Supervisors: Dr. Scott Petrie, Long Point Waterfowl Executive Director
Dr. Jack Millar, University of Western Ontario
Waterfowl produced in southern Ontario contribute substantially to regional sport harvest. From 1997-2000 Ducks Unlimited Canada conducted an assessment of the productivity of Mallards breeding in southern Ontario. Female Mallards were radio-marked and tracked throughout the breeding season to determine several important vital rates. Overall, recruitment of female ducklings to 30 days post-hatch was higher than reference areas in the prairie pothole region. However, breeding pair densities in southern Ontario are markedly lower. Because pairs returning from wintering areas settle to breed based on habitat availability, the assessment concluded that protecting, enhancing, and restoring wetlands would be the most cost- effective strategy to increase the southern Ontario Mallard breeding population.
The next logical step was to determine which wetland and upland habitat characteristics cause ducks to settle in a given area. Ducks respond to habitat cues viewed at several spatial scales (i.e., from very large landscapes all the way down to the characteristics of a particular wetland). Existing research in southern Ontario was limited by having only evaluated these associations at the wetland scale. Multi- spatial scale studies exist for the prairie regions, however habitats in southern Ontario differ considerably and it was determined that new region-specific information would be needed for effective conservation planning.
In 2008 and 2009 Dave used helicopter surveys of breeding waterfowl in sample sites that represented the various wetland and upland habitat configurations available in southern Ontario. A geospatial database of waterfowl pair locations, wetland types, and upland land use was compiled using survey results, aerial photography, and satellite imagery. From this database Dave sampled habitat characteristics around breeding pairs at several spatial scales (25 km2, 500 m and 250 m radius buffers, and wetland-scale).
Dave was also able to use this data to evaluate habitat associations of temperate-breeding Canada Goose pairs. Canada Goose populations have substantially increased
throughout southern Ontario and much of temperate North America, but to date there has been little research to determine what habitat associations exist. Although anecdotal evidence suggests that Canada Goose pairs are more influenced by smaller scale habitat features (e.g., muskrat huts for nest sites) and family philopatry than Mallards, Dave’s results still provide important information on larger scale associations. In addition, Dave was able to use the spatial dataset to test the hypothesis that Canada Goose pairs competitively exclude Mallards from breeding habitat.
For both species Dave developed mathematical models to represent settling patterns at each spatial scale. This process requires that the relationships between habitat variables and corresponding breeding pair responses be explicitly defined. This represents an important objective of conservation planning under the North American Waterfowl Management Plan. Dave’s models confirmed that spatial scale of analyses influences results, and that some habitat features are important at several scales.
Mallard breeding pairs were positively associated with densities of temporary open-water (sheet-water), emergent wetlands, and perennial field upland land use (pasture, hayland, idle land) at several spatial scales. These associations would not have been evident if analyses were only performed at the wetland scale as in previous studies. Additionally, riverine wetlands had a relatively large but highly variable effect on Mallard settling patterns.
Canada Goose results were more scale-dependent, but densities of riverine, forested and permanent open-water wetlands and the area of emergent wetlands and annually cropped fields were important. Dave’s results did not support the hypothesis of competitive exclusion of Mallards. In contrast there was a positive spatial association between the two species at the scales sampled.
By providing an explicit understanding of which habitat characteristics influence where Mallard pairs settle, managers can now predict pair densities based on habitat information and also assess what conservation actions are required to increase carrying-capacity in specific landscapes. Similarly, these results provide a first step in understanding the habitat associations of temperate-breeding Canada Goose pairs and will inform conservation planning in southern Ontario.
Dave completed his M.Sc. at the University of Western Ontario in May 2010. After completing this degree he worked for the U.S. Fish and Wildlife Service and Ducks Unlimited Canada, before beginning a Ph.D. program at the University of Saskatchewan in 2011. His research at the University of Saskatchewan will focus on modeling the North American scaup populations and furthering the understanding of wetland and aquatic invertebrate dynamics in the Western Boreal Forest of Canada.