Summer Position with John Confer

 RESEARCH ASSISTANT FOR GIS MAPPING: ~5 May to 25 July (12 wks) to develop protocol to assess vegetative characteristics of territories of shrubland birds, especially the Golden-winged Warbler.

A LOEWY FELLOWSHIP:  ENHANCING MANAGEMENT OF SHRUBLANDS IN NEW YORK

 
    This project will use Geographic Information System (GIS) to map the vegetative structure in several shrubland habitats. My Loewy Fellowship will finance a field assistant who will create these detailed maps. (For a general description of the fellowship and the preserve see <http://www.mohonkpreserve.org/index.php?jobs-internship#loewy >) Shrubland habitats are extremely heterogeneous, which leads to a large sampling variance and makes it very hard to determine the sample size sufficient to characterize the habitat. The first goal of this study is to create GIS vegetation maps of patches of herb, shrub and tree vegetation (minimum patch size of ~3m², but most patches are much larger) with attributes for density and height. This portion of the work will be completed summer 2008. As time and skill permit, the GIS assistant and subsequent workers will apply computer-generated sampling to these maps to determine the sample size that will be adequate to detect statistically significant differences between habitats with and without a particular nesting species.

Many shrublands with extreme differences in plant species support the same species of birds. As examples, both the Golden-winged Warbler (Vermivora chrysoptera) (GWWA) and the Common Yellowthroat (Geothlypis trichas) (COYA) nest in upland successional fields and in tamarack bogs and in some swamp forests. Presumably there are some shared features that attract these birds to these diverse habitats, but it is not clear which vegetative attributes are key indicators. The computer-generated sampling of the GIS maps will also test for attributes that are the best predictors of the presence of a species using multiple, log-linear, logistic regression analyses.

SAMPLING DESIGN 

My major research site provides an ideal opportunity to map and characterize GWWA territories as part of habitat management for them in Sterling Forest State Park, NY (SFSP). In SFSP, golden-wings nest in an experimental shrubland restoration site, and in managed utility rights-of-ways (ROW), and in swamp forests. Very preliminary sampling suggests that the density of herbs and shrubs are similar in the wet and dry habitats and the density and patch distribution of herbs and shrubs may be key indicators.
           
           My colleague, Dr. Jeff Larkin, is a faculty member of Indiana University of Pennsylvania and a member of the Golden-winged Working Group. He also is concerned with management of habitat appropriate for GWWA. He is studying several sites in PA including a large burn area with a scattering of successional aspen in PA state wildlife lands. His study sites have nesting GWWA and a variety of shrubland birds. Similar measurements at the three NY sites and one PA site with GWWA will help establish the vegetative characteristics that are consistent predictors of this species.

           Mohonk Preserve has both successional fields (160 ac) and a Pitch Pine-Oak Heath Rocky Summit (90 ac) ecosystems that support shrubland birds, although not GWWA. Larger areas of these ecosystems exist throughout the protected Shawangunk Mnts. This study will sample six ecosystems at three research areas where identical measurements and analyses of vegetation can be used to distinguish attributes that are merely correlated with the presence of a species at one site from attributes that are strong predictors for the presence of a species at all sites.

My previous studies of the influence of habitat attributes on nesting success for GWWA in various shrublands (e.g., Confer et al. 2003), which used techniques that were less rigorous than those proposed for this study, have helped develop a gestalt for critical and non-critical attributes. The species of plant is likely of secondary importance because many species nest in habitats dominated by quite different plant species. The presence of water is not a good predictor because GWWA commonly nest in both wet and dry habitats (Confer 1992).

I believe the patchy distribution of vegetation in shrubland habitats is particularly significant. Extreme patchiness is characteristic of shrubland vegetation, often due to the vagaries of dispersal. Shrubs, for example, often expand from a single plant by rhyzomous growth, which creates distinct patches, while trees are highly concentrated or exclusively located near a forest edge due to the short dispersal of propagules. It is important to realize that dense vegetative patches of, e.g., herbs that cover 30% of an area are not ecologically equivalent to uniformly dispersed herbs even if both cover 30% of a territory. To describe and compare the vegetative categories, we will consider the mean density but emphasize the frequency distribution of patch size for herbs, shrubs, and trees.

This project will create highly detailed GIS maps of vegetation patches in the territories of GWWA and one other species that occurs in all six habitats. Mapping will be determined both within the territorial boundaries and also in adjacent areas beyond the boundaries. GIS map layers will indicate boundaries for the following categories and some quantitative attributes of the categories:

A.  areas dominated by herbs and forbs

            A1. the density of such cover measured on a categorical scale as

                                have used to predict nesting success (Confer et al. 2003).

A2.  the dominate plant species. Although I don’t expect this to be a

        good predictor of presence/absence, others have suggested that

         some plant species are critical and the concept should be tested

         with field measurements

B.     areas dominated by shrubs

B1.  the density of shrub growth

                        B2.  the shrub height

                        B3.  the shrub species

C.     areas with any tree canopy

C1.  the density of tree canopy measured on a categorical scale.

C2.  the tree height

C3.  the tree species

            D.  soil moisture

D1.  presence of aquatic vegetation suggesting standing water for at least

        the spring season.

D2.  standing water during much of the nesting season.

PRODUCTS OF LOEWY FELLOWHSHIP

Enhancing the Manual for Shrubland Management in New York. As described in my recent State Wildlife Grant ($150,000 over three years) proposal, I will “Compile a Manual for Statewide Management of Wildlife.  The PI has worked with several utilities throughout the northeast on several ROW studies of shrubland management and the avian community. The PI will work with DEC personnel and utilities to develop a partnership call for a conference on shrubland management for wildlife in our state. Proceedings of this conference will be compiled and edited by the PI as a Manual for Statewide Management of Wildlife”. I anticipate that the compiled proceedings of this conference will include many descriptive studies of habitat and the supported species. Much can be gained by such descriptions. Yet management may be more effective if we know, not just a description of the species present in one set of conditions, but actually know those key attributes that predict a species presence and the range of those attributes that are suitable for a species.
 

The Loewy Fellowship will enable me to conduct the initial tests of a protocol that will help characterize key vegetative attributes of shrubland habitats. With fellowship support, this project will support sampling of six different ecosystems, two of which are at the Mohonk Preserve. Overly-ambitions plans could spread the effort to thinly and yield ambiguous and non-productive results.  As a developmental step for this protocol, I will focus on one species that occurs in all six habitats, probably COYA, and the GWWA. Although the GWWA is not known to nest at Mohonk Preserve, inclusion of this severely declining species justifies pooling with the majority of funding from my other grants and makes this study possible. If results of the first year suggest key attributes that predict GWWA and one other species, then I will validate these predictions with tests in other territories and initiate studies of more species in subsequent years. My existing funding allows me some freedom of assign future students to vegetative sampling that will allow me to test the predictive ability of the model developed in the year funded by the Loewy Fellowship.

Contribution to the International Golden-winged Warbler Working Group.  The PI is uniquely able to compile recommendations for management for GWWA based on 30 years of study of GWWA and BWWA throughout New York, participation with the international Golden-winged Warbler Working Group, co-authorship of the federal Status Assessment in relation to the Endangered Species Act, and several publications on habitat and GWWA nesting success. The Golden-Winged Warbler Working Group is considering what procedures to use so that each team can characterize nesting habitat at their site in ways that provide meaningful comparisons. However, there are major differences in these suggestions even in something as basic as sample size. I will develop a computer simulation of sampling that will be applied to the GIS maps. I will use a computer to generate multiple sample sets for each of several sample sizes. Using repetitive tests for a statistical difference between sample sets taken within and outside of the territorial boundary, I will determine the frequency of Type I and Type II statistical error associated with different sample sizes. I believe the first year’s results will be sufficient to document the sample size necessary to characterize GWWA territories. The GWWA working group has scheduled a conference for next fall in the center of the winter habitat, in Bogotá and San Vicente de Chucurí, Colombia and I anticipate presenting this aspect of the results of the Loewy Fellowship at that meeting.

Contribution to the Mohonk Preserve.  Sampling in successional fields and in the Pitch Pine Oak Heath Rocky Summit (as defined by the NY Natural Heritage Program)  ecosystems of the Mohonk Preserve will contribute to our understanding of why specific shrubland birds occur there and which shrubland birds may occur as succession or management changes the habitat. These studies are just in time to contribute to the formulation of management plans for all of the old fields in the Preserve, as supported by a different State Wildlife Grant. The proposed quantitative measurements combined with casual inspection of ROW and other shrublands in the Shawangunk Mountain landscapes may provide some guidance for avian management for the 40,000 acres of fire-influenced, successional habitats in the entire physiographic region. Quite significantly, the Loewy Fellowship will provide funding essential for development of a new protocol to characterize shrublands that will enhance the management of GWWA and A Manual for Shrubland Management in New York while collaborative among the PI, the Mohonk Preserve and the others groups with whom the PI is associated may have unpredictable benefits.

SCHEDULE

           During summer 2008, I will work full time for 14 weeks on shrubland birds with a majority of my time at Sterling Forest State Park. For this summer, I propose to have a supervisory role for a research assistant who will work for 12 weeks partially supported by the Loewy Fellowship. This assistant will being work in late April or early May, 2008. This person will need to be quite familiar with GIS, and be an experienced field worker with moderately good birding skills who is capable of working alone for part of the study. I will be at Mohonk Preserve for 1 week of field study with the field assistant, who will be at Mohonk for 2.5 weeks of field measurements and 1.5 weeks of data analyses. The field assistant will spend about 2.5 wks in PA and about 5.5 weeks at Sterling Forest.