Reproductive success of wood thrushes in forest fragments in Northern Indiana

Auk, The, Jan 2000 by Fauth, Peter T

ABSTRACT.-I monitored 278 Wood Thrush (Hylocichla mustelina) nests in 14 forest fragments (range 7 to 500 ha) in northern Indiana to explore patterns of brood parasitism by Brown-headed Cowbirds (Molothrus ater), nest predation, and reproductive success. Density of thrushes was negatively related to area of forest fragments. Cowbirds were common throughout the landscape, but I found no relationship between their abundance and forest area. Overall, 90% of the thrush nests were parasitized by cowbirds (x = 2.4 cowbird eggs per nest). The number of cowbird eggs per thrush nest was not related to forest area, abundance of host species, or distance to a forest edge but was positively related to thrush abundance. Nest predation rates averaged 58% and were lower than those reported in much of neighboring Illinois but also were unrelated to forest area and distance to a forest edge. The combination of cowbird parasitism and nest predation resulted in relatively low reproductive success (x = 0.6 thrush fledglings per nesting attempt). I estimated that Wood Thrushes in northern Indiana made an average of three nesting attempts per breeding season (based on 17 color-marked females) and had relatively low seasonal fecundity ff = 0.9 female fledglings per adult female per season). Nonetheless, considerable annual variation in seasonal fecundity suggested that some sites exceeded the source-sink threshold in some years. Regardless, the overall landscape appeared to be part of a regional sink for Wood Thrushes, although its negative influence on regional demography was not as severe as elsewhere in the midwestern United States. The poor demographic balance in much of the agriculturally dominated Midwest suggests that conservation efforts for Wood Thrushes and other Nearctic-Neotropical migrants should be directed at preserving and enhancing possible source habitats in regions where parasitism by cowbirds and nest predation are reduced. Received 9 October 1998, accepted 8 July 1999.

TWO GENERALIZATIONs have emerged from recent research on the breeding success of Nearctic-Neotropical migrant landbirds (hereafter, "Neotropical migrants"). First, population surveys of Neotropical migrants conducted on a local scale may be difficult to interpret without knowledge of breeding behavior and demography (Brawn and Robinson 1996). Second, effects of forest fragmentation on Neotropical migrants differ among geographic areas, particularly between the eastern and midwestern United States (Robinson et al. 1995).

Although bird surveys can be valuable tools for assessing regional population trends of Neotropical migrants (e.g. Robbins et al. 1989, Hagan et al. 1992, Peterjohn et al. 1995, James et al. 1996), such methods are based on observations of singing males that may not provide reliable indices of local breeding densities (Gibbs and Faaborg 1990, Gibbs and Wenny 1993, Porneluzi et al. 1993). In the extreme, Brawn and Robinson (1996) found no negative population trends based on local bird surveys conducted in the Midwest, yet reproductive success was below levels necessary to maintain viable populations in the absence of immigration. For many Neotropical migrant species, surplus individuals from highly productive source habitats may immigrate into sink habitats where reproduction fails to balance adult mortality (Pulliam 1988, Pulliam and Danielson 1991). One result of such source-sink dynamics is a time lag between the collapse of local breeding success and the detection of population declines based on local bird surveys (Brawn and Robinson 1996). Thus, the combination of measuring breeding success on a local scale and surveying populations on a regional scale may be necessary to evaluate long-term population viability of Neotropical migrants.

Several studies in the eastern United States have shown positive relationships between nesting success and forest fragment size (e.g. Porneluzi et al. 1993, Hoover et al. 1995). Reduced breeding success in small forest patches may result from edge effects because Brownheaded Cowbirds (Molothrus ater) and mammalian and avian predators occur at higher densities within edge habitats (Brittingham and Temple 1983, Wilcove 1985, Paton 1994). Nonetheless, breeding success in moderatesized fragments (>100 ha) and continuous forests in the eastern United States appears to be sufficient to maintain viable populations of Neotropical migrants (Holmes et al. 1992, Porneluzi et al. 1993, Hoover et al. 1995). In contrast, even relatively large forest fragments in the Midwest (200 to 2,200 ha) do not support viable populations of Neotropical migrants (Donovan et al. 1995, Trine 1998). As a result, Brawn and Robinson (1996) suggested that the agricultural portion of the Midwest acts as a sink in which all woodlot populations must be bolstered by constant immigration from areas with extensive forest, such as southern Indiana, northern Wisconsin, or southeastern Missouri.

I tested the regional-sink hypothesis (Brawn and Robinson 1996) by measuring the breeding success of Wood Thrushes (Hylocichla mustelina) in an agricultural landscape in northern Indiana that rivals the degree of forest fragmentation reported by Robinson (1992) and Brawn and Robinson (1996). 1 monitored Wood Thrush nests and conducted local bird surveys within 14 forest patches to (1) investigate relationships among densities of breeding birds, cowbirds, and reproductive success of Wood Thrushes; (2) measure the intensity of predation and cowbird parasitism on Wood Thrush nests; (3) describe variation in nesting success of Wood Thrushes among and within forest fragments; and (4) estimate the seasonal fecundity of Wood Thrushes within forest fragments. My study extends the geographic breadth of our knowledge about the reproductive success of Wood Thrushes, documents substantial annual variation in nesting success, and uses the renesting frequency of colormarked females to estimate seasonal fecundity.