Date of Graduation


Document Type


Degree Type



Davis College of Agriculture, Natural Resources and Design


Forest Resource Management

Committee Chair

James T Anderson


Amphibian populations have decreased in many parts of the world and the rate of decline has increased over the past 25 years. Much of the population decline can be attributed to habitat fragmentation, thus, possibly forming metapopulations. The semi-aquatic northern dusky salamander belongs to the family Plethodontidae and the genus Desmognathus. Amphibians, such as salamanders, may occupy undisturbed forest floors with biomass equal to, or exceeding, the biomass of other vertebrate groups and they can achieve their highest densities in ancient or undisturbed forests. Salamander densities can be estimated using cover items and this method has become a more common practice because of its relatively non-disruptive impact on the ecosystem, the ability to attract species that are difficult to trap in pitfalls, minimized observer biases and errors and reliability of developing a model estimating population size. The goals of this study were to 1) evaluate short-term changes in seasonal relative abundance of northern dusky salamanders; 2) determine if there was an increase in their relative abundance where cover boards were placed; 3) determine which habitat parameters influence relative salamander density; and 4) create a relative abundance model.;Four study areas each with 4 stream reaches were searched by using a 3--4 pass visual encounter survey and wooden cover boards (n = 50 boards/stream reach) during 2008 and 2009. A total of 2,287 salamanders from 7 species were captured and the northern dusky salamander comprised 87% of the total. Eighty--four percent of all the salamanders were captured under rocks while 9% were captured under cover boards. Salamanders were marked with a color coded visual implant elastomer and no northern dusky that was captured in one stream was then recaptured in a different stream indicating their strong site fidelity and limited dispersal ability. Of the 6 a priori models evaluated to estimate salamander density, the model using tree canopy cover, protective cover, and stream size proved to be the best fit (lowest corrected Akaike's Information Criterion). The best fit model was then slightly modified (post-hoc) to incorporate stream size differences and a constant was added. The post-hoc model was verified at an independent study area and able to estimate (with a percent error of ≤ 75% of the estimated number of salamanders/m) the salamander density 75% of the time. Additional habitat parameters (i.e., tree stand age, water quality, substrate embeddedness, and stream flow) may need to be measured to increase the accuracy of the post-hoc model.;The 3 critical habitat parameters, ranked in order, were tree canopy cover, protective cover, and stream size with the first 2 being directly related to salamander density and the latter suggesting that a stream may be too small or too large to support salamanders. Furthermore, we determined that streams with deficiencies in 1 of these 3 factors had relatively lower salamander densities and streams that were deficient in 2 or 3 of these factors had the lowest density estimates. Captured juvenile (snout to vent length [SVL] = 18.17 mm, mass = 6.99 g) and adult (SVL = 50.68 mm, mass = 14.15 g) northern dusky salamanders in this study were similar in size (length and mass) to northern dusky salamanders captured in other studies. Likewise, our salamander densities were similar (X¯ = 0.15, SE = 0.02 salamanders/m) to other studies.;Our study supported the idea that salamanders are indicator species sensitive to riparian habitat conditions and that a single poor riparian zone characteristic may significantly impact the salamander population. As the amount of logging and natural gas drilling continues to increase in Pennsylvania and northeastern United States, wildlife managers could use this information to help manage riparian habitat, especially if the habitat is scheduled to be impacted by some type of anthropogenic disturbance (i.e., road). Furthermore, this study provides wildlife managers with a model that can estimate salamander densities pre-- and post--disturbance. With this model, wildlife managers can use it to evaluate the quality of the stream with respect to the northern dusky salamander.;Keywords: Northern dusky salamander, Desmognathus fuscus fuscus, habitat fragmentation, Pennsylvania, salamanders, visible implant elastomer, visual encounter survey, metapopulation.