Wetland and Nest Scale Habitat use by the Four-Toed Salamander (Hemidactylium scutatum) in Maine, and a Comparison of Survey Methods

Chalmers, Rebecca

' Conserving amphibian populations requires knowledge of species and its habitat relationships. The Four-Toed salamander (Hemidactylium scutatum) is listed as Special Concern in Maine and 11 additional states and provinces, Threatened in Illinois, and Endangered in Indiana (Appendix A). Little is known of H. scutatum ecology despite the speciesÆ extensive range. Infrequent sightings of H. scutatum throughout its range may indicate either low numbers or that speciesÆ behavior make detection difficult. Records for H. scutatum in Maine existed from only 32 sites before my study, and the total number of occurrences of this speices in Maine is unknown (P. deMaynadier, Maine Department of Inland Fisheries and Wildlife, personal communication). I found 238 four-toed salamander nests in 36 wetlands, which were new sites. The survey protocol, natural history descriptions, and definitions of wetland and shoreline habitat presented will increase detections of this species and thus improve the monitoring and management of H. scutatum and the wetland habitat that this species occupies. In the first chapter, I compare monitoring techniques for the species, document new points in which I found the species, and present natural history information. I conducted surveys of adults on roads during rainy spring nights, surveys of nests, and surveys of larvae with dipnetting in wetlands, and I present incidental visual encounters and pitfall captures for comparison. Greatest numbers of salamanders were found with the nest surveys, which were conducted in palustrine wetlands by walking in the water and parting the shoreline vegetation to search for eggs and attendant females. I found H. scuatum nests in 35 of 92 wetlands intensively searched. In the second chapter, I investigate species-habitat relationships that predict H. scutatum presence at two scales: the wetland (and surrounding landscape) and the available shoreline points in which nests could occur. I collected data at wetlands with and without nests, and I collected data along the shoreline at points with and without nests. With these data, I created models that predicted wetlands with nests, and I created and evaluated models that predicted nest point selection within a wetland and available point characteristics between wetlands with and without nests. Wetlands with nests were best predicted by higher pH and were negatively associated with shrub scrub and unconsolidated bottom NWI classes. Wetlands with nests were also predicted by the availability of shoreline points that provided Sphagnum spp. for egg attachment, wood substrate, water flow, the presence of blue-joint reed grass (Calamagrostis canadensis), meadowsweet (Spirea alba), steeplebush (Spirea tomentosa), and sensitive fern (Onclea sensibilis), and the absence of sheep laurel (Kalmia angustifolia) and deciduous forest NWI class. Within wetlands with nests, shoreline points with nests were best distinguished from shoreline points without nests by steeper shores, greater near-shore and basin water depth, deeper nesting vegetation, presence of moss and winterberry (Ilex verticillata), and negative association with Spirea alba, leatherleaf (Chamaedaphne calyculata), and Kalmia angustifolia within 1 m of the shoreline point.'