Climate and Intraspecific Geographic Variation of Nests and Eggs in the Eastern Kingbird <i>Tyrannus tyrannus</i>

Abstract

The reproductive success of birds is ultimately driven by the traits of nests, eggs, and incubation behavior and the interactions among them. Avian reproduction varies on large geographic scales, most prominently between tropic- vs. temperate-breeding species, but also on finer scales, such as along latitudinal and elevational geographic gradients. However, despite the importance of eggs and nests, how their traits vary geographically in response to differences in regional climate remains understudied. We studied Eastern Kingbirds (Tyrannus tyrannus), which have been previously demonstrated to alter incubation length in response to local climate conditions. Our study sites, Kansas (KS), New York (NY), South Dakota (SD), and Louisiana (LA) were chosen to represent a range of latitudes and climates. Our analysis of study site climate during the breeding season (April - July) showed that NY had the coldest mean temperature and most frequent precipitation, while LA had the warmest mean temperature and least cloud cover. SD and KS represented the climate of the Great Plains with moderately warm temperatures and higher wind speeds. We collected the nests of kingbirds breeding in NY, SD, and LA to test our predictions that kingbirds breeding in colder, more northern climates would build larger nests with thicker walls, smaller nest cup openings, and greater insulative capacities than kingbirds breeding in warmer, more southern climates. We found that SD nests were the heaviest, with the thickest walls, greatest cup volumes, and smallest relative cup openings, while LA nests had the thinnest walls and largest relative cup openings. NY nests were intermediate. Nests at all sites were composed of mainly fine stems and rootlets, forb stems, and plant down, followed by coarse woody stems, though material composition varied greatly within populations. Experiments measuring rate of heat loss from nests in environmentally controlled chambers revealed that the insulative capacity of nests only differed across sites when nest cup openings were covered. LA nests were less well insulated than SD and NY nests. We used principal components analysis (PCA) to reduce dimensionality of nest structure and material composition variables to PC axes, which we included in best subsets regressions to model predictors of nest insulative quality. For all sites, nests with the greatest insulative capacity had small nest cup openings relative to nest cup volume and were larger overall, but insulative capacity was not significantly related to material composition of nests. When we considered only SD and NY nests, we found