Northern Prairie Wildlife Research Center
Many factors influence the composition of waterfowl diets. Coevolved patterns of habitat use and morphology set broad limits on potential food resources. Within these limits, spatial and temporal variation of food availability in dynamic wetland environments often favors opportunistic feeding behavior. Other factors affecting food choice include nutrient requirements, foraging efficiency, and competition within and among species.
Coevolved patterns of habitat use and morphology set general limits on the potential food resources available to each waterfowl species. Interspecific differences in habitat use and morphology are most evident in the partitioning of food resources among tribes (Fig. 1-2). Large body sizes and unusually long necks enable swans to exploit benthic food resources in the semipermanently flooded zone at depths that are beyond the reach of nondiving waterfowl. Swans use their powerful mandibles to pull loose aquatic plants and dig out roots and tubers not available to diving ducks. Geese feed primarily on herbaceous plants in the uplands, and in temporarily and seasonally flooded sites (Fig. 1-2). Species that feed on underground parts of plants (e.g., Snow and Graylag Geese) have robust mandibles suited for digging tubers and other foods from moist soils. At the other extreme are species with diminutive bills (e.g., Ross' and Cackling Canada Geese), which are adapted for pecking the short blades of closely cropped grasses.
 |
| Figure 1-2. Spatial segregation of the principal foraging habitats of waterfowl tribes with respect to water depth and wetland plant communities. |
Dabbling ducks (Tribe Anatini) forage in temporarily and seasonally flooded wetlands (Fig. 1-2) where the bottom is reached by dabbling or tipping. Breeding Wood Ducks (Tribe Cairinini) also forage in shallow wetlands, including flooded timber (Drobney and Fredrickson, 1979) and small "coarse sedge" marshes (Gilmer, 1971). The pochards (Tribe Aythyini) forage principally by diving in open waters of the semipermanently flooded zone (Fig. 1-2). Their short heavy bodies, legs positioned posteriorly, and large lobed feet are adapted for swimming underwater and feeding on a variety of benthic macroinvertebrates, zooplankton, and aquatic plant parts.
Sea ducks (Tribe Mergini) usually are found in the deepest and most permanent aquatic habitats used by waterfowl (Fig. 1-2). Sea ducks use their wings for propulsion under water, and some species (e.g., the mergansers, scoters, and eiders) have mandibles specialized for grasping, holding, or crushing large prey. The Ruddy Duck, the only species of stiff-tailed duck (Tribe Oxyurini) breeding in the United States and Canada, generally forages in semipermanently flooded wetlands (Fig. 1-2) by using mandibles with closely spaced lamellae to filter midge larvae and other macroinvertebrates from wetland sediments.
Within the habitats exploited by each species of waterfowl, food availability is subject to considerable short-term spatial and temporal variation. Availability of plant foods to prebreeding and breeding geese generally is more predictable than availability of macroinvertebrates to breeding ducks. Nevertheless, local foraging movements and seasonal migration patterns must track plant growth and phenology closely for geese to maximize rates of nutrient intake from foliage. In the Netherlands, Barnacle Geese stopped foraging on a dairy pasture and began using a salt marsh when availability of protein in Festuca rubra on the marsh equaled that of the pasture grasses (Prins and Ydenberg, 1985). During the spring staging period, Barnacle Geese and Brant revisit grazing sites in salt marshes in the Wadden Sea with a frequency that maintains the growth rates, succulence, and protein content of their plant foods (Prins et al., 1980; Ydenberg and Prins, 1981; Prop, 1991). Giant Canada Geese in Minnesota and Pink-footed Geese in Iceland begin feeding on grasses as soon as sufficient snow has melted to expose shoots that have initiated growth under the insulating cover of the snow (McLandress and Raveling, 1981a; Fox et al., 1991).
Food habits of ducks vary annually, seasonally, and diurnally in response to changes in food availability. In the Prairie Pothole Region, foods available to dabbling ducks are influenced by annual water conditions. In wet years, fairy shrimp, clam shrimp (Order Conchostraca), and earthworms (Class Oligochaeta) dominate the diets of laying female Mallards that feed in shallow, seasonally flooded habitats (Swanson et al., 1985). During drought years, these habitats are not available, and females that attempt to breed forage in semipermanent wetlands where benthic insects are important foods (Swanson et al., 1985).
Seasonal changes in food habits of ducks often are related to the population dynamics of their macroinvertebrate prey (Fig. 1-3). In North Dakota, fairy shrimp are a major food of laying female Northern Pintails in early spring (Krapu, 1974b) but are no longer available later in summer, so renesting females feed on chironomids and gastropods. Female Mallards using similar habitats eat earthworms, fairy shrimp, and terrestrial insects in April; water fleas, fairy shrimp, snails, and aquatic insects in May; and aquatic insects and water fleas in June (Swanson et al., 1985).
 |
| Figure 1-3. Temporal changes in the abundance of macroinvertebrate populations inhabiting a seasonal wetland in North Dakota (after Swanson et al., 1974). |
Diurnal changes in food use by ducks are influenced by macroinvertebrate behavior and by the effects of weather on food availability. In North Dakota, 93% of the diet of dabbling ducks feeding on a sewage lagoon during the day was immature insects and water fleas (Swanson, 1977). In contrast, 89% of the diet between sunset and midnight was adult insects, mostly chironomids. Emergence of chironomids in temperate latitudes occurs principally after sunset, whereas in the arctic chironomids generally emerge and are exploited by ducks at midday (Sjoberg and Danell, 1982).
Dabbling ducks are adapted to exploit the patchy and unpredictable food resources of temporary and seasonal wetlands and often respond opportunistically to weather-related changes in food availability. Early spring rains that saturate soils in the Prairie Pothole Region force earthworms to the surface where they are eaten by Northern Pintails (Krapu, 1974a) and Mallards (Swanson et al., 1985). Wind often concentrates macroinvertebrates and seeds along the shore of wetlands where they are consumed by breeding waterfowl (Swanson et al., 1974; Drobney and Fredrickson, 1979). Conversely, rain and cold temperatures can halt a midge emergence (Sjoberg and Danell, 1982) and reduce the availability of chironomids. In some cases, the impact of a single weather event, such as a torrential rain, can increase water levels markedly and change biotic communities and available food resources for several years. The dynamic nature of waterfowl food habits during the prebreeding and breeding periods emphasizes the importance of designing long-term feeding ecology studies to encompass the range of variation of food resources that a species is likely to experience.
Food selection by waterfowl is influenced by the need to satisfy nutrient requirements for reproduction. This hypothesis is supported by observed patterns of food selection by geese depositing nutrient reserves for reproduction (McLandress and Raveling, 1981a), by increases in the percentage of animal foods used by female dabbling ducks and pochards during egg formation (Table 1-5), and by experimental studies showing that reproductive performance is influenced by nutrient content of the diet (see section VII). In this section, we discuss the importance of selected nutrients to food choice by prebreeding and breeding waterfowl.
1. Protein
Protein is often considered the most limiting nutrient to reproduction among geese (Raveling, 1979b; Davies and Cooke, 1983) and certain ducks (Krapu, 1981; Drobney and Fredrickson, 1985; for an alternative view see Ankney and Afton, 1988; Afton and Ankney, 1991) because: (1) waterfowl are unable to store protein in a concentrated and labile form; (2) protein requirements increase during egg formation; and (3) availability of protein-rich foods fluctuates markedly during critical periods. Protein quality is also important. The biological value of protein is high only if all essential amino acids are contained in a ratio corresponding to the protein being formed (Scott et al., 1969, p. 57). The percentage of protein required in the diet varies with the amino acid composition of foods, with the ratio of protein to usable energy in the diet, and with the quantity of food consumed (Parrish and Martin, 1977). Generally, the protein requirement of waterfowl breeding in captivity is about 18% (Foster, 1976).
Geese obtain protein for reproduction primarily from the foliage of grasses, sedges, and forbs, but the roots of wetland plants also may be important (Owen, 1980; Prevett et al., 1985). Prebreeding Lesser Snow, Whitefronted, and Canada Geese acquire protein on staging areas in temperate regions from shoots of winter wheat, bluegrass, and sedges (Carex spp.) (Davies and Cooke, 1983; U.S. Fish and Wildlife Service, 1981; McLandress and Raveling, 1981a; Reed et al., 1977). On subarctic staging areas, Lesser Snow and Canada Geese acquire protein from the foliage and roots of grasses and sedges (Prevett et al., 1985). Dusky Canada Geese delay nesting until after plant growth has resumed on the Copper River Delta in southeastern Alaska, and obtain two-thirds of their protein requirements for egg production from sedges available on the breeding grounds (Bromley, 1984).
The protein concentration in shoots of winter wheat, bluegrass, and sedges is relatively high (Fig. 1-4), and protein intake often seems to be the primary factor affecting food choice by geese (Gauthier and Bédard, 1990). Pecking rates of Lesser Snow Geese feeding on an experimental pasture containing foliage of varying protein levels correlated positively with the crude protein content of plants at the feeding sites (Harwood, 1975; cited in Harwood, 1977). Similarly, White-fronted Geese foraged during winter on fertilized plots of grassland first and more intensively than on unfertilized control plots (Owen, 1975b). However, Buchsbaum et al. (1984) suggested that the protein content of vegetation may not be sufficient to explain food choice by geese when protein intake from foliage is constrained by high levels of secondary metabolites, which interfere with protein digestion. This may explain why female Canada Geese rejected certain plant species during egg laying when protein requirements were greatest (Murphy, 1988).
Food selection by geese also appears to be influenced by the digestibility of energy and organic matter in plants. Studies in Europe indicate that Brant select the most digestible plants and plant parts first and then shift to other feeding sites when more digestible forage becomes available (Boudewijn, 1984). Brant and Canada Geese digest only about 37% of the organic matter consumed (Buchsbaum et al., 1986). Rapid turnover rates of organic matter in the digestive tract enable geese to absorb 60-80% of the protein in grasses (Buchsbaum et al., 1986), but limit the time available for carbohydrate digestion (Parra, 1978). Determining whether protein intake is the primary factor affecting food selection by grazing geese is difficult because of correlations among the water, fiber, and available protein and energy content of foliage (Owen, 1976; Sedinger and Raveling, 1984; Buchsbaum and Valiela, 1987).
Most ducks avoid the potential constraints of food-processing time by eating protein-rich macroinvertebrates (Fig. 1-4) that can be digested rapidly (Swanson and Bartonek, 1970). The efficiency of converting animal protein and lipids to egg mass, and the quality of protein available in macroinvertebrates, probably explain why the proportion of macroinvertebrates in the diet of female ducks is particularly high during the egg-laying stage (Table 1-5).
Despite the quality and digestibility of protein in macroinvertebrates, female ducks must often feed intensively to obtain sufficient nutrients for egg production (Table 1-3). Drobney and Fredrickson (1985) estimated that female Wood Ducks converting dietary protein to egg protein with an efficiency of 100% would have to capture an invertebrate every 5.5 seconds during an 8-hour foraging day to obtain adequate protein for egg production. Maintaining intake rates of macroinvertebrates sufficient to satisfy nutrient requirements for egg production becomes difficult when suitable foods are scarce, fat reserves have been depleted, or frequent disruptions limit feeding time.
2. Fat and Energy
Fat is a major component of waterfowl eggs, and fat reserves are an important energy source for female waterfowl during the breeding cycle (see chapter 2 of this volume). Triglycerides, which comprise most of the lipids retained in adipose tissue, can be synthesized from many nutrient substrates. Linoleic acid apparently is the only fatty acid that cannot be synthesized by birds and must be obtained from dietary sources (Scott et al., 1969). Corn and other agricultural grains contain large amounts of digestible energy (Table 1-6). Lesser Snow Geese, White-fronted Geese, and some populations of Canada Geese apparently synthesize fat needed for migration and reproduction from corn and other waste grains available on wintering grounds (Giant Canada Geese) (McLandress and Raveling, 1981a) or migration stopovers in temperate regions (Atlantic Canada Geese, White-fronted Geese, and Lesser Snow Geese) (Reed et al., 1977; U.S. Fish and Wildlife Service, 1981; Alisauskas, 1988).
| Table 1-6. Digestibility of selected foods of prebreeding and breeding geese |
Geese that eat foliage during spring migration spend about 80% of the daylight hours feeding (Madsen, 1985a; Raveling, 1979a; Boudewijn, 1984), a much greater foraging effort than that required when feeding on barley grains (Madsen, 1985a). Cackling Canada Geese are able to meet their requirements from shoots of grasses and sedges because they stage at northern sites where day length is of sufficient duration to provide the required feeding time (Raveling, 1979a). Marked differences in foraging effort between populations of geese using agricultural and natural food sources apparently result from differences in the digestibility of their diets (Table 1-6). The effect of diet quality on rate of fat deposition was demonstrated by recent field studies of Greater Snow Geese staging in the St. Lawrence estuary. Geese foraging in bulrush marshes, where the digestibility of Scirpus americanus rhizomes was 50%, deposited greater fat reserves than segments of the same population using cordgrass marshes, where the digestibility of Spartina rhizomes was 25% (Gauthier et al., 1984a, 1988). Barnacle Geese feeding in agricultural areas during spring spent less time feeding each day but tended to have greater recruitment rates than geese feeding in natural sites (Black et al., 1991).
Ducks synthesize or obtain fat from a variety of food sources. Ruddy Ducks and certain sea, diving, and dabbling ducks acquire fat by feeding on the macroinvertebrates that dominate their diets throughout the year. Agricultural grains are a good source of carbohydrates (see NFE in Fig. 1-4 and Table 1-6), and probably provide most of the fat reserves of Mallards and Pintails (Jorde, 1981; Miller, 1987). Starchy aquatic plant parts such as sago pondweed tubers (Fig. 1-4) serve a similar function for Canvasbacks (J. Barzen and C. Korschgen, unpublished data), as do seeds of silver maple (Acer saccharinum), water shield (Brasenia schreberi), and elm (Ulmus spp.) for Wood Ducks (Drobney and Fredrickson, 1979).
3. Minerals
The highest calcium requirements of waterfowl (e.g., 1.5% of the diet of Mallards breeding in captivity [Foster, 1976]) are associated with egg production. With the exception of mollusks, crustaceans, horsetails (Equisetaceae), and certain algae, the calcium content of most waterfowl foods is relatively low (Table 1-7). Swans and geese apparently obtain calcium primarily from underground plant parts and/or associated soil minerals. Giant Canada Geese are thought to obtain minerals for eggshell formation from roots consumed before leaving their wintering grounds (McLandress and Raveling, 1981a). Horsetail stems probably are an important source of calcium for Interior Canada Geese at James Bay (Prevett et al., 1985; Table 1-7). Pink-footed Geese breeding in Iceland have been observed feeding from the bottom of unoccupied nest mounds and may be eating fragments of eggshell to obtain calcium (Gardarsson, 1976, cited in Inglis, 1977, p. 755).
| Table 1-7. Calcium and phosphorus concentrations (% of dry wt.) of selected foods consumed by breeding waterfowl |
Ducks obtain calcium for egg production by eating macroinvertebrates (Fig. 1-1, Table 1-7). The importance of gastropods and crustaceans in the diets of breeding ducks probably is related to the need for calcium; both foods are rich sources of this mineral (Table 1-7). Egg-laying female ducks often eat empty snail shells (Krapu and Swanson, 1975), whereas postlaying females sometimes remove the shell and ingest only the animal (Hohman, 1985). In environments where availability of calcium is limited, waterfowl seek atypical sources. In parts of the former USSR, for example, nesting King Eiders occasionally consume lemming and fish bones (Uspenski, 1972).
The phosphorus requirement for egg production in waterfowl is approximately 25% of the requirement for calcium (0.4% of the diet [Foster, 1976]). Natural foods of waterfowl seem to provide an adequate ratio of these minerals (Table 1-7), and there is little information to suggest deficiencies of phosphorus among free-ranging waterfowl. Similarly, evidence of sodium, potassium, and magnesium deficiencies in free-ranging birds is rare (Robbins, 1983).
In contrast, elevated concentrations of trace elements may pose a significant problem to some waterfowl. High selenium concentrations, for example, have caused embryo mortality and developmental abnormalities in waterfowl and other aquatic birds at Kesterson National Wildlife Refuge in California, where this element has become concentrated in irrigation drain water (Ohlendorf et al., 1986).
Until recently, most studies of food selection by waterfowl involved comparing food use with food availability (section III.A). Selection or avoidance of particular foods was attributed to intrinsic factors such as the nutrient requirements of breeding females, or to extrinsic factors such as the behavior of macroinvertebrate prey (Serie and Swanson, 1976; Hohman, 1985; Noyes and Jarvis, 1985). A disadvantage of this method of analyzing food selection is its dependence on a posteriori interpretations. Differences between food use and availability can always be attributed to the failure of food availability samples to reflect the morphological constraints of waterfowl or the behavior of macroinvertebrate prey. This increases the likelihood of accepting the hypothesis that food use is proportional to availability (i.e., opportunistic) and discourages consideration of alternative explanations.
Mathematical models of optimal food choice provide an alternative to traditional analyses of food selection. Models of food choice were among the first of many models of foraging behavior developed by theoretical ecologists. All foraging models are based on the premise that foraging efficiency (i.e., rate of energy or nutrient intake) is positively related to fitness, or equivalently, that efficient foragers are less likely to starve or to be killed by predators, and are more likely to produce offspring.
Among the conceptual advantages of formal models of feeding behavior are: (1) assumptions are stated explicitly and can be tested; (2) prior predictions are made regarding foraging behavior; and (3) the role of natural selection is clear--selection favors traits that increase rates of energy or nutrient intake. Models of foraging behavior generally include three components (Stephens and Krebs, 1986, pp. 5- 11): (1) decision assumptions or variables describe the choice being analyzed (e.g., whether or not to eat a particular food type); (2) currency assumptions include the units (e.g., energy) and criteria (e.g., maximizing average intake rate) used to compare alternative food choices; and (3) constraint assumptions describe limitations on the morphology and behavior of the predator and prey. For simplicity, we limited our review to the classic model of diet choice, wherein the constraints are that searching for and handling food items are exclusive activities; encounters with food items are random and sequential; and foraging animals know the encounter rates and profitabilities (i.e., energy or nutrient values relative to handling times) of each food type. One interesting prediction of the food choice model is that inclusion of a food in the optimal diet depends on the abundance of foods with greater profitabilities, but is independent of its own abundance.
Testing predictions from models of food choice requires that potential prey be ranked by profitability. Unless the diet is relatively simple, considerable effort may be needed to estimate the nutrient value and handling time of each food. Models of food choice also require estimates of abundance or encounter rates for each food and each foraging habitat or "patch." Because of difficulties associated with defining homogeneous foraging patches and estimating the abundance of potential foods, esophageal samples collected in field studies rarely are useful for testing models of food choice (Stephens and Krebs, 1986, p. 24).
The role of profitability in determining food selection by waterfowl is gaining acceptance (Owen and Black, 1990). In British Columbia, Wood and Hand (1985) showed that the preference of Common Mergansers for stocked Coho Salmon (Oncorhynchus kisutch) smolt over fry was independent of fry densities. Behavior of the mergansers was consistent with predictions of a diet model; food choice was related to the relative profitabilities of smolt and fry, and to the abundance of smolt, but not to the abundance of fry. In other studies, Clark et al. (1986) reported that preferences of Mallards for cereal grains were more consistent with a ranking of foods based on profitability than with a ranking based on relative energy content, and Draulans (1982) found that selection of freshwater mussels (Dreissena polymorpha) by Tufted Ducks agreed with qualitative but not quantitative predictions regarding relative profitability.
Testing model assumptions is as important as testing model predictions. Waterfowl exploit a diversity of food sources, and it is unlikely that constraints of the classic model of diet choice are consistent with the feeding behavior of all species. Sequential encounters with prey and exclusive search and handling times are typical of the feeding behavior of Common Mergansers (Wood and Hand, 1985), but are not typical of most other species of ducks or geese. Among filter-feeding ducks especially, searching and handling seem to occur concurrently whether food items are pecked individually (Kooloos, 1986) or filtered from water (Zweers et al., 1977).
Violating assumptions of the food choice model may alter model predictions. This can be illustrated with the principle of lost opportunity, a central concept in foraging theory. When large fish are abundant, the best foraging strategy for mergansers is to ignore small fish, because the opportunity to capture large fish is lost during the time required to catch and eat small fish. In contrast, little opportunity is lost by a filter-feeding duck that eats a low-quality food item if searching continues while other foods are being handled. There is some evidence that this occurs. Captive Mallards ate seeds of three sizes proportional to their availability, even when abundance of the largest, and presumably most profitable, was increased substantially (Strong, 1986). Similarly, Australian Pink-eared Ducks filtered more, rather than less, small plankton from experimental food samples when densities of the largest particles were increased (Crome, 1985). Handling and searching time may not be completely independent in filter-feeding ducks, however, because Tome (1989a) observed a decreased searching rate, and presumably an increased handling time, when Ruddy Ducks foraged in patches with greater food densities.
Theoretical models also have been developed to predict how efficient foragers should choose habitat patches and determine the length of time to feed in each patch. Predictions from these models are less sensitive to the diversity of feeding methods among waterfowl than are models of food choice. However, models of patch choice require that the cumulative energy gained while foraging in a particular patch is a negatively accelerated function of time (Stephens and Krebs, 1986, p. 28). Only one study of waterfowl has tested this assumption. Captive Ruddy Ducks trained regarding the location of experimental food patches exhibited a negatively accelerated energy gain function, and their patch residence times were consistent with predictions from a model of patch use (Tome, 1988, 1989b). In a field experiment that involved manipulation of fish densities (Wood, 1985), Common Mergansers allocated the greatest amount of foraging effort to the most profitable food patches, but several days of learning were necessary for the mergansers to recognize the location of profitable patches. In these studies, foraging Ruddy Ducks and Common Mergansers seemed to be using "area-restricted search," a behavioral strategy that increases feeding efficiency when exploiting prey with clumped distributions (Wood, 1985; Tome, 1989a).
Interest in theoretical foraging models and in the importance of foraging efficiency to waterfowl feeding behavior undoubtedly will increase in the future. At least in the case of models of food choice, however, the diversity of waterfowl feeding methods will require careful testing of: model assumptions, the robustness of model predictions to departures from assumptions, and alternative model constraints.
Interference and exploitation competition can influence where and when waterfowl feed, and what foods they eat. Interference competition occurs when one competitor uses aggressive behavior to prevent access by another competitor to a limited food supply. In exploitation competition, behavioral interactions are absent, but depletion of limited foods by one competitor negatively affects the welfare of another.
1. Intraspecific Competition
Intraspecific competition for food is likely to occur among geese during the prebreeding period, and probably affects species that depend on natural foods more than species that forage in croplands. In the Netherlands, increased agonistic interactions in flocks of Brant feeding on fertilized plots in salt marshes indicated that access to high-quality food was limited (Teunissen et al., 1985). Moreover, female Brant mated to dominant mates walked more slowly, spent more time feeding, and were more frequently accompanied by young during the subsequent autumn than were other females. In North America, Greater Snow Geese are apparently competing for food on their spring staging area in the St. Lawrence estuary. Recent studies have shown that bulrush marshes receiving heavy use by Snow Geese during autumn are avoided during spring (Giroux and Bédard, 1988); exploitation of rhizomes by Snow Geese has decreased the productivity of bulrush stands (Giroux and Bédard, 1987a); and segments of the Snow Goose population that have expanded into cordgrass marshes deposit less fat during the spring staging period (Gauthier et al., 1984a). Concern also is increasing that food limitations on staging or breeding areas are causing density-dependent decreases in recruitment rates and body weights of Lesser Snow Geese (Cooch et al., 1989, 1991). Most geese defend territories during the breeding season, but competition for food is not thought to be an important function of territoriality (Owen and Wells, 1979).
During spring migration, access to food resources in most duck species is determined by dominance hierarchies wherein pairs dominate unpaired birds, and unpaired males dominate unpaired females (e.g., Paulus, 1983). Females are able to avoid interactions with conspecifics and spend more time feeding when their mates are present (Ashcroft, 1976). On the breeding grounds, intraspecific territoriality is believed to be important for providing access to food and undisturbed feeding time in Blue-winged Teal (Stewart and Titman, 1980), Northern Shovelers (Afton, 1979), and Goldeneyes (Savard, 1984, 1988). In nonterritorial duck species, males guard moving areas around their mates rather than fixed sites. The primary function of mate guarding by nonterritorial duck species probably is protection of paternity and parental investment, but competition for food may be a secondary function if exclusion of conspecifics from profitable food patches prevents depletion of prey or depression of prey activity (e.g., Pöysä, 1985).
2. Interspecific Competition
Although past competition for food undoubtedly has contributed to interspecific differences in habitat use and morphology among geese (Owen, 1980; Owen and Black, 1990), there is little evidence of active interspecific competition during the prebreeding or breeding periods (e.g., Fox et al., 1992). Interest in competition is likely to increase, however, because geese often have a greater impact on their foraging habitats than other waterfowl species (Jefferies et al., 1979; Cargill and Jefferies, 1984; Giroux and Bédard, 1987a, b; Kerbes et al., 1990; Ruess et al., 1989; Hik and Jefferies, 1990; Patterson, 1991), and many North American goose populations are increasing (U.S. Fish and Wildlife Service and Canadian Wildlife Service, 1986). The coastal marshes of Hudson Bay are a likely site for future studies given the large numbers of Canada and Snow Geese that use this area for staging and breeding (Thomas and Prevett, 1982a).
Studies of interspecific competition in breeding ducks have investigated interference competition between Goldeneyes and Buffleheads, exploitation competition between ducks and fish, and mechanisms for partitioning food resources among closely related species of dabbling and diving ducks. Interspecific territoriality is rare in birds, but it occurs in all species of the genus Bucephala (Savard, 1984). Studies of the Barrow's Goldeneye suggest that the function of interspecific territoriality in these species is to reduce competition for food. The frequency with which male Barrow's Goldeneyes defend their territories against intruders is related to the degree of potential dietary overlap with the intruding species (Savard and Smith, 1987; Savard, 1988).
Exploitation competition between fish and breeding ducks has been documented in North America and Europe. Breeding pair densities of Mallards and Common Goldeneyes in Sweden, and of Common Goldeneyes in Ontario, were inversely related to densities of fish, especially Perch (Perca spp.) (Eriksson, 1979, 1983; Eadie and Keast, 1984; Pehrsson, 1984). The average size and abundance of macroinvertebrates important as food to breeding ducks were reduced in Swedish lakes with fish (Pehrsson, 1984), and Long-tailed Ducks feeding in fishless lakes on the Arctic Coastal Plain of Alaska obtained prey of larger average size than ducks collected from lakes with fish (Taylor, 1986). Effects of competition with fish on breeding densities of adult ducks are consistent with results from studies of ducklings. Competition with fish decreased the feeding efficiency and growth rates of Black Duck and Mallard ducklings (Hunter et al., 1986; Hill et al., 1987; Pehrsson, 1984). Taken together, these studies suggest that managers should be cautious about introducing fish to important duck-breeding habitats.
Many studies have identified interspecific differences in morphology, habitat use, and foraging methods that facilitate the partitioning of food resources among ducks during the breeding season. Nudds and Bowlby (1984) showed that variation in bill morphology among seven species of dabbling ducks could be explained by selection for resource partitioning along a prey size gradient. Siegfried (1976) reported spatial differences in use of foraging habitats among diving ducks; Ruddy Ducks and Redheads generally foraged in open water near shore, whereas Canvasbacks and Lesser Scaup fed primarily in the central parts of wetlands. Pöysä (1987) observed two major foraging strategies among dabbling ducks in Finland; Northern Shovelers, Garganeys, and Green-winged Teal fed at the water surface and used a variety of feeding sites, whereas Mallards, Northern Pintails, and Eurasian Wigeon used fewer feeding sites but a greater variety of feeding methods and water depths.
Two recent studies have tried to detect active competition for food in multispecies communities of breeding ducks. In Finland, differences in morphology, habitat use, and feeding methods were thought to be sufficient to account for the partitioning of food resources in a community of two dabbling and six diving duck species (Pöysä, 1986). In the United States, patterns of niche complementarity suggested that competition among seven dabbling duck species was greater in California during winter than in North Dakota during the breeding season (DuBowy, 1988). These studies provided little evidence that food was limiting or that interspecific competition occurred among ducks during the breeding season, and tended to support the hypothesis that under natural conditions competition for food among waterfowl species was more intense during winter (Owen and Black, 1990).
U.S. Department of the Interior |
U.S. Geological Survey
URL: http://www.npwrc.usgs.gov/resource/birds/ecomanag/foraging/factor.htm
Page Contact Information: Webmaster
Page Last Modified: Friday, 01-Feb-2013 19:18:18 EST
Menlo Park, CA [caww54]
