Avian removal experiments: do they test for habitat saturation or female availability?

Ecology, April, 1997 by Peter P. Marra, Richard T. Holmes

Territorial behavior during the breeding season has long been considered an important mechanism in determining the local distributions and abundances of many bird populations (Hinde 1956, Brown 1969, Fretwell and Lucas 1970, Watson and Moss 1970, Klomp 1972, Newton 1992). This assumes that habitats vary in quality and that, in years when bird density exceeds the carrying capacity of the more suitable habitats, some individuals are forced by already territorial individuals into suboptimal areas where reproductive success is compromised, or they become floaters and are prevented from breeding at all (Brown 1969). Territorial behavior has, therefore, been thought to act as a mechanism that limits local abundances (Fretwell and Lucas 1970, Newton 1992). Demonstrating habitat saturation, competitive exclusion, and the presence of a surplus population (i.e., floaters) has usually been attempted by experiments in which territorial individuals are removed and their vacated territories are monitored for replacements (see reviews by Klomp 1972, Newton 1992). In this paper, we present experimental results and theoretical arguments indicating that removal experiments, at least for species with particular life history and mating strategies, fail to demonstrate habitat saturation or even the existence of floaters. Alternative hypotheses, such as how the availability of females influences male behavior, need to be given more consideration.

The classical test for habitat saturation and the existence of floaters in avian studies involves removing territorial individuals and documenting whether or not replacement occurs. If new individuals settle into experimentally vacated territories, it is concluded that a lack of suitable habitats, combined with territorial behavior, limits breeding opportunities (Klomp 1972, Newton 1992). Most experimental manipulations, however, especially those with passerine birds, have removed only males from territories, leaving females (e.g., see Hensley and Cope 1951, Stewart and Aldrich 195 l, Thompson 1977, Arvidsson and Klaesson 1984, Hogstad 1989, Sherry and Holmes 1989; see Newton 1992 for additional references). Because females were left on the territories in these cases, an alternative explanation for the observed reoccupation by males is that they were attracted by the presence of these widowed females and not necessarily by the availability of habitat.

To test this hypothesis, we conducted removal experiments with Black-throated Blue Warblers (Dendroica caerulescens) breeding in a northern hardwoods forest, using three treatments: (1) removal of both the male and female, leaving the territory vacant; (2) removal [TABULAR DATA FOR TABLE 1 OMITTED] of only the male, leaving the female; and (3) removal of only the female, leaving the male. We predicted that if there are male floaters present due to habitat saturation, then replacement should occur upon removal of both the male and female from the territory, i.e., in a completely vacated territory. Alternatively, if the presence of floaters is due to a shortage of females, then replacement should occur only when the male is removed, leaving the female. Finally, if nonbreeding (floating) females are present in the population, then they should settle in territories where resident females are removed. The latter finding would indicate a surplus of females, i.e., that females are not limiting.

Methods

This study was conducted in the Hubbard Brook Experimental Forest in the White Mountain National Forest, West Thornton, Grafton County, New Hampshire, USA. The forest is second-growth northern hardwoods, with the canopy dominated by sugar maple (Acer saccharum), American beech (Fagus grandifolia), and yellow birch (Betula allegheniensis), with some white ash (Fraxinus americanus) and red spruce (Picea rubens). The forest understory is composed primarily of the saplings of the canopy trees, along with hobblebush (Viburnum alnifolium), striped maple (A. pensylvanicum), and mountain maple (A. spicatum) (Bormann and Likens 1979, Holmes et al. 1986). Removals were conducted in areas with a dense shrub layer, the preferred habitat of Black-throated Blue Warblers (Holmes et al. 1992, 1996).

Black-throated Blue Warblers were removed from territories that were located [greater than or equal to]300 m, and often [greater than]500 m, apart in an undisturbed section of forest in the west end of the Hubbard Brook valley. Neighbors of removed birds were not color-banded, which limited our ability to discriminate between true floaters and expansion by neighbors. However, falsification of our hypothesis only required us to determine that reoccupation did not occur when both males and females were removed. To aid in classifying replacement individuals as either territorial neighbors or floaters, we did a pre-and postremoval census of all singing males neighboring the manipulated territory. These data, combined with territory maps of replacement birds (i.e., neighbors could often be followed back to their territories), provided some indication of the status of any new birds that appeared in the vacated territory. All territorial pairs were randomly assigned to one of three treatments: (1) removal of both the male and female, leaving the territory vacant; (2) removal of only the male, leaving the female; and (3) removal of only the female, leaving the territorial male. In total, we permanently removed individuals from 27 spatially separated territories (see Table 1 for details on number removed per treatment per year). Birds were collected with a 0.410-gauge shotgun, and specimens were deposited at Dartmouth College. Removals took place in early- to mid-June, during the incubation period and after spring migration was complete (Holmes et al. 1992, Holmes 1994). This timing of removals ensured that any birds reoccupying vacant areas were not late migrants, and that replacement birds would still have time to nest prior to late-summer departure.