Assessing generalized egg mimicry: a quantitative comparison of eggs of Brown-headed Cowbirds and grassland passerines.
Abstract: Perceived similarity in appearance of eggs of the Brown-headed Cowbird (Molothrus ater) and several grassland passerine hosts has been suggested to represent a form of generalized mimicry evolved to impede egg discrimination. We tested this hypothesis by quantitatively comparing four parameters (ground color, maculation color, distribution of maculation, and maculation density) among cowbird eggs and those of six grassland passerines known to preferentially accept cowbird eggs over non-mimetic blue eggs. Cowbird eggs did not significantly differ in three of four parameters (mean ground color, and color and density of maculation) from the average grassland passerine egg measured for this community, in southern Saskatchewan, when all grassland passerine eggs were pooled. Cowbird eggs sufficiently overlapped grassland passerine eggs in these three parameters that 88% of cowbird eggs measured were statistically indistinguishable from among all grassland passerine eggs. The frequency at which cowbird eggs were misclassified as eggs of each grassland passerine ranged from 8 to 48%, which suggests that a single cowbird egg is capable of mimicking the eggs of more than one species. These results support the hypothesis that cowbirds have evolved a generalized egg appearance that mimics the eggs of multiple grassland passerine hosts within a single community.
Article Type: Report
Subject: Cowbirds (Behavior)
Mimicry (Biology) (Research)
Birds (Eggs and nests)
Birds (Properties)
Authors: Klippenstine, Dwight R.
Sealy, Spencer G.
Pub Date: 06/01/2010
Publication: Name: The Wilson Journal of Ornithology Publisher: Wilson Ornithological Society Audience: Academic Format: Magazine/Journal Subject: Biological sciences Copyright: COPYRIGHT 2010 Wilson Ornithological Society ISSN: 1559-4491
Issue: Date: June, 2010 Source Volume: 122 Source Issue: 2
Topic: Event Code: 310 Science & research
Geographic: Geographic Scope: Canada Geographic Code: 1CANA Canada
Accession Number: 228120720
Full Text: Evolution of egg mimicry by obligate brood parasites is regarded as one of the most common responses to development of egg discrimination and ejection by intolerant hosts (Underwood and Sealy 2002, Kilner 2006). Egg mimicry evolves because ejection of unlike eggs by hosts provides a selective advantage to parasitic eggs that match the clutch in which they are laid (Davies and Brooke 1991, Soler et al. 2003). The extent of similarity gradually escalates through a co-evolutionary race as hosts refine their egg-discrimination abilities over subsequent generations (Moksnes et al. 1990, Davies and Brooke 1991). The evolutionary equilibrium hypothesis posits that whenever the level of mimicry exceeds a species' ability to discriminate between eggs, the host accepts the parasitic egg (Davies and Brooke 1988, Lotero and Nakamura 1998). Presumably, acceptance occurs because of the greater cost of mistakenly ejecting the wrong egg (Davies and Brooke 1988, Lotem and Nakamura 1998).

Egg mimicry has been documented or suspected in several parasitic systems (Underwood and Sealy 2002). Most notably different groups of female Common Cuckoos (Cuculus canorus) have evolved eggs that closely mimic those of a preferred host (Davies and Brooke 1988, Marchetti et al. 1998, Gibbs et al. 2000). The Brown-headed Cowbird (Molothrus ater, hereafter cowbird) has been largely exempt from these speculations (Sealy et al. 2002, Mermoz and Ornelas 2004) because individual female cowbirds are known to parasitize multiple species (Fleischer 1985, Alderson et al. 1999). Rothstein (2001) suggested cowbird egg mimicry could not evolve due to the immense variability in egg appearance each female cowbird would encounter when parasitizing several species. He also argued that because cowbirds are known to parasitize hundreds of species it would be more beneficial to avoid hosts that discriminate among eggs than to enter a costly co-evolutionary race. A recent comparison of tolerant and intolerant hosts of the cowbird also found that egg appearance was not correlated with acceptance of cowbird eggs (Peer and Sealy 2004).

The close resemblance in color and maculation pattern between cowbird eggs and those of many grassland passerines, despite skepticism, has been suggested to be a form of generalized egg mimicry (Peer et al. 2000, Davis et al. 2002, Frontispiece in Klippenstine and Sealy 2008). Generalized cowbird egg mimicry is the idea that cowbird eggs have been adapted to match the appearance of eggs laid by multiple hosts. The eggs of a parasite in traditional mimicry evolve to resemble those of a single, intolerant host species (Underwood and Sealy 2002). Generalized egg mimicry, however, incorporates the selective forces provided by all hosts exhibiting egg discrimination. This causes the parasite's eggs to assume the average color and maculation pattern of eggs from all intolerant hosts in proportion to the selective pressure provided by each host. A single parasite egg can be laid randomly in any host nest in true generalized egg mimicry and still have a high probability of matching the host's eggs.

Indirect evidence of generalized cowbird egg mimicry has come from experiments that showed several grassland passerines accept cowbird eggs despite possessing an ability to eject less mimetic foreign eggs (Peer et al. 2000, Davis et al. 2002, Klippenstine and Sealy 2008). Therefore, the tolerance of parasitism observed in grassland passerines has, in part, been attributed to their inability to identify cowbird eggs among their clutch, i.e., evolutionary equilibrium hypothesis (Peer et al. 2000, Davis et al. 2002, Klippenstine and Sealy 2008). Generalized cowbird egg mimicry is plausible because most grassland passerine eggs are similar in color and maculation pattern (Peer and Sealy 2004, Klippenstine and Sealy 2008). Grassland passerines also are historic hosts of cowbirds and, presumably, have been exposed to parasitism sufficiently long for the required co-evolutionary race to have occurred (Mayfield 1965, Rothstein 1994, Igl and Johnson 2007).

We quantitatively compared color and maculation pattern of cowbird eggs to eggs of a community of grassland passerines known to accept cowbird eggs, but eject non-mimetic eggs, to examine the possibility of generalized cowbird egg mimicry (Klippenstine and Sealy 2008): Sprague's Pipits (Anthus spragueii), Vesper Sparrows (Pooecetes gramineus), Savannah Sparrows (Passerculus sandwichensis), Baird's Sparrows (Ammodramus

bairdii), Chestnut-collared Longspurs (Calcarius ornatus), and Western Meadowlarks (Sturnella neglecta). We predicted if generalized cowbird egg mimicry has evolved that: (1) cowbird eggs would match the average color and maculation pattern present among the eggs of all six grassland passerines; (2) cowbirds would have incorporated enough of the community-wide variation that a proportion of their eggs would match those of each grassland passerine; and (3) an individual cowbird egg would be capable of mimicking the eggs of more than one species.

METHODS

Fieldwork was conducted in southcentral Saskatchewan, Canada from 7 May to 1 July 2001 and 9 May to 15 July 2002. Sites were 60 km south of Regina within a 100-k[m.sup.2] area centered about the abandoned hamlet of Dummer (49[degrees] 50' N, 104[degrees] 49' W). Nests were found in cultivated and uncultivated land using predominantly a dragging rope as described by Klippenstine and Sealy (2008).

Photographs of Eggs.--Appearance of cowbird eggs and grassland passerine eggs was quantified using photographs taken with a 35-mm Pentax automatic focus camera and 100-speed color Kodak film for increased resolution (Fleischer and Smith 1992). Photographs of complete clutches were desired unless this was not possible because of partial predation, cowbird parasitism, or if eggs broke when handled. Cowbird eggs were photographed whenever a nest was naturally parasitized and were included in the photograph of the host's eggs.

All eggs were temporarily placed in an egg box, which measured ~20 by 50 cm and had a foam mat on the bottom with six indentations that held the eggs in place during each photograph. Three color chips (red, green, and blue), each with three shades of their corresponding color (light, medium, and dark), were placed on the mat and included in the photograph (Fleischer and Smith 1992, Villafuerte and Negro 1998). The egg box was placed in a portable hood to exclude outside lighting and a camera stand within the hood was used to ensure consistent photography. Films were developed as 4 x 6 color prints with a matte finish to decrease glare. Photographs in 2001 were developed throughout the field season, but all photographs in 2002 were developed on the same day to reduce variation in color, contrast, and brightness caused by developing at different times. Prints were scanned with a flatbed scanner at 300 dpi for digital analysis.

Measurements of Egg Parameters.--Four egg parameters were measured: ground color, maculation color, distribution of maculation, and total density of maculation. Eggs were isolated from their respective photographs using the magic wand tool of Adobe Photoshop 4.0 to enhance the accuracy of measurements. Maculation was measured using inverted black and white images of each egg. Markings were highlighted as bright white spots, leaving the ground color of the egg black. The proportion of maculation on the blunt and pointed ends of each egg was measured by centering a 20- x 30-pixel rectangular, selection box one-quarter the distance from the pointed end and one-quarter from the blunt end of each egg, and measuring the proportion of area covered in white compared to black using the Northern Eclipse Program (modified from Fleischer 1985). Density of maculation was calculated as the percent total area covered in maculation for both blunt and pointed ends of each egg on a scale of 0 to 100, i.e., percent x 100. The distribution of maculation was calculated as the percent of the total density of maculation occupying only the blunt end, i.e., blunt end maculation divided by total maculation multiplied by 100. Eggs with evenly distributed maculation had values close to 50 and values increased as the proportion of maculation on the blunt end of the eggs increased.

Ground and maculation colors were measured from the color images of each egg using the RGB program (Villafuerte and Negro 1998). Three samples of ground color were captured using Adobe Photoshop 4.0, i.e., three square 10 X 10 pixels, along the length of the egg, and avoiding the area of brightness caused by glare of the camera's flash. Three to six spots or scrawls were sampled in the case of maculation. All selected areas were transferred into the RGB program, which calculated the red, green, and blue value of each pixel on a scale of 0 to 256 (Villafuerte and Negro 1998). The average red, green, and blue values were assigned to the background color and color of maculation for each species' egg. Preliminary analyses revealed that for eggs of all species the proportion of red, green, and blue was highly correlated to each other in both ground color and maculation color, respectively. The amounts of red, green, and blue for both maculation and ground color were summed because they were not independent; the new value approximated the darkness or lightness of the color on a scale of 0 to 768 with an egg's color becoming darker as this number decreased. This was possible because ground color and maculation color of each species' eggs examined tended to be brown, differing only in intensity of color, which would be seen as lightening or darkening. The correlation between red, green, and blue would have been lower or absent had different colors, i.e., blue, pink, etc., been abundant on the eggs.

The extent to which the shade of each color varied within a photograph due to minor differences in illumination, development, and film was calculated by measuring the discrepancy between observed and theoretical values for the three shades of the color chips photographed with the eggs. A correction factor was computed that represented how much red, green, and blue was added or subtracted from the photograph during its exposure and development (Villafuerte and Negro 1998). This correction factor was calculated for each photograph by averaging the differences between theoretical and observed values for all three shades of each color.

Statistical Analyses.--Eggs of all six grassland passerines were pooled to calculate the average parameters for this community. The total number of eggs for each species was reduced to 15 by randomly eliminating additional eggs to ensure that all species were represented equally when pooled. All reductions were repeated twice and the results were compared to ensure that no bias occurred during the initial reduction. Only 14 Sprague's Pipit eggs were available for comparison, and the average of all 14 eggs was inserted to make up the missing value (Zar 1996). We opted to pool the eggs of each species in equal proportions rather than in proportion to their frequency of parasitism, relative abundance or frequency of cowbird egg ejection, i.e., as a measure of selective pressure, for two reasons: (1) these parameters do not vary significantly among species within this community (Davis 2003, Klippenstine and Sealy 2008); and (2) currently, levels of parasitism and relative abundance of each species varies between years and habitat structure (Davis 2003, Klippenstine and Sealy 2008). A single measure of these parameters may not accurately represent the selective pressure provided by each species currently or throughout their co-evolution with cowbirds.

We used Discriminate Analysis (DA) to examine the extent of discrimination possible between cowbird eggs and the average grassland passerine egg using the categories of maculation (density and distribution of maculation) and color (ground and maculation color). An ANOVA was used to identify which parameter or parameters created the difference if discrimination was observed in any of these categories (Manly 1986). All four parameters were recombined, based on the latter analyses, to acquire the correct collection of parameters that minimized discrimination between cowbird eggs and the average grassland passerine egg. The predicted group membership function following a DA was used to assess overlap between cowbird eggs and grassland passerine eggs using minimum discrimination parameters. This function works by attempting to re-assign each egg blindly to its correct species, listing the portion of eggs for each species that was misclassified as being the eggs of all other species (Manly 1986). The portion of misclassification between two species' eggs represented the extent of overlap or "mimicry" between their eggs.

The predicted group membership analysis was conducted on two scales to fully assess mimicry: (1) the frequency of misclassification that occurred when eggs of all six grassland species were compared simultaneously with cowbird eggs; and (2) the frequencies of misclassification that occurred when cowbird eggs were compared separately to eggs of each grassland species. We examined whether the sum of misclassification frequencies between cowbird eggs and eggs of each species was greater than the proportion of cowbird eggs misclassified when compared to all species simultaneously to learn whether individual cowbird eggs mimic those of multiple species. If, however, each cowbird egg only matches the eggs of one species, the sum of misclassification frequencies involving each species should be equal to the proportion misclassified when all eggs were compared together. Differences in the frequency of misclassification between species were assessed using a Chi-square test. There was a tendency for greater discrimination because the same data used in the initial DA were used in the predicted group membership (Manly 1986). This analysis was conservative, as it was more likely to separate species than place them together. Statistical significance was set at alpha = 0.05 for all analyses, and all average values are given as mean [+ or -] SE.

RESULTS

Cowbird Eggs versus Average Grassland Passerine Egg.--Cowbird eggs were indistinguishable in color (ground and maculation combined) from the average grassland passerine egg for this community (DA, [[chi square].sub.2] = 1.8, P = 0.40; Table 1, Fig. 1), but were readily discernable in maculation (density and distribution of maculation combined; DA, [[chi square].sub.2] = 13.5, P = 0.001). Dissimilarities in overall maculation resulted exclusively from cowbird eggs being far more evenly maculated than the average grassland passerine egg ([F.sub.1] = 12.3, P = 0.001), as no difference in the total density of maculation was observed between the two groups ([F.sub.1] = 2.5, P = 0.17, Table 1). Thus, minimum discrimination between cowbird eggs and the average grassland passerine egg was achieved by excluding distribution of maculation as a parameter (DA, [[chi square].sub.3] = 2.8, P = 0.43 vs. [[chi square].sub.3] = 13.5, P = 0.009 for all four parameters).

Overlap Between Cowbird Eggs and Grassland Passerine Eggs.--The predicted group membership analysis was significantly more likely to misclassify cowbird eggs on the basis of minimum discrimination parameters, i.e., color and total density of maculation, than eggs of all other grassland passerines ([[chi square].sub.1] > 4.4, P < 0.05 for all species; Table 2). Overall, 88% of cowbird eggs were misclassified compared to 4 to 70% of grassland passerine eggs (Table 2). Between 4 and 67% of grassland passerine eggs were misclassified as being the eggs of another grassland passerine, an indication of the close resemblance between grassland passerine eggs (Table 2).

Between 8 to 48% of cowbird eggs were misclassified, depending on the species, when cowbird eggs were compared to eggs of each species separately using minimum discrimination parameters (Table 3). The sum of all cowbird egg misclassifications for all six species greatly exceeded the proportion of cowbird eggs misclassified when all eggs were compared simultaneously (186 vs. 88%; Tables 2, 3). The proportion of cowbird eggs misclassified did not differ (28 to 49%; [[chi square].sub.4] = 6.0, P = 0.20) among eggs of Sprague's Pipit, Vesper, Savannah, and Baird's sparrows, and Chestnut-collared Longspurs (Table 3). A substantially lower proportion of cowbird eggs were misclassified as Western Meadowlark eggs (8%) than the other five species ([[chi square].sub.1] > 6.7, P < 0.05 for all species; Table 3).

[FIGURE 1 OMITTED]

DISCUSSION

Our results support the hypothesis that cowbird eggs have evolved to generally mimic the appearance of eggs from multiple grassland passerines. Cowbird eggs were indistinguishable from the ground and maculation colors, and total density of maculation as measured for the average grassland passerine eggs from this community in southern Saskatchewan (Table 1). All six grassland species examined have been experimentally demonstrated to eject non-mimetic eggs (Klippenstine and Sealy 2008). Thus, our results support the idea that ejection of discordant cowbird eggs has forced them to assume the average appearance of eggs from all intolerant hosts.

Cowbird eggs so closely approximated the color and maculation density of grassland passerine eggs that 88% (1 of every 1.1 eggs) of cowbird eggs were misclassified as belonging to another species (Table 2). Between 8 and 48% of cowbird eggs were misclassified depending on the grassland hosts, when compared to eggs of each species separately (Table 3). Thus, any cowbird egg laid randomly in a Sprague's Pipit, Vesper Sparrow, Savannah Sparrow, Baird's Sparrow, or Chestnut-collared Longspur nest has between a "one-in-three" and a "one-in-two" chance of matching the hosts eggs; the chance of matching for Western Meadowlarks is significantly lower at approximately "one-in-ten" (Table 3). Effective mimicry requires most cowbird eggs to match the eggs of at least one host. This is especially true considering female cowbirds are thought to exhibit little to no host selection (Fleischer 1985, Alderson et al. 1999). "Generalized" mimicry necessitates that color and maculation density of cowbird eggs overlap to a significant extent with the eggs of all six species.

Another important component of generalized mimicry is the idea that a single cowbird egg may mimic the eggs of more than one species. This contrasts with the elaborate mimetic system evolved by the Common Cuckoo. Different subgroups or "gentes" of female cuckoos in that system mimic the eggs of a preferred host (Davies and Brooke 1988, Marchetti et al. 1998, Gibbs et al. 2000). The benefit of gentes is that it allows cuckoos, as a species, to parasitize several species that exhibit egg discrimination while lowering competition for nests (Davies and Brooke 1988). Cowbird gentes are supported (Fig. 1), illustrating some of the remarkable similarities we discovered between cowbird eggs and the egg of each grassland passerine. However, the high frequency at which cowbird eggs were misclassified when compared to the eggs of each species (8 to 48%, sum total 186%; Table 3) is only possible if individual cowbird eggs are capable of matching the eggs of more than one species. We conclude that eggs of individual female cowbirds have not evolved to mimic those of a preferred species.

Cowbird eggs were more evenly maculated than the average grassland passerine egg despite the strong resemblance in three of four parameters (Table 1). This discrepancy may arise because an unforeseen cost exists for cowbird eggs that are more unevenly maculated, e.g., increased rates of depredation (Underwood and Sealy 2002, Kilner 2006). Cowbird eggs have similarly been shown to be rounder than host eggs to increase shell strength and reduce the potential for damage during laying or by intolerant hosts (Rohwer and Spaw 1988, Picman 1989). Alternatively, divergences in maculation distribution might reflect the parameters used by grassland passerines to discriminate cowbird eggs. Rothstein (1982) found that Gray Catbirds (Dumetella carolinensis) were more responsive to changes in color than maculation, and American Robins (Turdus migratorius) were more sensitive to color than size (also Underwood and Sealy 2006). Differences in host discrimination alter the selective pressure placed on cowbird eggs by each host and modify which parameters are mimicked (Rothstein 1982, Underwood and Sealy 2006). Consequently, the distribution of maculation may not be imitated by cowbirds because grassland passerines discriminate foreign eggs solely on the basis of color and density of maculation.

Our results show that cowbird eggs were most readily distinguished from eggs of Western Meadowlarks on the basis of color and maculation density (Tables 2, 3). This corresponds with the level of intolerance observed for this species compared to the other tire grassland passerines (Klippenstine and Sealy 2008). Western Meadowlarks ejected 67% of cowbird eggs and 91% of the non-mimetic eggs added to their nests. The five remaining grassland passerines, in comparison, accepted all or nearly all cowbird eggs and ejected about 20% of non-mimetic eggs (Klippenstine and Sealy 2008). Western Meadowlarks apparently eject cowbird eggs more readily because of the lower extent of matching between the eggs of the two species. Peer and Sealy (2004) rated Western Meadowlarks eggs similar in color and maculation to cowbird eggs. They also suggested egg appearance was not correlated with this host's intolerance of parasitism, but was related to their larger beaks. We suspect the difference arises because they compared the color and maculation of eggs using human observers. It is extremely difficult, however, for a human observer to assess the density of maculation independent of its distribution, which in this study had a profound affect. Consequently, discrepancies observed between the two studies emphasize the benefits of assessing egg appearance quantitatively.

Generalized cowbird egg mimicry is not the only possible explanation of the close resemblance between cowbird eggs and grassland passerine eggs. Concealment of eggs in the nesting habitat from depredation is a primary force directing the evolution of egg appearance in most species (Underwood and Sealy 2002, Kilner 2006). Therefore, the same forces promoting the evolution of color and maculation density in grassland passerine eggs may have selected the same pattern in cowbird eggs (Underwood and Sealy 2002, Kilner 2006). We propose two criteria to distinguish between resemblance due to nest concealment and cowbird egg mimicry. The first is that cowbird egg mimicry requires host species that exhibit egg discrimination and ejection. The presence of egg discrimination in all six species strongly supports the idea of cowbird egg mimicry (Klippenstine and Sealy 2008). A worthwhile test of this criterion would be to examine the extent of matching between cowbird eggs and eggs of grassland passerines that have not evolved egg recognition. The second criterion is that nest concealment would not necessarily make cowbird eggs better at matching grassland passerine eggs than eggs of any other grassland passerine. That cowbird eggs were significantly more likely to be misclassified than all other grassland passerine eggs (88% vs. 4-70%) supports generalized cowbird egg mimicry (Table 2).

The benefits of mimicking grassland passerine eggs are obvious for cowbirds. Despite having evolved egg discrimination, all six grassland species fully or partially accept cowbird eggs (Davis et al. 2002, Klippenstine and Sealy 2008). This tolerance has translated into reasonable breeding success for cowbirds within this community (Davis et al. 2002, Davis 2003, Klippenstine and Sealy 2008).). However, we believe the value of generalized cowbird mimicry is most likely limited to only a few host assemblages. Communities where host eggs are highly divergent in appearance are probably not conducive to generalized mimicry (Rothstein 2001). Generalized cowbird egg mimicry, even for grassland passerines examined in this study, is feasible only because each host exhibits a relatively crude level of egg recognition (Klippenstine and Sealy 2008). Greater selective pressure will be placed on cowbirds to evolve more species-specific egg mimicry as continued parasitism forces grassland passenes to develop their ability to discriminate foreign eggs (Moksnes et al. 1990, Davies and Brooke 1991). Cowbird egg mimicry of grassland passerine eggs may eventually resemble the type of mimicry exhibited by the Common Cuckoo. Alternatively, as Rothstein (2001) suggested, cowbirds may in the future choose to avoid such troublesome hosts.

ACKNOWLEDGMENTS

We are indebted to the landowners who permitted us to conduct our research on their land. We thank M. A. Talbot and N. L. Haalboom for their hard work in the field; Cornelius Klippenstine for use of his farmhouse; C. S. Bjomsson and Erwin Huebner for technical assistance; N. C. Kenkel, M. A. Klippenstine, D. M. Gillis, J. F. Hare, and M. V. Abrahams for statistical advice; M. D. Kujath, L. C. Graham for commenting on earlier drafts of the manuscript; and two anonymous reviewers for constructive comments that improved this manuscript. This study was funded by a Discovery Grant from the Natural Sciences and Engineering Research Council of Cariada to S. G. Sealy and through in-kind services provided by S. K. Davis and the Saskatchewan Watershed Authority (formerly Saskatchewan Wetlands Conservation Corporation). This paper is dedicated to the memory of B. A. Klippenstine.

Received 31 October 2008. Accepted 23 October 2009.

LITERATURE CITED

ALDERSON, G. W., H. L. GIBBS, AND S. G. SEALY. 1999. Determining the reproductive behaviour of individual Brown-headed Cowbirds using microsatellite DNA markers. Animal Behaviour 58:895-905.

DAVIES, N. B. AND M. DE L. BROOKE. 1988. Cuckoos versus Reed Warblers: adaptations and counter adaptations. Animal Behaviour 36:262-284.

DAVIES, N. B. AND M. DE L. BROOKE. 1991. Coevolution of the cuckoo and its hosts. Scientific American 264:92-98.

DAVIS, S. K. 2003. Nesting ecology of mixed-grass prairie songbirds in southern Saskatchewan. Wilson Bulletin 115:119-130.

DAVIS, S. K., D. R. KLIPPENSTINE, AND R. M. BRIGHAM. 2002. Does egg recognition account for the low incidence of cowbird parasitism in Chestnut-collared Longspurs (Calcarius ornatus)? Auk 119:556-560.

FLEISCHER, R. C. 1985. A new technique to identify and assess the dispersion of eggs of individual brood parasites. Behavioral Ecology and Sociobiology 17:91-99.

FLEISCHER, R. C. AND N. G. SMITH. 1992. Giant Cowbird eggs in the nests of two icterid hosts: the use of morphology and electrophoretic variants to identify individuals and species. Condor 94:572-578.

GIBBS, H. L., M. D. SORENSON, K. MAREHETTI, M. DE L. BROOKE, N. B. DAVIES, AND H. NAKAMURA. 2000. Genetic evidence for female host-specific races of the Common Cuckoo. Nature 407:183-186.

IGL, L. D. AND D. H. JOHNSON. 2007. Brown-headed Cowbird, Molothrus ater, parasitism and abundance in the Northern Great Plains. Canadian Field-Naturalist 121:239-255.

KILNER, R. M. 2006. The evolution of egg colour and patterning in birds. Biological Reviews 81:383-406.

KLIPPENSTINE, D. R. AND S. G. SEALY. 2008. Differential ejection of cowbird eggs and non-mimetic eggs by grassland passerines. Wilson Journal of Ornithology 120:667-673.

LOTEM, A. AND H. NAKAMURA. 1998. Evolutionary equilibria in avian brood parasitism: an alternative to the "arms race-evolutionary lag'" concept. Pages 223-235 in Parasitic birds and their hosts: studies in coevolution (S. 1. Rothstein and S. K. Robinson, Editors). Oxford University Press, Oxford, United Kingdom.

MANLY, B. F. J. 1986. Multivariate statistical methods: a primer. Chapman and Hall, New York, USA.

MARCHETTI, K., H. NAKAMURA, AND H. L. GIBBS. 1998. Host-race formation in the Common Cuckoo. Science 282:471-472.

MAYFIELD, H. 1965. The Brown-headed Cowbird, with old and new hosts. Living Bird 4:13-28.

MERMOZ, M. E. AND J. F. ORNELAS. 2004. Phylogenetic analysis of life-history adaptations in parasitic cowbirds. Behavioral Ecology 15:109-119.

MOKSNES, A., E. ROSKAFT, A. T. BRAA, L. KORSNES, H. M. LAMPE, AND H. C. PEDERSEN. 1990. Behavioral responses of potential hosts towards artificial cuckoo eggs and dummies. Behaviour 116:64-89.

PEER, B. D. AND S. G. SEALY. 2004. Correlates of egg rejection in hosts of the Brown-headed Cowbird. Condor 106:580-599.

PEER, B. D., S. K. ROBINSON, AND J. R. HERKERT. 2000. Egg rejection by cowbird hosts in grasslands. Auk 117:892-901.

PICMAN, J. 1989. Mechanism of increased puncture resistance of eggs of Brown-headed Cowbirds. Auk 106:577-583.

ROHWER, S. AND C. D. SPAW. 1988. Evolutionary lag versus bill-size constraints: a comparative study of the acceptance of cowbird egg by old hosts. Evolutionary Ecology 2:27-36.

ROTHSTEIN, S. I. 1982. Mechanisms of avian egg recognition: which egg parameters elicit responses by rejecter species? Behavioral Ecology and Sociobiology 11:229-239.

ROTHSTEIN, S. I. 1994. The cowbird's invasion of the far west: history, causes and consequences experienced by host species. Studies in Avian Biology 15:301-315.

ROTHSTEIN, S. L 2001. Relic behaviours, coevolution and the retention versus loss of host defences after episodes of avian brood parasitism. Animal Behaviour 61:95-107.

SEALY, S. G., D. G. McMASTER, AND B. D. PEER. 2002. Tactics of obligate brood parasites to secure suitable incubators. Pages 254-269 in Avian incubation: behaviour, environment, and evolution (D. C. Deeming, Editor). Oxford University Press, Oxford, United Kingdom.

SOLER, J. J., J. M. AVILES, M. SOLER, AND A. P. MOLLER. 2003. Evolution of host egg mimicry in a brood parasite, the Great Spotted Cuckoo. Biological Journal of the Linnean Society 79:551-563.

UNDERWOOD, T. J. AND S. G. SEALY. 2002. Adaptive significance of egg coloration. Pages 280-298 in Avian incubation: behaviour, environment, and evolution (D. C. Deeming, Editor). Oxford University Press, Oxford, United Kingdom.

UNDERWOOD, T. J. AND S. G. SEALY. 2006. Parameters of Brown-headed Cowbird Molothrus ater egg discrimination in Warbling Vireos Vireo gilvus. Journal of Avian Biology 37:457-466.

VILLAFUERTE, R. AND J. J. NEGRO. 1998. Digital imaging for colour measurement in ecological research. Ecology Letters 1:151-154.

ZAR, J. 1996. Biostatistical analysis. Third Edition. Prentice Hall, Upper Saddle River, New Jersey, USA.

DWIGHT R. KLIPPENSTINE (1,2) AND SPENCER G. SEALY (1,3)

(1) Department of Biological Sciences, University of Manitoba, Winnipeg, MB R3T 2N2, Canada.

(2) Current address: Faculty of Medicine, University of Manitoba, Winnipeg, MB R3E 3P5, Canada.

(3) Corresponding author; e-mail: sgsealy@cc.umanitoba.ca
TABLE 1. Mean parameter values for the average grassland passerine egg
and Brown-headed Cowbird eggs.

Species (n)                     Density of         Distribution of
                              maculation. %         maculation. %

Average grassland           23.5 [+ or - ] 2.0    75.8 [+ or - ] 2.2 *
  passerine (90)
Brown-headed Cowbird (50)   28.7 [+ or - ] 2.2    64.3 [+ or - ] 1.9

Species (n)                    Ground color        Maculation color

Average grassland           472.1 [+ or - ] 9.4   372.0 [+ or - ] 7.1
  passerine (90)
Brown-headed Cowbird (50)   453.6 [+ or - ] 7.2   362.1 [+ or - ] 6.4

* Differs (P = 0.05) from cowbird eggs.

TABLE 2. Proportion of correct identification and
misclassification observed for each species when
compared simultaneously using minimum discrimination
parameters (ground color, maculation color, and density
of maculation).

                                   Eggs (%) misclassified as ...

                       Eggs        Other
                     correctly   grassland   Brown-headed    Other
                    identified   passerine     Cowbird      species
Species (n)            (%)          eggs         eggs       overall

Sprague's Pipit
  (14)                  71          22            7           29
Vesper Sparrow
  (21)                  43          57            0           57
Savannah
  Sparrow (34)          41          56            3           59
Baird's Sparrow
  (37)                  54          30           16           46
Chestnut-collared
  Longspur(40)          30          67            3           70
Western
  Meadowlark
  (24)                  96           4            0            4
Brown-headed
  Cowbird (50)          12          88                        88 *

* Differs (P = 0.05) from all other grassland passerine eggs.

TABLE 3. Proportion of misclassification observed using minimum
discrimination parameters (ground color, maculation color, and
density of maculation) between cowbirds eggs and eggs of individual
species.

                                   Cowbird eggs misclassified
                                    as eggs of each grassland
Species (n)                              passerine (%)

Sprague's Pipit (14)                           28
Vesper Sparrow (21)                            38
Savannah Sparrow (34)                          36
Baird's Sparrow (37)                           28
Chestnut-collared Longspur (40)                48
Western Meadowlark (24)                         8 *
Total                                         186

* Differs (P = 0.05) from all other grassland eggs.
Gale Copyright: Copyright 2010 Gale, Cengage Learning. All rights reserved.