USGS - science for a changing world

Northern Prairie Wildlife Research Center

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

Influence of Age and Selected Environmental Factors
on Reproductive Performance of Canvasbacks

Results


Weather and Habitat Conditions

Spring phenology and environmental conditions varied widely among years (Table 1). May pond densities were highest in 1974 and declined to a low of 19/km² during the drought of 1977. Pond densities improved in 1978 and 1979 but declined again in 1980. May pond estimates for southern Manitoba in 1980 were similar to those that occurred in 1977 and were the lowest since 1961. Also, July pond conditions were at near-record lows in 1977 and 1980 (Novara and Voelzer 1987).


Arrival Date

Canvasbacks arrived on the study area from mid-April through mid-May (Table 1). Over all years, 50% of the marked birds arrived by 1 May and 90% by 6 May. The earliest spring arrivals occurred in 1977 when 50% of the birds arrived by 21 April. In 1979, the latest spring, 50% of the birds did not arrive until 7 May.

For all age classes combined, we found a negative linear relationship between arrival date and minimum April temperatures (F = 12.96; 1, 143 df; P < 0.001). Within age classes, however, the relationship was not apparent except for SY females (F = 3.72; 1, 143 df; P = 0.055). Over all years, SY females arrived an average 5.3 days later than older birds (F = 14.59; 1, 143 df; P < 0.001). Within years, this age-related difference was significant only in 1977, the warmest spring (F = 9.73; 1, 143 df; P = 0.002), when SY birds arrived an average of 7.7 days later than older birds.


Return Rates

Return rates of female cohorts (Table 2) were age-related (χ² = 70.45; 2 df; P < 0.001). Second-year females returned at lower rates than TY (χ² = 27.55; 1 df; P < 0.001) or ATY birds (χ² = 63.78; 1 df; P < 0.001). Overall, return rates averaged 21% for SY, 58% for TY, and 79% for ATY females (or 69% for ASY with TY and ATY combined). Annual return rates for ASY females ranged from 48% in 1977 under poor pond conditions to 84% in 1978 when pond conditions were good. By the time a marked cohort reached 3 years old, on average, only 8% of the birds originally marked returned to their natal area.

Table 2.  Yearly return rates (%) of female canvasbacks marked as juveniles on the Minnedosa Study Area, Manitoba, 1974-80.
Years marked n Years following marking a
1 2 3 4 5 6
1974 31 16
(5/31)
40
(2/5)
100
(2/2)
100
(2/2)
100
(3/3)d
75
(3/4)d
1975 58b 19
(9/48)
42
(8/19)
88
(7/8)
90
(9/10)a
73
(8/11)d
     
1976 39 23
(9/39)
78
(7/9)
67
(6/9)
63
(5/8)
          
1977 0                                                 
1978 63c 24
(13/54)
68
(15/22)
                                              

Mean of X (%)

191 21
(36/172)
58
(32/55)
79
(15/19)
80
(16/20)
79
(11/14)
75
(3/4)
a Number returned from those observed in the previous year is shown in parentheses.
b Return rates for greater-than or equal to years includes 10 females marked as SY's.
c Return rates for greater-than or equal to years includes 9 females marked as SY's.
d Includes marked females not observed in the previous year(s).

The best log-linear model incorporating age, return rates, and previous hen success (G² = 1.26; 3 df; P = 0.74) indicated that older females had higher return rates and were more successful in the previous year than younger hens. Return rates of ATY females varied according to hen success the previous year (χ² = 4.29; 1 df; P = 0.04) but return rates of TY females were independent of their success as SY birds (χ² = 0.50; 1 df; P = 0.48). Number of May ponds explained little of the variation in return rates of different age classes (r ² = 0.06; 0.11, and 0.06 for the 3 age classes, respectively; P > 0.50).

Of 70 marked individuals that returned during the study, 12 (17%) were not observed for periods of 1-3 years but were observed in later years. This indicates that a few individuals either did not return to the study area in some years, or were present but not observed. Eight of these 12 individuals (67%) were initially marked as juveniles (14% of 57 marked as juveniles).


Nesting

Nest Initiation Date. — Canvasbacks initiated nests as early as 29 April and in most years continued nesting into mid-June (Fig. 1). Of the environmental factors evaluated, only minimum April temperatures affected nest initiation date of early nests (H0: slope = 0, T = -11.32; P < 0.001). Regression slopes of minimum April temperatures and nest initiation dates did not differ (F = 0.28; 2, 65 df; P = 0.75) among age classes indicating that all age classes responded similarly to minimum April temperatures. Initiation of early nests averaged 2.36 days earlier for every 1 C increase in minimum April temperatures. When the effect of minimum April temperature was removed using covariance analysis, nest initiation date differed among ages (F = 5.73; 2, 67 df; P = 0.005). Second-year birds initiated nests later than did older birds (F = 6.56; 1, 67 df; P = 0.01).

Figure 1
Figure 1.  Mean canvasback nest initiation dates and range in the Minnedosa Study Area, 1974-80. Early nests (E) were initiated during the first 15 days of known nest initiations. Later nests (L) were initiated after this period. Too few nests occurred in 1977 to derive a meaningful average or range.

Clutch Size. — Canvasback clutch size averaged 8.15 0.19 (Mean of X SE, n = 133) eggs. A significant amount (F = 61.90; 3, 132 df; P < 0.001) of the variability in clutch size was accounted for by nest initiation date. When the effect of nest initiation date was removed using covariance analysis, clutch size did not differ (F = 2.55; 3, 132 df; P = 0.08) among age classes. Each 10-day delay in nesting resulted in a decrease of 1 egg in the clutch. This relationship did not differ (F = 0.72; 5, 132 df; P = 0.49) among age classes. Clutch size did not differ (F = 0.51; 1, 129 df; P = 0.48) between nests with or without eggs laid by parasitic hens (8.0 0.2 and 8.3 0.3 eggs, respectively).

Interspecific Nest Parasitism. — Of all canvasback nests parasitized by other species, 81 (96%) were parasitized by redheads, 3 (3%) by redheads and another species, and 1 (1%) by a ruddy duck (Oxyura jamaicensis). Rates of interspecific nest parasitism (Table 3) did not differ between SY and older birds (χ² = 2.33; 1 df; P = 0.13) nor were there differences among nests initiated early, mid, or late in the nesting season (χ² = 0.87; 2 df; P = 0.352). Parasitism rates did not differ between moderate (66%) and wet (73%) years (χ² = 0.11; 1 df; P = 0.74) but were lower in dry (28%) years (χ² = 8.10; 1 df; P = 0.004).

Table 3.  Rates of interspecific nest parasitism (%) of canvasbacks on the Minnedosa Study Area, Manitoba, 1974-80.
Year Age class Mean of X

n

SY na ASY n
1974     57 7 57 7
1975 100 4 67 3 86 7
1976 64 14 64 22 64 36
1977     0 3 0 3
1978     61 18 61 18
1979 92 14 68 34 75 48
1980     33 15 33 15
Mean of X 81 32 58 102 63 134
a Number of nests.

The annual average number of parasitic eggs in parasitized canvasback nests was 3.2 0.22 and ranged from 2.0 in 1974 to 4.0 in 1975. The incidence of nests with ≥4 parasitic eggs averaged 24% over all years and exceeded 30% only in 1975, when 43% of the nests were parasitized.

Hatching Success. — Hatching success of canvasback eggs found in canvasback nests did not differ by age (F = 1.74; 11, 53 df; P = 0.09), nor did the success of redhead eggs (F = 1.72; 9, 39 df; P = 0.12). Hatching success was 0.78 0.03 (n = 65) for canvasbacks and (0.77 0.05, n = 36) for redheads. In successful nests that contained eggs of both species, hatching success for canvasbacks (0.847 0.002) was higher (F = 6.13; 1, 48 df; P = 0.02) than for redheads (0.673 0.005). Hatching success of canvasback eggs was 0.87 0.05 in unparasitized nests and 0.77 0.04 in parasitized nests, respectively, but did not differ statistically (F = 0.01; 1, 63 df; P = 0.92).

Nest Success. — Nest success did not differ between SY and older birds (χ² = 2.12; 1 df; P = 0.15), but did differ among age-year groups when 3 age classes were considered (χ² = 26.84; 13 df; P = 0.01) (Table 4). Nest success did not differ (χ² = 2.31; 1 df; P = 0.13) between parasitized and unparasitized nests. Furthermore, it was independent (χ² = 4.39; 2 df; P = 0.11) of the number of parasitic eggs deposited. Nest success did not differ (χ² = 0.13; 1 df; P = 0.72) between moderately wet (54%) and wet (60%) years but was lower (χ² = 11.54; 1 df; P < 0.001) in dry years (17%) than in other years. However, neither May pond numbers (F = 0.08; 3, 5 df; P = 0.97) nor the decline in wetland numbers from May to July (F = 0.08; 3, 5 df; P = 0.51) were significant in predicting nest success.

Table 4.  Rates of apparent nest success (%) of canvasbacks on the Minnedosa Study Area, Manitoba, 1974-80.
Year Age class Mean of X n
SY na ASY n
1974     14 7 14 7
1975 25 4 33 3 29 7
1976 43 14 54 22 50 36
1977     0 3 0 3
1978     72 18 72 18
1979 57 14 71 34 67 48
1980     20 15 20 15
Mean of x 47 32 53 102 52 134
a Number of nests.

Of the 84 unsuccessful nests, 15 (18%) were abandoned and 69 (82%) were destroyed. Two (13%) of the abandonments were attributed to parasitism, and 13 (15%) were attributed to investigator disturbances during nest trapping. Of those destroyed, 41 (59%) were destroyed by mammals, 2 (3%) by avian predators, 9 (13%) by investigators, 1 (1 %) by a laying redhead, and 16 (24%) by unknown causes. Nest predation rates did not differ (χ² = 0.28; 1 df; P = 0.09) between SY and older birds. Predation rates in dry years (67%) were higher (χ² = 14.21; 1 df; P < 0.001) than those of other years (22%), but wet years had somewhat higher (χ² = 3.97; 1 df; P = 0.05) predation rates (34%) than moderately wet years (18%). However, neither May pond numbers (F = 0.01; 3, 5 df; P = 0.99) nor the decline in wetland numbers from May to July (F = 0.20; 3, 5 df; P = 0.89) were significant in predicting nest predation rates.

Of 12 renests identified during the study, 1 was by a SY female, 2 by TY females, and 9 by ATY females. No renesting was recorded in the drought years of 1977 or 1980.

Hen Success. — Hen success was 17-18% higher in older females than SY birds (χ² = 4.00; 1 df; P = 0.05) (Table 5). Hen success varied by year (χ² = 21.12; 6 df; P < 0.001) among ASY females and was lower (χ² = 15.20; 1 df; P < 0.001) in dry (15%) years than in other (64%) years.

Table 5.  Rates of hen success (%) of canvasbacks on the Minnedosa Study Area, Manitoba, 1974-80.
Year Age class Mean of X n
SY na ASY n
1974     25 4 25 4
1975 33 3 50 2 40 5
1976 50 12 67 18 60 30
1977     0 3 0 3
1978     76 17 76 17
1979 57 14 75 32 70 46
1980     20 15 20 15
Mean of X 52 29 59 91 58 120
a Number of nests.

No evidence of nesting was apparent during dry years in 1977 for 13 of the 17 birds (76%) and in 1980 for 4 of 17 (24%). During other years, all females were known to nest except one female in 1978, when no nesting effort was observed for this individual. When dry years were excluded, the overall nesting rate among females rose to 99%. Nesting in dry years (combined) appeared to increase with age; 0 of 2 SY birds, 4 of 10 TY birds (40%), and 13 of 22 ATY birds (59%) nested. Nest records of all marked birds support this trend during dry years: 12 of 15 nests (80%) in 1977 were ATY females, and 3 of 4 nests (75%) in 1980 were ATY females. No nests were found for SY females in 1977 or 1980.


Duckling Survival to ≥20 Days

Observations of marked females with broods in 1975 were insufficient for evaluation, and no broods were observed in 1977. Of 76 observations of marked females with broods in the remaining years, female canvasbacks raised an average of 4.7 0.3 ducklings to ≥20 days of age each year. The number of ducklings surviving to ≥20 days was unaffected by female age (F = 1.14; 2, 6 df; P = 0.33), years (F = 0.74; 4, 6 df; P = 0.57), or their interaction (F = 0.54; 2, 6 df; P = 0.71).


Nesting Experience

The influence of previous nesting experience on hen success was examined by using nesting histories (ranging up to 4 yr) of 24 individuals. We found hen success to be independent of the number of years of previous nesting experience (χ² = 2.877, 3 df, P = 0.411). However, these results may be confounded by yearly environmental conditions, which were shown to affect both nest and hen success among years.


Previous Section -- Study Area and Methods
Return to Contents
Next Section -- Discussion

Accessibility FOIA Privacy Policies and Notices

Take Pride in America logo USA.gov logo U.S. Department of the Interior | U.S. Geological Survey
URL: http://www.npwrc.usgs.gov/resource/birds/canvage/results.htm
Page Contact Information: Webmaster
Page Last Modified: Friday, 01-Feb-2013 18:36:54 EST
Sioux Falls, SD [sdww54]