Populations of some dabbling ducks have declined sharply in recent decades and information is needed to understand reasons for this. During 1982-85, we studied duck nesting for 1-4 years in 17 1.6 by 16.0-km, high-density duck areas in the Prairie Pothole Region (PPR) of Canada, 9 in parkland and 8 in prairie. We estimated nest-initiation dates, habitat preferences, nest success, and nest fates for mallards (Anas platyrhynchos), gadwalls (A. strepera), blue-winged teals (A. discors), northern shovelers (A. clypeata), and northern pintails (A. acuta). We also examined the relation of mallard production to geographic and temporal variation in wetlands, breeding populations, nesting effort, and hatch rate.

Average periods of nest initiation were similar for mallards and northern pintails, and nearly twice as long as those of gadwalls, blue-winged teals, and northern shovelers. Median date of nest initiation was related to presence of wet wetlands (contained visible standing water), spring precipitation, and May temperature. Length of initiation period was related to presence of wet wetlands and precipitation in May, June temperature, and nest success; it was negatively related overall to drought that prevailed over much of Prairie Canada during the study, especially in 1984.

Mallards, gadwalls, and northern pintails nested most often in brush in native grassland, blue-winged teals in road rights-of-way, and northern shovelers in hayfields and small (< 2 ha) untilled tracts of upland habitat (hereafter called Odd area). Among 8 habitat classes that composed all suitable nesting habitat of each study area, nest success estimates averaged 25% in Woodland, 19% in Brush, 18% in Hayland, 16% in Wetland, 15% in Grass, 11% in Odd area, 8% in Right-of-way, and 2% in Cropland. We detected no significant difference in nest success among species: mallard (11%), gadwall (14%), blue-winged teal (15%), northern shoveler (12%), and northern pintail (7%). Annual nest success (pooled by study area and averaged [unweighted] over all study areas) was 17% in 1982, 15% in 1983, 7% in 1984, and 14% in 1985.

We estimated that predators destroyed 72% of mallard, gadwall, blue-winged teal, and northern shoveler nests and 65% of northern pintail nests. In prairie, average nest success decreased about 4 percentage points for every 10 percentage points increase in Cropland, suggesting that under conditions of 1982-85, local populations of these species probably were not stable when Cropland exceeded about 56% of available habitat. We found recent remains of 573 dead ducks during 1983-85; most were females (Anas spp.) apparently killed by predators. In some years, mallards and northern pintails were more numerous among dead ducks than we expected. More females than males were found dead among mallards and northern shovelers, suggesting higher vulnerability of females. Of factors we examined, nest-success rate appeared to be the most influential factor in determining mallard production. Nest success varied both geographically and annually.