Search

The Mississippi Flyway supports millions of waterfowl during spring migration as individuals replenish vital nutrients en route to breeding locations. Green-winged teal Anas crecca consume natural plant and animal foods in shallow wetlands during autumn and winter, but little information is available to describe food use and selection during spring migration. We experimentally collected green-winged teal and quantified food use and availability to determine selection in the Illinois River Valley during February–April 2016–2018. We removed, dried, and weighed (±0.1 mg) food items by taxa from the upper digestive tract (proventriculus and esophagus) of birds and core samples for comparison. Additionally, we evaluated retention of common diet items between small (#60; 250 μm) and medium (#35; 500 μm) sieves and the effect of processing sieve size on selection coefficients. Seeds of moist-soil vegetation occurred in all green-winged teal diets, while invertebrates and vegetation material occurred in 67.4% and 25.8% of diets, respectively. Green-winged teal consumed 85.8% (CI95 = 81.2–90.3%) plant material and 14.2% (CI95 = 9.6–18.7%) invertebrates based on aggregate dry biomass. We failed to find support for selection of either plant or animal foods in general, but green-winged teal selected Cyperus spp., Ammannia spp., Leptochloa spp., and Potamogeton spp. and avoided Amaranthus spp., Ipomoea spp., Echinochloa spp., and Oligochaeta individual taxa. We found no support for a difference in selection ratios between sexes, but selection ratios differed among years and wetland connectivity regimes with changes in food availability. Sieve size had minimal impact on rank and selection intensity of most food items, but only small sieves captured Ammannia spp., which was an important diet item. We found no evidence that green-winged teal selected invertebrates in our midlatitude migration study area, as has been speculated for dabbling ducks in general during spring migration (i.e., spring diet-shift hypothesis). We encourage managers to provide shallowly flooded wetlands with desirable plant taxa (e.g., Cyperus spp., Ammannia spp., Leptochloa spp.) for green-winged teal by maintaining actively managed moist-soil wetlands that are made available during spring migration.

Federal, state, and private entities manage seasonal flooded, shallow wetlands to provide food and other habitat resources for wetland-dependent migratory birds, including migrating and wintering waterfowl. Individual National Wildlife Refuges managed by the U.S. Fish and Wildlife Service annually monitor seed production in moist-soil wetlands to track performance relative to regional foraging habitat objectives and to evaluate local habitat management activities. The National Wildlife Refuge System does not currently have a standard sampling protocol, and thus seeks a reliable rapid assessment method for estimating seed production to achieve standardized estimates and to avoid inconsistencies in data collection, metrics used, and usefulness of the monitoring efforts. We compared seed yield estimates derived from a suite of commonly used seed production assessment methods with those from soil core samples across six National Wildlife Refuges in the southeastern United States. The most parsimonious model included only common plant species and a single visual assessment of overall coverage (1–5) and seed quality (1–4) for each moist-soil unit (r2adj = 0.71). Generally, models that included only common plant species and a visual estimate of seed yield for moist-soil wetlands overall had greater support than models that included all plant species and those that included data from subplots (n = 10) nested within moist-soil wetlands. Experience level of observer had a moderate effect on accuracy (r2mar = 0.20) and geographic range increased variation in overall seed yield estimates within moist-soil wetlands. Notably, we found that similar indices developed in different geographic regions performed well across the Southeast, but a widely used index based on estimates of seed yield for individual plant species performed poorly in this study. Standardizing the use of a single, efficient, and reliable method to estimate seed abundance in moist-soil wetlands will provide wetland mangers the ability to consistently estimate performance relative to objectives, evaluate management actions, and track trends on National Wildlife Refuges in the southeastern United States.

Scientists estimate biomass of invertebrates to evaluate wetland management practices, estimate energetic carrying capacity for wildlife, assess habitat condition and disturbance, and quantify ecosystem services. For waterfowl and other waterbirds in North America, carrying capacity in migratory and wintering regions is estimated using food density, of which invertebrates can be a significant component. However, we are not aware of previous literature that has described the effects of reagents used during core sample processing on invertebrate biomass and abundance. We tested the effects of hydrogen peroxide on aquatic invertebrates to determine whether a reagent used to disassociate soils during core sample processing biased estimates of biomass and abundance. Wet masses of chironomid larvae were less (x̄ = 23.5% loss) in samples exposed to hydrogen peroxide than those exposed only to tap water and biomass decreased approximately 2.9% with each minute of exposure time. Dry mass of larvae was less in samples exposed to hydrogen peroxide than in those exposed only to tap water (x̄ = 2.5% loss), but we did not detect an effect of exposure time on mass lost. Hydrogen peroxide did not influence the abundance of macro- or microinvertebrates in test samples. Thus, bias associated with dry mass estimates of invertebrates from core samples treated with hydrogen peroxide is likely minimal in terms of application in energetic carrying capacity models. However, use of hydrogen peroxide during core sample processing may cause significant bias if biomass estimates are based on wet mass.