-We studied sexual and individual variation in habitat use by pine snakes, Pituophis melanoleucus (5 males, 5 females), in the New Jersey Pine Barrens. Both sexes were found in pine-oak forests. Habitat selection varied by sex and individual, with males using logs and bark extensively while females often were found under oak leaves. For over 50% of the observations snakes were near blueberry or pitch pine, although males were near blueberry twice as frequently as females. There were no sexual differences in activity patterns, but individuals differed in habitat use and activity. Some snakes remained in only one habitat, whereas others moved between habitats. Descriptions of habitat use are numerous for reptiles (e.g., Heatwole, 1977), but individual differences have seldom been quantified. Most studies of habitat use by snakes provide quantitative descriptions of where the snakes generally occur (Wright and Wright, 1957; Reinert and Kodrich, 1982; Weatherhead and Charland, 1985; Burger and Zappalorti, 1986) but do not compare where the sexes are. Reinert (1984a, b) performed a detailed analysis of habitat use by rattlesnakes (Crotalus horridus) and copperheads (Agkistrodon contortrix mokeson), but the topics of sexual or individual differences in habitat use by snakes remain essentially unstudied. In this paper we examine sexual and individual differences in habitat use and activities of 10 radiotracked pine snakes (Pituophis m. melanoleucus) in the New Jersey Pine Barrens. Our goal was to obtain detailed data on how individuals use habitat, whether males and females inhabit the same habitats, and if there is potential for resource competition between the sexes. Further, we examine the relative importance of specific habitat features for individuals and show that a feature may be important to one individual but not to others. Pine snakes occur in a variety of habitats in North America, but are nowhere common (Wright and Wright, 1957). In the New Jersey Pine Barrens, pine snakes occur in pitch pine (Pinus rigida) and scrub oak (Quercus marylandica) woodlands, but avoid white cedar (Thuja occidentalis) swamps and lowland pitch pine areas (Zappalorti and Burger, 1986). Thus, snakes generally choose moist, but not wet ground cover and some vegetation that provides shade (Burger and Zappalorti, 1988). METHODS AND STUDY AREA We studied pine snakes near Toms River, Ocean County, New Jersey in 1981 and 1982. The study area contained residential commuce competition between th sexes. Further, a ine the relative importance of spe ific 68 This content downloaded from 207.46.13.13 on Fri, 26 Aug 2016 06:17:05 UTC All use subject to http://about.jstor.org/terms HABITAT USE IN SNAKES nities and forested sections of pitch pine, scrub oak and white cedar. Ten snakes (5 males, 5 females) were located soon after emergence from hibernation in early spring (March, April) in 1981 (4 snakes) or while nesting (June-early July) in 1982 (6 snakes). Snakes were sexed by everting hemipenes (Gregory, 1983), measured (SVL), and weighed. Radio transmitters (AMV Instrument Company SM 1; 3 g, 2.5 cm long) were inserted surgically in the body cavity, following procedures in Reinert and Cundall (1982) and Burger and Zappalorti (1988). Prior to implantation in the study animals, transmitters were surgically implanted in other pine snakes that were then observed for two weeks to ascertain normal behavior. The study area was searched 3-4 times a week from June through December in 1981 and 1982. Transmitter batteries did not last through the winter, and male 2 disappeared three weeks after release. When a snake was located, the following data were collected: forest type (Forman, 1979), elevation, soil type (U.S.D.A., 1980), disturbance type, nearest vegetation (over 5 cm high), nearest ground cover, distance to water, cloacal temperature, air temperature (sun and shade) and snake activity. Forest type, nearest vegetation and soil type were not necessarily correlated since the same vegetation may grow on different soils. Vegetation type and ground cover type were noted for each snake. In this area of the pine barrens, pine trees are sometimes sparse, and other vegetation is scattered between the trees. We distinguished four forest types: (1) oakpine (over 50% oak), (2) pine-oak (50-80% pine), (3) pine lowland (> 80% pine), and (4) cedar bog (no pine). We recorded soil types (Atsion, Berryland, Lakehurst, Lakewood; U.S.D.A., 1980) because we suspected females might be found on a particular soil suitable for digging nests. A site was considered disturbed if it showed evidence of past human activity (clearing for agriculture, hunting lodges, railroads, or logging roads), undisturbed if it did not. Snake activity was recorded as: (1) inactive (in the shade or hidden), (2) basking in the sun, (3) basking in partial shade, and (4) moving. We defined basking as a snake thermoregulating on a clear day with 50% or more of its body exposed to direct sun (basking in sun), or with less than 50% exposed to direct sun (basking in shade; after Parker and Brown, 1980). Wilcoxon X2 tests and Contingency Table X2 tests were used to determine differences in the distribution of variables among groups, and stepwise multiple regression procedures with maximum R selection criterion were used to identify the environmental variables that contributed to the variation in the dependent variables (Barr et al., 1976). The procedure determines the model giving the highest R2 with the fewest variables, and gives the R2 and F values for the contribution of each variable to the dependent measures. The model's procedure eliminates the effect of one variable (e.g., date), before examining the effect of another variable. A variable that varies colinearly with another is not entered in the model, and these are not presented.