Orientation of Movements and Habitat Selection in a Spatially Structured Population of Marbled Salamanders (Ambystoma opacum)

—Most studies on orientation of movements of pond-breeding salamanders have considered only a single local population (or breeding pond) during 1 yr, or multiple populations during a single year. We quantified migratory patterns of Marbled Salamanders at nine breeding ponds during 5 yr in western Massachusetts. Based on captures at drift fences, movements were nonuniform at all breeding ponds. In addition, the direction of orientation differed among breeding ponds and changed slightly across years. Within ponds, orientation of adults and juveniles differed significantly in 52% of comparisons, and adult movements were more directionally concentrated than those of juveniles. In addition, migrating salamanders shifted slightly the orientation of their movements as they traveled into uplands, suggesting that migration routes are spatially complex and that determination of migration ‘‘corridors’’ based on concentrated captures at the pond periphery may be misleading. Although salamanders used migration routes with higher canopy cover, our models did not explain a large portion of the variation in orientation, and protecting areas of high canopy cover alone may not be sufficient as a protection strategy. Our results suggest that movement routes, though perhaps concentrated in the short term, are unpredictable in the long term. Thus, we can offer little evidence that distinct corridors can be identified and protected that would be used consistently over time by migrating or dispersing Marbled Salamanders. Consequently, until we better understand mechanisms governing movements in this species, a conservative conservation strategy would require protecting broad terrestrial areas around breeding sites. The scientific and conservation communities increasingly recognize the significance of upland areas in proximity to wetlands—both as terrestrial habitats for pond-breeding amphibians and as potential corridors for movement between wetlands (Semlitsch, 2000). A better understanding of movements between uplands and breeding sites is therefore critical to conservation of pond-breeding amphibians, especially as habitat loss and fragmentation increase. Movements among breeding sites may be especially important given complex spatial and temporal dynamics of amphibian populations and the likelihood that many species persist in constantly changing spatially structured populations (e.g., Skelly et al., 1999). Despite being such an important component of population structure and dynamics, little is known about the factors affecting such movements in pond-breeding amphibian populations (Pough et al., 2004). Previous studies on movements in pondbreeding amphibians suggest that individuals may use ‘‘corridors’’ (i.e., areas of concentrated use) as they migrate to and from breeding sites (Dodd and Cade, 1998; Johnson, 2003). In particular, a number of studies have documented nonuniform orientation of movements for mole salamanders (Ambystoma; Shoop and Doty, 1972; Douglas and Munroe, 1981; Stenhouse, 1985; Phillips and Sexton, 1989). Nonuniform orientation of movements during migration may indicate selection for certain habitats in the uplands surrounding breeding ponds (Stenhouse, 1985). Indeed, individuals of some species emigrate along relatively straight paths to upland areas containing small mammal burrows, which are utilized as terrestrial refuges (Madison, 1997; Jehle and Arntzen, 2000). In addition, there may be selection against movement in or toward disturbed or open-canopy areas (DeMaynadier and Hunter, 1998). Regardless of the mechanisms underlying nonuniform orientation of movements, the conservation implications are clear. If salamanders exhibit nonuniform orientation during migration, then these movement corridors can be targeted for protection. The majority of salamander orientation studies have considered only a single local population (or breeding pond) during 1 yr, or multiple populations during a single year (e.g., Shoop and Doty, 1972; Douglas and Munroe, 1981; Stenhouse, 1985; Phillips and Sexton, 1989; Dodd and Cade, 1998; 1 Corresponding Author. Present address: Greater Yellowstone Program, Wildlife Conservation Society, 1780 1st Street, Idaho Falls, Idaho 83401, USA; E-mail: [email protected] Journal of Herpetology, Vol. 40, No. 2, pp. 240–248, 2006 Copyright 2006 Society for the Study of Amphibians and Reptiles