Habitat associations of pool-breeding amphibians in eastern Massachusetts,USA

There is a pressing need to develop a sound conservation strategy for pool-breeding amphibians, which includes gaining a better understanding of the habitat and landscape-scale characteristics associated with populations. To investigate relationships between amphibian species richness and characteristics of breeding pools and surrounding landscapes, we surveyed 85 pools in eastern Massachusetts (USA) in 1996 and 1997. A total of 11 species was detected, with most pools having 2–5 species. Pools were typically small, 77.6% were <0.2 ha, but most pools (72%) had hydroperiods that persisted at least into August in most years. Based on linear regression analyses, species richness was positively associated with three within-pool variables (pool surface area, hydroperiod, and the amount of emergent vegetation), and a landscape-level variable (presence of another breeding pool within 1 km), while one within-pool variable (tree canopy cover) exhibited a significant negative association with species richness. These within-pool habitat variables and connectivity to other breeding pools are important characteristics to consider when attempting to identify breeding sites that could provide core habitat in conservation reserves designed for the conservation of pool-breeding amphibian species richness. Conservation reserves for pool-breeding amphibian populations should include pool complexes functioning as habitat for metapopulations. Core pool habitats within such reserves should be large (0.5–1.0 ha), with seasonal hydroperiods that persist into August or that dry in some years, and with sufficient emergent vegetation to provide diverse microhabitats and refugia.     

Double-observer approach to estimating egg mass abundance of pool-breeding amphibians

Interest in seasonally flooded pools, and the status of associated amphibian populations, has initiated programs in the northeastern United States to document and monitor these habitats. Counting egg masses is an effective way to determine the population size of pool-breeding amphibians, such as wood frogs (Rana sylvatica) and spotted salamanders (Ambystoma maculatum). However, bias is associated with counts if egg masses are missed. Counts unadjusted for the proportion missed (i.e., without adjustment for detection probability) could lead to false assessments of population trends. We used a dependent double-observer method in 2002–2003 to estimate numbers of wood frog and spotted salamander egg masses at seasonal forest pools in 13 National Wildlife Refuges, 1 National Park, 1 National Seashore, and 1 State Park in the northeastern United States. We calculated detection probabilities for egg masses and examined whether detection probabilities varied by species, observers, pools, and in relation to pool characteristics (pool area, pool maximum depth, within-pool vegetation). For the 2 years, model selection indicated that no consistent set of variables explained the variation in data sets from individual Refuges and Parks. Because our results indicated that egg mass detection probabilities vary spatially and temporally, we conclude that it is essential to use estimation procedures, such as double-observer methods with egg mass surveys, to determine population sizes and trends of these species.