To investigate the relationship between reproductive success (RS) and breeding—site fidelity in a transequatorial migrant, we studied two populations of marked Bobolinks (Dolichonyx oryzivorus) in upstate New York during 1982—1985. At these sites, 44% of the males and 25% of the females resident in one year returned in one or more subsequent years. There were more nests built, eggs laid, eggs hatched, and young fledged in territories of returning males in this polygynous species than in territories of males that did not return in subsequent years (P < .04). For females that returned, the number of eggs hatched, young fledged, young fledged on last nest attempt, and young fledged per egg laid were higher than for females that did not return (P < .02). Discriminant function analysis and data from 1982—1984 were used to predict correctly the return behavior of 30 to 42 individuals (sexes combined) in 1985, based on their reproductive success (RS) in 1984 (P < .01). Using data on RS and return rate for all years, the best single—variable model for males based on the number of young fledged from each territory classified correctly 59 of 85 individuals (69%) as returning or not (P < .005). The best single—variable model for females based on the number of yound fledged on their last nest attempt classified correctly 61 of 86 individuals (71%; P < .005). Using all data for 1982—1985, sex—specific models that included only whether an individual fledged one or more young classified correctly the same proportion as the best single—variable models above. However, among males that fledged one or more young, those that returned had fledged significantly more young the previous year than those that did not return (P < 0.5). Not all individuals that failed to return were dead, because five individuals that did not return to our study sites were captured in a subsequent year at another site. We concluded that male and female Bobolinks used information on their breeding success in one year to influence their choice of breeding site in the next year.

No abstract

We studied patterns of paternity in 840 nestling bobolinks (Dolichonyx oryzivorus) from 191 pairbonded adults and the chicks they reared in west central New York, USA, during 1983 1986, using allozyme data (4 variable loci). Thirty-six (4.3%) nestlings exhibited genotypes that excluded a putative parent as a genetic parent. Using the method of Westneat et al. (1987) we estimated that 14.6% of the nestlings were sired by extrapair copulations and that 38% of the nests contained >1 nestling resulting from extra-pair fertilizations (EPFs). Rates of EPFs did not appear to be related to population density among our bobolink populations. Younger males and older females were more likely to have young in their nests sired by EPFs than were older males and younger females. Primary females (i.e., firstmated) of socially polygynous males were more likely to have young sired by EPFs than were secondary females; socially monogamous females had intermediate rates of mismatched offspring. Nestlings resulting from EPFs were not more frequent in nests of socially polygynous males than in nests of socially monogamous males. Males responsible for EPFs were probably nearby (within 2 territories) residents.

This study tested whether inbreeding avoidance could influence the fre-qu~ncy and timing ofdispersal movements in a small microtine rodent, the meadow vole (~1~rotus pennsylva~Icus). Meadow voles released into experimental grassland plots with s1bhng~ were m.or~ hkely to disperse from these plots than were voles released into similar plots ~1th n~ns1bhngs. Furthermore, voles that dispersed from sibling groups did so sooner than d1spe~smg voles from nonsibli~g groups. Voles released with nonsiblings were recaptured w1thm the P.lots fo~ l~nger penods than were voles from sibling groups. Males from both .groups .remamed w1thm th~ P.lots for shorter times than did females, but the degree of t~IS sex b1as .was greater for s1bhng groups. Thus, it appears that inbreeding avoidance can mfluence dispersal movements of the meadow vole and may be partially responsible for patterns of male-biased dispersal in at least some mammals.

We estimated nesting success at real and artificial nests of grassland birds to test the influence of nest type, nest position, and egg size on predation rates. We distributed wicker nests and realistic woven-grass nests baited with a clay egg and either a Northern Bobwhite (Colinus virginianus) egg or a House Sparrow (Passer domesticus) egg in four grasslands that were part of the Conservation Reserve Program in east-central Illinois. Nesting success averaged 86.5% for 12 days of exposure for artificial nests. For real nests, nesting success was markedly lower, averaging 39% over the entire nesting cycle and 59% during approximately 12 days of incubation. Wicker nests were depredated more often than woven-grass artificial nests (18% vs. 8%), and nests baited with House Sparrow eggs were depredated more often than nests baited with Northern Bobwhite eggs (22% vs. 9%). Elevated and ground nests were depredated at the same rate. Patterns of nest predation on wicker nests were markedly different from depredation patterns on real nests over time and among fields. In contrast, patterns of nest predation on realistic woven-grass nests corresponded much more closely with predation rates of real nests over time and among fields. We suggest that future artificial nest studies use nests and eggs that mimic as closely as possible the real nests and eggs of target species. Use of unrealistic artificial nests and eggs, at least in grasslands, may result in patterns of predation that do not accurately reflect those of real nests. Artificial nests of any type appear to underestimate predation rates on nests of grassland birds, possibly because of a lack of snake predation on artificial nests.