SREL Reprint #0022




Density, Racial Composition, Sociality, and Selective Predation
in Nonbreeding Populations of Savannah Sparrows

R. A. Norris



In the Savannah River Plant area, on the Upper Coastal Plain of South Carolina, the Savannah  Sparrow (Passerculus sandwichensis) is the most abundant bird inhabiting old fields in the nonbreeding period (late September to early May).  Its density in winter may average four or five birds per acre, or more--up to 30 per acre--in particularly favorable fields. The short-term home range of a Savannah Sparrow was estimated to be about eight acres, but there was evidence that individuals exhibited greater agility over a several-week period. In such a period, then, a substantially larger number of sparrows would frequent a given area than would frequent it at any particular moment or short interval of time. The above-given density values refer to short-term populations.

The wintering populations include five subspecies or races: the "light races" mediogriseus, savanna, and nevadensis, and the "dark races" labradorins and oblitus. These breed in different geographic regions, the dark ones as far north as the Ungava Peninsula and Hudson Bay.  Data on the relative abundance of the races based on 1758 identifications (mostly of birds netted for banding and examined in the hand) made in the winters of 1955-56 and 1956-57, are presented for different seasonal periods. A case is made for the practice of making sight identifications, with binoculars, of the five races in question.  This practice, which seems quite valid, serves to supplement and extend the other studies in “field taxonomy,” which refers, in part, to the identification of infraspecific categories in live birds handled under field conditions.

From early January to May, 1957, special attention was given to a seminatural population of sparrows which, included all the races except nevadensis (the least common). The birds in this population were wing-clipped, color-banded, and kept inside a low plastic fence enclosing one acre of old-field vegetation (where natural food, chiefly seeds of the grass Digitaria, was particularly abundant). The area of the circular enclosure exceeded the home range" of any one of the confined birds, and it is proposed that the term beta-confinement be used to distinguish between this situation and that referring to birds in small enclosures or cages (alpha-confinement). In the enclosure the non-flying birds mingled freely with flying individuals of the same species. The density of fliers in the enclosure showed unusually sharp fluctuations which were thought to represent a "crowding effect." By means of the Petersen-Lincoln index (ratio of number of non-fliers to fliers as obtained by random counts) it was possible to estimate repeatedly the density of fliers in the enclosure and to compare results with those of other methods, notably a transect method, employed in adjacent areas.  Density values due to the two methods were very nearly the same.  Whereas measures of spatial relations of sparrows in the enclosure suggested near-random patters of distribution, it is possible that improved methods of investigation will reveal in many instances significant degrees of contagiousness involving twosomes, threesomes, or larger assemblages of birds.

Aspects of behavior of color-banded sparrows were observed from a blind near supplies of food and water, which were provided intermittently at the center of the enclosure. The presence of a rather loose social hierarchy soon become apparent; among the dominant individuals, certain males belonging to dark races were especially well represented. Many sparrows, however, displayed no signs of hostility, and nowhere, either within the enclosure or without, as there indication of social or ecological segregation among the different races. Of the confined birds, some of the measurable characteristics of individuals and racial groups were peripheral tendency {tendency to occupy relatively peripheral parts of the enclosure), conspicuousness (tantamount to tameness or "viewability" by the human observer), and ecological longevity (absolute longevity under beta-confinement weighted in accordance with degree of danger imposed by predators, during an individual's sojourn, on the surrounding, confined population).

The wing-clipped birds, numbering 30 at the end of February, suffered heavy predation by an owl or owls in early and middle March. In their capacity to survive for relatively long periods under such hazardous circumstances, individuals of the race labradorius showed marked superiority over those of the other races. There was no evidence that any race was more "protectively" colored than the others. The most "successful" race labradorius was characterized by individuals that were relatively large-sized, socially dominants and moderately conspicuous. There was evidence that the heaviest predation on the nonflying population was density-dependent and took place in moonlit periods during which many of the sparrows were engaged in northward-oriented, nocturnal, migratory unrest. The last of the flightless birds was taken about May 20. Two methods of describing the impact of a predator or predators on a prey population are considered; a formula expressing the "index of intensity of predation" is suggested inasmuch as it seems especially applicable to the predator-prey relations revolved in the present study.

It is believed that seminatural, beta-confined populations of birds--particularly small species that inhabit ground or low-herb strata---offer a wide range of possibilities for observation and experimentation. Such studies will help to bridge the gap which exists, at least in some ecologists' minds, between "natural" and "experimental" populations.


SREL Reprint #0022 

Norris, R.A. 1960. Density, racial composition, sociality, and selective predation in nonbreeding populations of Savannah sparrows. Bird Banding 31:55 57.


To request a reprint