## Habitat Evaluation: Do Use/Availability Data Reflect Carrying Capacity? > [!Cite]- > Hobbs, N. Thompson, and Thomas A. Hanley. “Habitat Evaluation: Do Use/Availability Data Reflect Carrying Capacity?” _The Journal of Wildlife Management_ 54, no. 4 (October 1990): 515. [https://doi.org/10.2307/3809344](https://doi.org/10.2307/3809344). > > [link](https://www.jstor.org/stable/3809344?origin=crossref) [online](http://zotero.org/users/local/kycSZ2wR/items/FC4BNXUK) [local](zotero://select/library/items/FC4BNXUK) ## Notes %% begin notes %% %% end notes %% %% begin annotations %% ## Annotations | evernote 2016.04.25 We advocate developing habitat evaluation systems that are based on a mechanistic understanding of relations between resource acquisition by individuals and the dynamics of populations. In most cases, carrying capacity (the capability of land to maintain and produce animals of a given species) is the stated or implied measure of habitat value (e.g., Flood et al. 1977, U.S. Fish Wildl. Serv. 1981, Cooperrider 1986, Verner et al. 1986). That assumption usually is based on the following reasoning: "It is assumed that (1) a species will select and use areas that are best able to satisfy its life requirements; and (2) as a result, greater use will occur in higher-quality habitat" (Schamberger and O'Neil 1986:9). Although there may be a logical relation between habitat preference and habitat quality at the level of the individual animal, the relation between habitat preference and carrying capacity at the level of the population is much less clear. Recently, Fagen (1988) dismissed the arguments of Van Horne (1983), viewing her results as a special case. Fagen (1988) contended that habitat value will be directly related to use/availability ratios whenever animals are free to choose the habitat that confers the greatest fitness. Specifically, he argued that when animal distributions are "ideal free" (sensu Fretwell and Lucas 1970, Fretwell 1972), the carrying capacities of habitats will be directly proportional to use/avail- ability indices, regardless of when those indices are obtained.  differences in population density among habitats. We assume further that limitations imposed by habitat resources on animal density can be described in at least 2 dimensions, a dimension defining the quantity of the resource and a dimension defining its quality. That is, we assume that both the quantity and quality of resources within a habitat deter- mine natality (b,) and survival (lx) of animals using that habitat. We define resource quality such that higher quality allows greater 1, and/ or b, per unit of available resource. Dimensions of quality and quantity that de- fine resource distributions may or may not be substitutable. A substitutable distribution of re- sources is one where deficiencies in resource quality can be directly offset by an increased abundance of the resource. For example, a predator may be able to compensate for reductions in prey size by eating a large number of prey, an outcome that is possible if sufficient prey are available. In contrast, other resources lack substitutability between quantity and quality. For example, a ruminant cannot fully compensate for low dry-matter digestibility by eating more dry matter. In these cases where quantity cannot substitute for quality, increasing resource abundance merely reduces intraspecific competition for re- sources, but cannot make up for inherent deficiencies in those resources Taken together, our simulations (Fig. 3) illustrate that animal distributions will reflect habitat carrying capacities only when (1) animals are spatially distributed in an ideal free manner, (2) environmental conditions permit long-term, stable equilibria between animal populations and limiting resources, and (3) use/availability data are obtained after equilibria are achieved. Given the importance of natural and man-caused disturbance in creating temporal variation in the relations between animal populations and their habitats (Pickett and White 1985, Turner 1987), we argue that the above assumptions will be rarely satisfied. Our results are congruent with those of Fahrig and Paloheimo (1988), whose simulations revealed that carrying capacities of habitat patches were unrelated to population densities of animals using those patches. Instead, the ability of animals to disperse and their ability to detect new habitats emerged as important regulators of realized population densities. Given such differences in resource distributions, we argue that the preference of individuals for habitats will change as population density changes. Even when assumptions of ideal free distributions are met, observations of patterns of habitat use may be unrelated to rates of population increase or to equilibrium densities. It follows that short-term measures of habitat use/ availability will reveal little about the value of habitat unless the underlying resource distributions limiting population growth are understood mechanistically. In evaluating habitats and in predicting the consequences of habitat change, however, such a mechanistic understanding makes the need for applying use/availability data unnecessary in the first place. We suggest that future approaches to habitat evaluation should focus on mechanisms linking the performance of animal populations to resources in the habitats they use.  First, habitat evaluation must be related to some objective for habitat. In our example simulations, habitat a offered the greatest population growth rates, habitat P the greatest equilibrium density. Which is better? By definition, evaluation is a judgement of worth. No matter how quantitative our appraisal of habitat characteristics may be, com- parisons of the worth of 1 habitat relative to another only have meaning in reference to some achieved result and the values we place on it.  Linking characteristics of habitats to objectives for populations can be achieved only when we understand the mechanistic relationship between those characteristics and the desired behavior in the ani- mal population.  Retrospective analyses of habitat relationships \[e.g., correlative models\] depend on the conditions prevailing when the analyses were conducted, however, and offer no predictive value when those conditions shift (Rotenberry 1986).  Forecasting the consequences of habitat change depends on understanding the processes that control animal response to change.  An understanding of the cause-and-effect relations linking the performance of animal populations to the resources in their habitats is fundamentally important to evaluating habitat. Correlations based on simple surrogate variables (e.g., density, habitat use/availability indices) offer un- reliable inferences on habitat value. %% end annotations %% %% Import Date: 2024-08-15T10:39:44.318-06:00 %%