[Ill.u.s.tration: PLATE 28]

FIGS. 1-3. Stages in courtship of _T. o. ornata_: male pursuing female and biting her sh.e.l.l; male lunging at female in attempt to mount; and, male just after mounting female (? ).

FIG. 4. _T. o. ornata_ smelling food (? 1).

FIG. 5. _T. o. ornata_ in att.i.tude of alertness after detecting intruder (? 3/8).

FIG. 6. Tracks of _T. o. ornata_ in muddy ravine (? 1/8) (June 5, 1956).

[Ill.u.s.tration: PLATE 29]

FIG. 1. A small, nearly-healed, injury on the carapace of an adult _T. o. ornata_ (? 2). Note regenerated epidermis at bottom of depression and two pieces of dead bone ("splinter scars") at upper right margin of depression.

FIG. 2. Injured area on the carapace of a juvenal _T. o. ornata_ (? 3) with dead bone removed and laid to the right, exposing regenerating epidermis in its early stages.

FIG. 3. Anterior edge of carapace (held away with forceps) of specimen of _T. carolina_ (KU 51461, Gulf Co., Florida) that had been badly burned (? 8/9). Nearly all the scutes of the sh.e.l.l had fallen off and large pieces of dead bone could be pulled away, exposing a gnarled ma.s.s of regenerating bone and epidermis.

FIG. 4. A fracture that has healed and joined the tibia (upper bone) to the fibula in a specimen of _T. o. ornata_ (KU 1877, ? 3).

[Ill.u.s.tration: PLATE 30]

_Top_: A sh.e.l.l of _T. o. ornata_ (? ) as it was found at the Damm Farm June 1, 1956. A serious injury (probably resulting from burns) had exposed a large area of dead bone on the carapace.

_Center_: Same sh.e.l.l with some of scutes removed.

_Bottom_: Same sh.e.l.l with dead bone removed to expose regenerating epidermis and bone. Note that the injury involved several of the neural bones; the turtle probably died as a result of this injury but not before regeneration was approximately one-half completed.

_Terrapene ornata_ seems to concentrate its breeding season (laying, incubation, and hatching of eggs) more nearly in the middle of the warm season than does _T. c. carolina_. This concentration probably is an adaptation for breeding in open habitats where, under environmental temperatures less equable than in forest, eggs would develop more rapidly and hatch sooner but would be less able to survive winter temperatures.

Males of _T. o. ornata_ become s.e.xually mature when younger and smaller than females and rarely grow as large as females. Nichols (1939a:20) indicated the reverse to be true of _T. c. carolina_; Nichols further indicated that growth continued some six to eight years after p.u.b.erty. Most individuals of _T. o. ornata_ attain maximum size within two to three years after p.u.b.erty.

Although it is difficult to be certain about the adaptive value of color and pattern, it seems that in box turtles, as in many other kinds of animals, patterns and colors most nearly blending with those of the habitat have some selective value in providing concealment from enemies. The pattern of linear radiations in _T. o. ornata_ closely resembles the patterns formed by light pa.s.sing through gra.s.ses and a.s.sociated vegetation and camouflages the turtle. In a similar manner, partial or complete loss of radial markings in _T. o. luteola_ seems to provide concealment in habitats where vegetation is spa.r.s.e and where blending with the substrate is of survival value. The patterns of blotches and broken radiations in most of the subspecies of _T.

carolina_ likewise provide camouflage by tending to match patterns formed by the light pa.s.sing through a leafy canopy.

Although ornate box turtles are omnivorous, they probably depend on insects as a dietary staple. In years when preferred kinds of insects were unusually abundant, the turtles grew more than in other years. A large proportion of the insects eaten is obtained by foraging in or near dung. Alteration of the dung community--at least in a physical sense, but presumably also by influencing the successional stages of the dung biota--is one of the few evident effects of box turtles on the environment. Although certain kinosternids (Carr, 1952:93), emyids (Deraniyagala, 1939:257; Loveridge and Williams, 1957:198), and testudinids (Loveridge and Williams, _op. cit._:247) eat mammalian feces, _T. ornata_ is seemingly the only chelonian that habitually seeks its staple diet in dung. The habit seems to be yet another specialization for terrestrial existence. The carnivorous habits of _T. ornata_ reverse the general trend toward omnivorous and herbivorous habits in other turtles that have become partly (emyids) or wholly (testudinids) terrestrial.

It seems remarkable that none of the species of true tortoises occurring in the gra.s.slands of the world has developed insectivorous habits or utilized the unique food niche (in regard to dung-foraging) filled by ornate box turtles in the Great Plains; tortoises are, as far as is known, strictly herbivorous. The ranges of _Gopherus_ and _Terrapene_ are now almost mutually exclusive and the two kinds do not compete with each other for food in the few places where they occur together. It is known, however, that box turtles (_T. longinsulae_, _ornata_-like, earliest known box turtle) and true tortoises (genera _Testudo_ and _Gopherus_, see Williams, 1950:25-26, Fig. 2) occurred together in what is now the Great Plains in early Pliocene times and probably for some time before and after this. a.s.suming that food habits of fossil representatives of these genera were somewhat like the habits of recent representatives, ornate box turtles may have developed insectivorous habits at a time when other food niches were filled by herbivorous tortoises. Box turtles possibly survived subsequent changes in habitat that made it impossible for populations of large tortoises to exist in the Great Plains.

SUMMARY

Box turtles of the genus _Terrapene_ are emyid turtles that are specialized for terrestrial existence. Two of the seven species now recognized--_T. ornata_ and _T. carolina_--occur in the United States.

_Terrapene carolina_ inhabits forested areas in the east whereas _T.

ornata_ is characteristic of open gra.s.sy areas in the west; the ranges of the two species overlap in the broad belt of prairie-forest ecotone in the central United States. _Terrapene ornata_ is considered to be the most specialized of living box turtles.

The natural history of _T. o. ornata_ Aga.s.siz was studied in the period, 1953 to 1957. Intensive field studies were made in Douglas County, northeastern Kansas, on a small area of prairie and on the University of Kansas Natural History Reservation. Field observations were made also in a number of other places in eastern Kansas.

Laboratory studies supplemented field studies.

Habitats occupied are chiefly open areas; they vary in regard to food supply, temperature, moisture, and kind of soil. The gra.s.sy prairies of Nebraska, Kansas, Oklahoma, and northern Texas seem to provide optimum habitat for ornate box turtles; in these areas box turtles are active on a majority of days from April to October. The subspecies _luteola_ is adapted to the more rigorous and arid environment of the southwestern United States, where activity may be possible for only a few weeks in the year. The remainder of the year is spent in a state of quiescence. Factors limiting the distribution of _T. ornata_ are: 1) the presence of a substrate too hard to permit digging of nests and forms (alt.i.tudinal distribution in southwestern United States and distribution at western edge of the range); 2) temperatures causing the ground to freeze deep enough (approximately 30 inches) to kill turtles in hibernacula (northern edge of range); and, 3) the lack of one or more relatively wet periods in the course of the warm season, preventing at least temporary emergence from quiescence (southwestern part of range). The activities of man probably have affected population density in local areas but limit the geographic range only in the north (Blanchard, 1923:19-20, 24) where intensive cultivation probably has excluded the species.

Preferred habitat in northeastern Kansas is open rolling gra.s.sland grazed by cattle; populations are most dense near natural breaks in the gra.s.sy vegetation such as fences, scattered rocks on hillsides, ravines, and stream-beds.

Mating occurs most commonly in spring and autumn; courtship behavior includes pushing and biting on the part of the male. In coitus the hind legs of the male are held tightly by the female; the male falls backward after coitus, still clasped by the female. A few sperm are stored in the oviducts; fertilization without reinsemination can occur. The spermatogenic cycle begins in May and reaches its peak in September, when large numbers of sperm and spermatids are present in the testes; the cycle is completed in October, when sperm pa.s.s into the epididymides. The testes are smallest in spring and largest in September. Females are inseminated with sperm produced in the preceding year. The ovarian cycle begins in midsummer, soon after ovulation, and continues up to the time of the next ovulation.

Follicular growth is rapid in the period from spring emergence to ovulation. Large follicles remaining after ovulation represent, in many instances, eggs that will be laid later in the same season.

Follicular atresia is never great enough to account for the destruction of all large follicles remaining after ovulation. All mature females lay at least one clutch of eggs per year. It is estimated that one-third of the females produces two clutches of eggs in a single season. Second clutches contain fewer eggs than first clutches. An alternation of ovarian activity occurs, whereby one ovary is more active than its partner in one season and less active in the next season. Alternating activity of ovaries accounts in part for the reduced number of eggs in young females, breeding for the first time, and in older, nearly senile females. Extrauterine migration of ova results usually in a more even distribution of eggs in the oviducts.

Corpora lutea const.i.tute an accurate record of the number of eggs produced by the ovary as well as the number of eggs laid.

Nesting occurs from May through July but is most common in mid-June; some of the females nesting early in the season lay a second clutch of eggs in July. Nests are dug in the earth by the female using her hind legs. Preferred nesting sites are open, well-drained places with a soft substrate. The nesting site is selected after a period of wandering, in which the female tests the substrate at a number of places; some females search for a nest site for more than a week. Nest digging begins in the evening and is usually completed after dark.

Captive females dug a preliminary cavity in which the body rested during the digging of the main nest cavity. The entire clutch of eggs is laid in one nest. The average number of eggs in 23 clutches was 4.7 (range, 2 to 8). The average size of eggs tends to be inversely proportional to the number of eggs in a clutch. Eggs increase in bulk by absorption of water in the course of incubation. Immersion in water for short periods does not harm eggs. The incubation period under favorable environmental conditions is approximately 65 days; cool, damp conditions prolong the incubation period and probably const.i.tute an important factor of prenatal mortality in certain years. Eggs that do not hatch before winter probably do not survive. Emergence of hatchlings from the nest may, however, be delayed until spring if the soil is dry in autumn. Hatchlings can probably escape freezing by burrowing into the walls of the nest. Infertility and prenatal mortality account for at least 40 per cent of the eggs laid, according to laboratory findings. Progeny of a single adult female (considering factors of mortality, multiple layings, and average age of p.u.b.erty) would number approximately 300 after 20 years. Reproductive processes probably continue throughout life, although possibly at a somewhat reduced rate in later life.

Young box turtles are active soon after hatching but become quiescent if allowed to burrow in soil or if they are covered with damp cotton.

Some captive hatchlings take live food in the first days of life but others do not eat until the following spring; initiation of growth is coincident with initiation of regular feeding. The yolk sac retracts mainly during hatching; it sometimes ruptures after hatching. The caruncle remains on the beak for a variable length of time, but never is present in the spring following hatching.

Major growth-rings on the epidermal laminae are formed regularly, one after each season of growth, in the first 10 to 14 years of life.

Minor growth-rings occur between major rings and are shallower. Growth of epidermal laminae results from the formation, in spring, of a new layer of epidermis beneath the existing scute. The peripheral projection of the new layer is distinct in texture and color from the older part of the scute and is separated from it by a major growth-ring. Minor growth-rings form when growth slows or temporarily stops during periods of quiescence; no new layer of epidermis is formed. Growth-rings const.i.tute an accurate record of growth that can be studied at any time in the life of the turtle; they are accurate indicators of age only as long as regular annual growth persists.

Growth in the season of hatching depends on early hatching and early emergence from the nest. Turtles that remain in the nest until spring probably do not grow. Slightly less than half of the free-living individuals studied grew in the season of hatching. Precociousness in early life often results in the attainment of s.e.xual maturity at an earlier than average age.

Growth is rapid at first (increments in plastral length average 68, 29, and 18 per cent, respectively, in the first three years) and then slows gradually until p.u.b.erty. Attainment of s.e.xual maturity is more closely correlated with size than with age. Males mature when smaller (76 per cent were mature when plastron 100 to 109 mm. long) and younger (average age, eight to nine years) than females (66 per cent were mature when plastron 110 to 119 mm. long, average age at maturity, ten to eleven years) but females grow larger than males. A few individuals of each s.e.x reach p.u.b.erty three to four years sooner than average.

The average number of growing days per season is approximately 160.

Amount of growth in any season depends on climatic factors that influence food supply and foraging conditions. Growth rate is directly correlated with precipitation, being highest when large populations of gra.s.shoppers and long periods of favorable weather occur in the same year. Zones of epidermis formed in years when growth was especially slow or especially fast const.i.tuted landmarks that were helpful in interpreting growth-histories. Growth stops two to three years after p.u.b.erty. The total growing period is estimated to be not more than 15 to 20 years. Longevity is estimated to be approximately 50 years.

A number of changes in structure and appearance occur in the period from hatching to p.u.b.erty. Fontanelles of the bony sh.e.l.l close at or before p.u.b.erty. Movable parts of the plastron are not functional until the fourth year. Markings on the carapace change from a series of dots to distinct, straight-sided radiations, and a similar pattern develops on the plastron. Markings on the heads of females resemble those of juveniles but males have greenish heads. Males further differ from females in having a red iris, more brightly colored antebrachial scales, and a turned in first toe.

a.n.a.lysis of some 500 body temperatures (Centigrade) obtained under natural conditions revealed the following: the optimum temperature for activity is near 30 degrees; box turtles emerge from cover usually when body temperature is 24 degrees or higher, and almost never when the body temperature is below 15 degrees; body temperature is raised to optimum by basking in open areas although activity begins at suboptimum temperatures if basking is impossible; cover of dens, burrows, or forms is sought when the body temperature rises above 30 degrees; and, maximum and minimum body temperatures that would be lethal to box turtles (for prolonged periods) are approximately 40 and zero degrees, respectively. Laboratory experiments showed speed of response to environmental temperature to be inversely proportional to bulk; hatchlings could be chilled or warmed more than twice as fast as adults and were active within a narrower range of temperature. Ornate box turtles in general are subject to a narrower range of thermal activity than are aquatic turtles that occur in the same areas.

Box turtles are dormant approximately five and one-half months of the year--from late October to mid-April. Warm weather in November and late March sometimes stimulates temporary activity but dormancy is uninterrupted from mid-November to early March. Forms, dens, and burrows are used as hibernacula. Depth of hibernacula is dependent on severity of temperatures and amount of vegetational cover; hibernacula in open gra.s.sland were seven to 18 inches deep whereas those in wooded areas were six inches or shallower. Box turtles are ordinarily solitary when hibernating. Injuries and deaths due to freezing probably occur in the coldest part of the winter. The lowest body temperature of a turtle that survived a winter was 2.7 degrees; an individual, the temperature of which was nearly zero for several days, subsequently died. Turtles burrow upward at the end of hibernation and remain just below the surface for a week or two before emerging. The primary stimulus for emergence seems to be a period of warm moist weather.

Populations of _T. ornata_ observed under natural conditions were chiefly carnivorous, although captives ate a variety of animal and vegetable matter. Insects, consisting chiefly of beetles, caterpillars, and one species of gra.s.shopper, comprised approximately 89 per cent (by volume) of the food present in stomachs. Beetles (chiefly scarabaeids and carabids) are obtained in or near dung and seem to const.i.tute the most important staple element of the diet.

Piles of dung, disturbed by turtles in the course of their foragings, were characteristic "sign" of _T. ornata_ in the areas studied.

Insects form the bulk of the diet for most of the year, although certain other foods, when especially abundant for short periods (mulberries for example), are eaten in large quant.i.ty or eaten to the exclusion of all other foods. Ornate box turtles occasionally eat the eggs and young of ground-nesting birds and slightly damage vegetables, but in no instance do these feeding habits significantly affect the economy of man. Box turtles probably benefit man by destroying large numbers of crop-damaging insects (locustids and noctuid caterpillars).

Box turtles were more numerous than most kinds of reptiles at the Damm Farm and were the most conspicuous of any kind of reptile. One hundred and ninety-four turtles were marked; one-fourth of these were recaptured at least twice. Population density in certain areas of favorable habitat ranged from 2.6 to 6.3 turtles per acre. The total number of individuals on the study area was estimated to be 286. The marked population consisted of 53 per cent adult or subadult females, 31 per cent adult males, and 16 per cent juveniles of undetermined s.e.x. Only six individuals had plastra shorter than 60 millimeters.

Small box turtles are not so rare as these samples indicate; they are infrequently obtained because their smallness and ability to blend with the substrate make them difficult to see. More females than males were found in all months of the season of activity, excepting April and August when more males were found; the preponderance of females was greatest in the nesting season (June and July).

Ornate box turtles walk with the sh.e.l.l held off the substrate. They are able to climb steep embankments or low barriers with some facility. Swimming ability is sufficient to permit survival in water and traversal of water-barriers but ornate box turtles almost never swim voluntarily.

Daily activity consists of periods of basking, foraging, and rest, the durations of which are influenced by temperature and humidity. There is no activity after dark except that of nesting females. After several days of activity there is a period of rest; rest periods seemed not to be correlated with climatic conditions. The average distance traveled per day in summer is 200 to 300 feet. Movements of gravid females are more extensive (average, 363 feet per day) than those of other members of the population; one individual traveled approximately one-fourth of a mile in a single day. Turtles removed from their normal home ranges traveled farther per day than any other group. Movements in autumn are less extensive (average, 152 feet per day) than at other times in the season of activity.

Individual box turtles tended to remain in small areas for long periods; these areas were interpreted as home ranges. The estimated average radius of 44 home ranges was 278 feet (average area, 5.6 acres). The average area of eight home ranges that were actually measured was five acres. General suitability of habitat and certain physical features of terrain (rock fences, ravines, barren fields) that acted as barriers were thought to be the most important factors governing size of home range. Of two turtles removed more than one-fourth of a mile from their home ranges, one homed and one did not. Home ranges of turtles of all ages and s.e.xes overlap broadly.

There was no indication that territoriality or social hierarchy existed in the population studied.

Box turtles are subject to injury from natural causes that include fire, cold, molestation by predators, and trampling by cattle.

Automobiles and farm machinery now const.i.tute major causes of mortality and serious injury. Capacity to recover after serious injury is great but there is increased chance for secondary injury, infection, and predation in the period of recovery. Pits on the sh.e.l.l from unknown causes ("sh.e.l.l pitting") are less common in ornate box turtles than in other kinds of turtles.

Ectoparasites infesting _T. ornata_ are larvae of chigger mites (genus _Trombicula_) and larvae of bot flies (_Sarcophaga cistudinis_).

Ectoparasites usually have little adverse effect on the turtles, although heavy infestations cause occasional injury or death.