Decision rules for egg-color-based rejection by two cavity-nesting hosts of the brown-headed cowbird

Abstract

Hosts of obligate avian brood parasites often evolve defense mechanisms to avoid rearing unrelated young. One common defense is egg rejection, for which hosts often rely on eggshell color. Most research has assumed that hosts respond to perceived color differences between their own eggs and parasite eggs regardless of the particular color; however, recent experiments have found that many hosts respond more strongly to brown foreign eggs than to equally dissimilar blue eggs. Yet, none of these prior studies tested a brown-egg-laying species and, with only one exception, all were conducted in open nests where light levels are considered sufficient for effective color-based egg discrimination. Here, we explored how two cavity-nesting hosts of the parasitic brown-headed cowbird (Molothrus ater) - the blue-egg-laying eastern bluebird (Sialia sialis) and the brown-egg-laying house wren (Troglodytes aedon) - respond to experimental eggs painted six distinct colors ranging from blue to brown. Rejection responses of both hosts were best predicted by perceived differences in color between the model egg and their own eggs. Specifically, we found that house wrens preferentially rejected eggs bluer than their own eggs. However, although we found that bluebirds relied on perceived differences in color for their egg rejection decisions, further tests are needed to determine whether they preferentially rejected brown eggs or simply responded to absolute perceived differences in color. These findings demonstrate that these cavity-nesting birds treat perceived color differences in distinct ways, which has important implications on the coevolutionary arms races and the interpretation of avian-perceived color differences.

Keywords: Brood parasitism; Color perception; Decision-making; Egg discrimination; Recognition.

Fig. 1.

Expected rejection probability of experimental egg models for eastern bluebirds and house wrens. Prior research has found that many hosts are more likely to reject eggs on one side of their phenotypic range than the other (i.e. the single threshold hypothesis), despite the degree of just-noticeable difference (JND) between their own egg and the foreign egg. (A) As eastern bluebird eggs are blue (arrow), we expected that they would reject (solid line) brown eggs at a disproportionately high rate. (B) By contrast, because house wren eggs are brown (arrow), we expected that they would reject (solid line) blue eggs at a disproportionately high rate. However, both hosts may actually reject based purely on the absolute perceived differences (dashed lines, i.e. the multiple thresholds hypothesis) and, therefore, remove any egg that they perceive as different (either bluer or browner than their own). (C) Experimental egg models illustrated for visual comparison. The chromatic and achromatic contrast between each egg model and each host egg is reported in Table 1.

Fig. 2.

The probability of egg rejection in relation to relative perceived chromatic and achromatic differences. (A,B) Eastern bluebirds and (C,D) house wrens will reject the experimental eggs used in this study depending on that egg's (A,C) directional chromatic and (B,D) achromatic contrast between those models and their own eggs (n=143 and n=97 for eastern bluebirds and house wrens, respectively). For clarity, chromatic contrasts explained (A) eastern bluebird responses marginally better than directional chromatic contrasts (P<0.01 versus P=0.02; Table 2), whereas directional chromatic contrasts explained (C) house wren responses better than chromatic contrasts (P<0.01 versus P=0.10; see Table 2 for more details). Discriminable differences are plotted using a model that accounts for color discrimination assuming that light is not limiting (see Materials and Methods and Table 1). A value of zero on the x-axis represents a color identical to the host's egg, and values of increasing magnitude are increasingly noticeably different from the host's own egg. The negative x-axis values represent bluer egg models and progressively more positive x-axis values represent browner eggs. Logistic fits (solid line) are shown for significant models. See Table 2 for parameter estimates from these models and Table 1 for rejection rates. Points are plotted with a jitter and in the color of the foreign egg to facilitate interpretation.