A Protocol for the Assessment of Biodiversity Hotspots for RI Municipalities

 
Nancy Cofer-Shabica, Aimee Mandeville, Alyson McAnn, Peter August, & Duane Chapman
Department of Natural Resources Science
University of Rhode Island
Kingston, RI 02881
 

Habitat loss is the single-most significant cause for the loss of local biodiversity. The legal authority to convert natural habitat to developed land occurs at the scale of the individual land parcel. The arena for these decisions is the Town Hall. Town zoning law dictates how land can be developed. Aquifers, recharge areas, wetlands and locations of regulated species are (or can be) clearly mapped, and therefore included in zoning regulations. Regions critical to support biodiversity are usually not mapped and are rarely (if ever) included in land use regulations. In theory, zoning regulations should reflect the need to protect critical natural resources such as groundwater, wetlands, rare and endangered species, and overall biodiversity.
 

We have developed a simple protocol to estimate lands critical to support regional biodiversity using "off-the-shelf" GIS data. The ultimate product of our analysis is a map showing general regions of high importance in support of biodiversity. Our "mapematical" definition of "biodiversity" includes data on wetlands, habitat for rare species, natural land covers, large tracts of forest, sources of disturbance (roads, residential areas), and lands adjacent to protected areas and critical habitats. We will illustrate our process using the Towns of Narragansett and West Greenwich, Rhode Island as examples.
 

COMPARISON OF THE POSITIONAL ACCURACY OF WETLAND DELINEATIONS USING DIGITAL ORTHOPHOTOGRAPHY AND AERIAL PHOTOGRAPHY

Jeffrey Barrette^1,2, Peter August¹, and Francis Golet¹

¹Department of Natural Resources Science
University of Rhode Island
Kingston, RI 02881
²ESRI, Inc. [Olympia, WA]
606 Columbia St. NW Suite 213
Olympia, WA 98501-1099

Our study compared the horizontal accuracy of forested wetland boundary delineations obtained from natural color aerial photography against delineations obtained from ‘heads-up’ digitizing of digital orthophotography. The distances from each of the derived boundaries (n=128) to the field-located ‘true’ wetland boundary were measured using global positioning system (GPS) and geographic information system software. The mean absolute value distance (expressed in feet + 1 standard deviation) between the field-derived ‘true’ wetland boundary location and the orthophotograph-derived wetland boundaries (11.26 + 11.31 feet) was significantly different (Z = -2.53, p < 0.05) from the mean distance between the ‘true’ boundary and the aerial photo-derived wetland boundaries (14.87 + 13.28 feet). Despite the statistical differences in horizontal accuracy between orthophotograph- and aerial photograph-derived wetland boundaries, three feet on the ground is not significant in terms of ‘true’ wetland boundary delineation. Procedurally, wetland delineation using heads-up digitizing with digital orthophotography was much faster and considerably simpler than the process required with aerial photography.

LANDSCAPE CORRELATES OF BREEDING BIRD DIVERSITY IN RHODE ISLAND
 

Aimee Mandeville & Peter August

Department of Natural Resources Science
University of Rhode Island
Kingston, RI 02881

The Atlas of Breeding Birds in Rhode Island (Enser, 1992) consists of a statewide grid network (24 km² per cell) with data on the level of certainty of breeding for each species comprising Rhode Island’s avifauna. In total, 165 grid cells occur wholly, or partially, in the state. We entered the Breeding Bird Survey data into GIS format and for each grid cell we computed a number of measures that reflect the quality of habitat available to breeding birds or levels of human disturbance. The fundamental purpose of this study was to examine relationships between breeding bird diversity and landscape-level measures of habitat quality. We computed a number of measures of bird diversity (e.g., total number of species, total number of passerines, shore birds, etc.). Our independent variables consist of two classes of measures -- aerial measures (e.g., total amount of forest, habitat diversity) and landscape/patch metrics (e.g., patchiness, amount of edge). For all analyses, we used residuals from linear regression analysis to standardize habitat measures and bird diversity data to the amount of area of each grid cell falling within Rhode Island (vs Connecticut, Massachusetts, ocean). We used correlation and regression methods to discern the relationships between habitat and bird diversity.

Patch Size and Scale Asymmetry: Limits for Detecting Wildlife Habitat Selection

Peter August, Matt Nicholson, and Nancy Andrew

Department of Natural Resources Science
University of Rhode Island
Kingston, RI 02881
 

Radio telemetry is frequently used to locate the position of animals in wildlife ecology research. In most circumstances, there is a certain amount of error about each telemetry location. Sources of error include distortion of radio transmitter signals from rough topography, human error in locating the radio signal, and errors associated with deriving a geographic coordinate for a fix (e.g., extrapolation from a map, LORAN, GPS). For example, the empirically-derived 90% circular error probability (CEP) radius for telemetry fixes of mule deer in chaparral habitat in southern California was 177m (Nicholson, 1994) and the 95% CEP for mountain sheep in the Sonoran desert was 1,000m (Andrew, 1994). Geographic Information Systems (GIS) is an obvious tool to assess habitat relationships using telemetry data. Commonly used habitat datasets in such analyses include terrain, vegetation, hydrography, land use, and distributions of key resources (e.g., water holes, nest sites, sources of disturbance). The minimum discernible patch size of habitat data can vary immensely. For example, aspect data derived from a 1:24,000 DEM using GRID will consist of 900 m2 patches; whereas, vegetation data taken from 1:24,000 aerial photography might yield habitat patches with minimum polygon sizes measured in square kilometers! The purpose of our study is to determine the extent to which scale differences (or asymmetries) in siting error and patch size render habitat analyses impossible. We are using ARC/INFO's GRID module to simulate conditions of varying patch size, CEP size, and levels of habitat selection. For each unique suite of parameters, we are measuring the point at which variation in the data overwhelm patterns resulting from habitat selection.

Very Large Scale Digital Orthophotography as a Foundation for a Municipal GIS
 
Mary Hutchinson¹, Margaret Pilaro², Jeffrey Barrette¹, Peter August¹, and Jonathan Stevens²

¹Department of Natural Resources Science
University of Rhode Island
Kingston, RI 02881

² Division of Planning
City Hall
Warwick, RI 02886

We created a prototype GIS database for the City of Warwick, Rhode Island using 1:1,200 true color orthophotography as a planimetric base. Assessors maps and zoning data were digitized and fit to the orthophoto base. A full suite of parcel-based attributes were appended to the plat/lot polygons. A number of new datasets were developed anew from the orthophotos and were included in the information system (e.g., building footprints, roads). An ArcView project was created to access the data and the software/data package was put on PC systems in the Planning Office, the Public Works Office, and the Sewer Authority. We are evaluating the utility of the system, and based on this evaluation, will implement an ortho-based GIS for the entire City.

The National Park Service Field Technical Support Center Program: Integration of Missions and Resources.
 

Peter V. August

Department of Natural Resources Science
University of Rhode Island
Kingston, RI 02881.

 
The Field Technical Support Center (FTSC) for Geographic Information System (GIS) assistance to the National Park Service New England Cluster resides at in the Department of Natural Resources Science at the University of Rhode Island. The purpose of the FTSC is to provide GIS and spatial analysis infrastructure for NPS technical staff and to provide assistance to the National Park Units in establishing and using GIS. The success of the FTSC model is contingent upon significant overlap of the missions of the academic host of the FTSC and the Park Service. The GIS Division of the NPS Northeast Field Area is responsible for the coordinated introduction of GIS technology among the region's 100+ Parks, Historic Sites, Battlefields, and Recreation Areas. The mission of the URI Department of Natural Resources is to conduct research and training using contemporary technologies and the scientific method to address environmental problems. The University has access to undergraduate and graduate students, excellent computing infrastructure, and immediate access to professional counsel in a variety of disciplines. The Park Service has a large number of projects requiring state-of-the-art approaches to land management, conservation of fauna and flora, regional planning, and historic preservation. The critical overlap between URI's academic mission and the NPS FTSC model is basic: the NPS provides real-world problems in need of attention and the University has technology and students willing and able to spend time exploring solutions to these problems.

Establishing GIS Capacity Within the National Park Service Field Units
 

Charles L. LaBash

Department of Natural Resources Science
University of Rhode Island, Kingston, RI 02881.
 

A primary objective of the National Park Service (NPS) and University of Rhode Island GIS Field Technical Support Center project is to develop GIS capacity within many of the NPS Field Units in the Northeast Region. The size and focus of NPS sites is very diverse and includes large natural areas (e.g., Acadia and Cape Cod National Seashore), historic parks (e.g., Minuteman and Saratoga Historical Parks), and major recreational parks (e.g., Gateway National Recreational Area). Some of the Parks have well-managed, functional GIS systems in operation, some have extensive databases but little in-house infrastructure or staff support, but most have no GIS capacity at present. We are developing GIS capacity for a number of Parks and this involves: (1) assembling existing data from federal, state, county, and municipal sources; (2) reformatting data to meet NPS GIS cartographic standards (e.g., datum, coordinate system); (3) developing metadata; (4) packaging the database with "user-friendly" software to access the data; and (5) providing training for new and prospective users. Developing GIS capacity for NPS Field Units has proven to be a complex and ambitious undertaking. However, it is an essential element in empowering NPS natural resource managers and planners with the information and technology required to exercise quality stewardship over the lands and historic or recreational resources under their care.

Relationships between geomorphological variation and biotic diversity: A multi-scale assessment.

 
August, P., K. Killingbeck, W. Nichols, M. Burnett, and J. Brown.

Department of Natural Resources Science
University of Rhode Island, Kingston, RI 02881, USA.

 
We tested the hypothesis that areas with high variation in geomorphological properties would show high biodiversity as compared to regions with low variation in geomorphology. We used changes in terrain (slope and aspect), and soils (drainage class) to represent geomorphological diversity. In one study, we compared the diversity of woody plants with geomorphological variation on 2 hectare study grids in a 500 hectare red oak/white pine forest complex. Mean species richness and diversity of shrubs and trees was significantly greater (t-test, p < 0.05) on study quadrats of high variation in slope and terrain as compared to quadrats characterized by low variation. Approximately 50-60% of the total variation in plant diversity can be explained by terrain and soil properties. In a second study conducted at a landscape scale, we measured plant species richness in 24 Audubon Society refuges located throughout the state of Rhode Island. Variation in plant species richness among refuges due to refuge size was removed with linear regression. Variation in soil drainage properties in refuges explained over 50% of the variation in plant species richness. These two studies demonstrate a significant relationship between geographic patterns of species richness and heterogeneous environments. Furthermore, they demonstrate a very tight linkage between edaphic characteristics and the biodiversity an area can support.

 
Applications of GIS in the National Park Service New England Support Office
 

Aimee Mandeville, Charles LaBash, Peter August, Nigel Shaw

NPS Field Technical Support Center for GIS
Department of Natural Resources Science
University of Rhode Island
Kingston, RI 02881

 
The University of Rhode Island Dept. of Natural Resources Science is the newly established primary cooperator for geographic information systems (GIS) for the New England Cluster of the National Park Service (NPS). The purpose of this relationship is to foster research and education utilizing GIS to support and promote natural and cultural resource preservation, recreation, and local/regional planning in partnership with state and local government agencies, as well as nonprofit conservation and historic preservation organizations. Innovative applications, new development methods, and technical transfer forums provide both researchers and students with opportunities to test and refine their methods using National Parks and other areas with complex land management issues as laboratories.

The NPS’s objective is to participate in the Field Technical Support Center (FTSC) for GIS at URI for all aspects of GIS research and practices as they relate to resource protection and recreation. The purpose of the FTSC is to provide an arena for land management agencies and organizations to coordinate data collection and analysis on landscape scales and for researchers, students, land managers and planners to work together on issues related to this evolving technology and to learn from each other.

SHORELINE CHANGE, THREATS, AND ANALYSIS, Fire Island National Seashore, New York

James R. Allen¹, Charles L. LaBash², and Peter V. August²
 

¹United States Geological Survey

Biological Resources Division
The University of Rhode Island
Graduate School of Oceanography
Coastal Park Research Unit
Narragansett, RI 02882-1197
 

²Department of Natural Resources Science
University of Rhode Island
Kingston, RI 02881

 
Introduction: Four major storms during the early 1990s caused much damage to coastal development in the northeastern U.S. At Fire Island National Seashore, over 70 homes have been lost and more than 50 have been severely damaged or relocated. There is public concern about drastic shoreline change and the potential for breaching of the barrier island because of the threats to both island and estuarine-margin communities in this densely developed area. To better understand where and when the erosion threats are significant, we have developed a digital database on Mean High Water shorelines, comprising nested temporal hierarchies, along the island. A previous study provided data from 1830 to 1979. We have expanded this database using dynamic-mode Global Positioning System (GPS) surveys of the wet/dry line along the beach during the late summer. At this time of year, fair-weather conditions have usually rebuilt beach width to a maximum. GPS surveys were also done at seasonal intervals, and after large storms in order to understand the spatial patterns associated with long and short-term processes, as well as individual events.
 

Analysis: The shoreline data are used in Arc/Info to display variations in the position through time and can be cross-referenced to other map data describing resources of value. Although the generation of map output is qualitatively useful and tabular associations of some quantitative utility, more statistically compatible datasets of shoreline change are required. To better understand where major erosion threats such as cross-island overwash or island breaching may be more probable, we sampled the shorelines at an alongshore spacing of 50 m, the most robust interval for statistically analysing shoreline change (Dolan et al., 1992). A base reference was created by buffering the 1979 shoreline by 150 meters to the south. A point was placed every 50 m along this base using route system and point event functionality within the Arc Dynamic Segmentation data model. The Arc NEAR routine was used to compute distances from each base reference point to the closest location on each the 1993, 1994, 1995 and 1996 GPS-derived shorelines. The measurement data were ported to MS Excel and quantitatively analysed.

 Results: Between 1979 and 1994 there was a mean accretion of 6.5m but with a standard deviation of 22m. A cyclic change with a wavelength of about 6 km separates erosional and accretionary zones in the western half of the island; a higher frequency periodicity of change dominates the eastern half. Furthermore, the area to the west of Moriches Inlet shows a large accretionary beach, likely due to the establishment of natural bypassing of sediment around the inlet to the island, and an erosional zone immediately westward where a sediment deficit may remain. The shorter duration changes suggest that a small amount of mean accretion between 1993 and 1994 was more than offset by substantial mean erosion from 1994 to 1995. Again, the dominant pattern is of high frequency alongshore change that mirrors the high frequency of coastal storms through February of 1996. The spatial stability of these cyclic patterns is now being studied to see where spatial coherence is high. High coherence would suggest a high probability of success in predicting erosion or accretion sites in the future as well as an isolation of specific, deterministic processes causing shoreline change.