Date of Award
2025
Degree Type
Thesis
Degree Name
Master of Science in Biological and Environmental Sciences (MSBES)
Department
Biological Sciences
First Advisor
Scott McWilliams
Abstract
Shrubland birds across North America have declined significantly in the last 40 years with the decline attributed to forest regrowth, natural disturbance suppression and anthropogenic pressures. As populations of these species decline, targeted conservation efforts have focused on maintaining and restoring the early successional habitats they depend on. Land management for American woodcock (Scolopax minor; hereafter woodcock) involves producing and maintaining a mosaic of early successional habitats including open fields and shrubland which can benefit a variety of songbirds that use similar breeding habitat. Eastern Towhee (Pipilo erythrophthalmus; hereafter towhee) and Prairie Warbler (Setophaga discolor; hereafter prairie) are two species that are assumed to benefit from management for woodcock, the designated umbrella species. Most assessments of such benefits from umbrella species management to non-target species only measure presence and abundance. My primary objective was to determine whether areas of high likelihood of use woodcock habitat, as identified by a resource selection function (RSF), were also high likelihood of use habitat for two songbird species, towhees and prairies. If woodcock are an effective umbrella species, then I expected high likelihood of use woodcock areas to support greater male densities, smaller territory sizes, higher nest survival, improved nestling growth rates, and less frequent adult provisioning for both species. The objectives of my thesis were to examine how likelihood of use (RSF) for woodcock influences key aspects of the breeding biology of towhees and prairies including (1) breeding male density and territory size, (2) nest success and daily nest survival, (3) nestling growth rates, fledging mass and size, (4) provisioning rates of adult towhees to nestlings, and (5) cowbird parasitism rates and how parasitism affected growth and fledging brood size of prairies.
In chapter 1, I describe the methods and results for towhees including how male density, territory size and nesting ecology are affected by likelihood of use for woodcock. During 2022 and 2023, 55 surveys of singing male towhees were conducted across six sites (3 high likelihood, 3 low likelihood) and we mapped 91 male towhee territories. RSF had no direct or indirect effect on towhee territory size but increased male density by 0.05 birds per hectare. I monitored 59 towhee nests with 36 succeeding (61%) to fledge young and 23 failing (39%) prior to fledging mostly due to predation (78%) or abandonment (22%). The mean daily survival probability was 0.96, leading to a 29% chance of surviving the 28-day nesting period. RSF had no effect on DSR (odds ratio: 1.11 95% CI: 0.64, 1.93). I estimated nestling towhee growth rates for body mass (n = 51 individuals) and tarsus length (n = 43 individuals) over the two years of the study. Nestlings in low-likelihood of use sites reached a lower asymptotic mass compared to those in high-likelihood sites. Similarly, nestlings in low-likelihood sites reached a lower asymptotic tarsus length and achieved this length later than those in high-likelihood sites. We used trail camera photos from 25 nests to assess provisioning rates. Provisioning rates of males and females did not differ for any chick age. Although total number of provisioning visits to nests by both adults increased with chick age, we did not find an effect of RSF on provisioning rate. Towhees largely benefited from forest management aimed at woodcock in that they occupied high-likelihood of use sites in greater densities, raised chicks with predicted higher fledging masses, potentially increasing recruits to the population.
In chapter 2, I describe the methods and results of how Prairie Warbler male density, territory size and nesting ecology are affected by likelihood of use for woodcock. During 2022 and 2023, 35 surveys for singing male prairies were conducted across five sites (3 high-likelihood, 2 low-likelihood) and we mapped 65 male territories. RSF had no direct or indirect effect on prairie territory size but significantly increased male density by 0.11 birds per hectare. I monitored 18 prairie nests for survival with 15 succeeding (83%) and 3 failing prior to fledging due to predation (16%). The mean daily survival probability was 0.98, leading to a 58% chance of surviving the 28-day nesting period. RSF was not a strong predictor of nest success (odds ratio: 1.14, 95% CI: 0.33, 3.96). Cowbird parasitism rates were slightly higher in high-likelihood of use (38%) compared to low-likelihood of use (30%), but Fisher's Exact Test showed no significant difference (odds ratio: 0.70, 95% CI: 0.08, 5.28, p = 1.0). We found no effect of likelihood of use on brood size at fledging age in unparasitized or parasitized nests, but final brood size was significantly lower in parasitized nests regardless of high or low likelihood of use. I estimated growth rates for nestling mass and tarsus length from seven nests (low-likelihood n = 5, high-likelihood n = 2) for 19 total individuals (low-likelihood, unparasitized = 9; low-likelihood, parasitized = 6; high-likelihood, parasitized = 4) over the two years of the study. Nestlings in parasitized nests at high-likelihood of use sites reached a lower asymptotic mass and exhibited a higher curve displacement, delaying growth compared to nestlings in parasitized and unparasitized nests at low-likelihood of use sites. Nestlings in parasitized and unparasitized nests at low likelihood of use sites had similar growth patterns reaching similar asymptotic masses. Nestlings in parasitized nests at high likelihood of use sites reached a smaller asymptotic tarsus length and exhibited a higher curve displacement, causing delayed growth compared to nestlings in parasitized nests at low-likelihood sites and unparasitized nests at low-likelihood sites. These nestlings reached a predicted smaller asymptotic tarsus length than unparasitized nestlings in low-likelihood sites due to later curve displacement and a higher growth rate coefficient, whereas nestlings in parasitized nests at low-likelihood sites showed a similar delayed growth pattern, ultimately reaching a smaller asymptotic tarsus. Even though the density of prairies was greater in high-likelihood sites for woodcock, prairies did not benefit from forest management aimed at woodcock in that several metrics of reproductive success were lower in high-likelihood sites for woodcock in part because of cowbird parasitism. My findings underscore the importance of considering multiple aspects of non-target species when assessing the umbrella species concept.
Recommended Citation
Gray, Megan, "CONTRASTING SUPPORT FOR AN UMBRELLA SPECIES: BREEDING ECOLOGY OF EASTERN TOWHEE AND PRAIRIE WARBLER IN HABITAT MANAGED FOR AMERICAN WOODCOCK IN RHODE ISLAND, USA" (2025). Open Access Master's Theses. Paper 2631.
https://digitalcommons.uri.edu/theses/2631