Discussion
A subset of the robins at our tree-dominated study site exhibit nest-placement flexibility, building their nests on the ground rather than in shrubs or on trees in the early half of the breeding season. The decision to use a ground nesting site could be the result of competition for suitable tree sites. Limited sites are a known driver of nest placement and a constraint on the number of breeding individuals, although this is a pattern most often seen in cavity-nesting birds rather than open-cup nesting birds (Newton 1994). Contrary to this hypothesis, most of the robin ground nests we observed were located directly under or within 5 m of seemingly suitable unoccupied trees, suggesting that competition for nesting trees was not the primary driver of the robins’ decision to build their nest on the ground at this site.
Because predation is the primary cause of nest failure in songbirds (Martin 1995), avoiding predators can also drive nest-placement decisions. While nesting on the ground may reduce the likelihood of depredation by avian predators and arboreal mammals, nesting in trees may make nests less accessible to mostly terrestrial predators like raccoons, rat snakes, and deer observed at nearby nest monitoring sites (Chiavacci et al. 2018), and nests placed higher are often safer overall (Burhans et al. 2002). Here, we statistically overlapping daily survival rates between ground- and tree-nests at the same study site, suggesting that flexibility in nest placement does not impact predation-driven nest failure overall.
While ground-nesting robins do not suffer increased predation risk overall, it is possible that the relative predation risk for ground- and tree-nesting birds varies across the season. Many robins at our site initiate nesting before the planted trees have leafed out, making the bulky robin nest structures built in barren trees clearly visible to human observers (Figure 1 inset), which might cause some robin females to show a seasonal preference for building nests on the ground. A seasonal leaf-limitation hypothesis predicts that ground-nesting should be beneficial relative to tree-nesting in the beginning of the breeding season when the leaves are not present. Accordingly, all of our ground nests had been initiated in April and early May, when tree leaves are absent or still growing; we found no ground nests initiated after 12 May despite our nest searching extending until July 4 each year. In contrast, we regularly located many arboreal nests initiated from April through early July (Figure 2). However, the ground nests in this early portion of the season were not more likely to survive than the exposed tree nests initiated during the same time period. An alternative hypothesis is that nest placement may be an important determinant of the immediate weather and microclimate conditions experienced by the incubating/brooding parents and the developing offspring (Ardia et al. 2006). Cup nests in trees in the open air experience a different microclimate than nests partially in the ground. The early half of the breeding season, when we found all of our ground nests, is considerably colder (average for Champaign County in April and May: low of 5oC, high of 23oC) than the second half of the breeding season (average for June and July: low of 17oC, high of 29oC, State Climatologist Office for Illinois). It is possible that nests in depressions on the ground retain heat better than nests in trees, benefiting ground-nesters over tree-nesters when the temperatures are cold, a prediction that could be readily tested with temperature monitors in nests throughout the season. These conditions may present a trade-off for the birds between choosing sites with a more suitable microclimate or lower predation risk, as is the case in some desert species (Tieleman et al. 2008), although the equivalent survival rates between tree and ground nests in our study site argue against this trade-off.
While it is not immediately clear why some of the otherwise overwhelmingly arboreal robins in our system are nesting on the ground within a tree farm, such behavioral plasticity in life history traits could be consequential for the robins’ ability to adjust to changing environments, including ongoing anthropogenic and climate change (Mainwaring et al. 2017). For example, dark-eyed juncos (Junco hyemalis ) colonizing novel, urban habitats have adjusted both their nest placement and nest reuse patterns, producing more successful offspring compared to forest dwelling populations (Yeh et al. 2007). Although our study sites are inside a strictly agricultural matrix, they are also within 10 km of dense urban robin population in Urbana-Champaign, suggesting that more flexibly-nesting urban robins may be the source of variation in nest site choice in our study. Future work should elucidate not only the environmental correlates of the ground-nesting behavior but also the fitness consequences of nest placement in this still common species. Indeed, ground nesting at our site may not be an adaptive response to local environmental conditions, but, for example, due to immigration of robins with preference for ground nesting due to previous experience, due to the breeding adult age or inexperience, and/or vertical transmission of nest-site preference from ground-nesting parents (Slagsvold et al. 2013).
Data availability: All data will be made publicly available (currently archived for confidential review during peer-critique at https://figshare.com/s/6c997b0afc2cbd9bca41 and https://figshare.com/s/b2360ecb5302c0fff734)
Acknowledgements: We are grateful for the permission of local landowners to work on their property. For funding we thank the Illinois Distinguished Fellowship program (to SKW), the BRIDGE program of the University of Illinois at Urbana-Champaign and the University of Birmingham (to MEH), and the Harley Jones Van Cleave Professorship, the University of Illinois’ Center for Advanced Study, and the Wissenschaftskolleg zu Berlin, Germany (to MEH). For assistance and discussions, we thank A. Luro, J. Enos, C. Goethe, C. Kim, J. Reynolds, and O. Rhodes.