Forest succession and prey availability influence the strength and scale of terrestrial‐aquatic linkages in a headwater salamander system

Abstract

1. Trophic linkages between terrestrial and aquatic ecosystems are common and sensitive to disruption. However, there is little information on what causes variation in the strength and spatial scale of these linkages. 2. In the highly aquatic adults of the headwater salamander Gyrinophilus porphyriticus (family Plethodontidae), use of terrestrial prey decreases along a gradient from early-to late-successional riparian forests. To understand the cause of this relationship, we tested the predictions that (i) terrestrial prey abundance is lower in late-successional forests, and (ii) G. porphyriticus adults cannot move as far from the stream to forage in late-successional forests, thus limiting access to terrestrial prey. 3. We established 100-m long study reaches on six headwater streams in the Hubbard Brook Experimental Forest, New Hampshire. Three reaches were in early-successional forests and three were in late-successional forests. We conducted pitfall trapping for invertebrate prey in June and July of 2005, with three traps at 0, 2, 5 and 10 m from the stream at each reach. In June, July and August of 2004 and 2005, nighttime salamander surveys were conducted at each reach along ten, 10-m long by 2.5-m wide transects perpendicular to the stream. 4. Abundance of terrestrial prey was consistently lower in late-successional forests, suggesting that consumption of terrestrial prey by G. porphyriticus is affected by prey abundance. Contrary to our prediction, G. porphyriticus adults moved farther from the stream in late-successional forests, suggesting that habitat conditions in late-successional forests do not limit movement away from the stream, and that lower abundances of terrestrial prey in these forests may cause salamanders to move farther from streams. 5. Our results provide novel insight on the extent of terrestrial habitat use by G. porphyriticus. More broadly, these results indicate that major habitat gradients, such as forest succession, can affect the strength and scale of terrestrial-aquatic linkages. Application of this insight to the design of vegetation buffers along headwater streams would have widespread benefits to freshwater ecosystems.

Can natural selection maintain long-distance dispersal? Insight from a stream salamander system

“…The generality of this hypothesis comes from the assumption that conditions determining the quality of empty sites are randomly distributed, but the conditions determining site quality are expected to be species-specific. For G. porphyriticus in Merrill Brook, where predatory fish are absent, these conditions may include availability of flow refugia (Baumgartner et al 1999;Lowe et al 2004), abundance of invertebrate prey (Burton 1976;Greene et al 2008), or abundance of terrestrial predators (Brodie et al 1979). Based on growth rates of G. porphyriticus individuals, quality of occupied sites appears to be randomly distributed in Merrill Brook (Lowe 2010), but we cannot test this assumption for empty sites without knowing the specific determinants of site quality.…”

“…2), and between leg PC2 and Cormack-Jolly-Seber estimates of survival, which confound mortality with permanent emigration (White and Burnham 1999;Lowe 2010). Extensive terrestrial dispersal would be surprising in light of what we know about the natural history, morphology, and behavior of G. porphyriticus (Brandon 1966;Petranka 1998;Greene et al 2008), but is possible. Also, Merrill Brook is close to the northern edge of the geographic range of G. porphyriticus, raising the possibility that dispersal-related phenotypes are maintained in part by spatial sorting-where traits enhancing dispersal rate accumulate at expanding range edges, and assortative mating between fast-dispersing individuals at the range edge results in an evolutionary increase in dispersal rates in successive generation (Shine et al 2011).…”

“…1; Lowe 2003Lowe , 2009). Second, G. porphyriticus is constrained to linear stream corridors; larvae are restricted to the stream channel, and although adults can move short distances into riparian forests at night (\3 m from the stream, on average), they return to the stream during the day (Greene et al 2008). The probability of detecting marked animals declines with movement distance in twodimensional landscapes (Koenig et al 1996), but when movement is along linear habitats like stream corridors and sampling is consistent throughout the available habitat (e.g., from bank to bank), detection probability is not affected by movement distance .…”

“…During the day, larvae and adults are found in interstitial spaces among the larger substrate particles of the streambed. In the northern Appalachians, larval size range is 26-80 mm snout-vent length (SVL), adults can reach 120 mm SVL, and both stages feed primarily on invertebrates (Greene et al 2008). …”

The Definition and Measurement of Individual Condition in Evolutionary Studies

“…; Reinhardt ), amphibians (Greene et al. ; Grayson & Wilbur ), and reptiles (Husak ; Joanne et al. ).…”