126 
Michael D. Stuart and Grover C. Miller 
this species in forested habitat throughout North Carolina, South Caro- 
lina, and Virginia up to an elevation of about 1220 m. All specimens 
collected in this study were adjacent to or in mixed hardwood or mixed 
hardwood-pine forests. 
During the hot, dry period from late June through early Sep- 
tember, turtles were often found active immediately after a rainstorm or 
very early in the morning. During these dry spells they were found rest- 
ing in or quite near standing pools of water. It seemed that the seasonal 
activity of the box turtle is temperature dependent. Ernst and Barbour 
(1972) described T. Carolina as emerging from winter hibernation in 
April and returning in October or November. Schwartz and Schwartz 
(1974) stated that T. c. triunguis in Missouri usually emerged from hi- 
bernation between the end of March and the end of April. This gener- 
ally coincided with the last severe frost in the spring. Return to hiberna- 
tion in mid-September to early November also coincided with the first 
killing frost of the fall. The earliest date of collection in our study was 
28 April with the latest date being 1 1 December (Fig. 4). 
Box turtles are opportunistic omnivores. Principal foods are mush- 
rooms, snails, and insects (Table 2). Of the 104 specimens examined in 
our survey, 72 contained recognizable food items. The following were 
identified, in order of percent of frequency of occurrence (Fig. 5): snails 
59%, insects 43%, Armadillidium vulgare 40%, plants material (mainly 
fungus) 32%, slugs 7%, rodents 5.5%, earthworms 3%, and millipeds 
3%. The low level of earthworm consumption agrees with Klimstra and 
Newsome (1960) in contrast to most generalized reports of feeding hab- 
its. The vertebrate material in all these reports is almost certainly the 
result of scavenging carrion, although Legler (1960) reported Terrapene 
ornata capturing and killing small chicks. 
Ernst and Barbour (1972) stated that juvenile box turtles were 
mainly carnivorous but became more herbivorous with age. However, 
no such difference in dietary preference was noted between juveniles and 
adults or between adult males and females in the 72 North Carolina 
specimens that contained recognizable food items. Both Stickel (1950) 
and Klimstra and Newsome (1960) reported seasonality and availability 
as important factors in determining the use of food items. The dietary 
differences between juveniles and adults reported by Ernst and Barbour 
(1972) may have been influenced by seasonality or availability of certain 
food items. 
Box turtles are reported to mate in the spring or occasionally in the 
fall, laying two to eight eggs in each of two or three clutches between 
late May and July (Smith 1961, Huheey and Stupka 1967, Mount 1975). 
Of the 43 females examined in this study, 14 contained shelled eggs. 
