BIRD REACTIONS TO OBSERVER CLOTHING COLOR - IMPLICATIONS FOR DISTANCE-SAMPLING TECHNIQUES

The species-confidence hypothesis asserts that birds prefer mates whos e colors match those of their species and avoid potential mates whose colors are atypical of conspecifics. Earlier point-count research indi cated this mate-choice hypothesis also may explain avian reactions to observer clothing color. In central Texas, during January and February 1993, we tested further whether the species-confidence hypothesis hol ds for reactions to observer clothing color by determining how close a n investigator could get to individuals before they flushed (approach distance) and how long individuals remained visible near their initial flush point (detectability period). The same investigator experimenta lly wore or did not wear a hunter-orange vest, which some wildlife bio logists wear as a safety precaution during hunting-season field studie s. The vest did not affect approach distance for species with red or o range on their bodies (American robin [Turdus,migratorius], P = 0.247; northern cardinal [Cardinalis cardinalis], P = 0.196; ruby-crowned ki nglet [Regulus calendula], P = 0.434; red-bellied woodpecker [Melanerp es carolinus], P = 0.113). In contrast, approach distances for species without red or orange were longer when the vest was worn than when it was not worn (Carolina chickadee [Parus carolinensis], P = 0.032; nor thern mockingbird [Mimus polyglottos], P = 0.033; yellow-rumped warble r [Dendroica coronata], P = 0.001). Differences in detectability perio d between vest conditions depended on air temperature for American rob ins (P = 0.003) and on the variation in visibility within habitats for northern cardinals (P < 0.001). Detectability period was shorter when the vest was worn than when it was not worn for Carolina chickadees ( P = 0.057) and yellow-rumped warblers (P = 0.060). Thus, for several s pecies this mate-choice hypothesis seems to apply to garment color as well. Some species' reactions demonstrated that, if birds respond sign ificantly to clothing color, important assumptions of distance-samplin g techniques used to estimate avian richness, density, and abundance w ill be violated, and these parameters will be underestimated. Wildlife scientists will be more likely to draw valid inferences and recommend effective management practices from avian data gathered via distance- sampling methods if reactions to clothing color are precluded.