USGS - science for a changing world

Northern Prairie Wildlife Research Center

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

Homing and Reproductive Habits of Mallards,
Gadwalls, and Blue-winged Teal


Relationship of Hen Age to Arrival Date, Nesting Period, Clutch Size, and Breeding Success

Older resident mallard and gadwall hens probably benefit by arriving on the breeding grounds early. Early arrivals have an opportunity to select breeding habitat that may have enhanced food resources or increased likelihood of reproductive success (Dzubin 1969, Findlay and Cooke 1982). Sowls (1955) thought that resident ducks preceded transients at Delta Marsh, Manitoba, and Sellers (1973) observed marked hens among the first arrivals in southwest Manitoba. We did not find marked resident hens to be the first arrivals of all migrants on the study areas. Gates (1962) also noted that resident gadwall hens were not the first arrivals in Utah. It seems likely that the first hens to arrive on the study areas may be older, successful hens that were homing to locations farther north and the second earliest arrivals were older, resident hens.

SY mallard hens returned at a later date than ASY hens, but we found no difference in productivity between the 2 groups. SY gadwall hens were not a productive segment of the breeding population. SY gadwalls started nesting later and produced few young in comparison to ASY and ATY hens. We believe that a portion of SY gadwalls may not nest under certain conditions such as during dry years, based on the following circumstantial evidence:

  1. SY hens arrived at the study areas late in the year when nesting was well underway.
  2. Many SY hens did not initially return until their third summer.
  3. There were significantly fewer nests found for known-age SY hens compared to ASY and ATY hens (Table 6).

We suggest that SY hens that stopped at locations away from the study areas summered in non-breeding flocks. However, we cannot substantiate this hypothesis, and SY hens might have avoided adverse conditions on the study areas and nested at other sites in dry years.

Krapu and Doty (1979) and Dane (1965) found completed initial clutches of SY mallard hens and SY blue-winged teal hens to be significantly smaller than completed initial clutches of ASY hens. We detected no difference in clutch size between ASY and SY mallard and blue-winged teal hens. However, we did record a significant decline in clutch size for all 3 species as the nesting season advanced (see also Cowardin et al. 1985).


Mallard and gadwall nest densities were highest in those habitats with the highest and densest plant covers. Seeded nesting cover contained the highest density of mallard and gadwall nests, but it was less important for blue-winged teal. Teal selected shorter cover types and did not move far from wetlands to nest (see also Duebbert and Lokemoen 1980). On the Koenig study area, the home range of most blue-winged teal pairs would not intersect seeded nesting cover, and this cover type contained only 8% of the blue-winged teal nests, Dry wetlands, canal-side, and roadside, conversely, were distributed throughout the breeding ranges of most blue-winged teal pairs, contained acceptable nesting cover, and received high use by blue-winged teal.

Nest sites of returning mallard and gadwall AHY hens and their female progeny were closely associated, presumably because hens were responding to the previous successful nest site. Nesting colonies might develop as returning successful adult and HY hens nested within 250 m of the previous nest site. Gates (1962) recorded an average of 334 to between nests of the same individual in 2 successive years but did not report the influence of gadwall hen success on the distance.

Return Rates and Homing Rates

Our overall AHY mallard hen return rate of 0.333 and homing rate of 0.494 were similar to overall estimates from other studies (Table 16). Previous return-rate estimates, however, are a simplification of rates that vary considerably depending on previous hen success, hen age, and wetland conditions. Only Doty and Lee (1974), who studied mallards at nest baskets, reported homing rates separately for successful and unsuccessful hens. Although our return rates generally were higher than those reported by others, they were minimal figures. We believe our return rates were not severely biased, but hens that had lost their nasal marker and were not remarked, were not seen, or were not identified, undoubtedly returned to the study areas.

Table 16.   Reports of AHY and HY mallard hen return rates and homing rates.
Hen age and marker used No. marked No. returned Return rate Homing rate Reference
AHY hens a
   Leg band 15 2 0.133   Sowls (1955)
   Leg band 24 10 0.417   Coulter and Miller (1968)
   Nasal saddle          
      Successful 94 45 0.479 0.811 Doty and Lee (1974)
      Unsuccessful 19 3 0.158 0.268  
      Total or unweighted 152 60 0.297 0.539  
HY hens a
   Leg band 20 1 0.050   Sowls (1955)
   Leg band 122 7 0.057   Coulter and Miller (1968)
   Nasal saddle 306 89 0.291 0.430 Lee and Kruse (1973)
   Nasal saddle 1204 76 0.063 0.280 Bailey (1979)
   Nasal saddle 537 35 0.065 0.333 Gatti (1981)
      Total or unweighted 2,189 208 0.105 0.348  
a All AHY hens were birds from the wild and all HY hens were hand-reared birds.

Our overall AHY gadwall return rate of 0.485 and homing rate of 0.618 were higher than the average rate reported by previous studies (Table 17). The rates previously calculated by Sowls (1955) and Gates (1962) probably were underestimated because they used markers that were difficult to observe or were lost. In contrast to our study, Blohm (1979) found no difference in homing rates of hens that were successful or unsuccessful.

Table 17.   Reports of AHY and HY gadwall hen return rates and homing rates.
Hen age and marker used No. marked No. returned Return rate Homing rate Reference
AHY hens a
   Leg band 16 6 0.375   Sowls (1955)
   Neck band 52 15 0.288   Gates (1962)
   Nasal saddle 54 22 0.407 0.566 Blohm (1979)
Total or unweighted 122 43 0.357 0.583  
HY hens b
   Leg band 8 1 0.125   Sowls (1955)
   Nasal saddle 28 2 0.071 0.216 Blohm (1979)
Total or unweighted 36 3 0.098 0.216  
a All AHY hens were birds from the wild.
b Those HY hens studied by Blohm (1979) were birds from the wild, but the HY hens studied by Sowls (1955) were hand-reared birds.

Compared to older successful hens, HY mallard and gadwall hens returned at low rates of 0.288 and 0.087, respectively. Although low, these rates were higher than or similar to the unweighted average return rates of 0.105 and 0.098 reported by others for HY mallards (Table 16) and HY gadwalls (Table 17). The HY mallard return rates calculated by previous workers were with hand-reared birds, which probably had lower survival than wild birds (Hickey 1952).

Blue-winged teal hens did not exhibit strong tendencies to return to a specific breeding site regardless of previous success or age. From our study and work by Sowls (1955) and McHenry (1971), we concluded that most blue-winged teal have a low incidence of homing compared to mallards and gadwalls. In spring, most blue-winged teal hens do not home but assemble on water in the prairie region, even in Illinois following flooding (Kennedy 1974).

Mallards and gadwalls are relatively long-lived compared to blue-winged teal and have large overlapping breeding territories. If mallards and gadwalls return to a breeding site with fewer wetlands, these species can more readily adapt by moving long distances to fill needed food, resting, and nesting requirements. Blue-winged teal have small, mutually exclusive breeding territories, and the nest is located within the territory (Titman and Seymour 1981). If this species returns to a habitat with fewer wetlands, pair density would be high and many pairs might not be able to obtain sufficient resources to nest. These factors may be responsible for the low degree of homing by blue-winged teal and a large amount of movement to wet areas each spring, if severe drought conditions are present, even mallards and gadwalls may prefer breeding. Gadwalls were more affected by the drought than mallards, but we noted negative impacts on both species during the severe drought of 1977. Cowardin et al. (1985) working nearby also reported a reduced nesting effort by mallards in 1977, Mallards may be less influenced by poor conditions on the breeding grounds because they arrive in spring with nearly sufficient nutrients for laying eggs (Krapu 1981).

In South Dakota, Duebbert and Lokemoen (1980) observed blue-winged teal nest success above 80% in 1972, but the breeding population declined 90% in the spring of 1973 when pond numbers decreased. Blue-winged teal populations also can increase rapidly when wetland areas increase. Numbers of blue-winged teal per wetland hectare on the Koenig study area increased sevenfold from 1977 to 1978, after a year of near zero production.

The tendency for successful adult mallard and gadwall hens to return at a high rate appears to be an excellent strategy that benefits birds using a familiar and favorable habitat rather than unknown situations. Conversely, the tendency for unsuccessful adult hens to change nest locations seems to be a response that enables hens to bypass unproductive sites and test new ones. This system tends to enhance nesting at successful locations such as predator-free islands and to depress nesting at locations where success is low.

Many authors have reported increased mallard and gadwall populations at a variety of specific situations. In New York, Cowardin et al. (1967) found an expanded mallard population adapted to nesting in crutches and stumps of trees. In central North Dakota, Krapu et al. (1979b) located a successful population of mallard hens nesting over water in marshes. On small islands in North Dakota, Hammond and Mann (1956) found 450 gadwall nests per hectare, and Lokemoen et al. (1984) found 507 mallard nests and 307 gadwall nests per hectare. None of these studies involved marked birds; thus, investigators could not explain specifically how these high populations occurred. We suggest that populations are enhanced when pioneers (probably unsuccessful ASY and ATY hens and some SY hens) nest at sites with good wetland conditions, preferred nesting cover, and high egg and hen survival. The first year that seeded nesting cover was available on the Koenig stud) area, it composed 2% of the land area but contained 27% of the mallard and gadwall nests. Kruse (1979) found 98 mallard nests in 39 ha of seeded nesting cover in central North Dakota in 1978 following 2 years of drought and low production. These 2 nesting populations could not have resulted from homing of successful hens but only from concentration of hens at attractive sites.

Once a mallard or gadwall population is established, it might be sustained by homing hens. On the Koenig study area, the nesting population was not being maintained by homing hens although the mallard population was probably maintaining itself through recruitment. For example, from estimates of hen success, clutch size, and survival rates in this study we calculated that a population of 100 hens in summer year t would increase to 103 hens by 1 May in year t + 1 (Table 18). Due to a low rate of return by HY hens, there would be an insufficient number of returning mallards to maintain the breeding population. We estimated that hen success would have had to reach 75% before the population at the breeding site would have increased from homing alone. Hen success of 75% is a rare event in the natural environment (Cowardin and Johnson 1979). Thus, pair populations at these breeding sites were either maintained by pioneers or mallard homing was higher than we observed.

Table 18.   Calculated population change and homing expected in year t + 1 from 100 mallard hens with 25% hen success breeding on the study areas in year t.
Success and age categories No. present late summer year t Survival ratea No. alive 1 May tear t +1 Homing rate No. alive and homing year t + 1 No. alive and not homing tear t + 1
Successful adult 25.0 0.681 17.0 0.766 13.0 4.0
Unsuccessful adult 75.0 0.670 50.2 0.258 12.9 37.3  
Recruits producedb 64.4 0.560 36.1 0.515 18.6 17.5  
Total     103.3     44.5 58.8  
a These survival rates are explained in Methods.
b For production calculations we used averages and estimates from our data; recruits produced = 8.7 eggs/clutch x 0.87 hatchability x 0.681 survival to near fledging x 0.50 of young are hens.

The values we used to estimate mallard survival were based on annual survival rates obtained from a large sample of birds banded over a long period of time (Anderson 1975). These survival estimates may have been lower for ASY hens and higher for HY hens than that experienced by this sample of hens. As a result, our homing rate estimate of 1.101 for successful ATY hens was too high and our estimate of 0.515 for homing for HY hens may have been low (Table 7). HY gadwall hens bad low homing rates in part because they had high direct recovery rates and hens known to be alive did not home as SY birds. Also, HY hens may return less precisely to the breeding area or may be excluded from some habitats by adults. Both Blohm (1979) and Lokemoen et al. (1984) observe mainly old hens nesting at preferred island sites.

Duckling Survival and Return Patterns

We calculated survival rates of 0.681 for mallard young and 0.828 for gadwall young between hatching and near fledging. Our survival rates were higher than the 0.350 fledging rate reported for mallards by Talent et al. (1983) in North Dakota and the 0.440 reported by Ball et al. (1975) in Minnesota. The higher survival we observed may be related to factors such as quality of the brood environment, or differences among years, or study techniques. Our estimate of survival was probably minimal because it was derived by comparing the return rates of web-tagged hens (which had to be recaptured to be identified) with those of nasal-marked hens, These data suggest that previous duckling mortality estimates may be overestimated.

More mallard and gadwall broods than expected had 2 or more female members that returned, possibly because some hens were more effective in fledging young. Perhaps, young hens from a brood had a strong chance of returning together because there was some association between them during migration and wintering.

Direct Band Recovery Rates

Higher direct band recovery rates for HY and SY gadwall hens indicate that the young birds were more vulnerable to hunting than older birds, Gadwall ASY hens had low hunting mortality and apparently low overall mortality. on the Koenig study area, 3 of 6 ASY female residents marked in 1976 and 6 of 25 ASY hens marked in 1978 returned to the study area through 1981.

Direct band recovery rates for resident gadwall hens that nested successfully may have been higher than unsuccessful hens because successful hens survived the critical nesting period (Johnson and Sargeant 1977). Most of these hens would be alive in fall and available to hunters. Conversely, a portion of resident gadwall hens that were unsuccessful (no successful nest or brood) were killed on nests by predators and were not alive in fall and available to hunters. The occurrence of indirect band recoveries for unsuccessful mallard and gadwall hens that did not return to the study areas indicates that some of these hens survived to later years and were shot in other locations. The lack of indirect recoveries of successful hens that did not return indicates that nearly all hens successful in year t that survive until year t + 1 returned to the study areas.

Previous Section -- Results - Direct Band Recovery Rates
Return to Contents
Next Section -- Conclusions

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 19:31:43 EST
Menlo Park, CA [caww55]