therefore, 2.3 mm per year is very likely an overestimate of average growth rate in smaller males. The data are insufficient to satisfactorily quantify growth rates of the various size classes, but the indication is that larger Chrysemys increase more slowly in plastron length than smaller ones. 89 Population Dynamics Sex Ratio The sex ratio of mature Chrysemys picta from Sher- riff's Marsh was close to a 1:1 ratio. This was true on the basis of actual captures and upon Lincoln Index ‘ estimates. 1- The slightly greater apparent abundance of mature 1 males is probably only a result of the method of determin- ing maturity in females. A plastron length of 115 mm was chosen to represent the size at which most females reach maturity. The length was arbitrarily picked from a rela- tively narrow size range. A length of 113 or 114 mm could just as easily have been used as a standard length for the size at which females reach maturity. In reality, then, the sex ratio of mature Chrysemys picta in Sherriff‘s Marsh probably does not vary significantly from 1:1. Age Structure and Survivorship According to the Lincoln Index technique, an average of 140 turtles were estimated to be in each of the first four age classes at Sherriff's Marsh in 1965 (Table 8). There was no trend towards reduction in numbers during suc- cessive years. One cohort (hatchlings from 1964) had fewer individuals than any of the others. Presumably before their second year this cohort suffered from a source of mortality 90 which did not affect older juveniles to the same extent. The collective estimate for immature females 6 to 11 years old was 90 individuals per age class. This is in accord with the estimates for the younger juveniles. There is no evidence to indicate that mortality increases with age in immature Chrysemys. Apparently, then, once a juvenile has completed his first year his chances of reaching maturity are extremely high. The survival pattern from the egg stage until matu- rity assumes that of a Type IV survivorship curve (Slobodkin 1963). Around 2 percent of the eggs laid at Sherriff's Marsh develop into juveniles which become recruited into the pOpulation. Most of these juveniles reach maturity. There are at least three possible causes of mortality in the nest. (l) Non-hatching of Chrysemys eggs was observed both in the field and laboratory. This may be due in some part to infertility, as has been observed for Malaclemmys (Hildebrand 1929). (2) Extreme environmental conditions such as excessive heat, cold, moisture, or dryness seem likely sources of mortality, although insufficient data- were collected to support this postulation. (3) Predation is a major mortality factor at the egg stage. The amount of predation is no doubt excessive under some circumstances, as when a large number of females lay their eggs in one area . II..I|1|I|It[['I.Il i.|lll {Illliilllr 91 Adults of both sexes eXperience a much higher mor— tality rate than is true of immature turtles. There appears to be a sudden reduction in the number of individuals per age class as maturity is approached. This phenomenon may be a result of the method of age determination. Growth rate was assumed to be 1 mm per year for all ages of both sexes. The ages in Table 10 were based upon this age-length relationship. However, old adults probably grow more slowly than young ones. If differential growth rates were taken into account, the smaller size ranges in Table 10 would in- clude fewer age classes than the larger ones. The effect of this would be to raise the number of adults per age class in the younger turtles. The number of individuals per age class would be reduced for older turtles. The re- sulting survivorship curves for both sexes would then ap- proach a Type III, as defined by Slobodkin (1963). Such a curve is characterized by a constant mortality rate for individuals of all ages. Once a Chrysemys reaches maturity at Sherriff's Marsh, its chances of survival presumably remain the same regardless of age. Many physiological and behavioral changes accompany maturity. Some of the more obvious effects are increased activity in the spring, expenditure of energy for reproduc- tive functions, and a greater likelihood of terrestrial activity. These and other changes associated with adulthood make mature Chrysemys susceptible to environmental hazards.. f . II (III . [ I'll-I I 3 (Ii 11 .. I . ‘ II 92 that do not confront immature individuals. The result is a much higher adult mortality rate. Longevity: Turtles are usually considered to be long-lived animals. Carr (1952) states that the age at— tained by turtles is "perhaps the grestest of any living vertebrate." Hildebrand (1929) gives a "conservative estimate” that Malaclemmys live 25 to 40+ years. Cagle (1950) states that Pseudemys under natural conditions may have a life span "possibly within the range of 50-75 years." Nichols (1939b) estimated that Terrapene carolina were fully grown in about 20 years but that some of the older ones were probably as much as 80 years old. Schneck (1886) reports an individual Terrapene which was marked in 1824 and recap- tured several times during the next 60 years. Turtles are always among the oldest vertebrates in zoos (Flower 1944; Conant and Hudson 1949). Once a Chrysemys hatchling is incorporated into the Sherriff's Marsh population the chances of reaching maturity appear to be excellent. Therefore, most males live at least six years, females at least twelve. In spite of the in- creased mortality of adults at least 20 percent of each sex probably reach an age of 20 years or more (Table 10). A few individuals were estimated to be over 40 years old. Due to the difficulty of aging old individuals, this is very likely a minimum estimate. There is no indication that Chrysemys at Sherriff's Marsh approach the "maximum life span" for the species(Deevey 1947). Mortality Predation on turtles and turtle eggs is a recog- nized source of mortality (Cagle, 1944c and 1950; Stophlet 1947; Woodbury and Hardy 1948; Carr 1952; Murphy 1964). Egg predation by mammals (presumably raccoons, skunks, or foxes) is undoubtedly of major importance in the Chrysemys populations of the present study. The observed predation on juveniles and adults, however, was extremely limited. Whether predation is a constant and important cause of mor- tality at this level is uncertain. Parasitism, external or internal, does not appear to be a mortality factor. Although numerous Chrysemys in- dividuals carried leeches (Placobdella sp.), no heavy in- festations were found. The turtles appeared unaffected by the parasites. Although several species of internal para- sites were found to be present in juvenile Chrysemys from Wintergreen Lake (Esch and Gibbons 1967), the turtles showed no ill effects. Adult Chrysemys harbored very few or' no internal parasites, whereas one to four year olds were often heavily infested. Death due to unknown causes was evident both in the nests and in the turtles at Sherriff's Marsh. Eggs and hatchlings died in some nests for no apparent reason, whereas others in the same nest survived. Infertile eggs are a possible cause for some non-hatching (Hildebrand 1929), 94 but the extent of infertility in natural populations of turtles is unknown. Several of the dead Chrysemys found at Sherriff's Marsh had not died as a result of predation, but no clue was given to the cause of death. ”Old age" has been described in Pseudemys (Cagle 1950), but this con- dition was not observed in any Chrysemys. Death from environmental causes resulting in freezing, desiccation, or over-heating does not seem likely in the adults from the marsh population. Seasonal Activity Aquatic Cagle (1954) stated that temperatures of 20 to 25° C are probably Optimum for Chrysemys. Temperatures around 40° C are usually fatal (Hutchison, Vinegar and Kosh 1966). Chrysemys are generally considered to become active at water temperatures about 8 to 11° C (Cagle 1954; Musacchia 1959; Sexton 1959a), which occurred in early spring at Sherriff's Marsh. Adults were first abundant in the middle of March, 1966, when water temperatures reached 10° C. They completely disappeared in late March when temperatures dropped to 5° C and did not reappear until temperatures reached 8° C in April. The effect of temperature in the fall was not as well defined. Water temperatures were generally above 10° C in September and early October of 1965 and 1966. Yet turtles became increasingly less abundant. 95 Activity at low temperatures in the fall is considerably reduced. Juveniles became active several weeks later than adults. Time of emergence in the spring varied among the juvenile age classes, with the older individuals appearing first. Although hatchlings were known to enter the water in April and May of 1966, none were seen in the marsh until late July. Younger turtles do not become active when water temperatures reach the minimum for activity. Hunger does not seem a likely motivation for the early spring activity of adults for three reasons: (1) only large turtles were active early. If hunger were the reason for vernal activity, smaller juveniles, having a higher metabolic rate, should be the first to appear, not the last. (2) Chrysemys are not known to feed extensively at temperatures below 15° C (Sexton 1959a). (3) Chrysemys were not collected in bait traps in Sherriff's Marsh until April 29, 1965, although turtles were active prior to that time. Since mating behavior was observed early in March and most males are in a testicular breeding condition by April, the early activity of Chrysemys seems related to re- production. Adult Chrysemys apparently are reporductively active in early spring. This leads to increased movement when water temperatures reach a minimum tolerance level. 96 Sexton (1959a) found that by June 1 most Chrysemys were associated with dense aquatic vegetation. This was also observed at Sherriff's Marsh. The increase in plant growth during May in conjunction with lowered water level provided many areas where aquatic plants were in heavy con- centration. On June 9, 1966, 60 heads were counted in an area with heavy duckweed cover, whereas no turtles were seen in a nearby area with no surface vegetation. Dense aquatic vegetation functions as a means of physical sup- port for the turtles as they carry out normal activities during the summer (Sexton 1959a). A search for food is apparently initiated by some individuals before the end of the breeding season as in- dicated by the heavy capture in bait traps during late April and early May of 1965. The reduction in the number of turtles collected in bait traps in May was apparently due to the increase in aquatic vegetation and abundant food in the marsh. Chrysemys entered Zone 5 in large numbers in late May and most departed by September. The association of Chrysemys with this area in the summer appears to be de- pendent upon the surface vegetation which provides an ideal habitat in regard to support and abundant food. The hard bottom and shallow depth, however, are not suitable for the winter. Apparently, then, Zone 5 serves as a feeding ground in which turtles congregate in the summer. 97 The observation of large numbers of Chrysemys swim- ming in 4° C water is perhaps explainable as thermoregula- tory behavior. The open water in April and clear ice in December afforded maximum light penetration and warming of- the dark substrate. By exposing themselves on the sub- strate surface, the turtles were able to obtain the maximum heat gain possible under the conditions. The effectiveness of solar radiation in raising body temperatures of turtles in shallow water is not known, although other species have been reported basking underwater (Cagle 1944). Boyer (1965) suggests that increased light intensity and rising water temperatures motivate basking behavior which raises body temperature in turtles. Terrestrial Chrysemys were collected on land during four periods: late April and the first week in May; late May; June and early July; and September. The early spring migrations oc- curred during the mating season and primarily involved mature individuals. However, the presence of juveniles would appear to mar an otherwise satisfactory explanation that this activity results from reproductive activity (Cagle 1944a). This is not necessarily so in the present study since the immature turtles found on land were very close to maturity and could conveivably have already been under adult hormonal influence. Terrestrial movement early in the 98 spring is possibly another expression of reproductive appe- titive behavior. All of the Chrysemys collected on land in early spring were directly between two proximate bodies of water. Turtles in Sexton's (1959a) population at this time of year were also traveling from one aquatic area to another. This indicates that terrestrial movement is not simply random wandering but is directed in that the turtle is trying to reach another body of water. Very few turtles were found on land in the vicinity of the marsh and in all instances these were between Zone 5 and Zone 15, the swampy area to the north. This small swamp represents the only area of Sherriff's Marsh which is isolated from the remainder. Perhaps the reason that turtles are seldom found on land around Sherriff's Marsh in the spring is because the dif- ferent habitats are connected by water routes and new areas can be reached without terrestrial movement. In conclusion, travel of Chrysemys over land in the spring occurs only when such land lies between aquaticareas which are not con- nected by a water route. A few turtles were collected on land in late May of 1965 and 1966. Two of these were mature females presumably seeking nesting sites. Theothers were three immature fe- males and a mature male. A search for a more favorable habitat might be a possible explanation for this terrestrial travel, although data are not available to substantiate this supposition. 99 All of the Chrysemys found on land during June and July were mature females presumably making temporary excur— sions onto land for nesting purposes. Terrestrial travel in the fall is widespread among turtles (Cahn 1937; Cagle 1944a; Carr 1952). These migra- tions have generally been attributed to movement to more suitable areas for hibernation. Further comment is not, warranted on the basis of the present study except to point out that no turtles were seen on land around Sherriff's Marsh during this season, suggesting that, as with Spring travels, Chrysemys travel across land only when a water route is not available. The effect on the population, of Chrysemys enter~ ing, or leaving, the study area appears to be minimal. A statement that migrations to, or from, any unenclosed area does not occur can hardly be made. However, indirect evi- dence supports the view that the population of Chrysemys at Sherriff's Marsh is relatively uninfluenced by migration. Several factors warrant such a conclusion. (1) More than 100 collecting trips to Sherriff's Marsh were made during the study. On many of these trips the roads encircling the marsh were traveled by car. The only turtles, dead or alive, were found on the road between Zones 5 and 15, two aqua- tic areas. Also, other than females laying eggs, no turtles were found on land around the marsh during the entire study. 100 (2) More than 1000 turtles were marked in Sherriff's Marsh. Almost 400 turtles were captured two miles away in Wintergreen Lake, the nearest sizable popu- lation of Chrysemys. None of these proved to be marked turtles from Sherriff's Marsh. Also, a large number (266) of Chrysemys from Wintergreen Lake were marked and released by Dr. M. M. Hensely in 1957. Although several of these were recaptured in Win- tergreen Lake during the present study, none were seen in Sherriff's Marsh. There is no evidence of interchange between the Sherriff's Marsh and Winter- green Lake populations. Local Travel in Sherriff's Marsh Several workers have investigated the home range of turtles. Terrapene carolina (Breder 1927; Nichols 1939a; Stickel 1950), gopherus.agassizi (Bogert 1937; Woodbury and Hardy 1948), and Trionyx ferox (Breckenridge 1955) have been shown to confine their activities to restricted areas. Cagle (1944a) studied home range and movement of Pseudemys and Chrysemys and cOncluded that they "live within selected areas" which "may include parts of two or more water bodies. Cagle states further that no territorial defensive behavior was observed. Pearse (1923a) considered Chrysemys in Lake Mendota to be "rather sedentary." Sexton (1959a) indicated that Chrysemys are restricted in their travel under conditions 101 of little vegetation change, but that under "varying condi- tions” they often change their areas of activity. Homing abilities have been noted in Chrysemys (Cagle 1944a; Wil- liams 1952) whereas Ortleb and Sexton (1964) investigated several cues which might be used in orientation. Travel during a single summer was limited in most Chrysemys from Sherriff's Marsh (Table 14). This behavior may be attributed to the several areas of surface vegeta- tion with which most of the turtles were associated. Such areas undoubtedly offer sufficient food and support, and there is little reason for turtles leaving such an area during the summer. Over 40 percent of the recaptured mature females were found to have changed locations within the marsh dur- ing successive summers. The change of area possibly occurs during the egg laying season. A female seeking a nesting site leaves the water and walks until a suitable place is found. This may come about only after considerable random wandering and be several meters from water. The return trip to water is probably more direct and not necessarily to the area previously inhabited. Upon reaching the water the turtle seeks an area with surface vegetation and prob- ably remains in the vicinity throughout the summer. Such travels would guarantee interchange among a population of otherwise sedentary animals. This could also result in the invasion of a new aquatic area if, after laying, the 102 female entered another body of water. A female entering a previously uninhabited area would probably continue to lay fertile eggs, at least in the following year as is true with Malaclemmys (Hildebrand 1929)° Such an effort would be adequate to establish a Chrysemys population within a few years. Mature males showed little change of location be- tween summers. This is particularly surprising in View of the high degree of travel in the spring time. One inter- pretation of this phenomenon may be that males have a ”sum- mer home range" to which the animal returns after long ventures during the breeding season. No explanation is readily available as to why males should return to former feeding areas, whereas, females do not. Sexton (1959a) in- dicated that adult Chrysemys tended to wander more than immature ones. This seemed true in regard to adult females at Sherriff's Marsh but not adult males. Most of the tur- tles from one to three years old were recaptured in the same places the following year. Large scale juvenils dis- persal does not appear to be a primary means of invading new areas . SUMMARY AND CONCLUSIONS (1) A quantitative and descriptive analysis was made of the life cycle, population dynamics, and ecology of the painted turtle, Chrysemys picta ,in two populations in southwestern Michigan. (2) Maturity of Chrysemys picta is a function of size. Males in the study populations reach maturity at a plastron length of approximately 80 mm, females at approximately 115 mm. (3) The male reproductive cycle is divisible into three seasonal phases. During the mating period from late March to early May the testes are small but the sperm ducts are enlarged and filled with sperm. Spermatogenesis occurs from July to October and results in an increase in the size of the testes which contain all of the sperm at this time. The winter phase is characterized by small testes with most of the sperm contained in the epididymis. Transitional periods occur between each of these phases. Courtship be- havior occurs in the spring and fall but effective mating during the latter period is unlikely, due to the lack of sperm in the ducts at this time. (4) Ovarian follicles begin to increase in size in September. By March, the largest follicles average 15 to 103 104 17 mm in diameter. Immediately prior to ovulation there is an increase to a diameter of approximately 18 mm. Ovu- lation occurs near the middle of May. Another group of enlarged follicles is invariably present after ovulation in May. These are presumably the follicles of a second clutch. The second ovulation probably occurs near the mid- dle of June. The first clutch of eggs is laid in late May and early June, the second clutch in late June and early . . ~ fl-1u+mi- y.” July. There is no indication of a third clutch. After {Q the second ovulation the follicles do not increase in size until September. Clutch size is about 6.5 eggs. The annual reproductive potential for individual females is around 13 eggs per year. (5) Growth rate, based on plastron length, is fastest during the first two years but is relatively similar through- out the juvenile stage. Growth decreases abruptly in both sexes when maturity is reached. Growth rate probably de- creases to some degree as mature turtles grow older. (6) An estimated 2328 individuals were in the marsh population. Of these, 1403 were immature. The mature ani- male were in a sex ratio of nearly 1:1. (7) Less than 2 percent of the eggs laid each year in the study population develop into juveniles which success- fully enter the population. Egg predation is a heavy source of mortality. 105 (8) Mortality is not consistently greater in any parti- cular age class of immature turtles. Thus, once a hatch- ling has completed the first growing season there is a high probability that maturity will be reached. (9) Upon reaching adulthood, both sexes experience abrupt increases in mortality rate. This is attributed to the combined effect Of behavioral and physiological changes which accompany maturity thus making adults more susceptible to certain environmental hazards. Some mani- festations of these changes are earlier vernal activity, greater terrestrial activity in some situations, and greater expenditure of energy for reproductive function. All of these factors tend to lessen the individual's chances of survival. (10) The survivorship pattern of the study population is a combination of a Type IV survivorship curve (egg- Juvenile) and a Type 111 (adult). The former is character- ized by an extremely high mortality rate initially followed by a low mortality rate. A population following a Type III curve has a constant mortality rate at all ages. (11) Adult Chrysemys become active in early spring when water temperatures reach the minimum tolerance limits. Re- productive activity appears to be responsible for this early emergence. Juveniles are seldom seen before late May. (12) The early spring activity is characterized by terres- trial movement in some areas where land separates proximate 106 bodies of water. Apparently the aquatic conditions pre- ferred for winter quarters may be unsuitable for breeding purposes, thus explaining the frequent overland migrations. Such terrestrial travel is seldom observed in areas where different aquatic habitats are connected by water routes. Travel by males is greatly increased in the spring. (13) By late May most Chrysemys are associated with areas of heavy vegetation. Local movement of most individuals is limited throughout the summer. Individual juveniles and mature males are usually found in the same general area in successive summers whereas mature females have often traveled. This is possibly a result of nesting females leaving the water from one area but returning to another. This would serve as an effective dispersal mechanism. Why males return to the same areas in successive summers after their springtime activity is uneXplained. (14) By September, the turtles begin to leave summer feeding areas which are unsuitable for winter. This again results in travel over land in some areas. The tolerance level to temperature during this period is less well de- fined than in the spring and responses of the population are more gradual. Activity at low temperatures in the fall is definitely reduced in adults. Juveniles remain active as long as adults in the fall. LITERATURE CITED Allee, W. C., A. E. Emerson, 0. Park, T. Park, and K. P. Schmidt. 1949. Principles of Animal Ecology. W. B. Saunders Company, Philadelphia. 1-837. Atland, Paul D. 1951. Observations on the structure of reproductive organs of the box turtle. J. Morphol. 89:599-616. Bogert, C. M. 1937. Note on the growth rate of the desert tortoise, Gopherus agassizi. Copeia 1937:191-192. Boyer, Don R. 1965. Ecology of the basking habit in tur- tles. Ecology 46:99-118. Breckenridge, W. J.* 1944. Reptiles and amphibians of Minnesota. University of Minnesota Press, Minnea- polis: 1-202. 1955. Observations on the life history of the soft-shelled turtle Trionyx ferox, with especial reference to growth. Copeia 1955:5—9. Breder, Ruth Bernice. 1927. ~Turtle trailing: a new tech- nique for studying the life habits of certain Testudinata. Zoologica 9:232-243. Cagle, Fred R. 1944a° Home range, homing behavior, and migration in turtles. Misc Publ. Mus. Zool. Univ. Mich. 61:1-34. .__ 1944b. Sexual maturity in the female of the tur- tle Pseudemys scripta elegans. Copeia 1944:149-152. 1944c. Activity and winter changes of hatchling Pseudemys. COpeia 1944:105-109. 1948. Sexual maturity in the male turtle, P. s. troostii. Copeia 1948:108-111. 1950. The life history of the slider turtle, Pseudemys scripta troostii (Holbrook). Ecol. Monogr. 107 108 - Cagle, Fred R. 1953. An outline for the study of a rep— tile life history. Tulane Studies in Zoology 1(3): 31-52. 1954. Observations on the life cycles of painted turtles (Genus Chrysemys). Am. Midl. Nat. 52:225-235. 1955. Courtship behavior in juvenile turtles. Copeia 1955:307. Cahn, Alvin R. 1937. The turtles of Illinois. Ill. Biol. Monogr. 16:1-218. Carr, Archie. 1952. Handbook of Turtles. Cornell Univer- sity Press, Ithaca, New York. 1-542. Conant, Roger. 1951. The reptiles of Ohio. University of Notre Dame Press, Notre Dame, Indiana. 1-284. , and Robert G. Hudson. 1949. Longevity records for reptiles and amphibians in the Philadelphia Zoological Garden. Herp. 5:1-8. Deevey, Jr., Edward S. 1947° Life Tables for natural popu-. lations of animals. The Quarterly Review of Biology, 22:283-314. Esch, Gerald R. and J. Whitfield Gibbons. 1967. Seasonal incidence of parasitism in a population of the painted turtle, Chrysemys picta marginata Agassiz. J. Parasit. In Press. Finneran, Leo Charles. 1948. Reptiles at Branford, Con- necticut. Herp. 4:123-126. Flower, S. S. 1944. Persistent growth in the tortoise Testudo graeca, for thirty-nine years, with other notes concerning that species. Proc. Zool. Soc. London 117:680-688. Hildebrand, Samuel F. 1929. Review of experiments on arti- ficial culture of diamond-back terrapins. Bull. U. S. Bureau of Fisheries 45:25-70° 1932. Growth of diamond4back terrapins. Size attained, sex ratio and longevity. Zoologica 9:551- 563. Hutchison, Victor H., Allen Vinegar, and Richard J. Kosh. 1966. Critical thermal maxima in turtles. Herp. 22:32-41. 109 Lagler, Karl F. 1943. Food habits and economic relations of the turtles of Michigan with special reference to fish management. Am. Midl. Nat. 29:257-312. Legler, J. M. 1960. Natural history of the ornate box turtle Terrapene ornata ornata Agassiz. Univ. Kans. Publ., Mus. Nat. Hist. 11(10):527-669. Marchand, Lewis J. 1942. A contribution to a knowledge of the natural history of certain freshwater tur- tles. Master's thesis. University of Florida 1-83. ' McDowell, Samuel B. 1964. Partition of the genus Clemm s and related problems in the taxonomy of the aquatic Testudinidae. Proc. Zool. Soc. London 143:239-279. Meseth, Earl H. and Owen J. Sexton. 1963. Response of painted turtles Chrysemys icta to removal of sur- face vegetation. Herp. 19: - 6. Murphy, Ted D. 1964. Box turtle, Terrapene carolina, in stomach of copperhead, Agkistrodon contortrix, Copeia 1964:221. Musacchia, X. J.‘ 1959. Viability of Chrésemys picta at low temperature. Physiol. Zoo . Nichols, J. T. 1939a. Range and homing of individual box turtles. Copeia 1939:125-127. 1939b. Data on size, growth and age in the box turtle, Terrapene carolina. Copeia 1939:14-20. Ortleb, Edward P. and Owen J. Sexton. 1964. Orientation of the painted turtle, Chrysemys picta, Am. Midl. Nat. 71:32-34. Pearse, A. S. 1923a. The abundance and migration of tur- tles. Ecology 4:24-28. 1923b. The growth of the painted turtle. Biol. Bull. 45(3):145-l48. , S. Lepkovsky, and Laura Hintze. 1925. The growth and chemical composition of three species of turtles fed on rations of pure foods. J. Morphol. 41:191-216. Pope, Clifford H. 1939. Turtles of the United States and Canada. Alfred A. Knopf, New York. 1-343. I'll] III .II ‘Illl'll" ‘ I‘lllll 110 Raney, Edward C. and Ernest A. Lachner. 1942. Summer food of Chrysemys icta marginata in Chataugua Lake New York. Copeia I942: - . ’ Risley, P. L. 1933. Observations on the natural history of the common musk turtle, Sternotherus odoratus (Latreille) Papers, Mich. Acadl Sci, Arts, and Let- ters 17:685-711. 1938; Seasonal changes in the testis of the musk turtle, Sternotherus odoratus L. J. Morphol. 63:301- 317. Schneck, J. 1886. The longevity of turtles. Amer. Nat. 20:897. Sexton, Owen J. 1959a. Spatial and temporal movements of a population of the painted turtle, Chr sem s picta marginata (Agassiz). Ecol. Monogr. zgzII3-I40. 1959b. A method of estimating the age of painted turtles for use in demographic studies. Ecology. 40:716-718. 1965. The annual cycle of growth and shedding in the midland painted turtle, Chrysemys picta marginata. Copeia 1965:314-318. Slobodkin, Lawrence B. 1963. Growth and regulation of ani- mal populations. Holt, Rinehart and Winston, New York. 1-184. Smith, Philip W. 1961. The amphibians and reptiles of Illi- nois. Bull. Illinois Nat. Hist. Survey 28:1-298. Stickel, Lucille F. 1950. POpulation and home range rela- tionships of the box turtle Terra ene c. carolina (Linnaeus) Ecol. Monogr. 20:351-37E. _ Stophlet, John J. 1947. Florida otters eat large terrapin. J. Mammalogy 28:183. Taylor, Edward H.‘ 1933. Observations on the courtship of turtles. Univ. Kanas Sci. Bull. 21(9):269-271. Tinkle, Donald W. 1958. Experiments with censusing of southern turtle populations._ Herp. 14:172-175. 1961. Geographic variation in reproduction, size, sex ratio and maturity of Sternothaerus odoratus (Testudinata: Chelydridae). Ecology 42:68-76. lll Tinkle, Donald W. 1962. Reproductive potential and cycles in female Crotalus atrox from northwestern Texas. Copeia 1962:306—3I3. Webb, Robert G. 1961. Observations on the life histories of turtles (Genus Pseudemys and Graptemys) in Lake Texoma, Oklahoma. Am. Midl. Nat. 65:193-214.. Wharton, Charles H. 1966. Reproduction and growth in the cottonmouth Agkistrodon piscivorus Lacepede of Cedar Keys, Florida. COpeia 1966:149-161. Williams, John E. 1952. Homing behavior of the painted turtle and muskturtle in a lake. C0peia 1952:76-82. Woodbury, Angus M. and Ross Hardy. 1948. Studies of the desert tortoise, Gopherus agassizii. Ecol. Monogr. 18:145-200. "I7'1111111111111111“