Foraging habits of spring migrating waterfowl in the upper mississippi river and great lakes region

GENERAL INTRODUCTION
Diets of most organisms are primarily determined by what foods are available to them, but also vary among species of closely related organisms, stage of growth, reproductive status, and stage of annual cycle; the latter 3 are likely most influenced by physiological changes in nutritional demand (Sedinger 1992). For example, diets of many animals coincide with changes in forage abundance and quality (Kennish 1996), and small changes in forage quality may influence fitness. Reindeer (Rangifer tarandus L.) in Alaska that fed in high-quality habitats (i.e., a diversity of forage) consumed high-energy foods disproportionately to availability, subsequently increasing their bodymass by 14% and were 35% more likely to conceive than reindeer that fed in low-quality habitats (White 1983). Cook et al. (2001) observed reduced pregnancy rates in elk (Cervis elaphus nelsoni) fed diets with low digestible energy, indicating reserves acquired during summer and early autumn are important for survival and reproduction later in the year. Results of these studies indicated some animals not only forage to satisfy daily nutritional and energetic requirements, but also to obtain nutrient reserves that influence survival and fitness.

Similarly, migratory waterfowl acquire and depend on a variety of foods throughout their annual cycle but have evolved to exploit unfamiliar feeding sites with variable forage abundance and quality. For example, waterfowl sustain long flights from breeding to staging areas via nutrients acquired from high-energy foods during late summer and early fall (Gruenhagen and Fredrickson 1990). Refueling and maintenance of body condition occurs on staging and wintering grounds by increasing consumption of high-energy foods (Heitmeyer 1985, Delnicki and Reinecke 1986). At the onset of spring, waterfowl migrate from wintering areas to their breeding areas. During spring migration, some waterfowl switch from high-energy diets to diets high in protein to prepare for reproduction (Lovvorn 1987, Miller 1987, Manley et al. 1992). It is unclear however, when the switch from a high carbohydrate winter diet to a high protein breeding diet occurs. This is important to understand because knowing when this switch occurs can guide habitat management practices to produce desirable food types depending on stage of migration.

Late-winter and spring are critically important to arctic-nesting geese that depend on fat and protein accumulated before reaching their breeding areas, as these reserves are used to meet nutritional requirements of reproduction (Ankney and MacInnes 1978). Some duck species also rely on reserves acquired during migration for reproduction (Krapu 1981, Afton and Ankney 1991, Ankney and Alisauskas 1991), and poor body condition in late-winter may lead to reduced fitness (Dubovsky and Kaminski 1994). Several field studies have shown that birds are capable of breeding earlier and achieving greater reproductive success if fed high-quality diets prior to nesting (Nager 2006), suggesting there is an endogenous nutrient threshold influencing the initiation of breeding (Reynolds 1972, Gorman et al. 2008). A female that rapidly exceeds this threshold after arrival should therefore experience greater reproductive success (Reynolds 1972).

Migratory waterfowl are likely limited by the quantity and quality of habitats available for foraging. Wetlands that migratory waterfowl depend on to acquire nutrients for survival and reproduction are being lost at a rate of 47,000 ha/yr (Tiner 1998), with the majority of this loss being anthropogenic (Howe et al. 1989). My study occurred in the Upper Mississippi River and Great Lakes Region (UMR/GLR), located in the middle of the Mississippi Flyway and annually used by millions of waterfowl (Bellrose 1980). The Upper Mississippi River and its nutrient-rich floodplain has been dramatically altered by the expansion of agriculture. Of the six states in North America that have lost > 80% of their original wetlands, five (Iowa, Missouri, Illinois, Indiana, and Ohio) are located in the UMR/GLR (Dahl 1990). Consequently, populations of waterfowl that breed in and migrate through the UMR/GLR must rely on fewer wetlands (most of which are degraded due to decreased water quality and invasive plant species), than historically available to meet nutritional requirements, potentially having negative impacts on populations (Krapu 1981). Because there is limited information regarding nutritional demands of ducks during spring, an important period for waterfowl populations, my objective in Chapter 1 was to document the diet of spring migrating waterfowl. I emphasized how diet varied among species and attempted to describe how some of the exogenous factors (e.g., latitude, longitude, year, date, foraging habitat, and reproductive status) of individuals may have influenced diet.

If specific nutrients (e.g., protein) are required during specific periods, and those nutrients are limited in the environment, then organisms should exhibit selection for resources high in required nutrients. Thus, researchers should be able to gain insight into nutritional requirements of organisms by determining what food sources are being selected, relative to their availability. Many previous diet studies were inadequate at evaluating resource selection because they did not assess or consider consequences of forage availability. Thus, little information is available regarding nutritional requirements of waterfowl during spring migration; a period of time when nutrient availability likely influences individual fitness. Insight into diet of waterfowl will allow habitat managers to manage habitat to provide resources that meet the needs of waterfowl during spring migration, potentially influencing the ability of individuals to successfully reproduce. My objective in Chapter 2 was to document if spring migrating waterfowl were selecting foods (i.e., consuming foods in greater proportion than what was available to them) high in specific nutrients (i.e., proteins or carbohydrates) from representative wetland habitats throughout the UMR/GLR during spring. Hereafter, if I observed a duck consuming a food type in greater proportion than available to them, I considered them to be selecting for that food type. Not only will this information provide a guideline for habitat management for spring migrating ducks but it will allow me to infer nutritional demands of ducks during this time period, as nutritional needs likely change as migration advances and reproduction nears.

CHAPTER 1: DIET OF MIGRATING WATERFOWL DURING SPRING IN

THE UPPER MISSISSIPPI RIVER AND GREAT LAKES REGION
INTRODUCTION
Diet in Relation to Annual Life Cycle of Waterfowl
Waterfowl have adapted to efficiently use foods in heterogeneous environments to acquire nutrients and energy for life-cycle activities (i.e., migration, courtship, and reproduction) while having limited nutrient-storage capacities (Barlein 2003). Because nutritional and energetic demands and food availability vary throughout the year, duck diets are diverse and variable (Krapu and Reinecke 1992). Assessment of forage availability and diet of ducks during breeding (Krapu 1979) and fall migration (Weller 1988, Stafford et al. 2006, Havens 2007, Kross et al. 2008) has been documented, enabling managers to identify, conserve, and restore habitats that provide critical forage during these seasons. Diet may impact reproductive success, as spring diet has been shown to influence reproductive habits of American black ducks (Anas rubripes) (Barboza and Jorde 2002), however little information is available on diet of spring migratory ducks and how it varies among species, with regards to availability, latitude and longitude, years, time within a season, habitat types, and reproductive status. Abundance of foods that waterfowl depend on varies within and among seasons and often have patchy distributions. Diet throughout the annual cycle is strongly correlated with availability, as invertebrates are most abundant during summer (Kaminski and Prince 1981) and seeds are most abundant during fall and winter (Gruenhagen and Fredrickson 1990). Food availability during spring is largely unknown, but is likely scarce when compared to other seasons of the year, especially considering that most managed wetlands are flooded in early fall for wintering waterfowl and these wetlands are more likely to have fewer seeds due to decomposition (Greer et al. 2006).

Although there is a general trend for duck diets to consist of high protein during breeding and carbohydrates outside of breeding, the proportions of proteinacious and carbohydrate foods consumed varies considerably among taxa (e.g., dabbling ducks vs. diving ducks). Some species such as mallards (Anas platyrhynchos) depend almost exclusively on carbohydrates while others such as blue-winged teal (Anas discors) consume a more varied diet that consists of both protein and carbohydrates. Additionally, some waterfowl species exhibit a switch in diet, from predominately seeds to predominately invertebrates, in late winter and early spring (Taylor 1978, Gruenhagen 1987, Lovvorn 1987, Miller 1987, Krapu and Reinecke 1992, Manley et al. 1992). The cause of this switch is not known, but is likely related to a reduction in availability of seeds or changes in nutritional demands, hence selection of proteinacious foods (Lovvorn 1987).

Because poor winter habitats may delay nesting in mallards (Kaminski and Gluesing 1987, Dubovsky and Kaminski 1994) and American black ducks (Barboza and Jorde 2002), it is likely that poor-quality spring foraging habitat may also negatively impact nesting waterfowl (Afton and Ankney 1991). Inadequate reserves acquired during spring-staging may decrease nest success through delayed nesting (Harris 1969, McNeil and Leger 1987, Rohwer 1992, Koons and Rotella 2003) or cause some hens to defer reproduction altogether (Newton 2006). Implicit in these findings is that, as spring progresses, diet should reflect reproductive needs. In particular, dietary needs of waterfowl at northern latitudes during spring may differ from waterfowl at southern latitudes during spring. There is strong evidence that late-winter and spring conditions have carryover effects on reproductive success of lesser scaup (Anteau and Afton 2004), however, there is sparse information pertaining to ecology of spring migrating waterfowl in the mid-latitude portions of the Mississippi Flyway, particularly with regards to how diet may change with date or latitude.

Existing information on diet during spring migration is sparse and conflicting (Newton 2006); some studies found that ducks consumed high-carbohydrate foods (Jorde 1981, LaGrange 1985, Gruenhagen and Fredrickson 1990, Strand et al. 2008), whereas others found ducks consumed high-protein foods (Manley et al. 1992, Badzinski and Petrie 2006a). Two recent studies on feeding ecology of scaup during spring reported high-carbohydrate foods were the main component of their diet (Smith 2007, Strand et al. 2008), but a third indicated high-protein foods as the main dietary component (Badzinski and Petrie 2006a).

STUDY OBJECTIVES

Previous spring feeding ecology studies of waterfowl are lacking in that they focused only on 1 or 2 species and usually only at a single location; therefore, my specific objective was to determine if diet during spring varied among selected dabbling (mallard, gadwall (Anas strepera) and blue-winged teal) and diving (lesser scaup (Aythya affinis) and ring-necked duck (Aythya collaris)) ducks. Additionally, I was interested in assessing if diets were influenced by availability of foods (including annual variation), collection date, habitat type, latitude, longitude, and reproductive status during spring.

METHODS

Study Site Selection

The Upper Mississippi River and Great Lakes Region (UMR/GLR) encompasses ten states and is located between important breeding and wintering areas of North American waterfowl (Figure 1.1). Study sites within the UMR/GLR were selected based on their presumed importance to migratory waterfowl during the spring (UMR/GLR Joint Venture 1998) and because they represent typical habitat within the region. The location of the sites exhibited considerable longitudinal and latitudinal variation to enable incorporation of spatial variation in diet as birds migrated northward during spring (Figure 1.2). The study region was divided into 2 transects, a western transect and an eastern transect, with each transect having three study sites distributed south to north. The western transect included (1) the Cache River region of southern Illinois, (2) the Illinois River region of central Illinois and, (3) the Horicon Marsh region of southeast Wisconsin, whereas the eastern transect was comprised of (1) the Scioto River region of southern Ohio, (2) the Lake Erie region of northern Ohio, and (3) the Saginaw Bay region of Michigan.

The southernmost study site (89^o3’ W, 37 ^o18’ N) of the western transect was located in Southern Illinois and included the Cache River and its floodplain. This region supports a variety of migratory birds, and was deemed an area of international importance by the RAMSAR convention. Bottomland forest habitat, identified as critical habitat for mid-migratory waterfowl (Heitmeyer 1985), represented 70% of available wetland area in the Cache River floodplain. This region also is home to extensive baldcypress and tupelo swamps that are unique to the Upper Mississippi River region. Because of the

Figure 1.1. Location of the Upper MS River and Great Lakes Region with respect to wintering and breeding grounds of ducks migrating through the Upper MS River and Great Lakes Region (image taken from United States Fish and Wildlife Service 2008).

Figure 1.2. Location of study sites in the Upper MS River and Great Lakes Region (image provided by Jake Straub, The Ohio State University).

uniqueness of this habitat to my area of interest (i.e., the UMR/GLR), it was not included in sampling efforts. Cypress-tupelo swamps did however represent approximately 16% of the wetland habitat in the Cache River floodplain. The Cache River floodplain also contained scrub-shrub (6.3%), open water (3.5%), moist-soil (2.0%), and emergent vegetation (1.3%) wetland habitat types (Havera 1999). The rivers floodplain is expansive and availability of wetland habitat in this region is largely dependant on winter and spring precipitation to produce flooding of riparian habitats. Due to restoration efforts in recent years through the wetlands reserve program (WRP), there are numerous managed wetlands that persist despite altered river conditions. Mean annual temperature is 13.7° C, with average winter temperatures ranging from 3.2 − 8.3° C, and spring temperatures ranging from 8.5 − 18.8° C (Illinois State Water Survey 2008). Mean annual rainfall is 122.7 cm; greatest precipitation occurs in late winter and spring (Illinois State Water Survey 2008).

The mid-latitude study site (90^o 12’ W, 40^o 12’ N) of the western transect was located along the central region of the Illinois River near Chandlerville, IL. In the early 1900’s, pristine bottomland water areas made the Illinois River one of the most important regions for migratory waterfowl in North America (Bellrose et al. 1983, UMR/GLR Joint Venture 1998). The Illinois River drains over half of the state, most of which is intensively farmed in row crops (Barrows 1910). Threats to this region include row crop and bank erosion (i.e., increasing nutrient and sediment loads in the river and causing the filling of lateral lakes) and navigation dams that increased low midsummer river levels, resulting in a deepening and extension of all water areas (Steffeck et al. 1980). As a result of the navigation dams, lakes that were previously separated from the river channel by bottomland timber are now connected and mast-producing timber is dead from inundation (Bellrose et al. 1983). Consequently, wetlands of this region (particularly lateral lakes) have been adversely impacted. Similar to some of my other study sites (i.e. the Cache and Scioto Rivers); the wetland area of this region is dictated by winter and spring precipitation and flood events of the river. Fortunately for waterfowl that depend on this region, the greatest precipitation occurs in winter and spring (Illinois State Water Survey 2008) and coincide with peak waterfowl migration (Havera 1999). Mean annual temperature is 10.8° C, with average winter temperatures ranging from -1.9 − 4.8° C, and spring temperatures ranging from 4.4 − 16.9° C (Illinois State Water Survey 2008).

The northernmost study site (88^o 50’ W, 43^o 48’ N) of the western transect was located in the Upper Rock River watershed in southeast Wisconsin near Horicon Marsh. This area has been identified as the region of Wisconsin that contains the majority of migratory habitat for waterfowl (UMR/GLR Joint Venture 1998). Representative wetland habitats in this area include riverine wetlands, lacustrine wetlands and a number of kettle ponds and prairie potholes. Agriculture represents 59% of the land use of this watershed and has resulted in substantial drainage of pothole wetlands (Wisconsin Department of Natural Resources 2002). A number of actively managed wetlands are present in this region with the majority of them being located on Horicon National Wildlife Refuge and State Wildlife Area and surrounding waterfowl production areas. One of the largest freshwater wetlands in the United States, Horicon Marsh covers 31,904 acres and is owned and managed by the Fish and Wildlife Service and the Wisconsin Department of Natural Resources. Soil erosion and siltation, invasion of exotic species (i.e. carp, purple loosestrife) and high inflow of nutrients from surrounding farms are the biggest threats to this wetland complex (Wisconsin Department of Natural Resources 2002). Mean annual temperature is 7.6° C, with average winter temperatures ranging from -6.3 − 1.9° C, and spring temperatures ranging from 0.2 − 14.3° C (Midwestern Regional Climate Center 2008). Mean annual rainfall is 83.9 cm; greatest precipitation occurs in late summer and early fall, while the driest periods are mid-late winter (Midwestern Regional Climate Center 2008).

The southernmost study site (82^o 59’ W, 39^o 40’ N) of the eastern transect was located near Circleville, Ohio and contained the Scioto River and its floodplain. The Scioto River Valley is recognized as an important area for migrating American black ducks and mallards, despite the fact that the area contains little wetland area (UMR/GLR Joint Venture 1998). Wetland area in the Scioto River Valley, as defined by the National Wetlands Inventory (Cowardin et al. 1979), was the least of the six study sites (i.e., 9.5 km²). The majority of wetland habitat at this study site was riverine with adjacent forested and scrub-shrub wetlands. Heavy rain and subsequent flooding from the Scioto River produces many acres of flooded hardwoods and agriculture that attract thousands of ducks (UMR/GLR Joint Venture 1998). Consequently, if the Scioto stays within its banks (i.e., in years of little precipitation), as experienced in the spring of 2006, there is little wetland habitat available to waterfowl. Mean annual temperature is 10.5° C, with average winter temperatures ranging from -3.06 − 5.67° C, and spring temperatures ranging from -1.11 − 15.5° C. Mean annual snowfall is 36.1 cm and mean annual rainfall is 99.1 cm; with highest snowfalls in January and greatest precipitation in summer and late spring (Midwestern Regional Climate Center 2008).

The mid-latitude study site (82^o 59’ W, 41^o 27’ N) of the eastern transect was located on Sandusky Bay Lake Erie, approximately 2 km southwest of Port Clinton, OH. Wetland habitat in this area consists of marshland, which separates the lake from agricultural farmland, and diked wetlands managed for migrating and wintering waterfowl. Agricultural practices have eliminated all but coastal marshes, which are now being impacted by urban encroachment (UMR/GLR Joint Venture 1998). This region still remains important to ducks that migrate through both the Mississippi and Atlantic Flyways despite the loss of historical Lake Erie marshes. The largest concentrations of staging American black ducks in North America can be found on Lake Erie marshes in this area (UMR/GLR Joint Venture 1998). Water levels in this region of Lake Erie are in constant flux because of varying wind direction and velocity. For example, a strong southwest wind may decrease the water level in this area, while a strong “northeaster” wind may cause water levels to rise (Farney 1975). Mean annual temperature is 9.89° C, with average winter temperatures ranging from -4.2 − 5.17° C, and spring temperatures ranging from -2.56 − 15.28° C. Mean annual snowfall is 55.1 and mean annual rainfall is 91.8 cm; with highest snowfalls in January and the greatest precipitation occurring in late summer (Midwestern Regional Climate Center 2008).

The northernmost study site (83^o 25’ W, 43^o 45’ N) of the eastern transect was located near Sebewaing, MI in Saginaw Bay of Lake Huron. Saginaw Bay is a large, shallow embayment of Lake Huron. Wetland habitat outside of the bay is limited and restricted to hunting clubs and state wildlife areas, which are all used in the spring by thousands of migrant tundra swans, Canada geese and various duck species (J. Straub, Ohio State University, personal communication). Agriculture is the dominant inland land use. Wetland degradation and loss in this area can be attributed to residential development (UMR/GLR Joint Venture 1998) and invasion of Common Reed (Phragmites australis). Mean annual temperature is 7.06° C, with average winter temperatures ranging from -6.11 − 3.11° C, and spring temperatures ranging from -5.33 – 11.5° C. Mean annual snowfall is 85.6 cm and mean annual rainfall is 66 cm with the greatest precipitation occurring in late summer (Midwestern Regional Climate Center 2008).

Study Species

I selected 3 species of dabbling ducks (Anatinae) for my study because of their diversity in body size, migration habits, and timing of reproduction. Dabbling ducks feed in shallow water by submerging their head or tipping up to reach submersed foods, whereas diving ducks feed in deeper water by diving underwater to feed. The mallard is the largest of the selected dabblers, travels the shortest distance to its wintering areas and initiates nesting within days after arriving on breeding areas. The blue-winged teal is the smallest of the selected dabblers, travels the greatest distance to its wintering areas and initiates nesting within days after arrival at breeding areas. The gadwall is a mid-sized dabbling duck, travels intermediate distances to its wintering areas and initiates nests 3 to 4 weeks after arriving on its breeding area (Arzel et al. 2006).

I selected the lesser scaup, hereafter may be referred to as scaup, and ring-necked duck to represent diving ducks (Aythyinae) for my study because they are similar in migratory and reproductive habits, but different with respect to diet and population trends. Scaup populations have experienced a substantial decline relative to the long-term average, while ring-necked duck populations have been stable or increasing during the same period (Wilkins et al. 2006). Additionally, scaup diets have previously been documented to include a large proportion of invertebrates during all stages of the annual cycle (Rogers and Korschgen 1966, Gammonley and Heitmeyer 1990), whereas ring-necked ducks appear to transition from seeds during fall and spring to invertebrates during breeding (Hohman 1985). Scaup and ring-necked ducks are similar in body size, wintering areas, and time between arrival and onset of incubation at breeding sites (Arzel et al. 2006).