Effects of Weather on Breeding Ducks in North Dakota
Peak Hatching Dates
A measure of the progress of the nesting season is the peak, or median, hatching date—the date by which 50% of all observed hatchings occur. We had adequate information on peak hatching dates for all four species at Salyer and for the gadwall and blue-winged teal at Woodworth.
We sought to determine relations between peak hatching dates and weather, as measured by mean temperatures during 23 April-3 June, the period of most active nesting. We also looked for correlations with arrival dates, and explored two hypotheses:
(1) That a late nesting season in one year will result in a delayed nesting season the following year. The reasoning behind this idea is that if ducks cannot breed until they attain a certain age, such as 9 months, then a late season one year can result the next year in a large proportion of yearlings not capable of breeding until later in the season. Thus, the nesting season will be delayed.
(2) That a year with high production of young will cause a delayed peak hatching the next year. This hypothesis is based on the observation that yearling ducks typically begin nesting later than adults (Courter and Miller 1968). Thus, a year with high production of young will, if the young return to the nesting area, be followed by a year with a high proportion of yearlings in the breeding population, and thereby a delayed peak hatching date.
Peak hatching dates were also closely correlated with the date of initiation of first nest. It is difficult to separate the effects of one variable from the other because both variables measure the phenology of the nesting season. For this reason, we considered the two variables in combination.
Mallard
The median peak hatching date for mallards at Salyer was 25 June (Table 2). The peak hatching date was late when mean temperatures during 23 April-3 June were low (Fig. 3a) and when initiation was late; the effects were jointly significant at P=0.005. The delay in peak hatching date averaged 0.99 day per Celsius degree difference in mean temperatures during that period, and 0.82 day for each day delay in initiating the first nest (Table 4).Arrival date, the previous year's peak hatching date, and the previous year's production did not significantly affect peak hatching date.
Gadwall
Median peak hatching dates for gadwalls were 9 July at Salyer and 29-30 June at Woodworth. Peak hatching dates of gadwalls were closely related to mean temperatures during 23 April-3 June (Fig. 3b). The coefficient of temperature was similar to the mallard, with 1.04 days delay per Celsius degree difference in mean temperatures, but the coefficient of initiation date was only 0.34 day (Table 4). Moreover, after the difference due to temperature was accounted for, peak hatching dates at Woodworth averaged 6.0 days earlier than at Salyer (P<0.05).Arrival date, the previous year's peak hatching date, and the previous year's production were not significantly related to peak hatching date.
Blue-winged Teal
Median peak hatching dates for blue-winged teal were 4-5 July at Salyer and 29 June at Woodworth. The blue-winged teal responded to mean temperatures during 23 April-3 June (P=0.05; Fig. 3c), but date of first nest initiation was not influential (P>0.10). The coefficient of temperature was -1.56 (Table 4).After adjustment for mean temperatures, the peak hatching date was about 8.1 days earlier at Woodworth than at Salyer (P<0.01). Effects of arrival date, and the previous year's production and peak hatching date were not significant with respect to peak hatching date.
Redhead
The median peak hatching date for redheads at Salyer was 12 July. Redheads exhibited a strong association with weather (Fig. 3d) and date of first initiation (P=0.0001). The peak hatching date was delayed 2.45 days for each Celsius degree decrease during 23 April-3 June, and 0.52 day for each day delay in initiating the first nest (Table 4).There was no evidence for any effect on peak hatching date due to arrival date, or the previous year's productivity or peak hatching date.
| Species | Mean temperaturea | Initiation dateb |
| Mallard | -0.99 | 0.82 |
| Gadwall | -1.04 | 0.34 |
| Blue-winged teal | -1.56 | 0.13 |
| Redhead | -2.45 | 0.52 |
| aDays
per Celsius degree difference bDays delay in peak hatching date per day delay in initiation date. |
||
General
All four species showed consistent relations between mean temperatures and peak hatching dates (Table 4). The four coefficients were roughly similar, varying from -0.99 to -2.45 days per Celsius degree change during 23 April-3 June.Delays were also associated with late initiations of first nests; coefficients suggested that, for every day initiation of the first nest was delayed in a nesting season, the peak would be set back by 0.13-0.82 day. The association between mean temperatures and date of first nest initiation made it impossible to determine the individual effects of those variables on peak hatching date.
The two species for which we have data from both Salyer and Woodworth showed that the more southern location had slightly but significantly earlier peak hatching dates, after temperature differences were accounted for. We have no evidence that arrival dates, the previous year's productivity, or the previous year's peak hatching date influenced peak hatching dates.
Other authors (Yocom and Hansen 1960; Newton and Campbell 1975) have observed delayed nesting peaks in years with cooler springs. An analysis of Yocom's (1950) data yielded consistently negative correlations between peak hatching dates of four species and mean temperatures during April and, especially, May. Evans and Black (1956) recorded somewhat later nesting peaks by gadwall and blue-winged teal in 1950, a year with below-normal temperatures during late April and May, but the mallard was not noticeably affected. Sorensen's (1978) data produced no relation between median hatch date and temperature. Sowls (1955) found that mallard and blue-winged teal nesting peaks were about 11 and 17 days later, respectively, in 1950 than in 1949. Temperatures during 15 April-30 May were 4 degrees Celsius lower in 1950 than in 1949. That difference was in the anticipated direction but was greater than would be predicted from the regression equation developed here.
Langford and Driver (1979) attempted to quantify relations between mallard nest initiations and temperatures. They estimated that a 1 Celsius degree change in minimum temperature during April resulted in a change in the mode of nesting dates of 2.0-2.4 days.
Delayed peaks of nesting can be caused by any of the earlier-described factors that influence the date of first nest. In addition, delayed peaks can result from interrupted nesting effort, sometimes ascribed to periods of low temperatures (Dane 1966; Dane and Pearson 1971; Dzubin and Gollop 1972).
Previous Section -- Nest Initiation Dates
Return to Contents
Next Section -- Span of Nesting
