Contribution of Nutrients and E. coli to Surface Water Condition in the Ozarks Part II. Using Landscape Ecology and Partial Least Squares Predictions to Map Watersheds That Are Vulnerable to Non-Point Source Pollution

Contribution of Nutrients and £ colt to Surface Water Condition in the Ozarks

Part II. Using Landscape Ecology and Partial Least Squares Predictions to
Map Watersheds That Are Vulnerable to Non-Point Source Pollution

Ricardo D. Lopez* and Maliha S. Nash

U.S. EPA, 944 East Harmon Avenue, Las Vegas, Nevada 89119

Corresponding author e-mail: lopez.ricardo@epa.gov

The U.S. EPA's Office of Research and Development and U.S. EPA Region 7 have collaborated to map and interpret
landscape-scale (i.e., broad-scale) ecological metrics among watersheds of the Upper White River (Figure 1) and have
produced the first geospatial models of water quality vulnerability in the Ozarks. These models have been developed
by using Partial Least Squares (PLS) analyses, existing field water quality monitoring data, remote sensing information,
a priori information about landscape conditions, and the water quality of streams and rivers in the associated
watershed(s). The analyses were conducted at multiple geographic scales, from site-specific water quality
measurements (fine-scale) to broader scale watershed analyses. The broad-scale results are reported for 8-digit U.S.
Geological Survey hydrologic units and 244 customized subwatersheds. The customized subwatersheds represent the
drainage area where water quality measurements were conducted and were used to increase the precision and
accuracy of water quality vulnerability predictions. Using PLS, we determined the following four different (surface)
water quality condition states among the 244 customized subwatersheds of the Ozarks that may be useful for
important management decisions in the region: (1) subwatersheds that have high concentrations of total ammonia,
high concentrations of total phosphorus, and high cell counts of Escherichia coli; (2) subwatersheds that have high
concentrations of total ammonia, low concentrations of total phosphorus, and high cell counts of E. coli; (3)
subwatersheds that have low concentrations of total ammonia, low concentrations of total phosphorus, and high cell
counts of E. coli; and (4) subwatersheds that have moderate concentrations of both total ammonia and total
phosphorus and moderate £. coli cell counts. The results of this project provide watershed managers with the first
broad-scale predictions that can be used to explain how land cover type, land cover configuration, environmental
change, and human activities may affect the chemical and biological characteristics of surface water in the Upper
White River region.

Methodology (Figure 2)

The 244 water quality sampling locations were
used to delineate 244 subwatersheds. The study
area can be depicted as a grouping of the 244
subwatersheds, each of which contains a single
hydrologic outlet (sometimes called a "pour
point"). The value of using this unconventional
view of the landscape is that the cumulative
effects of landscape condition on water quality can
be assessed, thereby increasing the predictive
power of any determined relationships between
land cover and water quality parameter. The
Missouri and the Arkansas land cover data sets
were created and aggregated to evaluate the
unified land cover map for 2003 and used for
statistical analyses of landscape metrics. For each
of the selected sites, the watershed support area
was delineated and a suite of landscape metrics
was calculated. Ecologically relevant (e.g.,
previous correlative relationships demonstrated in
other biophysical regions or at finer scales)
landscape metrics were calculated for 2003
landscape conditions using remote sensing and
geographic information systems data. Landscape
metrics were then compared to surface water
constituents, averaged over a period from 1997 to
2002, using Partial Least Squares analyses (see
accompanying poster Part 1 for details about PLS
analyses).

¦waiet Quality
Sampling Locations

Circa 1984
Landsai Image

Study Area Orientation

Figure 2. The Upper White River study area is in the Ozarks: of Missouri and
Arkansas, where 244 water quality sampling locations were sampled (A) and
used as "pour-points," from which 244 contributing subwatersheds were
delineated (B). A combination of multiple Landsat Thematic Mapper imagery
(C) and digital aerial photography was used to produce a 2003 land cover map
of the study area (D), which was used to calculate landscape metrics.

Results/Discussion

Figure 3. Study area water quality vulnerability prediction
maps for (A) total ammonia, (B) total phosphorus, and (C)

Escherichia coli.

Overview

Ammonia: The total ammonia PLS model resulted in one significant factor explaining 93% of the variability in the total

I ammonia. Percent forested land cover in close proximity to streams has a negative effect on total ammonia loading,
whereas urban has a positive effect (Figure 3A).

Phosphorus: The total phosphorus PLS model resulted in one significant factor explaining 59% of the variability in the
total phosphorous. The most significant contributors are the watershed percent barren and stream density. While the
stream density relates inversely with total phosphorous, percent barren enhances total phosphorous in surface water.
The forest-related variables contribute equally with a negative effect on the total phosphorous. Urban enhanced total
phosphorous but mostly in close proximity of streams effect (Figure 3B).

Escherichia coli: The E: coli PLS model resulted in two significant factors explaining 81% of the variability in the £.
coli cell count. Urban in close proximity to streams was positively correlated with E. coli cell count (Figure 3C).

Figure 1. White River Study Area, shown as four 8-digit
U.S. Geological Survey hydrologic unit code ("HUCs").

Total Ammonia Predicted (mgfl)
m 0.004 - 0.02
¦ 0-02 - 0.038
0.038 - 0.064
0.064 - 0.096
H 0 096 - 0.15

Toial Phosphorus Predicted (mg/l)

¦ 0 - 0.027
0.027 - 0.073

| | 0.073 -0.123

_ 0.123 -0.191
H 0.191 - 0.337

E. coll Predicted {bacterial colts/100 ml)

¦ 0.72 - 416.77
416,77 - 1267.64
i 1267.64 - 2979.47
3 2979.47 - 6659.83
¦I 6659.83 - 20764.66

Legend

Town or Ctty Center
I Major Urban Area
Surface Water Body
County Boundary
] HUC-Study An

J Stale and Boundary

The water quality vulnerability
maps and metrics associated
with this research will be
available in a compact disk
"browser" format during the
summer of 2006.

This research is a collaborative
effort between U.S. EPA's Office
of Research and Development
and U.S. EPA Region 7. We
thank our many local, state, and
regional partners in Arkansas
and Missouri, and U.S. EPA
Region 7 for their support during
this research.

epascienceforum

Your Health • Your Environment • Your Future

Notice: Although this work was reviewed by EPA and approved for publication, it may not necessarily reflect official Agency policy. Mention of trade names or commercial products does not constitute endorsement or recommendation by EPA for use.

113odc06SF-Lopez

-------