UVM Theses and Dissertations
Format:
Print
Author:
Shustack, Daniel P.
Dept./Program:
Wildlife and Fisheries Biology Program
Year:
2004
Degree:
M.S.
Abstract:
Bobolinks and savannah sparrows are grassland songbirds experiencing widespread population declines, in part because of intensifying agricultural practices on hayfields. Knowledge of the factors that influence selection of fields by these species early in the season and after widespread mowing is critical to devise management plans for these species. To assess temporal variation in vegetation, field, and landscape characteristics of habitat use, 4 point counts were conducted on 142 fields between 21 May 2002 and 16 July 2002. Counts were made in agricultural fields of perennial cover in 6 counties in the Champlain Valley of Vermont and New York. Using an information theoretic approach (Burnham and Anderson 2002), 10 candidate models (6 single process models and 4 combination models) were constructed, fit to the data, and ranked using Akaike Information Criteria for small sample sizes and with quasi-likelihood adjustments (QAICc). Bootstrap selection frequencies were also calculated based on the proportion of times a candidate model was ranked first out of 1,000 bootstrap trials.
Field selection by bobolinks appeared relatively constant throughout the breeding season, with landscape features being most important. The distant (2500 m) landscape features apparently drove field selection early in the season, but other factors contributed as well. Although there was little support for vegetation and field models alone, they were included in a highly supported combination model. After mowing began in visit 2, the pattern from visit 1 weakened, and closer landscape features (500 m) were more important than the distant landscape features (2500 m). There was still essentially no support for vegetation of field models alone. In visit 3, distant landscape features apparently once again drove selection. However, bobolinks avoided mowed fields and many bobolinks apparently exited the breeding population after mowing occurred on the fields they selected initially.
Field selection by savannah sparrows also appeared relatively constant over the breeding season with close (500 m) features of the landscape and characteristics of the field being most important, although the order of the model rankings varied across visits. The vegetation model was never supported and heavily parameterized combination models were over fit for savannah sparrows. Savannah sparrows appeared to have some attraction to mowed fields and likely attempted breeding in cut and uncut fields following mowing in their originally selected fields. Management for these species in agricultural grasslands should focus on areas of the landscape that are non-forested and undeveloped at broad (2000 ha) scales, yet provide clusters of perennial herbaceous cover, within 80 ha. Where possible, field sizes should be increased and woody field edges reduced. Although these features may attract these species to certain fields, differences in behavioral responses to mowing by the two species make general recommendations about mowing management difficult. However, within-season, temporal changes in landscape patterns must be specifically addressed in management plans.
Field selection by bobolinks appeared relatively constant throughout the breeding season, with landscape features being most important. The distant (2500 m) landscape features apparently drove field selection early in the season, but other factors contributed as well. Although there was little support for vegetation and field models alone, they were included in a highly supported combination model. After mowing began in visit 2, the pattern from visit 1 weakened, and closer landscape features (500 m) were more important than the distant landscape features (2500 m). There was still essentially no support for vegetation of field models alone. In visit 3, distant landscape features apparently once again drove selection. However, bobolinks avoided mowed fields and many bobolinks apparently exited the breeding population after mowing occurred on the fields they selected initially.
Field selection by savannah sparrows also appeared relatively constant over the breeding season with close (500 m) features of the landscape and characteristics of the field being most important, although the order of the model rankings varied across visits. The vegetation model was never supported and heavily parameterized combination models were over fit for savannah sparrows. Savannah sparrows appeared to have some attraction to mowed fields and likely attempted breeding in cut and uncut fields following mowing in their originally selected fields. Management for these species in agricultural grasslands should focus on areas of the landscape that are non-forested and undeveloped at broad (2000 ha) scales, yet provide clusters of perennial herbaceous cover, within 80 ha. Where possible, field sizes should be increased and woody field edges reduced. Although these features may attract these species to certain fields, differences in behavioral responses to mowing by the two species make general recommendations about mowing management difficult. However, within-season, temporal changes in landscape patterns must be specifically addressed in management plans.