Lesson in the Limitations of Field Experiments: Shrubsteppe Birds and Habitat Alteration

We manipulated the habitat composition and structure within territories of birds breeding in the Artemisia—dominated shrubsteppe of central Oregon in order to assess how closely individuals track habitat in features such as territory placement of size or behavioral budgeting and space. We removed 75, 50, 25, and 0% of the shrub individuals from 625—m² blocks in a checkerboard designs. Over the following 7 yr we monitored territorial locations and sizes and breeding densities of Horned Larks (Eremophila alpestris), Sage Sparrows (Amphispiza belli), and Brewer's Sparrows (Spizella breweri) on the manipulated area and adjacent unaltered control area. Activity and substrate use patterns and intensity of use of the different treatment blocks within the manipulated area were also determined for the two sparrow species. Sage Sparrows, which from our previous studies appear to be more closely linked to shrub (sagebrush) coverage than the other species, did not respond to the manipulation at a local population level. The manipulation apparently did affect territory placement and densities of Brewer's Sparrows and Horned Larks in 1980—1982, but these trends were not consistent over the entire postmanipulation period. Densities of all species varied among years. For Brewer's Sparrows and Horned Larks these variations did not parallel those in nearby census locations, but Sage Sparrow densities varied in the same ways over all the locations. These variations appeared to parallel variations in bioyear (October—April) precipitation, with a 1—yr time lag. Individuals of both sparrow species altered the details of their activity budgeting, but not their substrate—use patterns, in response to the manipulation. Both species clearly expressed an affinity for unmanipulated blocks within the manipulation area. This experimental manipulation is complicated by influences of time lags on individual and population responses that may stem in part from site tenacity by breeding adults, leading them to return to previous breeding locations in years following the manipulation despite the habitat changes. Further, the spatial scale on which the manipulation was conducted may have been inappropriate to gauge responses at the population level. We suggest that these complications may plague many field experiments in ecology.