Summary
We present the first experiment to assess band color effects in a natural bird population. 38 territorial male red-winged blackbirds (Agelaius phoeniceus) were randomly assigned to 1 of 2 banding treatments. They received either all red bands (to match their epaulets) or all black bands (controls). Over half the red-banded males lost their territories while all black-banded males retained their territories. Red and black-banded males did not differ morphologically. However, among red-banded males, those that lost their territories had larger epaulets and were in poorer condition than those retaining their territories. Red-banded males suffered much higher intrusion rates, particularly by neighbors, than black-banded males. We propose that red color bands exaggerated the males' natural aggressive signal beyond the point where the signal was reliable. Our results suggest that signal reliability is maintained by regular testing, particularly of those males most likely to be signalling dishonestly.
Methodology
We conducted this study in an extensive cattail marsh (Typha sp.) near Seeley's Bay, Ontario during the 1989 breeding season. Male red-winged blackbird territory owners were captured using decoy traps (Smith 1976) and mist nests, measured, color banded and released onto their territories in April and May. Body size and epaulet size have been shown to be important in aggressive interactions between captive male red-winged blackbirds (Searcy 1979b; Eckert and Weatherhead 1987b). Therefore, we recorded these morphological measures as estimates of male competitive ability. Body size was indexed by wing length, measured (to the nearest ram) from the bend of the wing to the tip of the longest primary (Searcy 1979b), body weight (to the nearest g), and tarsus length (to the nearest 0.1 ram). Epaulet size was indexed by the length of a male's epaulet measured as the distance from the bend of the wing to the furthest extent of yellow plumage (Searcy 1979 b).