313 Factors Associated with Fecal Glucocorticoids in Alaskan Brown Bears (Ursus arctos horribilis) Christina G. von der Ohe 1, * Samuel K. Wasser 2 Kathleen E. Hunt 2 Christopher Servheen 3 1 Wildlife Biology Program, University of Montana, Missoula, Montana 59812; 2 Department of Zoology, University of Washington, Seattle, Washington 98195; 3 U.S. Fish and Wildlife Service, Wildlife Biology Program, University of Montana, Missoula, Montana 59812 Accepted 5/2/03 ABSTRACT The aims of this study were to validate a radioimmunoassay (RIA) for quantifying glucocorticoid metabolite concentrations in the feces of Alaskan brown bears (Ursus arctos horribilis) and to investigate whether any of the following factors are associated with those concentrations: the presence of humans or other bears, fishing difficulty, sex-age class, diet, and season. We tested an established corticosterone RIA for assay sensitivity, similar- ity, precision, and sample matrix effects of brown bear feces, and it proved satisfactory. We collected fecal samples from brown bears along salmon-spawning streams and assessed fecal glucocorticoid (FG) concentrations. We observed that the fac- tors explaining the most variation in measured concentrations were date and diet type and that there was a significant inter- action between the two. We did not observe a significant effect of human and bear activities or sex-age class on FG concen- trations. This study demonstrates that although FG concentra- tions may be assessed in brown bears, complex dietary patterns and seasonal variations must be taken into consideration in the study design in order to make inferences regarding stress. Introduction An adverse stimulus is known to initiate a physiological cascade of responses, which provides resources necessary to cope with * Corresponding author. Present address: Department of Biological Sciences, 371 Serra Mall, Stanford University, Stanford, California 94305-5020; e-mail: vonderohe@stanford.edu. Physiological and Biochemical Zoology 77(2):313–320. 2004.  2004 by The University of Chicago. All rights reserved. 1522-2152/2004/7702-2072$15.00 a stressor (reviewed in Moberg 1987). One such response is the activation of the hypothalamic-pituitary-adrenocortical (HPA) axis, resulting in synthesis and secretion of glucocor- ticoids by the adrenal cortex. Concentrations of these hormones have been used by many studies as a physiological index of stress in animals. Traditionally, quantitative measurements of physiological stress have involved assessing glucocorticoid concentrations in blood plasma. The use of physiological stress measures on free- ranging animals has been limited due to the invasiveness and potential for bias of capturing and withdrawing blood from wild animals. Moreover, corticosteroid secretion into blood var- ies diurnally and has pulsatile secretory patterns (Monfort et al. 1993), causing sample plasma glucocorticoid concentrations to be highly variable. Fecal steroid measures now provide an appealing alternative to serum sampling. Samples for fecal glu- cocorticoid (FG) analysis are relatively easy to collect and can be gathered without disturbing study subjects. This approach also provides a smoothed estimate of glucocorticoid concen- trations over a longer time period than serum sampling, be- cause of the pooling effect of adrenocorticosteroids in feces (Harper and Austad 2000). This new technique is being applied to an increasing number of mammalian species (e.g., Monfort et al. 1998, African wild dog; Goymann et al. 1999, spotted hyenas; Strier et al. 1999, muriquis; Wallner et al. 1999, barbary macaques; Harper and Austad 2000, mice and voles; Wasser et al. 2000, multiple species; Millspaugh et al. 2001, elk). This study extends the technique to brown bears (Ursus arctos horribilis). For bears, environmental factors such as human activities, presence of other bears, and food availability are po- tential sources of cognitive stress that may affect FG concen- trations. In elk, human disturbance, as measured by road den- sity and use, has been shown to correlate with FG levels (Millspaugh et al. 2001). However, factors independent of cog- nitive stress, such as circannual physiological rhythms and diet type, may also affect these concentrations. Seasonal changes in serum glucocorticoid concentrations have been demonstrated in bears (Palumbo et al. 1980; Harlow et al. 1990) and other hibernators (Shivatcheva et al. 1988, European ground squirrel; Armitage 1991, yellow-bellied marmots; Boswell et al. 1994, golden-mantled ground squirrels). Similarly, diet type may alter both serum and FG concentrations through a variety of chan- nels. Diet can affect gut transit time, which in turn influences reabsorption of steroid hormones (Lewis et al. 1997). Also, FG concentrations can be augmented by ingestion and excretion of glucocorticoids from the diet (Cooper et al. 1996). Finally,