HIERARCHICAL SUMMER HABITAT SELECTION BY THE NORTH AMERICAN PORCUPINE IN WESTERN MONTANA
The North American porcupine (Erethizon dorsatum) has continually been a species of low priority for research and conservation efforts in the western United States. While many porcupine populations have been studied in North America, little to no research has been done on porcupines in western Montana or habitats of similar composition. More recently it has been noted that porcupine sightings have become rare in western Montana is areas where sightings were once common. Due to the overall lack of information about western porcupine populations it is imperative to gather basic natural history information in order to make informed management decisions. This study strives to quantify home range and hierarchical habitat selection at the second, third, and fourth order, while noting both mortality and reproductive rates. Second and third order habitat selection was analyzed by estimating resource selection functions using multiple logistic regression with a logit link, where matched case-control logistic regression with a logit link was used at the fourth order. Akaike’s Information Criterion for small sample size was used as a measure of goodness-of-fit to select between alternative models. Model evaluation at each scale was performed using k-folds cross validation. Public response to a sighting survey distributed throughout western Montana was used to quantify use points at the second order. Individual locations of twelve radio-collared porcupines were used to quantify use points at the third order and fourth order. Fourth order habitat selection was analyzed in three groups to isolate the influence of each group on selection. The groups are as follows: stand structure, dominant stand species, and species utilized. Home ranges were calculated using kernel density estimators (the 99% percentile) and minimum convex polygons. Mortality rates were estimated using Kaplan-Meier staggered entry methods. Reproductive rates were estimated as the proportion of marked females known to have reproduced. At the second order the top model showed porcupines avoided habitats dominated by grasslands as well as higher elevations. This model also showed porcupines selected for habitats dominated by mixed forests, pasture, wooded wetlands, and herbaceous wetlands. At the third order the top model showed avoidance of habitats that were further from permanent water sources, while it showed selection for habitats that were dominated by wooded and herbaceous wetlands. At the fourth order the top model showed selection for deciduous species (hawthorn, cottonwood, and willow). In contrast to some previous research porcupines were found highly selective across all scales as would be expected from a specialist species. There is evidence to suggest that this population is limited by the availability of lower elevation riparian areas and deciduous communities.
© Copyright 2008 Katie Ann Mally