Habitat amount and edge effects, not perch proximity, nest exposure, or vegetation diversity affect cowbird parasitism in agricultural landscapes

Abstract

Context : Prior research documented relationships between brown-headed cowbird (Molothrus ater) brood parasitism and edge effects, proximity of perches, and nest exposure. Those relationships have not been evaluated in agroecosystems containing extremes of fragmentation and vegetation diversity. Objectives: We compared three existing hypotheses on how cowbirds locate host nests with two new hypotheses regarding habitat amount and vegetation diversity to determine how the configuration and location of agricultural conservation practices affect grassland bird nest parasitism rates and predicted rates for eight common conservation practices. Methods : We assessed cowbird parasitism of grassland bird nests on corn and soybean farms in Iowa, USA, and measured perch proximity, nest exposure, edge effects, habitat amount, and vegetation diversity for each nest. We fit a global generalized linear mixed-effects model and compared importance of model parameters using odds ratios. We predicted parasitism likelihood for every subset model and averaged predictions to explore individual effects. Results: The variables that most influenced parasitism rates included main effects for nest initiation day-of-season (OR = 0.71, CI95 = 0.60–0.84) and the landscape variables of distance to nearest crop edge (0.63, 0.51–0.76) and proportion of grass land cover within 660 m (0.75, 0.57–1.00). We found little support that perch proximity, nest exposure, or native vegetation diversity affected parasitism. We also assessed parasitism likelihood by conservation practice and found no significant differences. Conclusions: Our results provide evidence to support the edge effect and habitat amount hypotheses, but not the nest exposure, vegetation diversity, or perch proximity hypotheses.