A Gis-Based Habitat Model Predicting Elk Nutritional Condition in the Pacific Northwest

Abstract

Declining habitat quality has been implicated in population declines in elk (Cervus elaphus) in the Pacific Northwest (PNW) although specific causes are unknown. Nutrition and summer-autumn body condition (percent body fat) have been linked to virtually every aspect of successful ungulate reproduction and productivity; however, many existing habitat evaluations discount the role of nutrition and are unlikely to detect changes in habitat nutritional quality. I applied an algorithm linking the biomass of suitable elk forage (BSF) and dietary digestibility (DD) to model habitat quality for elk at 3 sites in the PNW: 1 in western Oregon (Wendling), 1 in southwestern Washington (Pe Ell), and 1 in northwestern Washington (Nooksack). BSF was best predicted by overstory canopy cover (R2&ge;0.70) and I developed remote sensing-based maps of canopy cover (accuracy &ge;74%) to characterize DD levels at each site. In a GIS, I estimated habitat quality for the overall landscape, and within actual wild elk homeranges. I compared model predictions to autumn fat levels in free-ranging lactating elk. Wendling (DD=55.9%, n=14, SE=0.02) and Pe Ell (DD=55.8% n=19, SE=0.02) afforded only "marginal" habitat quality while Nooksack (DD=61.8%, n=19, SE=0.05) had significantly higher quality (P<0.0001). Fat levels in wild elk corresponded to "marginal" levels at Wendling (10.8% fat, n=10, SE=0.65) and "good" levels at Nooksack (13.6% fat, n=8, SE=0.45). Pe Ell elk (6.1% fat, n=9, SE=0.54) were below "marginal" levels but also had the highest density of wild elk (5 elk/km2) indicating a potential negative feedback between density and condition. I compared BSF/DD predictions to those of 2 forage based and 4 HSI/HEI-type elk habitat models and compared these to wild elk condition BSF/DD models detected marked differences in habitat capacity among sites and accurately predicted elk condition. Forage based models overestimated habitat capacity at all sites (P&le;0.001). HSI/HEI-type models did not did not reveal any reliable trends in habitat capacity or wild elk condition at these sites. Hence, habitat evaluation and management focused on road closures, forage-to-cover ratios, or total forage biomass are unlikely to address elk nutritional condition.