Mallard Recruitment in the Agricultural Environment of North Dakota

Hen Success

Hen success is probably the most important variable needed for evaluating recruitment because it is a function of both nest success and renesting rate. Radiotelemetry used in conjunction with direct observation is a powerful technique for estimating this important parameter because it allows determination of which hens hatch clutches in a sample of marked birds. Our estimates of hen success varied among years and between ages. Differences among means were tested by analysis of variance. The estimates of hen success (H) were weighted by N ÷ [H (1 — H)] because sample sizes varied among years and ages. The raw data were transformed by H = (N_s + 0.5) ÷ (N + 1), where N equals the number of hens in the population and N_s equals the number of successful hens. The transformation is necessary because H = 0 in 1977 for SY birds, making it impossible to calculate an appropriate weight. The difference between ages was highly significant; the older birds were almost twice as successful as the yearlings (Table 14). There were also significant differences among years. Hen success in 1977 was significantly lower than in any other year. This was a year of severe drought and warm spring weather. The highest success was in 1978, a year with a warm spring and excellent water conditions. The spring of 1979 was wet and cold and that of 1980 was the second driest and was again warm. Hen success in 1980 was second lowest.

**Table 14.** Estimated percent of hens with a successful nest (H) in one of their attempts on a study area in central North Dakota.
Age of hen^a	N	1977	1978	1979	1980	All years
SY	58	0.0	16.7	13.3	10.5	10.3 A^a
ASY	97	11.1	26.1	18.2	17.6	18.6 B
All ages^c	177	5.9 A^b	23.6 B	17.4 BC	13.6 BC	15.2 C
^aSY = second-year hens in first year of nesting; ASY = after-second-year hens in their second or later nesting years. ^bMeans followed by the same latter are not significantly different (P > 0.05). ^cThe all-age group includes AHY birds whose age could not be determined.

To test how well results obtained on our area reflected conditions over a wider geographic area, we related our hen success estimates to fall age ratios of wings of mallards shot in North Dakota (M. F. Sorensen, S. M. Carney, and E. M. Martin, unpubl. rep., U.S. Dep. Inter., Fish and Wildl. Serv., Off. Migratory Bird Manage., Laurel, Md., 1982). We used the raw data uncorrected for differential age vulnerability because in the wing collection survey there is no way to separate birds produced in North Dakota from birds produced outside the state. It is also probable that the wing sample from North Dakota includes some birds from more northern breeding areas. The correlation between these 2 independent estimates was excellent (Fig. 9), illustrating that yearly variation in our sample did reflect differences over a larger area.

GIF-Relation between fall age ratio and hen success

Relation of Hen Success to Habitat Conditions and Weather

We did not detect differences in nest success among years or ages; therefore, the differences in hen success are probably related to nesting effort. The number of nests per hen which can be estimated by hen success ÷ nest success can be used to compare nesting effort between ages. SY hens produced 0.92 nests/hen, and ASY hens produced 2.26 nests/hen.

We hypothesized that annual variation in nesting effort is a function of habitat conditions and possibly weather. Krapu et al. (1983) presented data that strongly supported this hypothesis. Cowardin and Johnson's (1979) model, H = fPe^{f(l - P)²} furnished a means of testing this hypothesis because it contains a single parameter (f) for nesting effort, and the data for nest success (P) and hen success (H) were available for each year. It was possible to solve for f. We developed an index to water conditions by assigning wetness codes (Table 15) to ponds on each set of aerial photographs and averaging codes for all ponds in the sample. Wetland indices for class II, III, and IV ponds (Stewart and Kantrud 1971) in the late May and early June flights were correlated with nesting effort. The correlation was strongest for class IV ponds (r = 0.98, P < 0.05) (Fig. 10). Correlations with class II (r = 0.83, P > 0.10) and class III (r = 0.91, P < 0.10) were not as strong and were not significant. Furthermore, the deviation of the data points from the regression for class IV ponds may be correlated with length of nesting season (r = 0.82, P = 0.18). Hen success is a function of nest success and nesting effort. Nest success is primarily a function of predation and nesting effort is a function of water conditions and age of the hen.

**Table 15.** Relations among nesting effort, wetland conditions, and length and timing of nest initiation period.
Year	Nesting effort (f)^a	Wetness index^b			Nest initiation
Year	Nesting effort (f)^a	II	III	IV	Length of season^c (Julian days)
1977	0.5918	0.035	0.123	0.697	48.3
1978	0.9108	0.463	1.129	1.720	48.2
1979	1.0444	1.021	1.631	1.896	49.7
1980	0.8700	0.106	0.474	1.369	52.2
^aBased on Cowardin and Johnson's (1979) model, H = fPe^f(1-P)². ^bWetness index = mean of codes; 0 = dry, 1 = one-half full, and 2 = full. Classes are after Stewart and Kantrud (1971): II = temporary, III = seasonal, IV = semipermanent. ^cLength = 90% quantile - 10% quantile of nest initiation dates.

GIF-Relation between nesting effort and wetness index

Previous Section -- Relation of Reproductive Effort to Condition of the Hen
Return to Contents
Next Section -- Survival of Hens During Spring and Summer