USGS Home
Contact USGS
Search USGS

Foraging Ecology and Nutrition

IX. Research Needs and Opportunities

Recent reviews of experimental design in ornithology (James and McCulloch, 1985) and community ecology (Diamond, 1986) have emphasized the value of integrating observational studies with laboratory, field, and natural experiments. Observational studies and natural experiments have greater realism and generality, whereas laboratory and field experiments provide better controls and stronger evidence of causation.

A. Observational Studies

About half the waterfowl species in North America have been the subject of recent studies of feeding ecology during the prebreeding or breeding period. Further research on the remaining species is needed to guide management and to provide more complete taxonomic coverage for comparative studies.

Data for the two species of whistling ducks (Tribe Dendrocygnini) would be of particular interest because of their relatively large clutch sizes and presumed herbivorous food habits (Bolen and Rylander, 1983). Data also are needed for both species of swans and several species or populations of geese. Although recent work on prebreeding geese, especially in the midcontinent and Hudson Bay areas, has improved our understanding of relationships between feeding ecology and nutrient storage (sections IV, V), the foods and staging areas important to geese in other areas of North America have received little study.

Except for Green-winged Teal and Greater Scaup, considerable data are available on the feeding ecology of perching, dabbling, diving, and stiff-tailed ducks in North America, at least during the breeding season. By comparison, little is known regarding the feeding ecology of sea ducks; only the White-winged Scoter, Longtailed Duck, and Common Eider have been studied during recent years. Data for the Mergini would be especially valuable because they generally inhabit less productive wetlands and have a lower reproductive effort than other ducks.

Observational studies of feeding ecology will continue to be hampered by problems associated with the measurement of foraging effort and foraging success. The foraging effort of ducks generally is determined with time-budgeting methods, which are limited by observer subjectivity regarding foraging "events" and "states" (Baldassarre et al., 1987), and by lack of nocturnal data (Swanson and Sargeant, 1972). Foraging success is studied by collecting actively feeding birds, and rates of food intake can be estimated only when foraging occurs in discrete bouts (e.g., Gray, 1980; Tome, 1981).

B. Laboratory Experiments

Laboratory experiments are needed to provide better data on the digestibility of foods of breeding waterfowl. The importance of macroinvertebrates as a source of protein for breeding ducks is well documented by field studies (Table 1-5), but there are no published estimates of the digestibility of protein in macroinvertebrates. Methods developed by Sibbald (1986, 1987) can be used to obtain these data. Another research need is to determine whether geese can digest cell structural compounds; high rates of food passage through the intestinal tract limit opportunities for digesting complex carbohydrates, but there is some evidence from experimental studies that hemicellulose can be metabolized by several species (Buchsbaum et al., 1986; Dawson et al., 1989; Sedinger et al., 1989).

Perhaps the best opportunity for laboratory research is to explore implications of the recent work by Murphy and King (1987), who showed that molting White-crowned Sparrows (Zonotrichia leucophrys) were able to regulate their intake of sulfur amino acids by selecting from two free-choice diets that provided no visual cues regarding nutrient content. If waterfowl are able to regulate nutrient intake in this way, experiments could be designed to estimate dietary requirements for protein and other nutrients, and to determine whether protein intake changes as a function of lipid reserves. The latter experiment might indicate why the concentration of protein in the diets of free-ranging female ducks (i.e., 30-50%) (Reinecke and Owen, 1980; Hohman, 1985; Noyes and Jarvis, 1985) is consistently greater than the protein requirement of females nesting in captivity (i.e., ≤ 20%) (Foster, 1976).

C. Field Experiments

Field experiments are needed to determine the effects of agricultural chemicals and other land use practices on the food resources and breeding success of ducks nesting in the prairie region. The results of these studies will be used primarily by land managers, but there may also be opportunities to study basic relationships such as effects of changes in macroinvertebrate abundance on the foraging behavior and nesting effort of breeding ducks.

Unfortunately, practical considerations limit options for designing field experiments on the feeding ecology of breeding waterfowl. Successful field experiments involving free-ranging waterfowl and their natural food sources have been limited thus far to the stocking of fish as prey for Common Mergansers (Wood and Hand, 1985), the exclosure or fertilization of feeding sites for geese (Lieff et al., 1970; Cargill and Jefferies, 1984), and the use of imprinting to facilitate the observation (Hunter et al., 1984; Pehrsson, 1984) or collection (Sedinger and Raveling, 1984) of foraging birds. Clearly, innovative methods of manipulating or supplementing (cf. Hill, 1988) natural food supplies are needed to advance field research on the feeding ecology of breeding waterfowl.

D. Natural Experiments

Most of the information that we currently have regarding proximate effects of food availability on the breeding success of free-ranging waterfowl has been derived from natural experiments (section VII). Unfortunately, opportunities for natural experiments on feeding ecology are difficult to anticipate; they often result from random variation of precipitation (Krapu et al., 1983; Davies and Cooke, 1983), temperature (Barry, 1962; Hammond and Johnson, 1984), or food supplies (Bengtson, 1971b), which are "replicated" at unpredictable intervals, if at all. Alternatively, spatial variation of food abundance, like that occurring among territories of Mute Swans (Scott and Birkhead, 1983), may provide opportunities for designing natural experiments.

The foraging behavior of geese using spring staging areas to store nutrient reserves presents opportunities for interesting natural experiments, particularly if the geese use two foraging habitats, one to obtain energy from foods such as grain or rhizomes, and the other to obtain protein from grasses or sedges. If rates of nutrient intake can be estimated for each habitat from defecation frequencies and from analyses of food and fecal samples (e.g., Boudewijn, 1984; Madsen, 1985b), modeling methods such as linear programming (Stephens and Krebs, 1986, pp. 119-121) can be used to predict the allocation of foraging effort between habitats that will maximize intake of: (1) energy, (2) energy subject to a protein constraint, (3) protein, or (4) protein subject to a constraint on protein quality. Comparison of observed and predicted foraging behavior relative to patterns of nutrient deposition should provide insights regarding food choice.

Natural experiments provide the best opportunities for understanding effects of individual variation in foraging success on reproductive performance and on the evolution of feeding strategies. This is illustrated by a study in the Netherlands, in which Teunissen et al. (1985) showed that female Dark-bellied Brant paired with dominant males had access to the best feeding opportunities on spring staging areas and were most likely to return during autumn accompanied by young. Recent research that attributes most of the variation in clutch sizes of Canada Geese (MacInnes and Dunn, 1988) and Northern Shovelers (Ankney and Afton, 1988) to proximate influences should stimulate additional research on the effects of food availability and individual foraging success on the reproductive effort of waterfowl.