USGS Home
Contact USGS
Search USGS

Foraging Ecology and Nutrition

VII. Proximate Effects of Food Resources on Reproduction

Assessing the significance of food as a proximate influence on waterfowl recruitment requires knowledge of the effects of diet on various components of reproductive effort. Many of the experiments necessary to test potential effects of food resources on reproduction have not been conducted because of the difficulty of manipulating natural food supplies. Nevertheless, considerable circumstantial evidence and some experimental data indicate that food resources available to prebreeding and breeding waterfowl influence reproductive performance in several ways. In this section, we consider the significance of food resources to: (1) timing of reproduction, (2) failure to breed, (3) clutch size, (4) egg mass and composition, (5) egg fertility and hatchability, and (6) occurrence and frequency of renesting.

A. Timing of Reproduction

Migratory birds that breed in temperate, subarctic, and arctic regions use lengthening photoperiod to provide a reliable stimulus for initiating migration and the physiological changes leading to nesting (Farner, 1961, 1964; Immelmann, 1971). Arrival of migratory geese on spring staging areas seems to be timed to coincide with foraging conditions that maximize nutrient deposition. Lesser Snow Geese and Ross' Geese initiate rapid follicle growth on subarctic staging areas and are capable of laying eggs shortly after arrival on the breeding grounds if nest sites are available (Raveling, 1978). Canada Geese and other species that initiate rapid follicle growth on departure from final staging areas continue to feed prior to nesting. Food availability is more likely to affect time of nest initiation in the latter species.

Among ducks breeding in temperate regions, the onset of laying is often linked to air temperatures (Langford and Driver, 1979; Krapu and Doty, 1979; Fredga and Dow, 1983; Hammond and Johnson, 1984). Temperature and food availability are correlated, however, so it is difficult to evaluate food availability as a separate influence on timing of nest initiation. In tropical regions where temporal variation of photoperiod and temperature is relatively small, some species of waterfowl breed at irregular intervals. Precipitation and/or changing water levels provide a reliable indication of access to necessary food resources for several species of ducks that breed in parts of Australia where rainfall is erratic (Braithwaite and Frith, 1969). Gray Teal nest during the early stages of flooding; their sexual cycles apparently are initiated by an increase in water levels (Frith, 1959). Pink-eared Ducks also breed irregularly, but unlike Gray Teal, they nest when declining floodwaters concentrate the small insects and crustaceans that are their principal foods (Frith, 1959, 1967, pp. 238-239). In contrast, the Musk Duck, Blue-billed Duck, and Hardhead utilize stable wetlands and have regular breeding seasons that coincide with predictable changes in photoperiod or water conditions (Braithwaite and Frith, 1969).

Interspecific differences in timing of nesting among waterfowl presumably are adaptive responses to the average availability of food to nesting females and/or their offspring. Intraspecific variation in timing of nest initiation probably results from differences in age and experience among females, and from spatial and temporal variation of food availability. In a nonmigratory Mute Swan population in England, breeding pairs with territories containing abundant aquatic vegetation initiated nesting earlier than other pairs (Scott and Birkhead, 1983). Establishing a cause-and-effect relationship between time of nest initiation and availability of food is difficult, however, because age and experience of the breeding pair often are confounded with effects of food availability. Among free-ranging Mallards and Pintails, for example, adult females generally nest earlier and lay larger clutches than young females (Krapu and Doty, 1979; Duncan, 1987), but probably also locate and/or dominate food sources more effectively. Dubovsky (1990) has shown experimentally that body weights, and therefore foraging conditions, of female Mallards affect the time of nest initiation. Breeding female Canvasbacks often use the same foraging sites from one year to the next (Anderson, 1985).

B. Failure to Breed

Food scarcity is probably a major reason why many temperate and arctic waterfowl fail to breed during certain years. The incidence of nonbreeding among midcontinent Lesser Snow Geese increased during years of drought on prairie staging areas (Davies and Cooke, 1983), presumably in response to decreased food intake and nutrient storage. Food shortages caused by inclement weather on arctic nesting grounds also can inhibit nesting. When snow cover prevents Brant from foraging for prolonged periods, fat reserves are depleted, follicles become atretic, and breeding activity ceases (Barry, 1962).

Food scarcity contributes to the failure of many prairie-nesting ducks to breed during years of severe drought (Smith, 1969, 1971; Krapu et al., 1983). Northern Pintails respond to drought in the Prairie Pothole Region by migrating to subarctic and arctic areas (Smith, 1970). Nesting by these birds probably is limited (Derksen and Eldridge, 1980), however, because fat reserves are depleted or food resources are insufficient for reproduction (Calverly and Boag, 1977). In Iceland, the percentage of nonbreeding pairs of Harlequin Ducks increased from 15 and 25% at Lake Mývatn during 1968 and 1969, respectively, to 89% in 1970 (Bengtson and Ulfstrand, 1971) when the abundance of their primary food source, black fly larvae (Family Simuliidae), was much below normal.

Thus, unfavorable foraging conditions on staging and/or breeding areas can cause waterfowl to forgo breeding attempts. Our understanding of the recruitment process is incomplete, but simulations indicate that recruitment rates of waterfowl are quite sensitive to the proportion of adults that attempt to breed (chapter 14).

C. Clutch Size

The magnitude of fat reserves that females possess at the onset of the breeding season is an important determinant of clutch size in certain arctic- and temperate-nesting waterfowl (Ankney and MacInnes, 1978; Krapu, 1981; Ankney and Afton, 1988). Clutch sizes of Lesser Snow Geese correlate positively with food availability on temperate staging areas used for nutrient deposition and correlate negatively with adverse weather conditions that cause depletion of nutrient reserves after arrival at breeding areas (Ankney and MacInnes, 1978; Davies and Cooke, 1983). Annual and seasonal differences in the nutrient reserves of females are thought to be the most important proximate factor affecting clutch sizes of Canada Geese breeding at McConnell River, Northwest Territories (MacInnes and Dunn, 1988). Annual variation in clutch size of 18-36% (i.e., 1-2 eggs) among Dusky Canada Geese nesting on the Copper River Delta in Alaska is thought to be determined by protein availability from both endogenous and local environmental sources (Bromley, 1984).

Ducks rely on foods available on the breeding grounds for a greater part of their nutrient requirements for producing clutches of eggs than geese (section IV). Three species of diving ducks (i.e., the Greater Scaup, Tufted Duck, and Black Scoter) experienced small but significant decreases in clutch size at Lake Mývatn in Iceland when densities of midge larvae declined by 82% during the egg-laying period (Bengtson, 1971b). The extent to which observed variation in food availability exceeds variation in clutch sizes among ducks reflects the important role of lipid reserves in compensating for the marked fluctuations in food abundance during the period eggs are being formed. Lipid reserves provide an energy source to the female when seeking protein-rich macroinvertebrates during rapid follicle growth and supply a major part of the lipid content of the initial clutches of eggs (Drobney, 1980; Krapu, 1981; Barzen and Serie, 1990). The relative importance of lipid reserves in contributing to energy needs and as a source of lipids for egg formation is likely to vary with species and period within the nesting season. Utilization of part of the lipid reserves to satisfy energy requirements is likely to be most important among those species of ducks that begin nesting in early spring shortly after snowmelt and before macroinvertebrate populations can build in the recently flooded wetland habitat. Among species that forage in wetland habitats where macroinvertebrates are abundant, access to dietary lipids appears to be more limiting to clutch size than is protein availability (Ankney and Afton, 1988; Alisauskas et al., 1990; Afton and Ankney, 1991). Renesting female Mallards must acquire most of their nutrient requirements for clutch production from dietary intake at the time the clutch is produced (Krapu, 1981).

D. Egg Mass and Composition

Substantial intraspecific variation exists in waterfowl egg mass (Krapu, 1979; Ankney, 1980; Birkhead, 1985; Rohwer, 1986; Hepp et al., 1987). Part of this variation is under genetic control (Batt and Prince, 1978), but diet also is an important contributing factor (Pehrsson, 1982,1991; Eldridge and Krapu, 1988; Murphy, 1988). When the supply of one or more amino acids in the diet is inadequate, synthesis of egg protein decreases (Scott et al., 1969, p. 387). Female Mallards eating natural foods produced eggs smaller than those of their daughters fed an enriched diet in captivity (Eldridge and Krapu, 1988), suggesting that food availability constrains egg size below the genetic potential of free-ranging Mallards. Large eggs are advantageous to waterfowl because the size of offspring at hatching is proportional to egg mass (Hepp et al., 1987; Eldridge and Krapu, 1988), and offspring from larger eggs have more efficient thermoregulatory ability (Rhymer, 1988), endure food deprivation longer than offspring from smaller eggs (Krapu, 1979; Ankney, 1980), and initially have a higher growth rate (Holmberg, 1991).

Diet also influences the composition of waterfowl eggs. Female Mallards fed ad libitum a balanced commercial ration containing 29% protein laid eggs with larger yolks than females fed a wheat diet with 18% protein (Eldridge and Krapu, 1988). Large eggs generally provide extra nutrients to offspring at hatching because the proportion of yolk in large eggs is equal to (Lesser Snow Geese [Ankney, 1980]; Wood Ducks [Hepp et al., 1987]) or greater than (Mute Swans [Birkhead, 1984]; Mallards [Birkhead, 1985; Eldridge and Krapu, 1988]; Blue-winged Teal [Rohwer, 1986]) the proportion of yolk in small eggs.

Although diet influences egg mass and composition in waterfowl, recruitment rate is likely to be less influenced by dietary effects on egg characteristics than the proportion of females attempting first nests or renests (see chapter 14 of this volume).

E. Egg Fertility and Hatchability

Egg fertility and hatchability are potentially limiting to waterfowl recruitment, and experimental studies have shown that both parameters can be influenced by the nutrient content of the diet available to breeding females (Holm and Scott, 1954; Krapu and Swanson, 1975; Foster, 1976). Egg fertility influences annual recruitment among Giant Canada Geese (Hanson, 1965, p. 171; Cooper, 1978), but no nutritional causes have been identified. Extensive data on hatching rates of duck eggs from the Prairie Pothole Region suggest that losses due to fertility or hatchability are relatively low. In most cases, nutritional deficiencies that are severe enough to cause low fertility and hatchability in free-ranging waterfowl probably are preceded by a failure to nest.

F. Occurrence and Frequency of Renesting

Northern-nesting geese and other waterfowl that rely primarily on endogenous nutrient reserves to produce eggs seldom renest. Renesting is common in temperate-nesting Canada Geese (Cooper, 1978), however, and may contribute substantially to annual recruitment in some populations (Geis, 1956; Brakhage, 1965). Effects of diet on renesting in geese probably are similar to effects of diet on first nest attempts. In captive Western Canada Geese, the proportion of females that initiated nests decreased when access to food was made more difficult and when the energy content of the diet was reduced (Murphy, 1988).

The importance of food availability to renesting is difficult to assess among free-ranging ducks but is well documented by experimental studies (Krapu and Swanson, 1975; Foster, 1976, p. 19; Swanson et al., 1986; Eldridge and Krapu, 1988). The total weight of eggs produced by female Mallards fed diets containing 10% and 18% protein was equivalent to 66% and 206% of their body weights, respectively (Foster, 1976). Female Mallards fed a wheat diet with 18% protein produced fewer clutches and had longer intervals between clutches than females fed a commercial diet with 29% protein (Eldridge and Krapu, 1988). Increased availability of essential amino acids rather than greater total protein intake apparently contributed to the higher frequency of renesting among female Mallards fed the commercial diet because female Northern Pintails fed commercial diets containing 14% and 29% protein (but with neither diet restricted to plant matter) renested at similar rates (Duncan, 1987). Among free-ranging ducks, studies of radiomarked Mallards (Krapu et al., 1983; Cowardin et al., 1985) and regional population surveys (Reynolds, 1987) have shown that renesting frequency is influenced by availability of wetland habitat during spring and summer. Favorable habitat conditions benefit ducks in many ways, but food availability is believed to be the major factor affecting frequency of renesting (Krapu et al., 1983).

For ducks, renesting is probably the most important reproductive parameter affected by diet. In environments such as the Prairie Pothole Region, where nest success rates are low, annual recruitment rates of ducks are sensitive to the amount of renesting that occurs.