Sex Ratios and Measurements of Migrant Goshawks

Abstract

In a recent paper (Mueller and Berger, 1967) we showed that it is the younger members of the population of Goshawks (Accipiter gentilis) that invade areas south of the breeding range during autumn and winter. We are now able to report that there also appears to, be a differential migration of the sexes. For some years we have been confident that an experienced worker could determine the sex of a living Goshawk simply by noting the relative size of the bird. We were reluctant to publish our findings based on sex determinations made by size alone because data from museum specimens usually indicated considerable overlap between the measurements of the sexes. Storer (1966) has recently provided an independent confirmation of the validity of our techniques of sex determination by showing that considerable sexual dimorphism exists in the size of Goshawks and that museum collectors often err in determining the sex of hawks. In the autumns of 1951 through 1964 we live-trapped, examined, and measured 105 Goshawks at the Cedar Grove Ornithological Station in southeastern Wisconsin. Wing chord was measured by placing the carpal joint of the closed wing on a metric rule placed on a table edge and pivoting the wing downward until the tip of the longest primary touched the rule. Pressing the wing flat provided a measurement of wing arc. The length of the tail was measured by inserting a thin metal rule between the central rectrices and sighting across the tips of the two longest rectrices. The birds were placed head downward in upright metal cylinders of appropriate diameter and weighed to the nearest gram on a triple-beam balance calibrated to 0.1 g. The weight of the contents of the esophagus (crop) was estimated and subtracted from the gross weight. An examination of the plumage permitted us to place birds into three age classes: (1) Juvenals, (2) Adult I, including birds more than one but less than two years old, and (3) Adult II, including all birds more than two years old (see Storer, 1966; Mueller and Berger, 1967). We determined the sex of each bird by a series of approximations. All but seven birds were tentatively assigned to a sex class by qualitative judgment at the time of capture. The measurements of wing chord and tail length for all Goshawks handled, irrespective of age class, were graphed. For both measurements the resulting distribution resembled two normal curves. The curve with the smaller mean was composed of birds qualitatively judged to be males; the larger, females. The seven undetermined birds fell largely into the zone of overlap. These seven birds were somewhat arbitrarily assigned to one or the other of the sexes, and the