Does nest placement affect the fate or productivity of California Gnatcatcher nests?

ABSTRACT.-I monitored 384 California Gnatcatcher (Polioptila californica) nests at five study areas in western Riverside County; California, from 1992 through 1995. Overall, 78.9% of the nesting attempts failed (52.9% from nest predation), and 21.1% fledged young. Vegetation measurements were taken at 207 of 384 nest locations and at 207 randomly selected locations within gnatcatcher territories. Nest placement was not random in that gnatcatchers placed their nests in locations with greater cover and height of perennial vegetation, increased horizontal structural homogeneity, and increased vertical structural homogeneity relative to random locations. The volume of gnatcatcher nest shrubs was lower than that of random shrubs, and concealment was higher at nests than at random shrubs. Nevertheless, variables that differed significantly between nests and random locations were not significantly associated with nesting success, nest predation, nest abandonment, or productivity. The use of specific perennial shrub species as nest substrates was not proportional to the availabilities of the shrub species within the study areas, yet I found no relationship between nesting success and the species of nest shrub. Nesting success appeared to be unrelated to the abundance of perennial shrubs at four of five study areas. My results suggest that factors affecting the fate of California Gnatcatcher nests operate at scales larger than the nest site or nest patch. Received 12 August 1998, accepted 29 January 1999.

NEST PREDATION is the primary cause of nest failure for a wide diversity of avian species (Ricklefs 1969, Martin 1988). Because nest predation affects fitness directly, natural selection should favor life-history or nest-placement characteristics that decrease predation. Furthermore, nesting success may be affected by factors operating at different or even multiple spatial scales (Martin 1992, Sedgwick and Knopf 1992). Therefore, understanding how nest placement affects nest fate is an important step toward understanding nest-site selection and potentially the structure of avian assemblages. Accordingly, many studies have found that nest failure, usually due to predation, will affect the nest placement of land birds (Peterson and Best 1985, Martin and Roper 1988, Holway 1991, Morton et al. 1993). However, studies have also found that nest placement does not always appear to influence nest predation or nesting success (Holway 1991, Filliater et al. 1994, Howlett and Stutchbury 1996, Wilson and Cooper 1998).

The California Gnatcatcher (Polioptila californica) is a small shrub-nesting passerine endemic to the coastal sage scrub of southern California and northern Baja, Mexico. The coastal sage scrub community is a Mediterranean-type habitat typified by facultatively drought-deciduous mesophilic shrubs from 0.5 to 2.0 m tall (Westman 1981, Mooney 1988). The California Gnatcatcher is listed as a federally threatened species because of habitat loss.

I examined the effects of nest placement on success, predation, abandonment, and productivity of California Gnatcatcher nests to test the hypothesis that an open-cup nester that experiences high nest failure, mostly due to predation, has nest-placement characteristics that increase nesting success or productivity and decrease nest predation. Specifically, I used nests and random habitat measurements to define variables that may be important to nest placement at two spatial scales and then used those variables to determine if nest placement affects nesting success and other productivity measures. In addition, I examined whether nesting success is related to the orientation of the nest within the shrub, slope aspect of the nest shrub, height of the nest, or species and abundance of the nest shrub.

STUDY AREA AND MEHTODS

I monitored 384 California Gnatcatcher nests at five study areas in western Riverside County, California, from 1992 through 1995. Thirty-five nests were monitored at the North and South Hills of the Domenigoni Valley in 1992 (hereafter referred to as the Hills study area), 47 nests were monitored at the University of California Motte Rimrock Reserve in 1993 and 1994, 53 nests were monitored at Lake Mathews Site I in 1993 and 1994, 60 nests were monitored at Lake Mathews Site 2 in 1994 and 1995, and 189 nests were monitored at Lake Skinner from 1992 through 1995. All study areas were dominated by the Riversidian coastal sage scrub subassociation (Westman 1981, O'Leary 1990), but vegetation cover and composition were heterogeneous among study areas. For a map and detailed descriptions of the study areas see Braden et at. (1997a, b).

Nest fates. -Nests were visited at intervals of three to five days. The frequency of nest checks was consistent among study areas and throughout the duration of the study. Nests were considered successful if one or more young fledged (i.e. left the nest). Because nestlings were banded when they were eight days old as part of a life-history study, the exact or at least approximate fledging dates were known. Each nest could be unambiguously attributed to specific pairs because all adults were uniquely color banded. Nests were considered lost to predation if some or all of the eggs disappeared, eggs were damaged, or eggs or egg shell fragments occurred in the vicinity of the nest, as long as the adult birds no longer attended the nest, Shredded or damaged nest structures were always interpreted as predation. Nests were classified as abandoned when there was no indication of predation and the nest was no longer attended by the adults. Because nest parasitism by Brown-headed Cowbirds (Molothrus ater) was shown to have little influence on nesting success of gnatcatchers at these study areas (Braden et al. 1997b), cowbird parasitism was not considered further in this study. The number of depreciated nests was probably underestimated because the disappearance of an egg during egg laying, but before the nest was checked, could have resulted in the nest being classified as abandoned.