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

CHAPTER 2: FOOD SELECTION BY MIGRATING WATERFOWL DURING

SPRING IN THE UPPER MISSISSIPPI RIVER AND GREAT LAKES REGION
INTRODUCTION
In the previous chapter, I identified several factors that may influence diet of spring migrating ducks, including interspecific, temporal, latitudinal (i.e., relative distance from breeding sites) and longitudinal (i.e., variation caused by encountering different wintering populations) variation, and variation in availability of food types. This information may be used to manage wetlands to produce highly-consumed food types, but it provides little insight regarding nutritional or physiological needs of ducks during spring. By assessing consumption relative to food availability, hence selection for a food type; biologists can gain insight on nutritional demands of ducks.
Food Selection

A number of previous studies established that ducks are capable of selecting specific food types (Pederson and Pederson 1983, Manley et al. 1992, Thompson et al. 1992, McKnight and Hepp 1998, Anderson et al. 2000, Smith 2007). During fall and winter, most ducks consume foods high in carbohydrates; however, the degree to which this selection occurs varies among species and molt intensity (Anderson et al. 2000). Although high-carbohydrate seeds and grains comprise a large portion of winter diets of dabbling ducks, and diving ducks consume varying amounts of animal matter during winter, few studies have evaluated diet in relation to availability during winter simultaneously (Pederson and Pederson 1983, Thompson et al. 1992, McKnight and Hepp 1998, Anderson et al. 2000). Alternatively, during breeding, dabbling and diving ducks forage almost exclusively on invertebrates (Swanson et al. 1974, Swanson et al. 1985, Ankney and Alisauskas 1991). Assuming breeding season diet is partially based on a physiological demand for protein, it is unclear when demands shift from high-energy foods during winter to high-protein foods during breeding. A number of factors may influence the timing of this transition in food types consumed (e.g., food availability, digestive physiology, amount of time spent at breeding areas before initiating rapid follicle development (RFD), use of endogenous/exogenous nutrients in clutch formation, proximity to breeding areas, etc.).

Birds that have been habituated to one food type (e.g., high-carbohydrate foods during winter) cannot immediately transition to an alternative digestive physiology that permits them to efficiently use other food types (e.g., high-protein foods during spring and summer); therefore, it is plausible that ducks may increase invertebrate consumption during spring in order to increase efficiency of high-protein diets at breeding areas (Afik and Karasov 1995). This digestive efficiency concept could be especially important to waterfowl that spend little time on their breeding areas before initiating nests (Toft et al. 1982).

Generally, diving ducks spend more time on breeding areas before initiating RFD than dabbling ducks (Alisauskas and Ankney 1992). This is important because waterfowl that immediately initiate RFD upon arrival (e.g., mallards and blue-winged teal) at breeding areas probably depend heavily on existing lipid reserves or somatic protein. Likewise, a species that spends 3 − 4 weeks at a breeding area before RFD (e.g., gadwall, lesser scaup, ring-necked duck) likely uses that time to acquire nutrients needed for RFD and subsequent incubation.

Because some species of waterfowl store nutrients before reaching breeding areas, conditions experienced prior to breeding must be considered when evaluating reproductive success (Raveling 1979). Poor-quality spring habitat and forage may negatively impact productivity of waterfowl (Afton and Ankney 1991, Dubovsky and Kaminski 1994, Barboza and Jorde 2002). Dubovsky and Kaminski (1994) found poor winter habitat conditions may have delayed nesting in mallards, presumably due to a lower rate of nutrient acquisition. 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), lead to a reduction in clutch size, or cause some hens to defer reproduction altogether (Newton 2006). The degree to which ducks depend on endogenous protein and lipid for breeding differs; some ducks relying more on lipid stores (wood ducks - Drobney 1980, mallards - Krapu 1981, ring-necked ducks - Hohman 1986), and others more on somatic protein (American wigeon and gadwall - Alisauskas and Ankney 1992). Understanding food selection during spring can provide insight on how lipid reserves and somatic protein are acquired before breeding. For example, reserves obtained during spring are central to reproduction in Arctic nesting geese (Ankney and MacInnes 1978, Raveling 1979). Despite evidence that late-winter and spring conditions have carryover affects on reproductive efforts of ducks (Kaminski and Gluesing 1987, Dubovsky and Kaminski 1994), little information is available regarding food selection of spring migrating ducks in mid-latitude portions of the Mississippi Flyway.

Lastly, proximity to nesting areas during spring migration may influence diet. Food consumption of ducks at northern latitudes during spring may reflect dietary needs for reproduction, whereas diet at southern latitudes during spring may reflect habitat conditions at respective wintering areas. Correspondingly, it is important to consider that “breeding areas” of blue-winged teal, gadwalls, and some mallards (e.g., Missouri Coteau area of the Prairie Pothole Region) are likely closer to the UMR/GLR than are “breeding areas” of lesser scaup and ring-necked ducks (e.g., Alaska and boreal forest region of Canada).

I selected the suite of species for this study because they represented a variety of foraging behaviors observed in spring-migrating ducks (i.e., dabbling and diving ducks). By examining diverse species, I expected to be able to detect differences in diet selection during spring based on different life-history strategies and traits. By determining if and where (i.e., at what latitude) diet transitions from one food type to another occur during spring, I will be able to recommend habitat management practices that maximize the productivity of foods and meet nutritional requirements of a wide variety of ducks.
STUDY OBJECTIVES

The goal of my study was to determine what 2 classes of foods, seeds or invertebrates, were selected by mallards, blue-winged teal, lesser scaup, and ringed-neck ducks during spring migration through the UMR/GLR. I was unable to evaluate food selection in gadwall because the largest component of their diet was vegetation, and I did not evaluate availability of vegetation. My specific objective was to determine if these species transitioned from selecting high-carbohydrate foods to high-protein foods prior to or during spring migration, and if the timing of this potential transition varied in a predictable manner as a function of life-history characteristics.

Based on life-history characteristics, I predicted that blue-winged teal and mallards would increase consumption of invertebrates as spring progressed, allowing them to efficiently transition to a high-protein diet and initiate RFD soon after arrival on breeding areas. Because gadwalls are herbivorous, protein-limited (i.e., depend on somatic protein for clutch formation; Ankney and Alisauskas 1991), and spend 3 − 4 weeks at breeding areas prior to nest initiation, I predicted they would increase invertebrate consumption as they migrated north during spring. Because of their herbivorous nature and dependence on green vegetation, however, I expected to see only small increases in invertebrate consumption with latitude. Scaup have been documented to consume > 50% animal matter at both wintering and breeding sites, therefore, I expected to see a heavy dependence on invertebrates by spring migrating scaup relative to other species in the UMR/GLR. I expected seeds to be an important component of scaup diets, however, because: (1) scaup consume a considerably large proportion of invertebrates throughout the annual cycle (Rogers and Korschgen 1966) and (2) scaup spend considerable time at breeding areas before initiating RFD and the habitats scaup depend on during breeding are highly productive for invertebrates, and (3) breeding female scaup do not further accumulate lipid reserves while on breeding areas (Afton and Ankney 1991). Similarly, I expected the diet of ring-necked ducks to consist largely of seeds, given that their time on breeding areas before initiating RFD is allocated to protein acquisition (Hohman 1985).


METHODS

Food Availability

I collected data used in these analyses in 2006 at the same study sites described in the previous chapter. I collected food availability samples during 3 time-periods of spring (early, middle, late) to quantify variation in forage abundance throughout migration. Timing of sampling varied among study sites to account for climatic differences and migratory stages of waterfowl (e.g., the difference in the timing of peak waterfowl abundance between northern and southern sites). For example, the early food sampling period began in early to mid-February at southern sites and mid-March at northern sites. As deemed by the latitude of the study area, early food samples were taken as soon as ice conditions permitted to determine food availability during early stages of migration. Mid-spring food samples were collected during the peak of waterfowl migration, whereas late food samples were taken after the majority of birds had passed through. Samples were collected using stratified random sampling from wetland types to estimate forage availability at each study site and to provide an index of food availability in different wetland types (i.e., palustrine forested, palustrine emergent, and lacustrine open-water/riverine wetlands). I determined sample locations by overlaying a grid of 400m x 400m (16 hectares) cells and excluding cells with < 2 ha of wetland habitat as identified by the national wetlands inventory (NWI) (Cowardin et al. 1979) or < 8 ha of soil that held moisture, as identified by soil moisture index data (Ducks Unlimited 2005). I categorized remaining cells as forested wetland, non-forested wetland, riverine/lacustrine, or agricultural according to dominant vegetation and wetland types present. I selected agricultural habitats based on soil moisture index data as areas that had wet or very wet soils. Within each selected cell, I identified and sampled each wetland type. For example, if cell 231 at the Cache River study area was identified as forested wetland based on NWI data, but the cell contained palustrine forested and palustrine emergent wetland, I identified and sampled each wetland type.

I sampled 60 wetland cells at each of the 6 study sites, 20 of which were agricultural wetlands and the rest were sampled in proportion to abundance of wetland type. For example, forested wetland represented approximately 80% of wetland area in the Cache River study area; therefore, 32 of the 60 cells selected for sampling were dominated by forested wetland. I did not select blocks proportionally in flooded agricultural habitat due to the high abundance of ephemeral pools in agricultural fields that may have resulted in over-sampling if considered proportionally. Additionally, reliable data were not available to estimate the total proportion of flooded agricultural crops in an area. Because waterfowl use of agricultural land may be substantial (LaGrange and Dinsmore 1989, Krapu et al. 2004), I allocated 20 cells as flooded agriculture regardless of the estimated proportion of flooded agriculture at a study site.

Within a wetland basin, I collected food samples along randomly selected transects. Once in the center of a wetland basin, I used the time (in seconds) on a handheld PDA to determine the direction in which the transect occurred (00 corresponding to straight north, 15 to straight east, 30 to straight south, 45 to straight west, and so on). Then, I took 2 samples along the transect; a deep sample taken at the first location encountered along the transect that was approximately 45 cm deep and a shallow sample taken at a randomly selected depth provided by the PDA between 45 cm and 1 cm. Deep food samples were used to estimate food availability to diving ducks whereas shallow food samples were used for dabbling ducks.

A wetland sample consisted of a d-frame sweep net sample taken along the length of a drop box for 3 sweeps, (33 cm diameter, 500 µm mesh) and a core sample (7 cm diameter, ~5 cm deep) taken from within the drop box (100 cm x 50 cm x 75 cm, 500 µm mesh side panels). The drop box was used to ensure consistency in sampling area of each sample. I rinsed samples through a 500 µm mesh sieve bucket to remove clay material and unwanted debris, placed core and sweep samples in separate bags, and preserved the sample in 10% buffered formalin solution.
Laboratory Analysis

I washed food availability samples through 1mm, 750 µm and 500 µm sieves to facilitate processing of sample contents by size. All seeds and invertebrates were recovered from samples in the laboratory at Southern Illinois University Carbondale (SIUC). Animal food items were identified at SIUC (Schultheis, Southern Illinois University, dissertation in progress; Merrit and Cummins 1996), whereas seed identification was conducted at the Ohio State University (Straub, The Ohio State University, thesis in progress). Foods recovered from availability samples were identified similarly to esophageal contents, but I did not record plant material (i.e., algae, Lemna sp., etc.). I dried food items for ≥ 48 hours at 60^oC and then weighed them on a top-loading balance.

Statistical Analysis

There was evidence that both invertebrate and seed availability did not significantly increase or decrease throughout spring (Straub 2008, Schultheis, dissertation in progress), therefore I calculated means and standard errors of invertebrates and seeds in both diet and availability samples (PROC UNIVARIATE; SAS Institute, Inc., Cary, NC) and compared them using a Z-test in program CONTRAST to investigate if the proportion of food items in duck diets differed from the proportion of food items available at the study site in which the ducks were collected (i.e., selecting for a food type) (Hines and Sauer 1989). I assumed that a diet that was significantly different (P < 0.05) than availability indicated selection of that food item. Additionally, I assumed a diet was moderately significant if 0.05 < P < 0.10. I did not evaluate selection by a species if < 5 ducks of a particular species were collected at a study site. I considered all seeds and invertebrates recorded from shallow and deep samples as available to foraging dabbling and diving ducks, respectively.

I estimated forage availability for diving ducks by including deep availability samples taken in lacustrine open-water and palustrine emergent habitats, but did not include samples from palustrine forested habitats because I considered it unavailable to diving ducks (i.e., diving ducks were never observed using this habitat and I considered it to be inaccessible to diving ducks). Likewise, I estimated forage availability for dabbling ducks by including all shallow availability samples that were collected. Ducks collected in agricultural habitats were not included in these analyses because food sampling methods in these habitats were not replicable among sites in 2006. Additionally, to eliminate possible bias in diet estimates, I removed 5 mallards collected in emergent wetlands because corn was the predominant food item in esophageal contents. With regard to gadwall diet and selection analyses, I was unable to assess selection because vegetative items composed the largest proportion of their diet and vegetative items were not sorted from availability samples (i.e., I could not determine availability, hence selection, of vegetative food items).
RESULTS

Availability

Generally, there were more seeds available at each study site than invertebrates (Table 2.1). Availability estimates were not calculated for diving ducks at Wisconsin because < 5 ring-necked ducks and lesser scaup were collected in 2006. Mean seed availability was highest for dabbling ducks at the Wisconsin site (309.8 ± 51.4 kg/ha) and diving ducks at the Illinois River site (131.1 ± 38.7 kg/ha). Mean invertebrate availability was highest for dabbling ducks at the Saginaw Bay site (116.4 ± 37.6 kg/ha) and diving ducks at the Cache River site (68.7 ± 20.8 kg/ha).

Blue-winged Teal Food Selection

I included data from 94 blue-winged teal collected in spring 2006 in selection analyses. Of these, 22 were collected at the Cache River, 21 at the Illinois River, 19 at Wisconsin, 20 at Lake Erie, and 12 at Saginaw Bay. Only 2 teal were collected at the Scioto River, therefore I did not include these in analyses. Using the mean available forage and associated standard error from shallow food samples in program CONTRAST, I found blue-winged teal consumed a significantly higher percentage of invertebrates than

Table 2.1. Mean food availability (kg/ha) and standard error (SE) of seeds and invertebrates found in shallow (for dabbling ducks) and deep (for diving ducks) habitats during spring 2006.

________________________________________________________________________'>________________________________________________________________________

Species Type Site Mean SE ________________________________________________________________________

Dabbling Ducks Seeds Cache River 163.1 44.7

Invertebrates Cache River 49.7 14.3

Dabbling Ducks Seeds Illinois River 97.2 19.8

Invertebrates Illinois River 33.9 6.8

Dabbling Ducks Seeds Wisconsin 309.8 51.4

Invertebrates Wisconsin 50.6 16.3

Dabbling Ducks Seeds Scioto River 115.4 33.9

Invertebrates Scioto River 12.1 3.4

Dabbling Ducks Seeds Lake Erie 177.3 42.9

Invertebrates Lake Erie 31.4 9.5

Dabbling Ducks Seeds Saginaw Bay 145.5 32.5

Invertebrates Saginaw Bay 116.4 37.6

Diving Ducks Seeds Cache River 124.6 37.3

Invertebrates Cache River 68.7 20.8

Diving Ducks Seeds Illinois River 131.1 38.7

Invertebrates Illinois River 22.3 9.3

Diving Ducks Seeds Wisconsin N/A N/A

Invertebrates Wisconsin N/A N/A

Diving Ducks Seeds Scioto River 70.0 32.3

Invertebrates Scioto River 58.2 35.1

Diving Ducks Seeds Lake Erie 120.6 44.3

Invertebrates Lake Erie 12.2 4.0

Diving Ducks Seeds Saginaw Bay 25.9 12.9

Invertebrates Saginaw Bay 20.5 8.3

________________________________________________________________________

were available at Wisconsin (P < 0.001) and Lake Erie (P < 0.001) (Table 2.2). Blue-winged teal consumed food in proportion to availability at the Cache River, Illinois River, and Saginaw Bay (P > 0.05).

Mallard Food Selection

I included data from 84 mallards collected in spring 2006 in selection analyses. Of these, 13 were collected at the Cache River, 11 at the Illinois River, 8 at Wisconsin, 11 at the Scioto River, 24 at Lake Erie, and 17 at Saginaw Bay. Using the mean available forage and associated standard error from shallow food samples in program CONTRAST, I found mallards consumed food in proportion to availability at all study sites (P > 0.05) (Table 2.2).

Lesser Scaup Food Selection

I included data from 46 scaup collected in spring 2006 in selection analyses. Of these, 10 were collected at the Illinois River, 20 at Lake Erie, and 16 at Saginaw Bay. Only 2 scaup were collected at Wisconsin and 1 at the Scioto River, therefore I did not include these in analyses. Using the mean available forage and associated standard error from deep food samples in program CONTRAST, I found scaup consumed a significantly higher percentage of invertebrates than were available at Lake Erie (P < 0.001) (Table 2.2). Scaup consumed food in proportion to availability at Illinois River and Saginaw Bay (P > 0.05).

Table 2.2. Results of selection analyses for ducks collected at study sites in the Upper MS River and Great Lakes Region (CA = Cache River, IR = Illinois River, WI = Wisconsin, SR = Scioto River, LE = Lake Erie, and SB = Saginaw Bay) during spring 2006. An “I” indicates selection of invertebrates, “=” indicates consumption in proportion to availability, and “S” indicates selection of seeds.

_______________________________________________________________________

CA IR WI SR LE SB

________________________________________________________________________
Blue-winged teal = = I I =
Mallard = = = = = =
Ring-necked duck S = = = =
Lesser scaup = I =

________________________________________________________________________

Ring-necked Duck Food Selection

I included data from 83 ring-necked ducks collected in spring 2006 in selection analyses. Of these, 13 were collected at the Cache River, 11 at the Illinois River, 7 at the Scioto River, 37 at Lake Erie, and 15 at Saginaw Bay. Only 3 ring-necked ducks were collected at Wisconsin, therefore I did not include these in analyses. Using the mean available forage and associated standard error from deep food samples in program CONTRAST, I found ring-necked ducks consumed a significantly fewer invertebrates than were available at the Cache River (P = 0.007) (Table 2.2). Ring-necked ducks consumed food in proportion to availability at the Illinois River, Scioto River, Lake Erie, and Saginaw Bay sites (P > 0.05).

DISCUSSION

Food Selection

For species I observed consuming higher percentages of a food type than was available to them, I interpreted this as selection. Likewise, I considered ducks to be consuming food in proportion to availability if diet was not significantly different than availability (e.g., P > 0.10). I observed all species except mallards, exhibit selection for either invertebrates or seeds (Table 2.2). The mallard was the only species with large enough sample sizes (e.g., ≥ 5) to evaluate selection at all study sites; therefore, my inferences regarding diet trends for some species in spring 2006 were limited. In the previous chapter, I considered southern sites to be the Cache River and Scioto River, the mid-latitude sites to be the Illinois River and Lake Erie, and the northern sites to be Wisconsin and Saginaw Bay, based on their location within their respective transect (e.g., western or eastern transect). For purposes of detecting a diet trend according to the latitude in which they were collected, I considered transects jointly and hereafter refer to the Cache River as the southern site (37 ^o18’ N), Scioto (39^o 40’ N) and Illinois River (40^o 12’ N) and Lake Erie (41^o 27’ N) as mid-latitude sites, and Saginaw Bay (43^o 45’ N) and Wisconsin (43^o 48’ N) as northern sites. It is important to consider, however, that this is only for conceptual purposes, as both the Saginaw Bay and Wisconsin site would be “southern sites” to ducks breeding in Alaska.

Blue-winged teal.- Previous spring studies of blue-winged teal indicated selection of invertebrates (Swanson et al. 1974, Manley et al. 1992). Another diet study of blue-winged teal also demonstrated a heavy reliance on invertebrates during spring, but availability data was not collected (Taylor 1978). Interestingly, breeding blue-winged teal rely heavily on somatic lipid, whereas somatic protein remains relatively constant through clutch formation, suggesting exogenous resources are used to meet protein demands (Rohwer, unpublished data). Considering this, and the fact that teal initiate nesting shortly after arrival at breeding areas (Toft et al. 1982), I expected lipid acquisition (i.e., consumption of high-lipid seeds) to be relatively important to spring migrating blue-winged teal.

Contrary to my expectation, blue-winged teal appeared to rely more on invertebrates than seeds during spring as teal exhibited selection for invertebrates at both Wisconsin and Lake Erie study sites (Table 2.2). My results are similar to previous spring studies of blue-winged teal, indicating selection of invertebrates (Swanson et al. 1974, Manley et al. 1992).

Mallard. - Mallard diet studies during spring provided mixed results, with one reporting selection for Chironomidae larvae (Pederson and Pederson 1983) and others documenting a heavy reliance on seeds and agricultural grains (Jorde 1981, Heitmeyer 1985, LaGrange 1985). Only one of these studies, however, collected data on food availability to assess diet selection (Pederson and Pederson 1983). Breeding mallards rely heavily on somatic lipids acquired prior to arrival on breeding grounds (Krapu 1981) and initiate nesting shortly after arriving at breeding area (Toft et al. 1982); therefore I expected mallards staging in the UMR/GLR to consume large amounts of seeds. Because digestive physiology of ducks, however, restricts them from making abrupt changes in diet composition (i.e., from seeds to invertebrates), I expected to see a heavier reliance on animal foods as they approached reproduction in late spring (Barlein 2003).

Although I was unable to detect selection of food items, seeds seemed to be most important to spring migrating mallards in the UMR/GLR in 2006, as mallards consumed > 78% seeds at all study sites in 2006 (e.g., a diet would have to be exclusively seeds to indicate any kind of seed selection because invertebrate availability was so low at these sites) (Table 2.3). A recent food selection study conducted at one site during spring in the UMR/GLR found similar results with mallards selecting moist-soil seeds (Smith 2007).

I was unable to detect a diet transition, as I observed mallards consuming food in proportion to availability at all sites (i.e., latitudes) (Table 2.2). I did, however, observe the highest proportions of invertebrates in the diet of mallards collected at Saginaw Bay and Wisconsin (i.e., northern sites) (Table 2.3). Young (1993) suggested that somatic protein is not used by mallards during egg laying; rather, protein was obtained from

Table 2.3. Mean percentage of food items and standard error (SE) in diet of dabbling ducks at study sites in the Upper MS River and Great Lakes Region during spring 2006.

________________________________________________________________________

Food Type Site Mean % SE

________________________________________________________________________
Mallard Seeds Cache River 91 7

Invertebrates Cache River 9 7

Mallard Seeds Illinois River 80 11

Invertebrates Illinois River 20 11

Mallard Seeds Scioto River 90 9

Invertebrates Scioto River 10 9

Mallard Seeds Lake Erie 86 6

Invertebrates Lake Erie 14 6

Mallard Seeds Wisconsin 78 15

Invertebrates Wisconsin 22 15

Mallard Seeds Saginaw Bay 78 7

Invertebrates Saginaw Bay 22 7

Blue-winged teal Seeds Cache River 67 8

Invertebrates Cache River 33 8

Blue-winged teal Seeds Illinois River 72 8

Invertebrates Illinois River 28 8

Blue-winged teal Seeds Lake Erie 46 9

Invertebrates Lake Erie 54 9

Blue-winged teal Seeds Wisconsin 33 10

Invertebrates Wisconsin 67 10

Blue-winged teal Seeds Saginaw Bay 30 12

Invertebrates Saginaw Bay 70 12

________________________________________________________________________

exogenous sources. This, along with the fact I found mallards consuming large amounts of seeds, emphasizes the importance of spring-staging areas for lipid acquisition.

Lesser scaup. - It has been suggested that decreased body condition of breeding scaup, resulting from poor spring-habitat conditions, may be partly responsible for declines in scaup breeding populations (Anteau and Afton 2006, Anteau and Afton 2008). Current spring-staging populations of scaup encounter habitats that do not provide historically preferred foods and may therefore be unable to acquire nutrients that were available previously (Anteau and Afton 2006). I found scaup in the UMR/GLR either consumed more invertebrates than were available or in proportion to availability (Table 2.2). These invertebrates, however, were not the “historically preferred” invertebrates (e.g., amphipods) consumed by scaup during spring, but primarily Gastropoda and Chironomidae larvae (see previous chapter). Despite the selection of invertebrates I observed in staging scaup, their diets consisted of higher proportions of seeds (Table 2.4) than reported in another recent scaup spring-diet study (Anteau and Afton 2008). Other previous studies indicated scaup fed primarily on invertebrates during spring (Rogers and Korschgen 1966, Gammonley and Heitmeyer 1990, Afton et al. 1991, Anteau and Afton 2006, Badzinski and Petrie 2006a, Anteau and Afton 2008); however, 2 recent studies reported scaup spring-diet consisted primarily of moist-soil plant seeds (Smith 2007, Strand et al. 2007), and 1 suggested scaup selected this food type (Smith 2007). I believe the contradictory result of food selection by spring-migrating scaup in recent studies is likely a result of geographic location. For example, the most southern sites of Anteau and Afton (2006) were at an equivalent latitude to my northern sites. Additionally,

Table 2.4. Mean percentage of food items and standard error (SE) in diet of diving ducks at study sites in the Upper MS River and Great Lakes Region during spring 2006.

________________________________________________________________________

Food Type Site Mean % SE

________________________________________________________________________
Lesser Scaup Seeds Illinois River 77.0 13.0

Invertebrates Illinois River 23.0 13.0

Lesser Scaup Seeds Lake Erie 42.0 8.0

Invertebrates Lake Erie 58.0 8.0

Lesser Scaup Seeds Saginaw Bay 26.0 10.0

Invertebrates Saginaw Bay 74.0 10.0

Ring-necked Duck Seeds Cache River 99.5 0.4

Invertebrates Cache River 0.5 0.4

Ring-necked Duck Seeds Illinois River 91.0 9.0

Invertebrates Illinois River 9.0 9.0

Ring-necked Duck Seeds Scioto River 88.0 7.0

Invertebrates Scioto River 12.0 7.0

Ring-necked Duck Seeds Lake Erie 88.0 5.0

Invertebrates Lake Erie 12.0 5.0

Ring-necked Duck Seeds Saginaw Bay 36.0 13.0

Invertebrates Saginaw Bay 64.0 13.0

________________________________________________________________________

Anteau and Afton (2006) examined scaup diets at breeding sites (Anteau and Afton 2006), whereas no scaup bred at any sites in my study (i.e., macroscopic examination of internal reproductive organs did not indicate follicle development in any of the collected scaup).

I did not collect scaup at the southern study site in 2006 and was unable to determine scaup diet at southern latitudes; however, I did detect selection of invertebrates at a mid-latitude site (e.g., Lake Erie). Interestingly, scaup consumed foods in proportion to availability at the Illinois River site, yet selected for invertebrates at Lake Erie, even though invertebrate and seed availability was similar at these sites (Table 2.1). Also

unique to the Illinois River site, I regularly observed large concentrations of scaup foraging in unharvested cornfields soon after inundation. Considering scaup spend 3 − 4 weeks on breeding areas to acquire and maintain fat reserves before initiating RFD, and their diets during this time consist of high-protein foods that are inefficient for building somatic lipid (Afton and Hier 1991), the pattern I observed in scaup diet staging on the Illinois River is what I expected to observe in spring migrating scaup (i.e., consuming high-carbohydrate seeds to lessen their dependence on high-carbohydrate foods at breeding areas).

Ring-necked duck. - To my knowledge, this study was the only food-selection study of spring-migrating ring-necked ducks conducted. Previously, feeding ecology of spring-migrating ring-necked ducks could only be inferred from pre-laying female ring-necked ducks presumably at their breeding area in Minnesota (Hohman 1985). I found ring-necked ducks staging in the UMR/GLR in spring 2006 consumed substantially less animal matter than pre-laying female ring-necked ducks in Minnesota (Hohman 1985) at all sites except Saginaw Bay, and either selected for seeds or consumed foods in proportion to availability.

My data indicated spring-migrating female ring-necked ducks were heavily dependant on seeds but may have transitioned to a diet consisting mostly of invertebrates at northern latitudes (e.g., Saginaw Bay). Because of small sample sizes at the Wisconsin site in 2006, however, I was unable to test these results at an additional northern latitude site. Female ring-necked ducks typically acquire lipid reserves that are used to meet reproductive requirements before reaching nesting areas (Hohman 1986). Therefore, it is not surprising that I found ring-necked ducks selecting seeds at the Cache River site during spring, as these high-carbohydrate foods are likely used to accumulate or restore somatic lipid. Because the proportions of invertebrates and seeds available for ring-necked ducks collected at Saginaw Bay was similar to Scioto River (approximately 50% seeds and 50% invertebrates available; see Table 2.1), yet ring-necked ducks at the Scioto River consumed approximately twice the amount of seeds as ring-necked ducks at Saginaw Bay, I suggest that ring-necked ducks at Saginaw Bay were in a different physiological state. It is possible that ring-necked ducks staging at Saginaw Bay during spring were in close proximity to breeding areas, perhaps explaining why I observed an increase in invertebrate consumption at this site.

CHAPTER 3: IMPLICATIONS FOR WETLAND MANAGEMENT FOR SPRING-MIGRATING WATERFOWL IN THE UMR/GLR
Management Implications

Unlike the diet of ducks during breeding, when all ducks depend heavily on invertebrates or during fall migration and winter when seeds and agricultural grains compose the majority of the diet of ducks, spring diet is much more variable and differs considerably among species. At least some species appear to transition to invertebrates later in spring migration, leading to intraspecific variation of diet. The variability I discovered in spring diets emphasizes the importance of providing a variety of habitats (e.g., food types) during spring. Even though ducks are likely capable of searching for, and selecting specific food types from within their environment (see chapter 2), the variability I discovered in diet among sites and transects indicates that the ability of ducks to modify food intake is limited. Thus, food availability still plays a large role in diet, hence nutrient acquisition.

Management of wetland habitats specifically to benefit spring-migrating waterfowl is uncommon. Current wetland management practices typically intend to maximize seed (i.e., moist-soil and agricultural) abundance for fall-migrating and wintering waterfowl. Although this approach likely benefits fall-migrating and wintering waterfowl, it may not yield quality foraging habitat in spring (Greer et al. 2006). My research indicated that wetlands managed for moist-soil plant species provide important foraging habitats for some spring-migrating waterfowl (e.g., mallards, ring-necked ducks), and attract others that consume invertebrates and seeds (e.g., blue-winged teal, gadwall, lesser scaup). Thus, I recommend managers provide shallow and deep water habitats during spring with abundant moist-soil seeds and invertebrates.

Managing Wetlands for Invertebrates During Spring Migration

Differing water regimes will affect macroinvertebrate taxa available to foraging waterfowl. For example, Neckles et al. (2006) found that semipermanent flooding (standing water present through the growing season) in marshes in Manitoba, Canada reduced total invertebrate densities. The taxa that were negatively impacted by semipermanent flooding are very important to foraging waterfowl (e.g., Cladocera, Ostracoda, and Culicidae). Neckles et al. (2006) suggested that semi-permanent flooding may eliminate cues necessary for oviposition and hatch among dominant taxa. Under seasonally flooded wetlands (standing water present only through mid-summer), however, macroinvertebrate densities were not reduced, regardless of the availability of detritus. A flooding regime that would likely benefit spring-migrating and wintering waterfowl is deep-flooding impoundments that were kept flooded shallowly during winter, as these newly flooded deep wetlands will expand into previously dry habitat as water levels rise, resulting in deep habitat for diving ducks and shallow habitat for dabbling ducks (Fredrickson and Reid 1988). Conversely, wetlands that undergo spring water drawdown, likely concentrate invertebrates as they follow receding water levels and consequently improve foraging conditions for invertebrates. Gray et al. (1999) found moist-soil wetlands that were mowed during winter, rather than disked or tilled, supported diverse invertebrate communities, likely because of detritus that served as substrate for invertebrate production (Kaminski and Prince 1981). Therefore, late-winter flooding of moist-soil wetlands with mowed areas would likely benefit spring-migrating waterfowl by maximizing invertebrate abundance. Invertebrate abundance is also higher on wetlands lacking predators, such as fathead minnows and other fish (Cox et al. 1998, Hornung and Foote 2006); therefore wetlands managed for waterfowl should experience complete annual drawdown and be protected from flood events that can establish such fish populations. Preventing flood events will also likely increase water clarity, hence improving foraging conditions for ducks.

We found invertebrate production was highest in shallow palustrine forested wetlands during spring (Schultheis, dissertation in progress). Although this habitat type (e.g., forested wetlands) is likely inaccessible to diving ducks, it supported a diversity of foods beneficial to spring-migrating dabbling ducks, particularly blue-winged teal that rely heavily on invertebrate foods. Thus, forested wetland habitat should be maintained throughout spring, particularly wetlands predominated with button-bush (Cephalanthus sp.) and willow trees (Salix sp.) because of their high flood tolerance. Water levels in green-tree reservoirs (GTR) (e.g., bottomland hardwood forests intentionally flooded to produce habitat for waterfowl), in particular, should be held as long as possible to provide rich invertebrate sources to late-migrating dabbling ducks. Because the integrity of GTR’s depends on the survival of early successional mast-producing oak trees and these trees are susceptible to disease when inundated during the growing season, late-winter flooding with a pre-growing season drawdown may be beneficial to both the GTR and spring-migrating waterfowl.
Managing Wetlands for Seeds During Spring Migration

To optimize seed abundance for spring-migrating ducks, I suggest late-winter flooding of moist-soil wetlands at varying depths, as this wetland management practice provides abundant seeds (Greer et al. 2006). Seed loss from depredation and decomposition from inundation is minimized by a delayed flooding regime. By delaying flooding of GTR’s, not only is seed availability maximized, but survival of important mast-producing hardwood species is encouraged.

Challenges to Providing Habitat for Spring-Migrating Waterfowl

As I previously stated, I believe the greatest opportunity to manage wetland habitats for spring-migrating waterfowl is on state and federal refuges. These, however, are often managed to accommodate public waterfowl hunters, and delayed flooding of these areas may meet heavy criticism from the hunting constituency. To avoid this, wetlands could be kept at low pool during winter, still providing habitat for hunters, with water coverage allowed to increase (newly flooding the perimeter) throughout spring. Minimally, this strategy should be used on non-huntable wetlands that are true ‘refuges’ during winter.

Another challenge to maximizing forage availability for spring-migrating waterfowl, particularly in the Midwest, is convincing private landowners and area managers to provide wetland habitat other than flooded agricultural fields (i.e., moist-soil wetlands). While these habitats may be heavily utilized by spring-migrating waterfowl (LaGrange and Dinsmore 1989), these food-types are lacking in important amino acids (Buckley 1989). Additionally, water must be removed from these wetlands in early spring to prepare for planting next years crop, rendering these wetlands useless to many mid- and late-migratory species during spring. In contrast, wetlands managed for native, moist-soil plants can remain flooded without negatively impacting conditions for the subsequent year.

LITERATURE CITED

Affik, D., and W. H. Karasov. 1995. The trade-offs between digestion rate and efficiency in warblers and their ecological implications. Ecology 76:2247-2257.

Afton, A. D., and C. D. Ankney. 1991. Nutrient-reserve dynamics of breeding lesser scaup: a test of competing hypotheses. The Condor 93:89-97.
Afton, A. D. and R. H. Hier. 1991. Diets of lesser scaup breeding in Manitoba. Journal of Field Ornithology 62:325-334.
Afton, A. D., R. H. Hier, and S. L. Paulus. 1991. Lesser scaup diets during migration and winter in the Mississippi flyway. Canadian Journal of Zoology 69:328-333.
Alisauskas, R. T., and C. D. Ankney. 1992. The cost of egg laying and its relationship to nutrient reserves in waterfowl. Pages 30-61 in B. D. J. Batt, A. D. Afton, M. G. Anderson, C. D. Ankney, D. H. Johnson, J. A. Kadlec, and G. L. Krapu, editors. Ecology and management of breeding waterfowl.University Minnesota Press, Minneapolis, Minnesota, USA.
Anderson, J. T., L. M. Smith, and D. A. Haukos. 2000. Food selection and feather molt by non-breeding American green-winged teal in Texas playas. The Journal of Wildlife Management 64:222-230.
Ankney, C. D., and C. D. MacInnes. 1978. Nutrient reserves and reproductive

performance of female lesser snow geese. Auk 95:459-471.

Ankney, C. D., and R. T. Alisauskas. 1991. Nutrient-reserve dynamics and diet of

breeding female gadwalls. The Condor 93:799-810.

Anteau, M. J., and A. D. Afton. 2004. Nutrient reserves of lesser scaup (Aythya affinis)

during spring migration in the Mississippi flyway: a test of the spring condition

hypothesis. The Auk 121:917-929.
Anteau, M. J., and A. D. Afton. 2006. Diet shifts of lesser scaup are consistent with the

spring condition hypothesis. Canadian Journal of Zoology 84:779-786.

Anteau, M. J. and A. D. Afton. 2008. Diets of lesser scaup during spring migration

throughout the upper-midwest are consistent with the spring condition hypothesis.

Waterbirds 31:97-106.
Arzel, C., J. Elmberg, and M. Guillemain. 2006. Ecology of spring-migrating Anatidae:

a review. Journal of Ornithology 147:167-184.

Badzinski, S.S., and S. A. Petrie. 2006a. Diets of lesser and greater scaup during autumn

and spring on the lower Great Lakes. Wildlife Society Bulletin 34:664-674.

Badzinski, S.S., and S. A. Petrie. 2006b. Satellite tracking lesser scaup and greater scaup

from the lower Great Lakes. Pages 100–101 in Proceedings of the 4^th North

American Duck Symposium and Workshop. Integrating science and duck

management, 23-26 August 2006, Bismarck, North Dakota, USA.

Barboza, P. S., and D. G. Jorde. 2002. Intermittent fasting during winter and spring

affects body composition and reproduction of a migratory duck. Journal of

Comparative Physiology B 172:419-434.
Barlein, F. 2003. Nutritional strategies in migratory birds. Pages 321-332 in P. Berthold,

E. Gwinner, and E. Sonnenschein, editors. Avian Migration. Springer-Verlag Berlin, Heidelberg, New York.

Barrows, H. H. 1910. Geography of the Middle Illinois Valley. Illinois Geological

Survey Bulletin 15.

Beason, R. C. 1978. The influences of weather and topography on water bird migration in

the southwestern United States. Oecologia 32:153-169.

Bellrose, F. C. 1980. Ducks, geese, and swans of North America. Third edition.

Stackpole Books. Harrisburg, Pennsylvania, USA.

Bellrose, F. C., S. P. Havera, F. L. Paveglio, Jr., and D. W. Steffeck. 1983. The fate of

lakes in the Illinois River Valley. Illinois Natural History Survey Biological Notes

119.
Buckley, C. E. 1989. The nutritional quality of selected row crop and moist-soil seeds for

Canada geese. Thesis, University of Missouri, Columbia, Missouri, USA.

Cook, R. C., D. L. Murray, J. G. Cook, P. Zager, and S. L. Monfort. 2001. Nutritional

influences on breeding dynamics in elk. Canadian Journal of Zoology 79:845-

853.
Cowardin, L. M., V. Carter, F. C. Golet, and E. T. LaRoe. 1979. Classification of

wetlands and deepwater habitats in the United States. U. S. Dept. Interior, Fish &

Wildlife Service, FWS/OBS-79/31.
Cox, R. R., M. A. Hanson, C. C. Roy, N. H. Euliss, D. H. Johnson, and M. G. Butler.

1998. Mallard duckling growth and survival in relation to aquatic invertebrates.

Journal of Wildlife Management 62:124-133.
Dahl, T. E. 1990. Wetland losses in the United States 1780s to 1980s. U. S. Dept. of

Interior, Fish and Wildlife Service, Washington, D.C.

Delnicki, D., and K. J. Reinecke. 1986. Mid-winter food use and body weights of

mallards and wood ducks in Mississippi. Journal of Wildlife Management 50:43-51.
Dirschl, H. J. 1969. Foods of lesser scaup and blue-winged teal in the Saskatchewan

River Delta. Journal of Wildlife Management 33:77-87.

Drobney, R. D. 1980. Reproductive bioenergetics of wood ducks. Auk 97:480-490.
Dubovsky, J. A., and R. M. Kaminski. 1994. Potential reproductive consequences of

winter-diet restrictions in mallards. Journal of Wildlife Management 58:780-786.

Ducks Unlimited. 2005. Development of a Potential Wetland Restoration Layer for Research and Planning in the Great Lakes. GLARO/_documents/library/_gis/GLPotentialWetland.pdf> Accessed 1 May 2005.

Elmberg, J., K. Sjoberg, H. Poysa, and P. Nummi. 2000. Abundance-distribution relationships on interacting trophic levels: the case of lake-nesting waterfowl and dytiscid water beetles. Journal of Biogeography 27:821-827.
Farney, R. A. 1975. Fall foods of ducks in Lake Erie marshes during high water years. Thesis, The Ohio State University, Columbus, Ohio, USA.
Fredrickson, L. H. and F. A. Reed. 1988. Invertebrate response to wetland management.

Fish and Wildlife Leaflet 13.3.1, Waterfowl Management Handbook. U. S. Fish

and Wildlife Service, Washington, D.C., USA.
Gammonley, J. H., and M. E. Heitmeyer. 1990. Behavior, body condition, and foods of

buffleheads and lesser scaups during spring migration through the Klamath Basin,

California. Wilson Bulletin 102:672-683.
Gorman, K. B., D. Esler, P. L. Flint, and T. D. Williams. 2008. Nutrient-Reserve dynamics during egg production by female greater scaup (Aythya marila): Relationships with timing of reproduction. The Auk 125:384-394.
Gray, M. J., R. M. Kaminski, G. Weerakkody, B. D. Leopold, and K. C. Jensen. 1999. Aquatic invertebrate and plant responses following mechanical manipulations of moist-soil habitat. Wildlife Society Bulletin 27:770-779.
Green, A. J. 1998. Comparative feeding behaviour and niche organization in a Mediterranean duck community. Canadian Journal of Zoology 76:500-507.
Greer, A. K., B. D. Dugger, D. A. Graber, and M. J. Petrie. 2006. The effects of seasonal

flooding on seed availability for spring migrating waterfowl. Journal of Wildlife Management 71:1561-1566.

Gruenhagen, N. M. 1987. Feeding ecology, behavior and carcass dynamics of migratory

female mallards. Thesis, University of Missouri, Columbia, Missouri, USA.

Gruenhagen, N. M., and L. H. Fredrickson. 1990. Food use by migratory female mallards in northwest Missouri. Journal of Wildlife Management 54:622-626.
Harris, M. P. 1969. Effect of laying date on chick production in oystercatchers and herring gulls. Britain Birds 62:70-75.
Havens, J. H. 2007. Winter abundance of waterfowl, waterbirds, and waste rice in

managed Arkansas rice fields. Thesis, Mississippi State University, Mississippi State, Mississippi, USA.

Havera, S. P. 1999. Waterfowl of Illinois: status and management. Illinois Natural History Survey Special Publication 21.
Heitmeyer, M. E. 1985. Wintering strategies of female mallards related to dynamics of lowland hardwood wetlands in the upper Mississippi delta. Dissertation, University of Missouri, Columbia, Missouri, USA.
Hines, J. E. and J. R. Sauer. 1989. Program CONTRAST – A general program for the analysis of several survival or recovery rate estimates. US Fish and Wildlife Service, Fish and Wildlife Technical Report 24, Washington, D.C., USA.
Hohman, W. L. 1985. Feeding ecology of ring-necked ducks in northwestern Minnesota. Journal of Wildlife Management 49:546-557.
Hohman, W. L. 1986. Changes in body weight and body composition of breeding ring-

necked ducks. Auk 103:181-188.

Hoppe, R. T., L. M. Smith, and D. B. Webster. 1986. Foods of wintering diving ducks in

South Carolina. Journal of Field Ornithology 57:126-134.

Hornung, J. P. and A. L. Foote. 2006. Aquatic invertebrate responses to fish presence and vegetation complexity in western boreal wetlands, with implications for waterbird productivity. Wetlands 26:1-12.
Howe, M.A., P. H. Geissler, and B. A. Harrington. 1989. Population trends of North

American shorebirds based on the International Shorebird Survey. Biological

Conservation 49:185-199.
Illinois State Water Survey [ISWS]. 2008. ISWS home page.