Geoscience Reference
In-Depth Information
number of EPT taxa (richness). After restricting dates of interest to those within
the range of the land cover information, and filtering out repeated sampling at the
same location (keeping the most recent), we had a sample of 32 sites with NAWQA
data in southern New England (Fig.
4.1
). No sampling locations for sensitive taxa
were located within Rhode Island.
4.2.3 HBI
Our research made use of the Hilsenhoff Biotic Index (HBI) for data collected
across 83 stream reaches in Rhode Island and Southeastern Massachusetts.
The HBI, developed in the late 1970s by William Hilsenhoff at the University
of Wisconsin, rates stream biologic integrity on a scale of 0-10, with the lower
values representing an excellent ranking and the higher values representing
poorer conditions (Hilsenhoff
1987
). The HBI scale incorporates the tolerance
of each taxa identified during stream reach inspections to organic pollutants
(Robinson
2004
). The HBI is one of several metrics from the Wadeable Stream
Condition Index used for stream health monitoring in Rhode Island, and a Rapid
Biomonitoring Protocol (RBP) index used to rate stream health in Massachusetts.
The Rhode Island portion of these data (n
39) were part of an archive of
11 years of biomonitoring data, collected by Sara da Silva (University of Rhode
Island) (
2003
), the Rhode Island Department of Environmental Management and
Environmental Resource & Wetlands Assessment, and Nelson, Pope & Voorhis
LLC of Melville, New York (da Silva,
personal communication
). The HBI metric
information for Massachusetts (n
¼
44) was provided by the Massachusetts
Department of Environmental Protection, Division of Watershed Management.
These data were concentrated on the Taunton River watershed area of south-
eastern Massachusetts, but also included samples from the Blackstone and
10-mile watersheds (see Fig.
4.1
).
¼
4.2.4 Land Cover Metrics
The land cover metrics used in this analysis included maps of percent impervious
surface cover, percent tree cover, and categorical (presence/absence) of grassland
and crop cover. These data sets were derived from National Land Cover Database
2001 (NLCD), available through the U.S. Geological Survey's EROS Data Center.
The NLCD data set, produced using 30 m Landsat Thematic Mapper satellite
imagery, covers watersheds across the nation (including our entire study area)
using a consistent classification scheme (Homer et al.
2004
). From this categorical
map product, we extracted Grassland/Herbaceous and Pasture/Hay land cover
information (NLCD classes numbered 71 and 81) to identify areas dominated
by grass vegetation cover. Corresponding crop cover was derived using NLCD
