Summary
Seasonal, habitat, and sex-specific food habits of red-winged blackbirds were studied by stomach analysis from March through October 1977 in eastern Ontario. The overall diet is nonspecialized, but within seasons, habitats, or sexes, specializations do occur. In early spring, males eat waste grain and seeds, with birds in agricultural habitats specializing more on the former, and in nonagricultural habitats on the latter. In the breeding season, both sexes in both habitats switch to insects, although waste grain remains an important component of the diet of birds in agricultural areas. Following breeding, birds move into agricultural areas and exploit both standing oats and corn, and then waste grains again late in the fall. Some of the insects and weed seeds in the birds' diet are potentially damaging to agriculture, so consumption by birds may have direct benefits to agriculture. A cost-benefit analysis, focusing on details of the quantities and economics of consumption by red-winged blackbirds of some of these insect and weed pests should be done before widespread control of blackbirds is attempted.
Methodology
Study areas
The study was conducted in eastern Ontario in Frontenac and Leeds counties, both of which sustain extensive blackbird damage to their corn crops (Tyler and Kannenberg 1980). In this region red-winged blackbirds occur in two major habitats, agricultural and nonagricultural. Nonagricultural samples were collected in the vicinity of the Queen's University Biological Station at Lake Opinicon. While not devoid of agricultural influence, most land in the area sampled is typified by mixed deciduous woodland with the red-winged blackbirds generally associated with marshes along lake and pond perimeters. Agricultural areas sampled were centered around Kingston where the majority of land is farmed. Agricultural sites were subdivided during the breeding season on the basis of whether the birds were nesting in upland or marsh vegetation and birds from both habitats were regularly sampled.
Methods
Samples were collected from March through October 1977 with sampling periods during this time designated according to the migratory and reproductive activities of the birds. Sampling sites within a given habitat were selected randomly other than for the obvious constraint that there be red-winged blackbirds present. An effort was also made to randomize sampling times throughout the day so as to avoid biases associated with diurnal feeding patterns (Hintz 1968). Each habitat was visited two to four times per week, with between five and eight birds collected on each visit. An effort was made to obtain sample sizes of 15 to 20 birds per habitat per week, subdivided about equally amongst sex and to a lesser extent age categories. Birds were collected using a 12-gauge shotgun. The esophagael-proventicular region (crop) and gizzard were immediately removed and stored separately in ethanol. For analysis, contents of each were placed in a petri dish and examined using a dissecting microscope. The crop or gizzard contents were separated into categories and the percentage of the total volume comprising the material in each category was visually estimated (Hyne 1950; Crase and DeHaven 1978; Robertson et al. 1978). Weed seeds and insect specimens were categorized by "type," with representative individuals of each type later identified with the assistance of experts in the appropriate fields. Following separate analysis, the values for the crop and gizzard contents of each individual were averaged, in order to reduce biases due to differential retention rates of different food types (Crase and DeHaven 1978).