Northern Prairie Wildlife Research Center
Freshly laid Mallard eggs from three clutches averaged 49.3 g (Table 1). After one week of incubation, eggs averaged 46.3 g (n = 18), with weights declining another 0.8 g at pipping. One-day-old young (42% of which were dry and 58% of which were damp) weighed an average of 32.4 g. That weight loss of mainly shell, membrane, and moisture represented an average decline of 34% from newly laid eggs and 29% from pipped eggs.
Table 1. Average weights (in g), sample sizes (n), and standard deviations (S.D.) of Mallard eggs and ducklings, by sex and age. | |||||||
Class^{a} | Age | Females | Males | ||||
Mean | n | S.D. | Mean | n | S.D. | ||
Fresh egg^{b} | 49.3 | 27 | 3.5 | ||||
Pipped egg^{b} | 0 | 45.5 | 302 | 3.9 | |||
Ia^{b} | 3.5 | 32.4 | 36 | 2.4 | |||
Ib | 9.5 | 115.3 | 6 | 37.3 | 92.2 | 4 | 11.5 |
Ic | 15.5 | 265.0 | 2 | 91.9 | 215.0 | 3 | 5.0 |
IIa | 22.0 | 288.9 | 14 | 60.5 | 343.2 | 11 | 75.3 |
IIb | 30.5 | 401.2 | 20 | 92.2 | 460.3 | 30 | 93.4 |
IIc | 40.5 | 575.0 | 22 | 152.9 | 648.4 | 19 | 128.4 |
III | 50.5 | 774.3 | 38 | 124.9 | 863.9 | 31 | 102.1 |
Flying | 56.0 | 740.0 | 5 | 114.9 | 817.1 | 7 | 91.4 |
^{a} Age classes as defined by Gollop
& Marshall (1954). ^{b} Not separated by sex. |
Gadwall egg weights declined 10% from freshly laid eggs to pipping (Table 2). After hatching, young in the nest (25% dry, 75% damp) weighed an average of 30.4 g, indicating losses of 29% from fresh eggs and 22% from pipped eggs. We did not obtain fresh egg weights for Blue-winged Teal, but there was a 33% decline in weight between pipped eggs and one-day-old young (all dry) in the nest (Table 3).
Table 2. Average weights (in g), sample sizes (n), and standard deviations (S.D.) of Gadwall eggs and ducklings, by sex and age. | |||||||
Class^{a} | Age | Females | Males | ||||
Mean | n | S.D. | Mean | n | S.D. | ||
Fresh egg^{b} | 43.1 | 20 | 1.7 | ||||
Pipped egg^{b} | 0 | 39.0 | 439 | 3.5 | |||
Ia^{b} | 3.5 | 30.4 | 75 | 2.3 | |||
Ib | 10.5 | 99.3 | 40 | 24.2 | 93.7 | 27 | 13.0 |
Ic | 16.5 | 186.6 | 40 | 45.5 | 188.0 | 63 | 48.1 |
IIa | 23.0 | 287.6 | 94 | 50.8 | 294.0 | 89 | 49.9 |
IIb | 33.0 | 380.7 | 91 | 74.9 | 406.6 | 131 | 83.0 |
IIc | 41.5 | 484.5 | 55 | 60.5 | 556.1 | 55 | 81.5 |
III | 47.5 | 573.0 | 33 | 60.0 | 616.9 | 50 | 69.8 |
Flying | 50.0 | 552.9 | 7 | 44.2 | 606.4 | 14 | 41.4 |
^{a} Age classes as defined by Gollop
& Marshall (1954). ^{b} Not separated by sex. |
Table 3. Average weights (in g), sample sizes (n), and standard deviations (S.D.) of Blue-winged Teal eggs and ducklings, by sex and age. | |||||||
Class^{a} | Age | Females | Males | ||||
Mean | n | S.D. | Mean | n | S.D. | ||
Pipped egg^{b} | 0 | 27.2 | 22 | 5.2 | |||
Ia^{b} | 3.0 | 18.1 | 21 | 3.2 | |||
IIa | 17.5 | 162.0 | 5 | 40.2 | 166.2 | 8 | 40.0 |
IIb | 26.0 | 203.5 | 13 | 22.1 | 206.7 | 9 | 25.5 |
IIc | 33.5 | 261.3 | 41 | 30.5 | 269.0 | 56 | 37.2 |
III | 38.5 | 281.9 | 93 | 29.4 | 301.3 | 97 | 34.4 |
^{a} Age classes as defined by Gollop
& Marshall (1954). ^{b} Not separated by sex. |
Although male ducklings weighed more than female ducklings beginning with the Class II stage (Tables 1-3) variability was considerable and sex differences were not significant (P > 0.05).
We fit three growth curves to the data for flightless young of each species and sex: logistic, Gompertz, and Richards (e.g. Ricklefs 1973, Richards 1959). We found that the logistic curve provided a better fit than the Gompertz curve, and was as good as the Richards curve, which requires estimation of an additional parameter for shape. We accordingly used the logistic model:
Where W(t) is the weight at age t, A is the asymptote, r is the growth rate, and b reflects the weight at hatch (t = 0) relative to the asymptotic weight.
Initially we fitted the logistic model for each species and sex. We found that, within a species, the sexes had similar parameter estimates, except for the asymptote. We then fitted a single model for each species, with asymptote depending on sex. Estimates of the parameters obtained are presented in Table 4, and fitted curves are shown in Figure 1.
Table 4. Estimates of parameters of the logistic equation fitted to weights of Mallard, Gadwall, and Blue-winged Teal flightless ducklings. A is the asymptote, r is the growth rate, and b reflects the weight at hatch (t = 0) relative to the asymptotic weight. Standard errors are in parentheses. | |||
Parameter | Mallard | Gadwall | Blue-winged Teal |
A (females) | 1076 (10.0) | 785 (6.7) | 365 (3.2) |
A (males) | 1210 (7.2) | 840 (5.7) | 384 (2.7) |
r | 0.0692 (0.0030) | 0.0700 (0.0014) | 0.0750 (0.0045) |
b | 13.45 (1.58) | 10.40 (0.47) | 5.14 (0.78) |
For Mallard and Gadwall ducklings, residuals from the logistic model varied by year (P < 0.0001), but effects due to sex and the sex-year interaction were not significant (P > 0.10). For Blue-winged Teal, the main effect due to sex was not significant, but the year effect and sex-year interaction were significant (P < 0.005). Some, but not all, of the annual variation could be attributed to variation in wetland area or wetland change. Mallard and Gadwall tended to be heavier than expected in years when wetland area was extensive (P = 0.048, P = 0.0007, respectively) and when wetland area increased from the previous year (P = 0.003, P = 0.025). No such trend was evident for female Blue-winged Teal, and for male Blue-winged Teal only wetland change was significant (P = 0.001). Blue-winged Teal ducklings, however, were weighed in only three years.
Figure 1. Logistic curves fitted to weights of flightless Mallard, Gadwall, and Blue-winged Teal ducklings and projected to age 100 days, by sex. Weights are in grams and age is in days. Solid lines represent females and dashed lines represent males. |
The variation in weight of individuals within the same feather-age class was high (Tables 1-3). The coefficients of variation (standard deviation/mean) often exceeded 20% for Class II Mallard and 15% for Class II Gadwall and Blue-winged Teal. Potential sources of variation include measurement error in weighing, inaccuracy of ageing the birds from plumage, variation of ages within feather-age class, and true variation in weight among individuals of the same age.
Figure 2. Weights of known-age female Gadwall ducklings plotted against fitted logistic growth curve. Weights are in grams and age is in days. |
Some insight into the importance of this last source of variation was obtained by examining weights of 62 female and 50 male Gadwall ducklings of known age (numbers of the other species were too small to analyze). Ducklings of the same age and sex varied greatly in weight as shown by the display of female weight variation in Figure 2. High weight variability even occurred within broods. For broods recaptured 19 to 41 days after hatching, the average difference among female broodmates was 24 g (range 0-140 g), and among male broodmates was 46 g (range 0-95 g). The difference between broodmates regardless of sex averaged 57 g, and ranged from 0 to 180 g. Weights of known-age Gadwall ducklings led to another unexpected finding. Actual weights were significantly lower than values predicted from the logistic curve based on ages estimated from feather-age classes (P = 0.003 for females, P = 0.053 for males, Fig. 2).
Flying Mallard and Gadwall young of both sexes weighed less than the younger Class III young and less than the logistic model predicted. The differences between actual values (Tables 1 and 2) and the predicted values (Mallard female = 841.2 g and male = 945.4 g; Gadwall female = 597.7 g and male = 639.5 g) were consistent, but the overall test not significant (P = 0.12), probably because of small samples.
ASY Mallard females weighed significantly (P < 0.001) more ( = 1020 g, n = 173) than SY females ( = 973 g, n = 177). Average weights of females (ASY and SY were combined because patterns were similar for both ages) changed significantly through the breeding season (P < 0.001, Fig. 3). From a peak of 1104 g during the 30 March period, females declined steadily to 895 g during the 17 July period. During those 109 days, Mallard females lost an average of 209 g (19% of their initial weight), or 1.9 g per day. Weights of females captured on nests (n = 109) averaged 64 g lower (P < 0.001) than those of females captured during the same times but not on nests (n = 241). There was no relationship between female weights and wetland change or wetland area (P > 0.05), but there was significant annual variation in weights (P < 0.001, Fig. 4). Females were lightest in the drought year of 1977 ( = 951 g) and heaviest in 1976 ( = 1061 g) and 1978 ( = 1046 g).
Figure 3. Average weights (estimated population marginal means) of adult ducks by 10-day period of the breeding season, for three species. Circles denote females and squares denote males. |
Male Mallard weighed an average 1206 g (n = 660) and outweighed females during each period of the breeding season (Fig. 3). Male weights declined significantly (P < 0.001) during the early portion of the breeding season until the 20 May period. From the peak weight of 1277 g during 31 March period males lost an average of 132 g in 50 days (10%), or 2.6 g per day. After 20 May males generally gained weight through the end of June. Male weights showed significant (P < 0.004) annual variation (Fig. 4), much of which could be explained by the fluctuation in wetland condition between years. Male weights tended to be greater in years when wetland area increased from the previous year (P < 0.001). Male weights increased an average of 0.14 g for each added hectare of flooded wetland.
Figure 4. Average weights (estimated population marginal means) of adult ducks by year, for three species. Circles denote females and squares denote males. |
Weights of female Gadwall did not differ significantly by age (P > 0.05); the overall average was 722 g (n = 346). However, average weights of females did change significantly through the breeding season (P < 0.001). Their average weights increased from 747 to 781 g from April through the end of May, after which they declined to 651 g in the 9 July period (Fig. 3). From their peak weight near the end of May to the low on 9 July, average weights declined 130 g (17% of the average peak weight), or 3.3 g per day. Gadwall females captured on nests were significantly lighter (58 g less) than those captured at the same time but not on nests (P < 0.001). The annual weight variation of Gadwall females was largely associated with the change in wetland area from the preceding year (P = 0.004). For each hectare increase in wetland area, female weights increased an average of 0.15 g.
Male Gadwall averaged 835 g (n = 344), outweighing females by 16%. Average weights of males changed significantly (P < 0.001) through the breeding season (Fig. 3). From earliest captures through the 20 April period male weights averaged 873 g. Average weights increased to 904 g during the 30 April period and then declined to 783 g by the 9 June period, an average loss of 121 g (13% ) or 3.0 g per day. Weights of male Gadwall varied significantly from year to year (P < 0.001), with higher weights in 1976 and 1978 than in other years (Fig. 4). The annual weight fluctuation was not accounted for by variation in wetland area or wetland change.
Weights of female Blue-winged Teal did not differ significantly by age (P = 0.13); ASY and SY females had an overall average of 348 g (n = 255). Like Mallard and Gadwall, Blue-winged Teal females showed significant weight variation through the breeding season (P < 0.001, Fig. 3). Average female weights increased from 339 g during the 19 April period to 374 g in the 10 May period and then declined to 323 g in the 29 June period. Females thus lost an average of 51 g (14%), or 1.0 g per day between early May and late June. Female teal began gaining weight again during the 10-day interval in mid-July, after the nesting season. Females captured on nests weighed an average of 34 g less than those captured elsewhere at the same time (P < 0.001). Female weights were positively related to wetland area (P = 0.001); for each added hectare of wetland in a year hen weights averaged 0.20 g greater.
Male Blue-winged Teal weighed an average of 397 g (n = 313), or 14% more than the average female. Although male weights varied significantly through the breeding season (P = 0.006), changes were small and rather erratic (Fig. 3). Average weights increased from the 20 April period to the 30 April period, and generally declined until late June. From a peak average weight of 403 g during the 30 April period, male weights declined for the next 50 days but losses averaged only 13.8 g or 3%. Male weights were significantly related to wetland change (P = 0.021), but there was an additional effect due to year (P = 0.004, Fig. 4). Each additional wetland hectare was associated with an average weight increase of 0.16 g.