USGS - science for a changing world

Northern Prairie Wildlife Research Center

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

Description and Identification of American
Black Duck, Mallard, and Hybrid Wing Plumage


Key for the Identification of Hybrid Wings

We developed a key from our analyses (Table). We are confident the key facilitates discrimination among wild Black Ducks, Mallards, and their hybrids. In general, most wings will be obvious forms of domestic Mallards, Mallards, or Black Ducks; only a few wings will require closer scrutiny to determine if they are from hybrids.

Table.  A dichotomous key for identification of wings from American Black Ducks, Mallards, and hybrids between the two. Like-numbered couplets will permit only one choice; numbers in parentheses indicate the next couplet to compare.
  1. Wing distinctly unlike those figured for Black Ducks and Mallards by Bellrose (1976: plates 6 and 7) including but not limited to: gray underwing (Smithe 1974a: colors #83-#86); white in areas other than bounds of speculum (colored portion of secondaries) on the dorsum; extremely faded feather areas or colors not usually observed in American Black Ducks and Mallards, especially shades of black mixed with white for a piebald effect. Actual size of the wing is immaterial, although in many domestic breeds the ratio of length to width seems out of proportion compared with wild birds, and the wing may be extremely large in some breeds. In sum, resembling any bird figured by Delacour (1964:154-166)...DOMESTIC MALLARD OR DOMESTIC CROSS
  1. Not as above...(2)
  1. Wing Mallard-like on dorsal surface including two complete white bars bounding the speculum. This white forms a broad white terminal bar on the secondaries and a similar subterminal bar on each greater covert...(3)
  1. Wing not Mallard-like on dorsal surface. One of the two white bars bounding the speculum may be lacking, incomplete, or indistinct; wing usually darker than most Mallard wings...(6)
  1. Wing Mallard-like; white bar anterior to speculum terminates at the proximal edge of the speculum: male...(4)
  1. Wing Mallard-like; white bar anterior to the speculum extends onto the greater tertial coverts: female...(5)
  1. Dark brown coverts absent on the undersurface (lining) of the wing. Gray vermiculated feathers are permissible; the brown in question does not appear on the Mallard wing but is similar to that observed on Black Duck dorsal wing surfaces...MALLARD
  1. At least one dark brown feather on lining, usually on leading edge and often near bend of wing...HYBRID
  1. Dark brown feathers (<10, small) present on the underwing, usually along the leading edge...MALLARD
  1. Dark brown feathers (>10, small or medium-sized) on the underwing, usually along the leading edge and near the alula... HYBRID
  1. Wing Black Duck-like; white subterminal bar present on >1 greater covert anterior to speculum; white terminal bar on secondaries may be present or lacking...HYBRID
  1. Wing Black Duck-like; white subterminal bar on greater coverts lacking. White on posterior edge of speculum (tips of secondaries) may be complete, incomplete, or lacking...(7)
  1. Dark brown feathers (< 10, small) present on underwing...HYBRID
  1. Dark brown feathers (>10, small to medium) present on underwing...BLACK DUCK

Identification of Domestic Mallards

Wings from domestic Mallards, even if the birds are feral and free-flying for several generations, confound attempts to estimate hybrid incidence in the wild Black Duck and Mallard population through parts collection surveys. Unfortunately, complete descriptions of domestic forms of the Mallard are not available, even though useful discussions and some illustrations may be found in Jaap (1934), Lancaster (1963), Delacour (1964), Batty (1985), and Kear (1990). For identifying wings of otherwise unavailable birds, it is easier to state what a wild Mallard wing looks like (i.e., Palmer's 1976 definition) than what it is does not. With reference to Palmer's definition, this means that all Mallard-type wings with large amounts of white, unusual colors, or overall dark underwings, etc., must be removed at the first step in a hybrid key. Persons familiar with domestic Mallards with whom we discussed this step concurred with us that common domestic Mallard strains are easily identified as are many uncommon ones, such as the recessive melanistic form known as the "Dusky Mallard." Characteristics listed above for the domestic Mallard can invariably identify domestic forms, particularly if care is taken to examine the underwing for overall dark coloration and dark spotting (F. B. Lee, Emeritus Biologist, Northern Prairie Wildlife Research Center, personal communication.)

Limitations of the Wing Key

Despite many attempts to qualitatively and quantitatively differentiate plumages in the Mallard group, the sexual dichromatism in the Mallard and the similarity of the southern forms to one another and to female Mallards have largely defeated such efforts. For example, Palmer (1976:314) described both sexes of the Mottled Duck as superficially alike and quite similar to Black Ducks, but with general coloring not quite as dark as Black Ducks (though darker than female Mallards). He emphasized the presence of a white bar only at the trailing edge of speculum, and even then often obscure or lacking, especially in Florida birds. Differences from Black Ducks were described as "a matter of degree." Similarly, Palmer (1976:309) described both sexes of the Mexican Duck as resembling female Mallards in general appearance but with "darker, richer browns," especially on the breast. Palmer described white on the dorsal wing surface of both sexes of the Mexican Duck as a bar at the leading and another at the trailing edge of the speculum, or variably, an indistinct "brownish or dusky" bar.

These and similar descriptions are inadequate unless one has a reference collection. Thus, keys for identifying members of the Mallard group (e.g., Bellrose 1976:72) usually suggest that knowledge of the geographic origin of the specimen is necessary to make positive identification. To demonstrate the confusion that might result from applying the hybrid key to other members of the Mallard group, we inspected 237 specimens collected from 1872 to 1929 and originally identified by the collectors as Anas fulvigula, A. f. fulvigula, A. f. maculosa, Anas diazi, or A. d. novimexicana. Many of the specimens examined met the hybrid criteria if we ignored other characters of the plumage. This hybrid identification was especially true regarding Mottled Ducks, which in this sample had a mean of 17 dark feathers in the underwing, and many Mexican Ducks that did not have the expected two white bars bounding the speculum. We conclude that our key should be used only for specimens obtained from portions of the continent where Mexican and Mottled Ducks are unlikely to be encountered: North Carolina and north in the Atlantic Flyway and Arkansas and Tennessee north in the Mississippi Flyway (unless birds obtained from the Louisiana and Mississippi coastal areas where Mottled Ducks are found can be separated from birds shot elsewhere in these latter states).

What Proportion of the Hybrids in a Sample Can Be Identified?

The rank order of abundance of hybrids in any freely-mixing population of Black Ducks, Mallards, and their hybrids will be P1 > F1 > F2 > F3 ... > Fn. This order is dictated because F2, F3 ...Fn hybrids are very rare as a consequence of the F1 itself being rare compared to the parental populations. In other words, most F1 matings would be a backcross to one or the other of the parental forms in subsequent years instead of inter se (Merrell 1981). As a result, hybrids detectable in a wild population of Black Ducks and Mallards would be F1 birds and a large but unknown fraction of the F1 backcrosses to P1. Both the absolute number and proportional representation of F2, F3 ...Fn, and other combinations, whether detectable or not, would be quite small.

Our data showed that 82-96% of the male progeny and 10-80% of the female progeny of backcrosses to P1 will be detectable, depending upon the P1 in the cross. Because wings from many of the backcrosses of hybrids to either P1, especially females, defy identification with any plumage key, inability to detect all backcrosses prevents determining the true proportion of hybrids in the wild from even large samples of wings. It seems imprudent to generate a correction factor to account for these undetectable hybrids resulting from F1 × P1 matings since we lack information on the proportion of F1 that mate with Black Ducks versus Mallards. Thus we do not know whether misidentification is as a Black Duck or a Mallard. Regardless, the wing characters employed are capable of identifying a large and constant fraction of the hybrids present in a sample, namely 100% of the F1, at least 82% of male progeny of backcrosses to either P1, and 80% of the female progeny of backcrosses to Black Ducks, but perhaps as few as 10% of the female progeny of backcrosses to Mallards. The hybrid key and the North American parts collection surveys certainly can be used to generate an index to error in mate choice of the two parental populations. We recommend it for this use as well as for identifying birds in hand during banding operations.

Reinterpretation of the 1977 U.S. Parts Collection Survey

We reexamined the 4,608 wings obtained in the 1977 U.S. Parts Collection Survey for the Atlantic Flyway originally identified as Black Duck, Mallard, or hybrid between the two. With the new criteria presented in our key (Table), we found that the number of hybrids had been underestimated by at least a factor of 2.37 (237 wings versus 100 wings). This underestimation would change the proportion of hybrids/(hybrids + Black Ducks) from 0.056 to 0.132, the proportion of hybrids/(hybrids + Mallards) from 0.034 to 0.078, and the ratio of hybrids/(hybrids + Black Ducks + Mallards), which is the incidence of overall hybridization, from 0.022 to 0.052. When we applied our new criteria, we identified 34 (34%) of wings classified as hybrids as domestic strains, Mallards, or Black Ducks. We reclassified as hybrids 136 (8%) of the wings originally identified as Black Duck. Similarly, we identified as hybrids 34 (1.2%) of the wings originally classified as Mallards. The most important conclusion for management is that hybridization between the Black Duck and Mallard occurred at about twice the rate heretofore reported for the Atlantic Flyway in 1977.

We could not make a similar comparison in any subsequent year. Our work generated substantial interest among participants in the "wing bees" (cooperative evaluations of parts collection survey receipts) in identifying Black Duck × Mallard hybrids. Managers of the U.S. Parts Collection Survey modified the criteria used to identify hybrids after 1977, and we confirmed that subsequent wing bees have identified larger percentages of hybrids among the available wings. Managers of the Canadian Parts Collection Survey likewise emphasized recording Black Duck-like and Mallard-like hybrids.

Comparison with Another Quantitative Measure of Hybridization

Barnes (1989) presented another key for identifying F1 hybrids between the Black Duck and Mallard. He concluded that the area of the color band on the fifth greater secondary covert (whether white or some other color) was a measure that discriminated between the P1 forms and the F1 hybrids, and the length of the white band at the tip of the fifth secondary correctly identified 19 of 22 hybrids from a sample of P1 and F1 birds.

As is generally the case with F1 hybrids, Black Duck × Mallard F1 are uniform in phenotype and intermediate between the parental forms in most characters. Our review has shown that subterminal white on the secondary coverts is not a Black Duck character. Barnes's criterion for area of "color" on the secondary coverts is thus the same as our presence/absence criterion for white in these feathers except that he requires a measurement. We believe that a presence/absence criterion is preferable because dimensions and amount of white in this color band will be distributed approximately normally (as are most other morphological characters) in a sample. This normal distribution can be seen in the intermediate frequency distribution of this character for hybrids between the frequency distributions for Mallards and Black Ducks in Barnes (1989:Fig. 2).

Further, we have shown that the amount of terminal white on the secondaries of Black Ducks is variable and is a poor criterion of genetic origin. It is particularly poor as a criterion when applied to F2 and subsequent generations that are quite mutable because of Mendellian segregation and recombination, and which may theoretically include the extremes of the P1 populations. The frequency distributions of white on the wings of birds raised in the experimental flock in Jamestown (Fig. 8) clearly show the intermediate nature of the F1, the expected broader range of values of the F2 (because of genetic segregation and recombination), and the shift in F1 × P1 crosses in the direction of the P1 phenotype. Barnes's key may be relied upon for experimental work with F1 and P1 known-parentage birds, but our key (Table) should be used for all birds of unknown genetic origin.

Fig. 8. Frequency distribution of the width of white bars bounding the speculum on the wing dorsum of Black Ducks, Mallards, and hybrids from an experimental flock reared at Northern Prairie Wildlife Research Center. All measurements are compared with Mallard =100%; actual range in width 0-10 mm.

Previous Section -- Results
Return to Contents
Next Section -- The Need for Additional Research

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 19:11:29 EST
Reston, VA [vaww54]