PARENT SESSION
Poster Session 28: Avian Ecology
Thursday, August 11, 5:00 PM - 6:30 PM, Exhibit Hall 220 A-E, Level 2, Palais des congrès de Montréal

Modeling forest bird species' habitat with extant presence points and generated pseudo-absence points in Utah.

Zarnetske, Phoebe^*,1, Edwards, Thomas², ¹ Utah State University, Logan, UT² USGS Utah Cooperative Fish and Wildlife Research Unit, Logan, UT

ABSTRACT- Modeling management indicator species' habitat and estimating their likelihood of occurrence across an extensive landscape can assist managers and ecologists in assessing a species' status on a broad scale. In order to create a statistically valid habitat classification model, both presence and absence points are necessary. Unfortunately, available occurrence data for landscape-scale modeling is often lacking and usually only in the form of observed presences. Combining generated pseudo-absence points and extant presence points can improve species habitat modeling. We are assessing several habitat classification models in Utah forests at three different resolutions (30, 90, and 250-meter), each for Northern Goshawk (Accipiter gentilis) and for American Three-toed Woodpecker (Picoides dorsalis) based on the species' extant presence and generated pseudo-absence points. Habitat parameters include spatially-explicit data from: Forest Inventory and Analysis (FIA), USDA Landfire vegetation models, Daymet climatological summaries, digital elevation models, and U.S. Forest Service fire, beetle-kill, and timber surveys. Sets of species' pseudo-absence points were generated within habitat variable envelopes, encompassing nest site habitats for A. gentilis and foraging habitats for P. dorsalis. Use of pseudo-absence points to accompany extant presence points in regression analysis of habitat models can be a powerful tool in habitat modeling and conservation of species.

Key words: habitat model, pseudo-absence, Accipiter gentilis, Picoides dorsalis