Investigating terrestrial isopod abundance in sandplain grassland using a multiple linear regression

Most North American species of terrestrial isopod (Isopoda) have been introduced from Europe. Sandplain grassland is a globally rare habitat that is abundant on Nantucket Island, Massachusetts and the abundance of terrestrial isopods in the habitat has never been studied. The objective of this project was to develop a model to explain isopod abundance based on vegetation characteristics within Sandplain grassland and use this model to test for land management effects (prescribed burning and mowing) on isopod abundance. I counted terrestrial isopods from 175 pitfall traps set for one week and used multiple linear regression with several selection algorithms to select the best model. The vegetation characteristics I used as regressors do not appear to explain terrestrial abundance well and the final model only contains the percent grass coverage as a regressor. The model suggests that terrestrial isopods decrease in abundance with increasing grass coverage and it explains 29 percent of the data. When management effects are incorporated, the model suggests that mowing significantly increases isopod abundance.

Funding for this project came from the Nantucket Islands Land Bank, Nantucket Land Council, and the Nantucket Biodiversity Initiative.

Associated vegetation data is in the published "Effects of Sandplain Grassland Management on Spider Richness and Abundance on Nantucket Island" dataset.  Sampling methods are in the thesis linked from that dataset.

allisopodData.csv - isopod counts by trap
dataDictionary.csv - descriptions of variables
mckenna-foster_2009.pdf - a report submitted to NBI and used as part of a statistics class at the University of Wisconsin-Green Bay