Macroinvertebrate Response to Land Use and Stream Chemistry n

the Mid-Atlantic Coastal Plains

1. Purpose of Study

Megan Mehaffey, Mai ilia Nash, Ann Pitchford
U S E PA/O R D/N E R L/E SD/LEB

2. Design and Clustering Method

/ps*
's~*

4. Benthic and Water Quality Measure
Comparisons

We used non-parametric analysis to test differences in
benthic metrics and water chemistry data between clusters.
We found that typical water chemistry measures associated
with urban runoff such as specific conductance and dissolved
chloride were significantly higher in the more urbanized
watershed group while intolerant benthos were lowest (Figures
4D to 4G). In the highly cropped group, we found variables
commonly associated with fanning such as nutrients and
pesticides significantly greater than the other two groups. The
tolerant benthic metrics in the cropped watersheds were in the
middle, while the intolerant metric values were close to that of
the urban group (Figures 4A to 4C).

Acknowledgement: The authors thank the U.S. EPA's Eco-
logical Exposure Research Division for collecting and
analyzing the benthic macroinvertebrate samples and the U.S.
Geological Survey for performing the water quality sampling
and analysis for this study.

5. Within Cluster Step-Wise

Fores ted/Wetland watersheds
were less polluted and had more
intolerant species, probably due to
better water quality. Tolerant benthos
in urban and cropped watersheds did
not show measurable response to
differences in pollutants, while intol-
erant benthos decreased in abun-
dance with increasing pollutant con-
centration. We hypothesize that the
proportion of tolerant taxa increases
rapidly in less polluted sites as pollu-
tion increases (i.e., ag. chemicals,
urban runoff, sediment, nutrients,
etc.) and begins to enter the stream,
making these metrics sensitive indi-
cators of changes in the landscape
detrimental to water quality. In sites
with more human disturbance, how-
ever, metrics of intolerant taxa might
provide a better means of monitoring
continued water quality degradation.

Linear Regressions

The Landscape Indicators for Pesticides Study in Mid-Atlantic Coastal
Streams (LI PS-MACS) is a research collaboration between the U.S. Envi-
ronmental Protection Agency's Office of Research and Development and the
U.S. Geological Survey's National Water Quality Assessment Program. An
objective of this research is to develop models to relate land use, geology,
and other geographic variables to water quality and the biotic integrity of
small streams.

Pesticides, nutrients, and other chemicals can dramatically affect water
quality and biota in streams in areas with substantial agriculture, industry, or
urban development. Thus, the landscape setting (i.e., the location of a
stream within its valley and the relative proportions of various land uses
combined with the topography and related physical features, Figure 1 A) can
provide crucial information about a water body's condition. The problem,
however, is that measurements are not possible in every watershed because
of cost and other practical constraints. We examined the applicability of
potential landscape indicators to efficiently estimate the biological integrity
of streams in the Mid-Atlantic Coastal Plains Region.

From a population of 10,144 first-order watersheds in the region, we
chose 174 small (typically first-order) streams that represented the regional
hydrogeologic and land use gradient. Of these 174 sites sampled during base
flow in the late winter and spring of 2000, we selected a subset of 82 having
benthic riffle habitat for analysis.

We used principal components and cluster analyses to group the first-
order streams by land use to capture the greatest variability across the region.
The analyses resulted in three distinct watershed groupings. The first group
was dominated by natural cover (forest and wetland), the second by urban,
and the third by crops (Figures 2 A and B).

stonefly
?

Landscape Variables

The graph to the left shows the
range of percentages within each
cluster group and where there is
overlap (Figure 2C). The red and
blue bars represent areas used for
ranking all 10,144 first-order water-
sheds in the Mid-Atlantic Coastal
Plain (Figure 6A).

6. Expanding Methods to all First-Order Watersheds
in the Mid-Atlantic Coastal Plains

We used the ranges of land cover/land use from the cluster analysis to
categorize the 10,144 first-order watersheds delineated for the Mid-Atlantic
Coastal Plains. Cropped watersheds dominated the Delaware and North
Carolina regions and likely have the highest nutrient and pesticide levels
(Figure 6A). The western most ridge areas, particularly in Maryland, tended
to have a larger number of urbanized watersheds where runoff and dissolved
chloride would be expected to be greater.

From data collected in this study, we ranked all watersheds based on
land use so that estimates of regional stream biological integrity could be
made. We used cluster overlap areas to define cutoffs (Figure 2C); those
watersheds with percentages above the blue line for forest and wetland were
given the highest rank and those below the red line the lowest, while for
human use ranking was reversed. The final rank for each watershed was de-
termined by summing the ranks given to each land use (see Figure 6B).
Using this method, we can identify a range of conditions from reference to
at-risk watersheds. Furthermore, using the major land use group and ranking
maps together provides a way to relate land use type to major pollutants and
those pollutants likely to have a continued effect on biotic integrity. We dem-
onstrate with this study the applicability of using landscape models for effi-
ciently estimating the biological integrity of small streams in the Mid-
Atlantic Coastal Plains Region.

Order, PTV

3. Macroinvertebrate Distribution

Tubificids (leeches and blood worms), Veneriods (clams and snails), and Amphipods and
Isopods (crustaceans) dominated the tolerant taxa (pollution tolerance value or PTV — 7.0
or more). The largest portion of intolerant taxa (PTV ~ 4.0 or less) was Ephemeroptera
(mayfly), Plecoptera (stonefly), and Tricoptera (caddisfly).

Urban streams had greater proportions
of leeches and blood worms
(Tubificida) (Figure 3B), versus
forest/wetland streams that had greater
proportions of caddisflies,
stoneflies and

caddisfly

mayflies (EPTs)
(Figure 3C).


-------