EFFECTS OF ROADS ON AMPHIBIAN COMMUNITY STRUCTURE AT BREEDING PONDS IN RHODE ISLAND

15 March 2001

Dr. Peter Paton
Associate Professor, Department of Natural Resources Science, University of Rhode Island, Kingston RI 02881

Scott Egan
Graduate Assistant, Department of Natural Resources Science, University of Rhode Island, Kingston RI 02881

EXECUTIVE SUMMARY
(1) We investigated the potential impact of roads and habitat fragmentation on amphibian community structure at breeding       ponds. This study represents one of the few attempts to quantify these relationships on pond-breeding amphibians in       North America.
(2) We subdivided the state into 1258 ha hexagons, then randomly selected 10 hexagons within each of three road densities       (low [rural], medium [suburban], and high [urban]). Within each hexagon, we located potential breeding ponds by visually       inspecting orthophotoquads.
(3) After gaining permission, we surveyed 119 ponds (56 low density, 45 medium density, and 18 high density ponds) during       the 2000 field season.
(4) At each pond, we conducted egg mass counts for wood frogs and spotted salamanders, used minnow traps to assess fish       presence, and conducted three rounds of dip net surveys to determine tadpole presence. Therefore, each pond was       visited 5-6 times from March through mid-August.
(5) We detected 11 species of amphibians during fieldwork; wood frogs and spotted salamanders were the most common       species (>72% of ponds), while marbled salamanders (<5% of ponds) were the least common species.
(6) We analyzed species occurrence patterns at three spatial scales (macrohabitat <200m from ponds, landscape <2000 m       from ponds, and regional (1258 ha hexagons). Within each scale, we quantified road density, hydroperiod, and habitat       characteristics (e.g., amount of forest, wetlands, barren lands, and urban lands).
(7) At all spatial scales, hydroperiod tended to be among the most important variables that explained species occurrence       patterns. Only wood frogs tended to occur most frequently in ponds with shorter hydroperiods, whereas other species       preferred longer hydroperiods.
(8) At macrohabitat scales (<200 m from ponds), four species (wood frog, red-spotted newt, spotted salamander, marbled       salamander) tended to occur in more forested landscapes, whereas most other species were more closely linked to open       habitats. At this scale, road density was not a significant variable for any species, although forested habitats tended to       occur in areas with fewer roads.
(9) At the landscape scale (<2000 m from ponds), wood frog, red-spotted newt and spotted salamander presence was       positively associated with forests, while American toads and green frog presence was positively associated with       urbanization. In fact, green frogs were positively associated with roads, which was the only species for which roads       showed any relationship, either negative or positive.
(10) When species occurrence patterns were analyzed at the hexagon scale (1258 ha cells), we found that pickerel frogs,        green frogs, bullfrogs, and American toads were not affected by urbanization, whereas gray treefrog, red-spotted newt,        spring peeper, wood frog, and spotted salamander detection probabilities declined in urban environments.
(11) Results from this study suggest that some species (red-spotted newt, wood frog, and gray treefrog) may be more        vulnerable to extirpation in urban environments, whereas other species appear to be less susceptible. Urban areas appear        to have three major impacts on amphibians, (1) they tend to have fewer ponds with shorter hydroperiods than adjacent        rural areas, (2) they have less forested habitat available, which provides wintering habitat, and (3) roads may act as       dispersal barriers to some species.