Local and landscape level variables influencing migratory bird abundance, diversity, behavior, and community structure in Rainwater Basin wetlands

The Rainwater Basin (RWB) region in south-central Nebraska, which currently contains approximately 400 wetlands, is one of the most threatened and least studied wetland complexes in the Great Plains. Early soil surveys indicate the RWB originally included more than 4,000 natural wetland basins, totaling approximately 38,000 ha; however, conversion of wetlands to agricultural fields has resulted in the destruction of over 90% of the wetlands and 88% of the original wetland area. Even in their reduced, degraded condition, RWB wetlands provide essential stopover and staging areas to migratory wetland birds in the Central Flyway. Consequently, the RWB region has been identified as containing waterfowl habitat of major concern by the North American Waterfowl Management Plan. Five to 7 million waterfowl pass through the RWB region every spring, including virtually all of the 600,000 mid-continental white-fronted geese (Anser albifrons), 500,000 Canada geese (Branta canadensis), 50% of the mid-continental mallard (Anas platyrhynchos) population, and 30% of the continental northern pintail (Anas acuta) population.

Staging areas and migratory stopovers often function as geographic bottlenecks; entire populations within a flyway can be affected by the quality and quantity of available wetland habitat at stopover sites. This common dependence among migratory birds on staging sites has major implications for wetland conservation and restoration. In an effort to guide wetland stopover site restoration and conservation efforts, there has been an increased effort to determine what factors influence wetland habitat quality for migrating shorebirds and waterfowl. Although previous studies have shown that local and landscape-level factors influence abundance, diversity, and behavior of breeding birds on wetlands, few studies have assessed the impact of these factors on wetland birds during migration. Moreover, examination of wetland bird habitat use, behavior, and community structure provides the opportunity to test hypotheses about the mechanisms that allow wetland birds to coexist while migrating through a region with unpredictable resources. My objectives were to 1) examine local (within wetland and immediate watershed) factors influencing abundance, species richness, and diversity of waterfowl, shorebirds, and wading birds; 2) evaluate landscape-level factors influencing abundance, species richness, and diversity of waterfowl, shorebirds, and wading birds at various spatial scales; 3) determine the effects of hunting disturbance, time period, and migration chronology on bird behavior; and 4) examine community patterns and species associations to assess the importance of assembly rules in structuring wetland bird distribution and communities during migration.

I conducted biweekly avian surveys in 36-40 wetlands from mid-February through mid-May, 2002-2004. Concurrently, I recorded behaviors of wetland birds during 1 of 6 time periods, using a flock scan technique. I measured 16 local wetland habitat variables in mid-May of each year, including percent emergent vegetation, water depth and organic matter depth. Using ArcView and various spatial datasets, I calculated landscape variables for each study wetland within 5 and 10 km. Landscape variables included area, number of patches, and mean patch size of agricultural land, grassland, wetlands, riparian area, and land with high sedimentation potential. Using multiple linear regression, I tested hypotheses about how local and landscape variables influenced species richness, diversity, and abundance of geese, dabbling ducks, diving ducks, shorebirds, and wading birds. The best models containing local and landscape variables for each dependent variable were selected using Akaike’s Information Criterion.

I observed an estimated 945,333 individual migratory wetland birds representing 61 species during 2002 surveys (n = 637), 517,650 birds representing 62 species during 2003 surveys (n = 873), and 1,097,950 birds representing 64 species during 2004 surveys (n = 955). Wetland area had a positive influence on goose abundance in all years, whereas percent emergent vegetation and hunting pressure had negative influences. Models predicting dabbling duck abundance were variable among years; however, number of wetlands within 5 km was the best predictor of dabbling duck abundance. Area of agriculture within 10 km of a wetland had a positive influence on dabbling duck abundance in years with low precipitation (2002 and 2003), whereas wetland vegetation was important for dabbling ducks in the wetter year (2004). Wetland area and percent of vegetation composed of inner marsh (drawdown and aquatic bed species) were the best and most consistent predictors of diving duck abundance. Shorebird abundance was best predicted by wetland area and area of agricultural land within 10 km. Wetland area was the only consistent predictor of wading bird abundance. Models predicting species richness contained wetland area as a positive predictor and water depth as a negative predictor. In addition, percent emergent vegetation was a positive predictor of species richness, indicating that species richness was greatest in wetlands with intermediate levels of vegetation.

To determine for what purpose birds are using wetlands during migration, I examined the effects of time period, hunting pressure, and year on behavior. I used a multivariate analysis of variance to test for differences in behavior among time periods, hunting categories (hunted within season, hunted out of season, and closed to hunting), and years for geese, dabbling ducks, diving ducks, and shorebirds. I also compared behaviors of early (before March 17) and late (after March 17) migrating dabbling ducks to determine whether a greater proportion of early migrants were observed feeding in wetlands than later migrants. Goose behavior did not differ among time periods; however, a smaller proportion of geese was observed feeding on hunted wetlands than unhunted wetlands in 2004. A smaller proportion of dabbling ducks was observed feeding on hunted wetlands compared to unhunted wetlands and in 2004 compared to other years. Diving duck behavior did not differ among time periods, years, or hunting categories. Shorebirds spent most of the time feeding; however, proportion of birds feeding was lower in 2004 than other years. Early migrating dabbling ducks fed less and rested more than did later migrants.

I used null model analysis to examine spatial and temporal co-occurrence patterns of geese, dabbling ducks, diving ducks, and shorebirds. I also calculated species association values to determine if dabbling ducks were avoiding wetlands occupied by snow geese (Chen caerluscens). Goose species co-occurred less often than expected in all years of the study, whereas co-occurrence patterns of dabbling ducks were not different than expected by chance in all years. However, when evaluated at a weekly scale, dabbling ducks co-occurred less often than expected during weeks of peak migration (high abundance). Diving ducks co-occurred less often than expected in 2002 and 2004 but not in 2003, whereas shorebirds co-occurred less often than expected in all years. The majority of association values among snow geese and dabbling species were positive, indicating that dabbling ducks are not avoiding wetlands with snow geese. However, negative associations among snow geese and dabbling ducks species might be observed in high precipitation years, when greater wetland availability would permit segregation and increase the chances of detecting co-occurrence patterns.

Wetland conservation in the RWB should focus on providing wetland complexes for migratory birds, particularly dabbling ducks. In addition, allowing wetlands to go through their natural hydrologic cycle should promote intermediate levels of emergent vegetation, which will increase use by dabbling ducks, shorebirds, and wading birds, while discouraging goose use of these wetlands. Spring hunting, especially in low precipitation years, causes a re-distribution of geese and dabbling ducks as well as decreasing dabbling duck feeding opportunities. During dry years, flooding temporary and seasonal wetlands and closing them to hunting will help offset the effects of hunting on non-target species by providing more protected feeding areas. In addition, a clearly defined, comprehensive plan for measuring the success of these objectives, as well as standardized monitoring protocols, are necessary for incorporating adaptive management into the conservation goals of the RWBJV. In particular, the ability to quickly and accurately assess wetland availability in the RWB will be crucial to making the best possible management decisions on where and when to distribute water on the landscape.