USGS - science for a changing world

Northern Prairie Wildlife Research Center

  Home About NPWRC Our Science Staff Employment Contacts Common Questions About the Site

Factors Associated with Duck Nest Success
in the Prairie Pothole Region of Canada

Components of Mallard Production

We evaluated mallard reproduction during 1983-85 in relation to variation in component parameters and partitioned this variation into geographic and temporal factors. Production of ducks requires that pairs are available to populate the breeding area and that wetlands are available in spring to attract and support the breeding pairs and ducklings produced. Females must nest, some nests must hatch and produce young, and some of those young must survive to fledge.

This process can be represented as follows:

Number Fledged                                                       
                  Pairs    Nests   Hatched Nests     Fledged        
    = Wetlands x ------- x ----- x ------------- x ------------  (1)
                 Wetland   Pairs       Nest        Hatched Nest     
By determining which variables on the right-hand side of equation (1) most closely correlate with Number Fledged, we can hypothesize which ones are most influential in their production. The method is similar in principle to key-factor analysis (Varley and Gradwell 1960, Podoler and Rogers 1975).

We used data for the mallard for this exercise, because it was of special interest in other research that our study supported (Brace et al. 1987) and was generally the most numerous duck species on all area-years. We lacked information on the survival of ducklings on areas we studied, so the final component could not be determined. This left

Hatched nests                                         
                  Pairs    Nests   Hatched Nests     
    = Wetlands x ------- x ----- x -------------  (2)
                 Wetland   Pair        Nest          

By estimating, for each half-area-year, the number of Hatched Nests and the 4 components on the right-hand side of equation (2), we determined the relative importance of each of them in relation to the variability of Hatched Nests.

An alternative formulation is

Hatched Nests                           
             Nests   Hatched Nests     
    =Pairs x ----- x -------------  (3)
             Pair        Nest          
This has the disadvantage of ignoring the impact of wetland numbers but shows more directly the effect of the number of pairs on production.

Not all of these variables were directly estimated in our field study; some had to be calculated from other variables. The number of Hatched Nests was determined by taking the number of successful mallard nests found on the area searched within each half-area-year and scaling it upwards to account for the amount of nesting habitat on that half-area-year that was not searched.

Pairs and Wetlands were measured directly. For this analysis we used the total number of temporary, seasonal, and semipermanent wetlands that were wet in May.

The Nest-per-Pair value required an estimate of the total number of nests initiated on a half-area-year, as follows:

  1. Nest success rate estimates the number of successful nests divided by the total number of nests initiated.
  2. Nest success was independently estimated by the Mayfield method (Johnson 1979).
  3. The total number of nests initiated thus can be estimated as the number of successful nests divided by the nest success rate (Miller and Johnson 1978).
  4. The resulting value is divided by the number of Pairs to yield the Nests-per-Pair value.
The Hatched Nests-per-Nest component is the nest success rate.

Some derived values were clearly out of line because mallard populations on each half-area-year were not closed (i.e., birds could freely move into and out of each area) and, possibly, because of errors in estimating the quantities. For example, on the west half of the Ceylon Study Area in 1984, only 1 successful mallard nest was observed. The estimated hatch rate of mallard nests was only 0.002, so the single successful nest is estimated to represent 1 ÷ 0.002 = 500 initiated nests. This is far too many for the 20 mallard pairs estimated to be on that area. Accordingly, we constrained the number of nests per pair to be no greater than 4. This constraint was imposed for only a few (n = 7) half-area-years.

We determined the relations between Hatched Nests and each component by examining bivariate plots and calculating correlation coefficients. Because of the multiplicative form of equation (2), logarithms were taken to yield an additive model:

log(Hatched nests)                           
    = log(Wetlands) + log(Pairs/Wetland)     
      + log(Nests/Pair)                      
      + log(Hatched Nests/Nest)           (4)
This transformation precluded the use of half-area-years with no hatched nests because the logarithm of zero is undefined. Correspondingly, for a half-area-year with no hatched nests, the analogous model for equation (3) is
log(Hatched Nests) = log(Pairs) + log(Nests/Pair)    
                     + log(Hatched Nests/Nest)    (5)

Previous Section -- Adult Mortality
Return to Contents
Next Section -- Geographic and Temporal Effects on Mallard Production

Accessibility FOIA Privacy Policies and Notices

Take Pride in America logo logo U.S. Department of the Interior | U.S. Geological Survey
Page Contact Information: Webmaster
Page Last Modified: Friday, 01-Feb-2013 19:39:09 EST
Sioux Falls, SD [sdww55]