Pitfalls and Snap Traps for Sampling Small Mammals and Herpetofauna

Abstract

Experiments were conducted to compare the efficiency of drift-fence pitfalls with three types of snap traps for sampling small mammals and to determine the usefulness of pitfalls for sampling herpetofauna. During winter 1984, small mammal capture rates were lower in pitfalls and Victor rat traps than in Victor mouse traps (P<0.01). During summer and autumn 1984, capture rates in pitfalls were greater than in Victor mouse traps (P<0.01). During winter and summer 1985, capture rates of pitfalls and Museum Specials were similar. Pitfalls with drift fences also captured 220 individual reptiles and amphibians. INTRODUCTION Some form of trapping is often used to sample the members of a local fauna. Most authors agree that a combination of trapping methods yields the most complete data on community composition and reduces biases associated with various traps (Getz, 1961; Smith et al., 1971; Weiner and Smith, 1972). Boonstra and Rodd (1984) and others have compared conventional live traps with pitfalls for a single species of small mammal. However, we know of no studies comparing the effectiveness of snap traps and pitfalls for a large assemblage of species. We compare three snap trap types and pitfalls with drift fences as methods for collecting small mammals, and examine the effectiveness of pitfalls for sampling herpetofauna. STUDY AREAS Data were collected from Abbeville, Edgefield, Greenwood, McCormick and Saluda counties in the Piedmont Plateau Region of western South Carolina. Twenty-eight sites within young (0-21 years old), even-aged loblolly pine (Pinus taeda) stands were sampled. Criteria for site selection included seed-tree or plantation regeneration, chop-and-burn site preparation and access (Smith, 1962). Stands previously thinned or with a recent history of fire were not selected. Study sites were located on the Long Cane Division of the Sumter National Forest and lands owned by International Paper Company. Average tree age was 6.6 years (SE = 1.2 years) and tree density was 3078 pines/ha (SE = 489.0 pines/ha). Average annual temperature and precipitation in the study counties was 16.7 C and 150 cm (NOAA, 1983), respectively. Stand size varied from 4-200 ha and averaged 58.1 ha (SE = 9.95 ha). MATERIALS AND METHODS Trap locations were determined by drawing grid lines on a map of each study site and randomly selecting five lines. Pitfalls were arbitrarily placed in the center of the trap lines and equidistant from the nearest two lines. Field placement of lines and pitfalls was achieved by using a compass and pacing from the stand boundary or road intersection. By preselecting trap locations, we reduced any bias associated with subjective placement of trap lines and pitfalls. Pitfall traps. -We used pitfalls with drift fences to sample small terrestrial vertebrates. A single cross-shaped array with four fence sections was used on each site (Campbell and Christman, 1982). All fences were constructed of 5-m sections of aluminum flashing placed in a ditch 8-12 cm deep. All sections of fence were 15 m apart. Sites were randomly assigned a fence height of either 25.4 cm or 50.8 cm. A 19-liter plastic bucket was buried flush with the ground at either end of each fence section. 'Present address: Division of Life Sciences, Ferrum College, Ferrum, Va. 24088.