Oecologia (Berlin) (1987) 73:60-70 Oecologia 9 Springer-Verlag1987 Temporal and spatial variations in the behavior of shrubsteppe birds J.A. Wiens l*, B. Van Horne 1'**, and J.T. Rotenberry 2 1 Department of Biology, University of New Mexico, Albuquerque, NM 87131, USA 2 Department of Biological Sciences, Bowling Green State University, Bowling Green, OH 43403, USA Summary. We documented patterns of activity and micro- habitat use by breeding male Sage Sparrows (Amphispiza belli) and Brewer's Sparrows (Spizella breweri) in the semi- arid shrubsteppe of northwestern U.S.A. to determine how their behavior varies in space and time and how these varia- tions relate to environmental features. In analyses of several sites with multiple plots that were studied for 2-4 yr, the frequencies of activities and substrate use varied significant- ly with year, plot, and location. The major activities, forag- ing and singing, varied inversely with one another. Thus, although the general structure of the activity budgets of these species was stable (i.e. foraging and singing were al- ways the most frequent activities), there was considerable variation in the quantitative details of their behavior in space and time. Using stepwise regression, we related these variations in behavior to variations in environmental fea- tures such as the densities of the same or other species, the number of species in the community, and coverages of the different substrate types. Few patterns emerged, and those that did were generally not in accord with our expec- tations. In particular, patterns of substrate use were not related to variations in coverages of those substrates. Sage Sparrows, for example, increased their use of sagebrush and grass with increasing grass coverage and with increas- ing densities of Sage Sparrows. Responses to both variables suggested nonlinear threshold patterns. We interpret the variations in behavior as indicating a loose coupling between behavior and proximate environ- mental conditions for these species. In such situations, the variability of behavior complicates attempts to test ecologi- cal hypotheses (e.g. behavioral niche partitioning) with data from a limited number of locations or years. Key words: Behavior Birds- Shrubsteppe Determining how organisms behave and how they use mi- crohabitats is important to testing hypotheses in various areas of ecology. Behavioral differences among species, for * Present address and address for offprint requests: J.A. Wiens, Department of Biology and Natural Resource Ecology Labora- tory, Colorado State University, Fort Collins, CO 80523, USA ** Present address." Department of Biology, Colorado State Uni- versity, Fort Collins, CO 80523, USA example, were a primary focus of MacArthur's (1958) work on niche partitioning among Dendroica warblers, and many studies since then have described how organisms partition resources through differences in activity patterns or sub- strate use. Optimal foraging theory has been tested by ex- amining how individuals apportion their time among habi- tat patches (Krebs et al. 1983), and physiological studies have related the activity patterns and microhabitat selection of individuals to energy demands and heat exchanges (Bryant and Westerterp 1980, Mahoney 1976, Wiens 1986). To address questions in these areas, one must document behavioral patterns in quantitative detail. Although many workers now recognize this requirement (e.g. Morrison 1984; Altmann 1984), the need to determine the extent to which behavioral patterns are invariable, species-specific at- tributes or exhibit flexibility and variation is less widely appreciated. Which of these alternatives holds in a given situation may have profound effects on how one goes about testing ecological hypotheses with behavioral data. Whether one finds behavior to be relatively stable or variable also depends on one's perspective or the questions one asks. If, for example, one's focus is on the general, qualitative structure of activity budgeting, the finding that a species always spends more of its time foraging than singing and more time singing than in aggressive displays may indicate that there is little overall variation in behavior. If instead one focuses on the particular behaviors, the fact that the amount of time spent in foraging varies substantially be- tween years or locations may be of considerable impor- tance, even though foraging is always the dominant behav- ior in the activity budget. Here, we consider the behavior and substrate use of Sage (Amphispiza belli) and Brewer's ( Spizelta breweri) spar- rows, focusing on the frequency with which particular be- haviors occur and the extent to which this varies in time and space. These species are widespread and abundant in the Artemisia-dominated shrubsteppe of North America and are usually the numerically dominant bird species in this system (Rotenberry and Wiens 1980b). They are gener- ally similar in patterns of habitat association (Wiens and Rotenberry 1981, pers. observ.) and occupy broadly over- lapping territories where they are syntopic. In this paper, we assess (1) the degree of stereotypy or variability in activi- ties or substrate use by these species in relation to location and time and (2) how these patterns relate to habitat fea- tures, population densities, and community composition.