Thesis Proposal
Mark Breunig
College of Natural Resources, UWSP
Background
Phosphorous is of major concern to water quality managers in agricultural watersheds because it can
cause eutrophication in surface waters. Stream phosphorus concentrations are important to the stream
ecosystem and to the loading to downstream water bodies. Agricultural fertilizers and animal waste
contain phosphorus. This chemical has a tendency to absorb to soil particles, and is transported to
surface water primarily during runoff events. In cases of chronic over-application, it can be transported
independently of soil particles (USEPA 2007).
The Wisconsin DNR, in conjunction with the USGS, is currently working to develop total phosphorous
standards for Wisconsin streams. Based largely on their interpretation of the results of a recent study
by Dale Robertson of the USGS, median stream concentrations above 74 µg/L and median river
concentrations above 103 µg/L have a greater impact on the overall biotic integrity of a water body
(Robertson et al. 2006). Reference concentrations throughout Wisconsin, which represent the natural
condition of a water body, were estimated to be between 30 and 40 µg/L. Using a different statistical
method, the United States Environmental Protection Agency (USEPA) estimated reference conditions in
the north central hardwood forest subecoregion, which spans across central Wisconsin, to be 29 µg/L
(Agency 2000)
Statewide Data
The Robertson study, which considered over 200 sites
across Wisconsin, did not explain the specific
mechanisms that cause randomly sampled wadeable
stream total phosphorous concentrations to increase,
but found percent agriculture to be the best
explanatory variable (R2 = 44%). While many basin
characteristics were compared to the phosphorous
concentrations, it may be possible to improve the
percent of variation in
median total phosphorous
concentrations that can be
explained by one or more
spatially-based variables.
Preliminary analysis has
been done on the data set using different expressions of flow distance (example images shown left and
above), which have offered no no ticeable improvement in the regression. However, many possibilities
to improve this technique or explore other spatial characteristics exist.
Montello River Watershed Data
The Montello River Watershed Project, which is just beginning the second of the two years of the study,
contains a data set that may be useful in explaining the mechanisms that influence randomly sampled
median total phosphorous concentrations. This project consists of a much smaller scale than the statewide data set, but still may be useful. An opportunity exists to explain the mechanism behind the very
strong percent cultivated cropland – median phosphorous concentration relationship observed in the
data set. Regressions comparing cultivated cropland and log median baseflow total and soluble reactive
phosphorous (TP, SRP respectively) concentrations at the 6 primary monitoring sites are shown below.
log10(B_medTPconc) = 1.413 + 1.324 Cultivated Crops
100
log median baseflow TP
90
80
70
60
50
adj R2 = 80%
p = 0.009
40
0.20
0.25
0.30
0.35
0.40
0.45
Cultivated Crops (proportion of total land use)
log10(B_medRPconc) = 1.029 + 1.899 Cultivated Crops
log median baseflow SRP
70
KW02
60
KW03
50
UN01
40
WC02
30
TC01
20
WC01
adj R2 = 97%
p = 0.001
0.20
0.25
0.30
0.35
0.40
0.45
Cultivated Crops (proportion of total land use)
It is hypothesized that a metric known as the stream sediment equilibrium phosphorous concentration
(EPC0), controls baseflow phosphorous concentrations. When EPC0 is greater than the stream SRP,
sediments are a potential phosphorous source to the stream. When EPC0 is less than the stream SRP, it
is suggested that stream sediments will sorb phosphorous from the water column (McDaniel et al. 2009).
A time series of baseflow total phosphorous concentrations during the 2008 monitoring season is shown
below. These data show an interesting temporal pattern. The highest values were observed after the
major flood and early June, which steadily decreased until the end of the sampling season in November.
It appears that phosphorous release rates from streambed sediments deposited in the flood were higher
directly after the event; the release rate decreased over time as the stream sediment and water
approached equilibrium. Seasonality may have also influenced this trend, as biologic activity is higher in
the summer. Sites on Klawitter Creek, KW02 and KW03, also had higher concentrations in early spring,
which decreased until the flood in early June. Phosphorous enriched sediments were deposited into the
stream during spring runoff, and release rates decreased until the next major input in June. Samples
representing winter baseflow concentration were collected 2/22/08, which were significantly lower at
KW02 and KW03 than the spring baseflow concentrations. This suggests that the groundwater
discharging into Klawitter Creek did not have elevated total phosphorous concentrations.
4/1/2008 7/1/2008 10/1/2008
300
KW02
KW03
MR01
TC01
UN01
WC01
150
TP (µg/L)
0
300
150
300
0
WC02
4/1/2008 7/1/2008 10/1/2008
150
0
4/1/2008 7/1/2008 10/1/2008
DateTime15
SITE
KW02
KW03
MR01
TC01
UN01
WC01
WC02
The Montello Dataset has a good distribution of phosphorous concentrations across sites. The effect of
total phosphorous on the biotic integrity of each stream site, according to the DNR proposed standards,
are shown below.
130
120
117
110
median TP (µg/L)
100
91
90
80
70
60
90
effected
70
minimally effected
58
least impacted
50
40
30
40
natural conditions
KW02
KW03
TC01
UN01
WC01
WC02
SITE
This data set will also include phosphorous index (PI) estimates calculated by Snap-Plus nutrient
management software, which have potential benefits over generalized spatial datasets. This study
represent a pioneering effort in the application of the PI to stream monitoring data at the watershed
scale.
Thesis Objectives
The specific scope of this thesis is yet to be determined, but it will consist of one or more of the
following objectives.
1. Improve the statewide cropland vs median total phosphorous regression by incorporating a
more efficient explanatory variable. It is likely that this will require the derivation of a spatiallybased metric.
2. Use the EPC0 as evidence behind the mechanism controlling baseflow total phosphorous
concentrations in the Montello River Watershed.
3. Link the PI or some other export/delivery term to the EPC0. This could involve applying the
spatially-based metric derived in objective 1, or it could be completely independent of this.
If all three objectives could be realistically accomplished, a very beneficial insight describing the factors
that control randomly sampled median phosphorous concentrations. This information could be used to
more effectively manage the landscape to achieve desired total phosphorous concentrations. It also
could be used to develop a predictive model, offering great benefit to the scientific community and
others.
Works Cited
Agency, U. S. E. P. (2000). Ambient Water Quality Criteria Recommendations. O. o. S. a. T. Office of
Water, Health and Ecological Criteria Division. Washington, D.C., U.S. Environmental Protection
Agency: 92.
McDaniel, M. D., M. B. David, et al. (2009). "Relationships between Benthic Sediments and Water
Column Phosphorous in Illionois Streams." Journal of Environmental Management 38: 607-617.
Robertson, D. M., D. J. Graczyk, et al. (2006). "Nutrient Concentrations and Their Relationships to the
Biotic Integrity of Wadeable Streams in Wisconsin." USGS Professional Paper(PP 1722).