Evaluation of the habitat suitability index models from the black-capped chickadee and downy woodpecker.

Abstract

Habitat Suitability Index (HSI) models for the black-capped chickadee (Parus atricapillus) and downy woodpecker (Picoides pubescens) were used to evaluate the model's abilities in predicting species response to habitat disturbances. HSI models identify habitat variables assumed to limit the success or occurrence of a species. After quantifying these habitat variables, a measure of species response was compared with the model output to determine the relationship between habitat and species abundance. Avian censuses and habitat sampling techniques were conducted at 25 impact and 25 control sites to evaluate the effects of stream channelization on the populations of the black-capped chickadee and downy woodpecker. Censuses were conducted between 1 June and 11 July, 1988. Habitat sampling was conducted between 31 May and 31 September, 1988. Variables representing the food component of the chickadee model are the average height of the overstory and percent canopy cover. Snag availability represents the reproduction component of the model. No relationship was found between black-capped chickadee densities and the HSI (r = -0.01, P > 0.5). The lack of relationship between HSI and chickadee densities was largely due to the ineffectiveness of the sampling technique used to estimate the reproduction component of the model. However, the food component of the black-capped chickadee model was able to closely predict the upper limits of species response (r = 0.93,P < 0.005). These results suggest that the assumptions concerning the positive correlation between chickadee densities and canopy volume were valid. Black-capped chickadee densities (P = 0.06) and habitat units (P = 0.08) were found to be higher at control sites than at impact sites. It appears that these channelization projects do affect the populations of the black-capped chickadee. continued review of future channelization projects is suggested to minimize the extent of channelization and the amount of riparian vegetation removed. Variables for the downy woodpecker are basal area and snag availability, representing the food and reproduction components, respectively. Downy woodpecker densities showed no relationship with model output values. The method of sampling snags was not adequate in estimating snag availability. The amount of time and area used in the censusing procedure did not allow adequate detection of the downy woodpecker because of its larger home range.