Portal de Periódicos da CAPES

Drift Fence-Associated Sampling Bias of Amphibians at a Florida Sandhills Temporary Pond

1991; The Society for the Study of Amphibians and Reptiles; Volume: 25; Issue: 3 Linguagem: Inglês

10.2307/1564587

1937-2418

C. Kenneth Dodd,

Animal Behavior and Reproduction

Five years of field observations suggest that amphibians inhabiting a sandhills community are able to circumvent a drift fence-pitfall trap enclosure as they move toward and away from an ephemeral pond. Trespass rates varied depending on species and showed no tendency to increase or decrease as the study progressed. Laboratory trials confirmed that frogs easily crossed the fence by walking up the side or hopping over it. Frogs crossed the fence readily regardless of sex or whether the frog was an adult or juvenile. Although striped newts did not climb over the fence in the laboratory, they may use tunnels to go under fences under field conditions. Other species may burrow directly under the fence. A priori assumptions about a species' ability to climb a fence, or that trespass rates do not vary temporally or among sites, are unwarranted and may lead to misinterpretations of the results from studies using drift fences and pitfall traps. The use of a drift fence with pitfall traps is a common sampling technique for studies of individual species or herpetofaunal communities (e.g., Dole, 1971; Gibbons and Bennett, 1974; Gibbons and Semlitsch, 1982; Semlitsch and Moran, 1984; Semlitsch and Pechmann, 1985; Gibbons, 1990). The fences may be constructed of many types of materials, including metal, plastic, or tarpaper, and can be arranged to completely enclose a specific area or sample habitats in a variety of non-continuous configurations (Campbell and Christman, 1982; Gibbons and Semlitsch, 1982; Vogt and Hine, 1982; Jones, 1986). The results of studies employing drift fences with pitfall traps have provided valuable insights to population and community ecology and behavioral patterns of secretive and difficult to study species. One underlying assumption of studies employing drift fences with pitfall traps is that they effectively sample the community or population in question. However, several authors have noted that some species may not be amenable to accurate census using this technique, particularly large snakes, large turtles, or treefrogs (Gibbons and Bennett, 1974; Gibbons and Semlitsch, 1982; Jones, 1986). Few investigators in amphibian studies address the importance of trespass, i.e., climbing over or under the fence (Semlitsch, 1985; Semlitsch and Pechmann, 1985). For example, Pechmann and Semlitsch (1986) stated that they had no reason to suspect diel variation in trespass rates, whereas Pechmann et al. (1989) stated that trespass rates probably did not vary among sites and years. Some investigators do not even mention trespass (e.g., Shoop and Doty, 1972). From January 1985 until September 1990, I conducted a study of the herpetofaunal community inhabiting an ephemeral wetland in the high region of north-central Florida (Dodd and Charest, 1988). Initially, large numbers of unmarked animals were caught inside the drift fence, and I assumed that these animals were within the enclosed area when the fence was installed. However, relatively large numbers of unmarked animals were still captured within the enclosure during the fourth and fifth years of monitoring. During this period, north-central Florida experienced a pronounced drought. Because of the drought, Gastrophryne carolinensis successfully reproduced only during the summer of 1985 at Breezeway Pond. Juveniles left the pond area during the fall of 1985 and during the spring of 1986 (Dodd, unpubl. obs.), but movements back and forth to the pond and immigration from other breeding sites complicated the interpretation of pitfall trap catch results in 19851986 for juveniles of this species. Only 42 unmarked recently-metamorphosed Notophthalmus perstriatus were found leaving the pond in 1987. No reproduction occurred for these species during the other years of the study, and no other species successfully reproduced in Breezeway Pond from 1985 through 1990. All of the unmarked animals may have been within the enclosure at the start of the project. For example, Semlitsch (1983) noted adult and metamorphosing Ambystoma tigrinum used a 15 m wide strip of habitat between a drift fence and pond perimeter during a long-term study on the Savannah River Plant, South Carolina. As the study progressed, his sampling efficiency increased so that by the second year 90% of the breeding population was censused. AlternaThis content downloaded from 157.55.39.113 on Sun, 15 May 2016 06:18:10 UTC All use subject to http://about.jstor.org/terms DRIFT FENCE SAMPLING BIAS tively, some unmarked animals may have entered the enclosure by crossing the fence. This paper reports the results from my examination of field capture data on the eight most common amphibians and laboratory observations on the ability of five common amphibian species to cross a drift fence. These results have important implications for the interpretation of the results of studies using drift fence enclosure sampling regimes. STUDY AREA AND METHODS Field data were recorded at Breezeway Pond, a 0.16 ha depression marsh (Florida Natural Areas Inventory, 1990) located in a shallow 1.3 ha basin on the Katharine Ordway Preserve/ Swisher Memorial Sanctuary, Putnam County, Florida. The pond is surrounded by a longleaf pine (Pinus palustris)-turkey oak (Quercus laevis)wiregrass (Aristida stricta) community to the south and west, a maidencane (Panicum hemitomon) meadow to the east, and a xeric oak hammock dominated by sand live oak (Q. geminata) and laurel oak (Q. laurifolia) to the north. The distance to the nearest forest cover is generally 20-30 m, but extends to about 50 m behind the Panicum meadow. No vegetation overhung the drift fence; the fence and pitfalls were exposed to direct sun. Breezeway Pond is a depression marsh located in an uplands sandhill community. As such, it does not have a flowing water source. Instead, the duration of hydroperiod is dependent upon localized rainfall and recharge from groundwater, which in turn is dependent upon the level of the water table in the surrounding uplands. From 1985-1990, little rain fell in the vicinity of Breezeway Pond, resulting in short hydroperiods at various times of the year rather than hydroperiods which could form a dependable source of water corresponding with amphibian breeding requirements. Although the maximum recorded depth was 75 cm, the pond held water only 14 mo from January 1985 through September 1990. A 230 m drift fence made of aluminum flashing (36 cm above ground, 10-15 cm below the surface) encircled the pond. Within the enclosure, herbaceous hydrophytic vegetation dominates the basin although a few shrubs, including buttonbush (Cephalanthus occidentalis), myrtle holly (Ilex myrtifolia), and wax myrtle (Myrica cerifera), are present. Several sapling longleaf and slash pines (P. elliottii) grow within the enclosure. Maidencane and carpetgrass (Axonopus furcatus) comprised 76% of the ground cover on vegetation transects (LaClaire and Smith, unpubl. obs.). Pitfalls (19 L buckets) were placed on opposite sides of the fence at 10 m intervals. The pitfalls were checked 5 days per week between 0700 and 0900 h depending on season from February through April 1985, and from October 1985 through September 1990. A board covered the buckets on days when the pitfalls were not checked. Vegetation was kept cut and away from the fence, exposing bare white sand for about 40 cm from the base of the fence in either direction. Animals were marked by toe-clipping using a year-specific cohort number, and released on the opposite side of the fence. Animals were carefully examined for toe regeneration. Dodd and Charest (1988) provide additional details, including methodology and site description, of the project at Breezeway Pond. An experimental chamber (74.5 x 30 x 53 cm) was set up in the laboratory to test the ability of five common amphibian species (Acris gryllus, Gastrophryne carolinensis, Hyla femoralis, Limnaoedus ocularis, and Notophthalmus perstriatus) to climb or cross over a drift fence. The bottom of the chamber was covered with moist paper towels, and a piece of aluminum drift fence (80 cm length, 35 cm height) was wedged securely diagonally in the chamber, dividing it into two equal compartments. I coated the glass above the top of the fence and at the contact between fence and glass with a thick strip of petroleum jelly to discourage animals from walking up the glass and from using the glassfence contact to cross over the fence. Thus, to go from one side of the chamber to the other, an animal had to walk up the fence or hop over it. A clear plastic lid covered the experimental chamber to prevent escape. Individuals for laboratory trials were caught at the drift fence at Breezeway Pond or by hand at nearby Breezeway Sandhills Pond. Animals were tested individually or in small groups of 2-10 individuals immediately after capture; different species were not tested together. An individual or group of individuals was placed on one side of the chamber in the afternoon (16001800 h). Location was recorded the next morning (0600-0900 h). Records were made of the sex or life stage (adults vs. juveniles) of individuals that crossed and did not cross the fence. After testing on one side, the individual or group was placed on the other side of the chamber and the observations were repeated to see if prior experience or the orientation of the chamber with regard to its position in the room influenced behavior. Overnight experiments were supplemented by direct observation. Testing was conducted at room temperature (24-26 C). I used the X2 test of independence with Fisher's Exact Test (two-tailed) where appropriate for statistical analyses. Except for 1985, a year was defined as extending from October through 297 This content downloaded from 157.55.39.113 on Sun, 15 May 2016 06:18:10 UTC All use subject to http://about.jstor.org/terms C. KENNETH DODD, JR.