Foraging Thresholds of Spring Migrating Dabbling Ducks in Central Illinois

The Upper Mississippi River and Great Lakes Region Joint Venture (hereafter, JV) endeavors to model energetic carrying capacity to inform conservation planning in the JV region. Currently, carrying capacity models use estimates of food production (e.g., moist-soil plant seeds) and habitat availability (area). However, estimates of the amount of food exploited by ducks with respect to availability are lacking. The JV currently assumes a conservative foraging threshold of 50% of gross food abundance can be exploited by foraging ducks. Giving-up densities (GUDs), which express the amount of food that remains after organisms cease foraging, can be used to estimate foraging thresholds. We endeavored to provide information to refine the JV’s foraging-threshold estimate through field experiments. We used experimental foraging patches, placed in wetlands used by spring-migrating dabbling ducks (Anas spp.) along the central Illinois River valley (IRV), to estimate the GUD in relation to experimentally manipulated seed density, seed size, seed depth in the substrate, substrate type, and predation risk. We conducted 7 foraging trials in 2010(March and April) and 10 in 2011(February–April), beginning immediately following spring ice-out. Trials were comprised of a series of plastic pans (foraging patches) filled with a combination of substrate (e.g., sand, clay) and seed (Japanese millet and red rice) and placed in wetlands near dabbling duck concentration areas. We monitored trial plots daily for duck use and conducted behavioral observations of ducks near trial plots. Once plots were abandoned by foraging ducks, we removed experimental patches, sorted seed from substrate, and dried and weighed remaining seed to estimate the GUD. Our results differed greatly between years. We had difficulty attracting ducks to trial plots in 2010, and use and seed exploitation was correspondingly low. On average, 521.4kg/ha (20%removed) of seed remained following duck abandonment in 2010. We had greater success attracting ducks to plots in 2011, and this was reflected by lower average GUD(35.8kg/ha, 94% removed). Ducks foraged more efficiently in sand than clay substrates, and better exploited shallowly buried over deeply buried seeds; however, we only collected data on the latter in 2010. Initial seed density decreased the GUD in 2010, but not in 2011, whereas predation risk increased the GUD in 2011 but not in 2010. Finally, ducks favored small seeds in 2010, but large seeds in 2011. Although our annual results contrasted, several of these differences may be explained by foraging theory and variation in migration chronology. Indeed, local food abundance likely varied considerably between years. Other food sources represent missed opportunities to ducks; thus, we expect the GUD to vary with respect to missed opportunity costs. When missed opportunity costs are high(i.e., high local food abundance outside of our test plots), the GUD in experimental patches should be correspondingly high, whereas the GUD will be lower when missed opportunity costs are also lower(i.e., relatively low local food abundance). Additionally, ice-out was nearly 1 month later than average (15 March) in 2010, and approximately average (15 February) in 2011. This difference may have shortened the stopover duration of large-bodied dabbling ducks (e.g., mallard [Anas platyrhynchos]) at our study sites, potentially altering the GUDs. Despite these interannual differences, our results demonstrate that ducks can remove substantially more seed from wetland habitats than the estimate currently used by the JV. Therefore, we suggest the JV consider incorporating the GUD estimates generated by this study into future energetic carrying capacity models. However, revising carrying capacity models would lead to revision of habitat protection and enhancement goals and should be approached cautiously. Perhaps carrying capacity estimates based on the results of our study could be considered as alternate or competing models to the current approach. In this scenario, consideration of formally revising the estimates based on lowered foraging thresholds might be framed in the context of adaptive resource management, whereby support for formal revision could be based on the weight of evidence as our study is replicated or results otherwise supported or refuted.unpublishednot peer reviewedOpe