66 
John M. Hranitz et al. 
lands than natural populations on the adjacent mainland, and Gib- 
bons et al. (1979) attributed this disparity in size to higher quality 
diets and warmer temperatures on the islands. The eastern hognose 
snake ( Heterodon platyrhinos) is smaller on Assateague Island, Vir- 
ginia, than on the adjacent mainland (Edgren 1961, Scott 1986). 
The basis for the smaller H. platyrhinos size on Assateague Island 
compared to the size of animals on the mainland is not known. 
In typical amphibians, the effects of mainland versus island 
ecology on amphibian biology are more difficult to predict because 
environmental factors might differentially affect the aquatic and ter- 
restrial phases of the life cycle. We studied the size and abundance 
of Bufo woodhousii fowleri in three different habitats on Assateague 
Island compared with a population in three different habitats on the 
adjacent Delmarva peninsula. 
MATERIALS AND METHODS 
We captured Bufo woodhousii fowleri in drift fences and pit- 
fall traps (Gibbons and Semiitch 1981) randomly placed in three 
habitats on the island (146.78 ha) and mainland (30.45 ha) study 
areas. There was one replicate for drift fences and pitfall traps in 
each habitat at each location in 1988 and one replicate for each 
location in 1989. Upon capture, toads were measured (SVL, mm), 
weighed (g), sexed, toe clipped (Clarke 1972) and released. Recap- 
tured toads were measured and weighed before release, but the sizes 
of recaptured toads are not presented because periods between re- 
capture events were too brief to accurately assess growth in either 
toad population. 
In 1988, toads were trapped in coniferous forests, meadows, 
and primary dunes on Assateague Island (37°56'N; 75°2CTW) from 
2 to 29 June and in deciduous forests, coniferous forests, and 
meadows on the adjacent mainland (37°56"N; 75°29'W) from 2 to 
16 July. We stopped trapping on the mainland when it was appar- 
ent that about 50% of the population in the area was marked (Davis 
1982). In 1989, toads were trapped concurrently from 10 June to 
10 July in coniferous forests and meadows on the island, and in 
deciduous and coniferous forests on the mainland. The Schnabel 
method (Schnabel 1938, Smith 1980) was used to estimate the mean 
abundance of toads in habitats at each location. The assumptions 
inherent in mark-release-recapture studies, constant population size 
and random samples (Schnabel 1938), were met because we could 
detect recruitment, represented by juveniles, and because the place- 
ment of drift fences and pitfall traps at each location was random. 
Abundance and sex ratios were analyzed with the Systat™ 
ANOVA procedure. Adult SVL and mass were analyzed with a 
