- Laura Monti
Masters Student, University of Maine, Orono, Maine
The redback salamander, Plethodon cinereus, is one of the most abundant vertebrates in eastern North American forests, with densities often exceeding 1 per sq. m. Given its numbers and sensitivity to habitat conditions, it may be useful as a bioindicator of environmental degradation. Prior to using redbacks in this way, their habitat preferences in a variety of locations needs to be assessed, and an effective method of monitoring their density needs to be developed.
At the Holt Research Forest (HRF) in Arrowsic, ME, 60 redback salamander census stations were established in 1988. Each station consists of six l0x25 cm cedar shingles under which redbacks retreat to avoid desiccation. The stations were checked for salamanders every other week from May through September. With data from 1989 through 1995. I used multiple regression to correlate spatial patterns of salamander abundance with characteristics of the vegetation and soil around the stations. I also performed experiments in 1996 to address relationships between salamander abundance and coarse woody debris, O horizon pH and moisture, and overstory cover.
I found that the relationships between these habitat variables and redback numbers were complex. Short term changes in moisture conditions do not appear as important as the average yearly precipitation in influencing density patterns. The pH of the O horizon at the HRF was generally at or below the purported chronically lethal level for this species, yet laboratory experiments revealed that HRF salamanders are as sensitive to acidity as those from a forest with more basic soils. Gaps in the canopy correlate with areas of low redback density; related variables such as O-horizon depth may also influence this relationship. Although the initial multivariate model suggested a negative relationship between coarse woody debris and redback density, subsequent experiments failed to substantiate this. In sum, redback densities did not correlate well with the habitat variables I investigated.
I also addressed the robustness of the censusing methodology used at the HRF against possible sources of bias. I evaluated the effects of natural cover objects near the stations, social interactions between redbacks, and methodological variables (such as shingle age and placement) on salamander use of the shingles. I established new stations to determine if the animals at the stations were drawn from natural cover objects and the relative quality of natural vs. artificial cover.
This method was relatively unaffected by territorial behavior, age of shingles, and presence of natural cover objects. However, the length of time the shingles had been in place, the depth of the shingles, and seasonal changes in salamander abundance created biases in station counts. Despite these drawbacks, I concluded that the use of artificial cover is superior to two other censusing techniques, transect walks and quadrat searches, for long-term monitoring of redback population densities.