Northern Prairie Wildlife Research Center
Food habits of Lesser Scaups consisted of gizzard shad during periods when the lakes were iced covered, usually during January and February (Assel et al. 1983). Waterfowl were forced to leave or move into the remaining areas of open water (CMC, unpubl. data) at the mouths of rivers, power plant water discharge points, and ice leads. Shad were plentiful and may have been the only food available in the power plant discharge areas (CMC, pers. obs).
Only about 25% of Bufflehead diets was zebra mussels in Lake Erie; amphipods dominated (47%). Buffleheads in California consumed mainly seeds of Scirpus robustus, Polygonum lapathifolium and Potamogeton pectinatus (34% of diet), corixids (water boatman, 29%) and chironomids (22%); molluscs comprised <11% of the diets (Gammonley and Heitmeyer 1990). Cottam's (1939) compilation of waterfowl diets (mainly from gizzards) indicated that insects (41%), crustaceans (17%) and molluscs (16%) were the main foods in the diet of Buffleheads.
Nearly 80% of Common Goldeneye diet was zebra mussels in Lake Erie. Prior to the invasion of the zebra mussels, Common Goldeneye diets in the nearby Detroit River were comprised of 37% wildcelery winterbuds, 9% oligochaetes and 7% decapods; molluscs were found in only trace amounts (Jones and Drobney 1986). Cottam (1939) indicated that crustaceans (32%) and insects (28%) were the dominant components in Common Goldeneye diets (mainly from gizzards) with molluscs comprising <10%.
Canvasbacks fed primarily on wildcelery winterbuds in Lakes Erie and St. Clair, probably because this aquatic vegetation was widely available (CMC, pers. obs.). Few individuals consumed zebra mussels. Canvasbacks switched to a diet of Baltic macoma (Macoma balthica, 82%) (data mainly from gizzard samples, Perry and Uhler 1988) when the submersed aquatic vegetation declined in Chesapeake Bay (Bayley et al. 1978, Haramis and Carter 1983, Orth and Moore 1981). In North and South Carolina aquatic vegetation predominated in the diet of Canvasback when available, otherwise Baltic macoma dominated (gizzard data, Perry and Uhler 1982). On the upper Mississippi River Canvasbacks consumed ≥99% wildcelery winterbuds and Sagittaria rigida tubers (Korschgen et al. 1988) even though invertebrate foods were available (C. E. Korschgen, pers. comm.).
The Canvasback that was collected while feeding alongside a Lesser Scaup was eating 100% wildcelery winterbuds whereas the Lesser Scaup was eating 100% zebra mussels. This may suggest a preference for wildcelery winterbuds by Canvasbacks and a preference for zebra mussels by Lesser Scaups. The one Canvasback that was eating zebra mussels was apparently taking advantage of zebra mussels that were exposed due to low water conditions created by a seiche.
In Lakes Erie and St. Clair eight of 10 Redheads consumed mainly above-ground portions of Potamogeton spp. and incidently ingested zebra mussels which were attached to the vegetation stems. In two of six Redheads in Lake Erie, however, zebra mussels were the only food in the GI tract. This dominance of vegetation in the diet is consistent with other studies of Redhead food habits. Redheads in North Carolina ate predominately shoalgrass (Halodule wrightii) (from gizzard data mainly, Perry and Uhler 1982) as did Redheads in Louisiana (83% & 67% of diet as determined from esophagi and proventriculi) (Michot and Nault 1993) and Texas (Koenig 1969). On the Mississippi River in Wisconsin, 65% of Redhead's diet was wildcelery winterbuds (Korschgen et al. 1988).
Zebra mussels consumed by ducks, except for the eight Redheads mentioned above, were present as individual mussels. We did not find clumps of zebra mussels still held together by their byssal threads nor zebra mussels still attached to a substrate such as vegetation or rock. This was true even for Buffleheads that consumed very small zebra mussels. This observation suggests that zebra mussels were either plucked individually from the substrate or if pulled off in clumps then the clumps were separated in the duck's mouth prior to swallowing.
Waterfowl from both lakes that were specifically foraging on zebra mussels consumed zebra mussels that averaged 8-12 mm long. Mussels that were attached to vegetation, primarily from Lake St. Clair, and incidentally consumed were smaller ( = 3 mm). The size distribution of zebra mussels in the areas where we collected waterfowl is unknown, but waterfowl consumed sizes close to the size deemed most profitable for female (9-10 mm) and male (13-14 mm) Tufted Ducks (de Leeuw and Van Eerden 1992). The modal size of zebra mussels eaten by six ducks at Point Pelee, Ontario was 11-13 mm (Hamilton et al. 1994). This size was larger than the modes in our study; however, Hamilton et al. (1994) disregarded all mussels < 5 mm in length when calculating modes, so their modal value is biased high. In our study, 38% of zebra mussels eaten by Lesser Scaup and 50% of zebra mussels eaten by Bufflehead were ≤5 mm.
Diving duck numbers, especially Lesser Scaups and Canvasbacks, and their duration of stay have increased on Lake Erie and Lake St. Clair since the arrival of zebra mussels (Wormington and Leach 1992; E. N. Kafcas and J. Weeks, unpubl. data). Canvasback seem to have benefitted indirectly from the arrival of zebra mussels because the increased water clarity (Holland 1993) has allowed wildcelery to proliferate (E. N. Kafcas, pers. commun.). This alteration of distribution patterns could increase survival rates for wintering diving ducks because some do not travel the additional 500-1300 km to the Atlantic or Gulf Coasts and some migrate late, after the hunting seasons have closed. In contrast, more than 15,000 diving ducks died of starvation in Switzerland and the Netherlands when their feeding areas froze over for an extended period in late winter (Suter and Van Eerden 1992). Body reserves of these birds were not sufficient to survive a 29-d period when ice covered zebra mussel beds and physiological condition prevented them from moving south so late in the winter.
Body masses of Lesser Scaups and Buffleheads in this study (CMC, unpubl. data) were similar to body masses reported from spring migrants in California (Gammonley and Heitmeyer 1990) and wintering waterfowl in New York (Ryan 1972). Lipid content of Lesser Scaups from Lake Erie (5-46%, CMC, unpubl. data) was also similar to lipid levels in fall migrant Lesser Scaups in Manitoba (Austin and Fredrickson 1987). These mass comparisons may indicate that zebra mussels are an adequate winter food. Energetic constraints of zebra mussels as a food should be examined experimentally, however. Waterfowl may need to consume greater volumes of and spend more time feeding when consuming zebra mussels, a low-energy food (4389 cal/g dry mass [Stanczykowska and Lawacz 1976] ) compared with other higher energy foods (Driver et al. 1974, Jorde et al. 1995, Perry et al. 1986). The low energy content of zebra mussels may be partially compensated for, however, by their high abundance and easy availability.
Because zebra mussels are known to accumulate contaminants such as polychlorinated biphenyls (Brieger and Hunter 1993), it will be important to determine whether waterfowl are being affected by increased contaminant exposure, especially in areas of known pollution such as the Detroit and Maumee Rivers (Hoke et al. 1990, Smith et al. 1985). Reduced reproduction has been identified in Tufted Ducks fed zebra mussels from a contaminated basin in the Netherlands (de Kock and Bowmer 1993).