Logistic regression is an important tool for wildlife habitat‐selection studies, but the method frequently has been misapplied due to an inadequate understanding of the logistic model, its interpretation, and the influence of sampling design. To promote better use of this method, we review its application and interpretation under 3 sampling designs: random, case‐control, and use‐availability. Logistic regression is appropriate for habitat use‐nonuse studies employing random sampling and can be used to directly model the conditional probability of use in such cases. Logistic regression also is appropriate for studies employing case‐control sampling designs, but careful attention is required to interpret results correctly. Unless bias can be estimated or probability of use is small for all habitats, results of case‐control studies should be interpreted as odds ratios, rather than probability of use or relative probability of use. When data are gathered under a use‐availability design, logistic regression can be used to estimate approximate odds ratios if probability of use is small, at least on average. More generally, however, logistic regression is inappropriate for modeling habitat selection in use‐availability studies. In particular, using logistic regression to fit the exponential model of Manly et al. (2002:100) does not guarantee maximum‐likelihood estimates, valid probabilities, or valid likelihoods. We show that the resource selection function (RSF) commonly used for the exponential model is proportional to a logistic discriminant function. Thus, it may be used to rank habitats with respect to probability of use and to identify important habitat characteristics or their surrogates, but it is not guaranteed to be proportional to probability of use. Other problems associated with the exponential model also are discussed. We describe an alternative model based on Lancaster and Imbens (1996) that offers a method for estimating conditional probability of use in use‐availability studies. Although promising, this model fails to converge to a unique solution in some important situations. Further work is needed to obtain a robust method that is broadly applicable to use‐availability studies.

During the past 2 decades, the grizzly bear (Ursus arctos) population in the Greater Yellowstone Ecosystem (GYE) has increased in numbers and expanded in range. Understanding temporal, environmental, and spatial variables responsible for this change is useful in evaluating what likely influenced grizzly bear demographics in the GYE and where future management efforts might benefit conservation and management. We used recent data from radio-marked bears to estimate reproduction (1983)(1984)(1985)(1986)(1987)(1988)(1989)(1990)(1991)(1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002) and survival (1983)(1984)(1985)(1986)(1987)(1988)(1989)(1990)(1991)(1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001); these we combined into models to evaluate demographic vigor (lambda [k]). We explored the influence of an array of individual, temporal, and spatial covariates on demographic vigor.We identified an important relationship between k and where a bear resides within the GYE. This potential for a source-sink dynamic in the GYE, coupled with concerns for managing sustainable mortality, reshaped our thinking about how management agencies might approach longterm conservation of the species. Consequently, we assessed the current spatial dynamic of the GYE grizzly bear population. Throughout, we followed the information-theoretic approach. We developed suites of a priori models that included individual, temporal, and spatial covariates that potentially affected reproduction and survival. We selected our best approximating models using Akaike's information criterion (AIC) adjusted for small sample sizes and overdispersion (AIC c or QAIC c , respectively).We provide recent estimates for reproductive parameters of grizzly bears based on 108 adult ( .3 years old ) females observed for 329 bearyears. We documented production of 104 litters with cub counts for 102 litters. Mean age of females producing their first litter was 5.81 years and ranged from 4 to 7 years. Proportion of nulliparous females that produced cubs at age 4-7 years was 9.8, 29.4, 56.4, and 100%, respectively. Mean (6SE) litter size (n ¼ 102) was 2.0 6 0.1. The proportion of litters of 1, 2, and 3 cubs was 0.18, 0.61, and 0.22, respectively. Mean yearling litter size (n ¼ 57 ) was 2.0 6 0.1. The proportion of litters containing 1, 2, 3, and 4 yearlings was 0.26, 0.51, 0.21, and 0.02, respectively. The proportion of radio-marked females accompanied by cubs varied among years from 0.05 to 0.60; the mean was 0.316 6 0.03. Reproductive rate was estimated as 0.318 female cubs/female/year. We evaluated the probability of producing a litter of 0-3 cubs relative to a suite of individual and temporal covariates using multinomial logistic regression. Our best models indicated that reproductive output, measured as cubs per litter, was most strongly influenced by indices of population size and whitebark pine (Pinus albicaulis) cone production. Our data suggest a possible density-dependent response in reproductive output, although perinatal mortality could...

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org..Wiley-Blackwell and Nordic Society Oikos are collaborating with JSTOR to digitize, preserve and extend access to Oikos.