EPA/600/R-95/173
Assessment of Nitrogen Loads
to Aquatic Systems
by
A.S. Patwardhan and A.S. Donigian, Jr.
AQUA TERRA Consultants
Mountain View, CA 94043
Contract Number 68-CO-0019
Project Officer
Thomas O. Barnwell, Jr.
National Exposure Research Laboratory
Athens, GA 30605
National Exposure Research Laboratory
Office of Research and Development yt-nv---
U.S. Environmental Protection Agency DRINKING ^^^r
Athens, GA 30605
-------
NOTICE
The U.S. Environmental Protection Agency through its Office of Research and Development funded
and managed the research described here under Contract No. 68-CO-0019 to AQUA TERRA
Consultants. It has been subjected to the Agency's peer and administrative review and has been
approved for publication as an EPA document.
-------
FOREWORD
The Ecosystems Research Division of EPA's National Exposure Research Laboratory conducts
research on organic and inorganic chemicals, greenhouse gas biogeochemical cycles, and land use
perturbations that create direct and indirect, chemical and nonchemical, stressor exposures and
potential risks to humans and ecosystems. Comprehensive models based on fundamental studies of
stressor behavior are developed to predict exposures in mutimedia environments, to simulate the
interactions of the climate system and the terrestrial biosphere, and to evaluate the aggregate causes
of ecological stress, including land use change/management, within a watershed/regional context.
Field and laboratory experiments are conducted to quantify and model greenhouse gas fluxes between
the atmosphere and the terrestrial biosphere and to understand abiotic and biotic pollutant fate
processes in soils, sediments and water.
In the study reported here, work by the Environmental Defense Fund to estimate the various sources
of nitrogen to Chesapeake Bay is reviewed and replicated. An improved methodology was then
developed for estimating atmospheric nitrogen contributions to aquatic systems by refining some of
the EDF procedures and assumptions used in estimating the contributions from various nitrogen
sources within the watershed. The refined procedures were applied to the Chesapeake Bay
watershed, to Galveston Bay (Texas) and to Tampa Bay (Florida) using commonly available
databases.
Rosemarie C. Russo, Ph.D.
Director
Ecosystems Research Division
Athens, Georgia
111
-------
ABSTRACT
Nitrogen is ubiquitous in the environment, released by many mobile and stationary sources, and
is a major plant nutrient applied as fertilizers and animal wastes. EPA has recognized the
importance of nitrogen in the environment for both beneficial (e.g. soil fertility, plant nutrition)
and deleterious effects, ranging from the control of ozone distribution, greenhouse gas impacts,
air pollution, acidification and eutrophication of surface waters, and contamination of
groundwater drinking supplies.
In 1988 the Environmental Defense Fund (EDF) estimated the various sources of nitrogen to the
Chesapeake Bay; these included 23% from point sources, 39% from atmospheric deposition
(including both nitrate and ammonia), 34% from fertilizers, and 4% from animal wastes. They
further projected that with atmospheric nitrogen emissions to increase by 44% by the year 2030,
atmospheric nitrogen contributions to aquatic systems are likely to become even more significant.
The objective of the study is to estimate the contributions of atmospheric nitrogen deposition to
Chesapeake Bay, Galveston Bay, and Tampa Bay with respect to the nitrogen contributions from
nonpoint and point sources. The EDF methodology was reviewed and refined to develop the
current project methodology. A Lotus 123 spreadsheet was developed based on the new
methodology, and the spreadsheet was used in estimating nitrogen loadings from the various
nitrogen sources found in each of the three study watersheds.
The results obtained from the application of the methodology to the Chesapeake Bay watershed
were in agreement with the EDF results. Sensitivity analysis was performed on three
spreadsheet assessment parameters — ratio of wet to dry atmospheric deposition, first-order
riverine nitrogen decay rate, and amount of fertilizer application. During the sensitivity analysis
we analyzed the sensitivity of the ratio of wet to dry deposition because dry atmospheric
deposition data is not readily available, and changes to this ratio effectively change the total
deposition. It was concluded from the sensitivity analysis that, as atmospheric deposition is an
important source of nitrogen input on all the land use categories in the watershed, it has a
significant effect on the nitrogen loadings that are delivered to the aquatic systems.
The results obtained from the application of the methodology to the three study sites indicate that
over 40% of the total nitrogen load delivered to aquatic systems may originate from atmospheric
deposition. Recommendations to improve the current methodology estimates and further
research issues are provided.
This report was submitted in fulfillment of Contract No. 68-CO-0019, Work Assignment No.
27(11), by AQUA TERRA Consultants of Mountain View, CA under sponsorship of the U.S.
Environmental Protection Agency. This report covers the period from October 1992 through
September 1993, and work was completed as of 30 September 1993. Final report revisions were
performed under Purchase Order No. 5D2086NASA, and were completed September 30, 1995.
iv
-------
CONTENTS
Page
Disclaimer ii
Foreword iii
Abstract iv
Figures vii
Tables viii
Acknowledgments x
1.0 INTRODUCTION 1
1.1 Study Goals and Objectives 1
1.2 Summary of Study Results 2
1.3 Study Conclusions and Recommendations 4
1.4 Format of Report 5
2.0 REVIEW OF EOF ANALYSIS 6
2.1 EDF Procedures Used for the Chesapeake Bay Watershed 6
2.1.1 Equal Retention Procedure 8
2.1.2 Differential Retention Procedure 10
2.2 EDF Study Conclusions 14
2.3 Assumptions Used in the EDF Study 14
3.0 METHODOLOGY USED IN ESTIMATING NITROGEN LOADS TO AQUATIC
SYSTEMS 17
3.1 Calculational Steps in Current Methodology 17
3.2 Define NPS Loads by Land Use 18
3.3 Define Point Source Loads 20
3.4 Subbasin Loads 20
4.0 ASSESSMENT OF NITROGEN LOADS FROM STUDY SITES 25
4.1 Application of the Methodology to Chesapeake Bay
Watershed 25
4.1.1 Comparison of New Methodology Results with
EDF Results 28
4.1.2 Comparisons of Percent Loadings from Individual
Nitrogen Sources 35
4.1.3 Conclusions from Chesapeake Bay Comparison 36
-------
4.2 Application of Methodology to Galveston Bay 37
4.2.1 Available Data for Analysis 37
4.2.2 Results Obtained from Galveston Bay 43
4.3 Application of Methodology to Tampa Bay 47
4.3.1 Available Data for Analysis 47
4.3.2 Results Obtained from Tampa Bay 50
4.4 Sensitivity Analysis 53
4.4.1 Sensitivity of Wet and Dry Atmospheric Deposition
Ratios 53
4.4.2 Sensitivity to Riverine Decay Parameter 56
4.4.3 Sensitivity to Fertilizer Application Amounts 58
4.4.4 Conclusions from Sensitivity Analysis 61
5.0 CONCLUSIONS AND RECOMMENDATIONS 63
6.0 REFERENCES 66
7.0 APPENDIX A - Description of the Lotus 123* Spreadsheet A-l
VI
-------
FIGURES
2.1 Chesapeake Bay Watershed 7
2.2 EDF's Differential Retention Procedure Flowchart 13
3.1 Methodology Flow Chart for Estimating Nonpoint and Point
Source Loads 22
3.2 Example of NPS Load Computations from each Subbasin 23
4.1 Chesapeake Bay Watershed with Subbasin Boundaries 26
4.2 Galveston Bay Watershed with Subbasin Boundaries 38
4.3 Tampa Bay Watershed with Subbasin Boundaries 48
vn
-------
TABLES
1.1 Comparison of Percent Nitrogen Contributions from Various
Sources to Chesapeake Bay, Galveston Bay
and Tampa Bay 3
2.1 Calculated Nitrogen Loadings to the Land and Water Surfaces
of the Chesapeake Bay Watershed in 1984 10
2.2 Nitrogen Loadings to the Chesapeake Bay (US EPA, 1983) 11
2.3 Equal Retention Estimates of Nitrogen Sources to the Bay 11
2.4 Retention Factors and N Loadings Using Differential
Retention Procedure 15
4.1 Comparison of Current Nitrogen Inputs to EDF Nitrogen Inputs 27
4.2 Estimated Mean Annual Total Nitrogen Loads from the Subbasins
in the Chesapeake Bay Watershed (from Donigian et al., 1994) 30
4.3 Land Use Areas in the Chesapeake Bay Watershed (from
Donigian et al., 1991) 31
4.4 Above Fall Line Nitrogen Point Loads (from Donigian et al., 1994) 31
4.5 An Example of the Input Requirements for the Lotus 123 Spreadsheet .... 33
4.6 Comparison of Nitrogen Loads to the Chesapeake Bay 34
4.7 Comparison of Percent Nitrogen Contributions from Various
Nitrogen Sources to the Chesapeake Bay 35
4.8 Land Use Areas in the Various Subbasins of the Galveston Bay
Watershed (modified from Newell et al., 1992) 39
4.9 Total Edge-of-Field Nitrogen Loads from the Subbasins of the
Galveston Bay Watershed (modified from Newell et al., 1992) 42
4.10 Unit Area Nitrogen Loads from the Various Land Use Categories
in the Galveston Bay Watershed 44
4.11 Point Sources, Wet Atmospheric Deposition, and Travel Time for
the Galveston Bay Watershed Subbasins 45
4.12 Estimated Total Nitrogen Loads to the Galveston Bay from
Various Nitrogen Sources 46
4.13 Percent Loadings from Various Nitrogen Sources to the
Galveston Bay 46
4.14 Total Edge-of-Field Nitrogen Loads from the Subbasins of the
Tampa Bay Watershed 49
4.15 Point Sources, Wet Atmospheric Deposition, and Travel Time for
the Tampa Bay Watershed Subbasins 51
vin
-------
4.16 Estimated Total Nitrogen Loads to the Tampa Bay from
Various Nitrogen Sources 52
4.17 Percent Loadings from Various Nitrogen Sources to the
Tampa Bay 52
4.18 Retention Factors for the Base Case for the Chesapeake Bay
Watershed 55
4.19 Sensitivity of Ratio of Dry to Wet Atmospheric Deposition
Rates to Nitrogen Loadings to Chesapeake Bay 57
4.20 Percent Contribution to the Chesapeake Bay with Varying
Ratios of Wet to Dry Atmospheric Deposition Rates 57
4.21 Sensitivity of Riverine Decay Rates to Nitrogen Loadings
to Chesapeake Bay 59
4.22 Percent Contribution to the Chesapeake Bay with Varying
Riverine Decay Rates 59
4.23 Sensitivity of Fertilizer Application Rates to Nitrogen
Loadings to Chesapeake Bay 60
4.24 Percent Contribution to the Chesapeake Bay with Varying Base
Scenario Fertilizer Rates 60
4.25 Effect of Changing Atmospheric Deposition and Fertilizer
Application of Total Nitrogen Load from the
Chesapeake Bay Watershed 62
IX
-------
ACKNOWLEDGMENTS
This document was prepared under Work Assignment No. 27(11) of Contract No. 68-CO-0019
by AQUA TERRA Consultants for the U.S. Environmental Protection Agency, Environmental
Research Laboratory in Athens, GA. Mr. Thomas Barnwell of the EPA-Athens Laboratory was
the EPA Project Manager providing technical and administrative guidance, and project direction.
The EPA Office of Air Quality and Standards, through Mr. John Bachmann and his staff, is
acknowledged for its interest, support, and funding of this work.
A number of individuals provided data and information vital to the completion of this work. Dr.
Charles Newell and Elaine Higgins of Groundwater Services, Inc., provided all of the nonpoint
loading data and information about the Galveston Bay watershed. Dr Neil Armstrong provided
point source loading data for the Galveston Bay watershed. Herbert Hudson from the Galveston
Bay National Estuary Program (NEP) program provided initial information about the Galveston
Bay. Ms. Holly Greening of the Tampa Bay NEP provided general background on the Tampa
Bay watershed, and Mr. Hans Zarbock of Coastal Engineering, Inc. provided nonpoint and point
source data, and travel time estimates for the subbasins in the Tampa Bay watershed. Ms. Gail
Huff of Soil Conservation Service provided estimates of fertilizer application rates in the Tampa
Bay watershed.
For AQUA TERRA Consultants, Mr. Avinash Patwardhan acted as the Project Manager, and
applied the developed methodology to the three study sites, analyzed results, and wrote the
report. Mr. Anthony Donigian provided input in the conceptualization of the project
methodology, and provided technical review and guidance.
-------
SECTION 1.0
INTRODUCTION
1.1 STUDY GOALS AND OBJECTIVES
Nitrogen is ubiquitous in the environment, released by many mobile and stationary sources, and
is a major plant nutrient applied as fertilizers and animal wastes. EPA has recognized the
importance of nitrogen in the environment for both beneficial (e.g. soil fertility, plant nutrition)
and deleterious effects, ranging from the control of ozone distribution, greenhouse gas impacts,
air pollution, acidification and eutrophication of surface waters, and contamination of
groundwater drinking supplies. EPA has mandates to evaluate and regulate these effects of
nitrogen under the Clean Air Act, the Clean Water Act, the Safe Drinking Water Act, and under
numerous international agreements.
In 1988 the Environmental Defense Fund (EDF) (Fisher et al., 1988) estimated the various
sources of nitrogen to the Chesapeake Bay; these included 23% from point sources, 39% from
atmospheric deposition (including both nitrate and ammonia), 34% from fertilizers, and 4% from
animal wastes. They further projected that the atmospheric component would increase to 55%
by the year 2030. In EPA Region III, the Chesapeake Bay Program is currently evaluating
strategies for attaining a 40% reduction in nutrient loads (including nitrogen and phosphorous)
reaching the Chesapeake Bay. The modeling systems developed as part of that study include
performing mass balance modeling of nitrogen on various land categories to explore and quantify
both the sources of nitrogen and potential reduction alternatives.
The objective of this work was to assess the applicability of the EDF approach and associated
findings to other parts of the country, in particular to the Gulf Coast region. The approach was
to review and evaluate the EDF methodology, to improve and refine the EDF methodology and
assumptions, and to implement these refined procedures to reproduce their results (in a Lotus
123 spreadsheet format) using commonly available data sources (e.g. EPA/State waste
discharges, emission inventories, soils and meteorologic data bases). The refined procedures
were then to be applied to the Chesapeake Bay watershed; Galveston Bay, Texas; Tampa Bay,
Florida; and the Ohio River Drainage Basin. However, due to lack of nonpoint and point source
data for the Ohio River Drainage Basin, this site was not included in the study.
The first component of this study deals with reviewing the EDF analysis and assumptions, and
then replicating the EDF results for the Chesapeake Bay watershed. The second component of
the study deals with developing an improved methodology to estimate atmospheric nitrogen
contributions to aquatic systems by refining some of the EDF procedures and assumptions used
in estimating the contributions from various nitrogen sources within a watershed.
-------
To formalize and facilitate the refined project methodology, we developed a Lotus 123*
spreadsheet to estimate the contributions of atmospheric nitrogen deposition to the study sites
listed above with respect to the nitrogen contributions from nonpoint and point sources. The
land use categories from which nonpoint source loadings were calculated consist of forest,
cropland, pastures, and urban area. The point source loadings consist of contributions of
nitrogen from industrial and sewage treatment plant discharges. The annual wet atmospheric
nitrogen deposition data was obtained from the National Atmospheric Deposition Program
(NADP).
The Lotus 123 spreadsheet, which was developed by refining some of the assumptions that were
used in the EDF study, was applied to the Chesapeake Bay watershed. The total nitrogen loads
from each land use obtained by the current methodology were compared to the results obtained
from the EDF analysis. The spreadsheet was then used to calculate the contributions of
atmospheric nitrogen deposition as it relates to other sources of nitrogen for the Galveston Bay
and Tampa Bay study sites.
1.2 SUMMARY OF STUDY RESULTS
The percent contributions from various nitrogen sources in the watershed for all three study sites
are presented in Table 1.1. For the Chesapeake Bay, the results for the current analysis are
based on dividing the watershed into fourteen subbasins, whereas the EDF results (also shown
in Table 1.1) were obtained by considering the watershed as a single basin. As can be seen
from Table 1.1, the current methodology indicates that 43% of the total nitrogen load to the
Chesapeake Bay results from atmospheric deposition, which is in close agreement with the 39%
estimate obtained by the EDF study; the total nitrogen loads of 140 million kg per year from the
EDF study was similar, but slightly higher than the 115 million kg per year from the current
study. It can also be concluded from Table 1.1 that the percentage contributions calculated from
point sources and manure loadings to the Chesapeake Bay using the current methodology and
the EDF estimates are also in close agreement. The current methodology estimated slightly
higher contributions of nitrogen from point sources as compared to the EDF analysis. This
could be due to the data used to calculate point source loads. In the current study we used an
average (1984-1987) point source loadings as estimated by the U.S. EPA Chesapeake Bay
Program Office (Donigian et al., 1994), while the EDF analysis used point source data for the
year 1985 only. The percent contribution from manure was 3.2% higher than the EDF estimate.
In the EDF analysis manure was only applied to the pastures, and the authors used a retention
factor 97.5% which resulted in very low nitrogen loadings to the Bay as a result of manure
application. However, in our analysis, manure was applied to both cropland and pastures, and
as cropland has lower retention of nitrogen than pastures, a slightly higher percentage nitrogen
load to the Bay was estimated.
The percentage nitrogen loadings to the Chesapeake Bay from fertilizer applications as estimated
by the two procedures do not agree. In the EDF analysis the authors used the county fertilizer
sales data to estimate the total amount of fertilizer applied to the entire watershed and also used
a single constant retention factor for all the agricultural cropland in the Chesapeake Bay
-------
Table 1.1 Comparison of Percent Nitrogen Contributions from Various Sources to
Chesapeake Bay, Galveston Bay, and Tampa Bay.
Nitrogen EDF Percent Using Current Methodology
Source Analysis
Chesapeake Chesapeake Galveston Tampa
Bay Bay Bay Bay
Point Source 23% 30% 48% 19%
Atmospheric Deposition 39% 43% 49% 67%
Fertilizer 34% 20% 3% 14%
Manure 4% 7%
watershed. However, in our analysis we used fertilizer application data, estimated by state and
county extension officials (Donigian et al., 1994). In the present analysis the total fertilizer
application was 13 million kilograms greater than the amount applied in the EDF study.
However, using variable retention parameters (computed from unit area loads as reported by
Donigian et al., 1994) for all 14 subbasins resulted in lower fertilizer loads to the Chesapeake
Bay as compared to the EDF estimated load.
The results obtained from the application of the project methodology to the Galveston Bay study
site support the findings that are obtained from the Chesapeake Bay study. The nonpoint
nitrogen loadings that were estimated by Newell et al. (1992) and used in this project indicate
that over 50% of the total nonpoint load is due to the land use area occupied by urban dwellings.
In the Galveston Bay watershed we estimated the total nitrogen load to the Bay to equal 36
million kilograms. From the total nitrogen load that was delivered to the Galveston Bay we
estimated the following distribution: 48% to originate from point sources, 49% due to
atmospheric deposition, and only 3% from fertilizer application (Table 1.1). The direct
atmospheric deposition onto the Galveston Bay water surface accounted for 39% of the total 49%
atmospheric nitrogen deposition load; and the watershed contributed the remaining 10%.
Fertilizer loadings to the Bay were quite low as only 22% of the watershed was under crop
production.
The Tampa Bay watershed results support the findings that were obtained from the Chesapeake
Bay and Galveston Bay study. The nonpoint and point source nitrogen loadings for the
watershed were provided by Greening (1993, personal communications). In the Tampa Bay
watershed we estimated the total nitrogen load to the Bay to equal 4.5 million kilograms. From
the total nitrogen load that was delivered to the Tampa Bay we estimated the following
distribution: 19% to originate from point sources, 67% due to atmospheric deposition, and 14%
from fertilizer application (Table 1.1). The direct atmospheric deposition onto the Tampa Bay
water surface accounted for 40% of the total 67% atmospheric nitrogen deposition load, while
the watershed contributed the remaining 27%. The relatively large contribution due to
-------
atmospheric deposition may be due to the total nitrogen load being considerably lower than the
loads in the other two studies. Also, the relatively small watershed area and large water surface
area of Tampa Bay will also magnify the relative contributions from atmospheric sources.
1.3 STUDY CONCLUSIONS AND RECOMMENDATIONS
In summary, the study results indicate that atmosphere deposition of nitrogen over portions of
the United States may be a significant contribution of nitrogen to aquatic systems. The
application of the project methodology to Chesapeake Bay, Galveston Bay, and Tampa Bay
supports this conclusion; we estimated that over 40% of the total nitrogen load that is delivered
to these water bodies is due to atmospheric nitrogen deposition. Although not confirmed in this
study, increased emissions of nitrous oxides to the atmosphere may be a source of the increased
nitrogen inputs to aquatic systems through the deposition pathway.
One salient point that the user should be aware of while using the estimates derived in this report
is that the results obtained from the project methodology are based on certain assumptions, and
the computations are performed for mean annual flow conditions. The assumptions used in the
application of the current methodology are:
1. Dry atmospheric deposition is equal to wet atmospheric deposition;
however, users of the spreadsheet developed in this effort can
apply a different relationship if desired and supported by dry
deposition data.
2. The nitrogen inputs to the land use categories of pastures, urban,
and wetlands is only from atmospheric deposition. Normally
pastures receive manure application, urban areas receive some
fertilizer application, and the nitrogen inputs to wetlands can also
be due to runoff from neighboring land uses. These are assumed
to be small, negligible, sources in this study.
3. The riverine nitrogen loss parameter is constant for all the
streams/rivers in all the three study sites. Again, spreadsheet
users can employ different loss rates if supported by the available
data.
4. Biological fixation was ignored for crops and forests. It has been
reported that symbiotic fixation in forests can range from 30 to
100kg/ha/yr(Tarrant, 1983). However, there is no indication that
fixation should have a significant effect on nitrogen retention.
5. The retention parameter for fertilizer application and atmospheric
deposition on the cropland is assumed to be the same. In reality,
-------
fertilizers are applied two to three times in a year; and atmospheric
deposition occurs on a daily basis. It is unclear what impact, if
any, this has on the retention parameter.
There is a need to further investigate the assumptions that were used in this report. Following
is a list of some of the areas which need further investigation:
1. Evaluate nitrogen loadings under transient conditions.
2. Investigate and better characterize total nitrogen deposition, and
the relationship between wet and dry atmospheric deposition.
3. Evaluate possible changes in the first-order nitrogen loss rate with
the travel time associated with various river/strearns found in the
watershed.
4. Investigate how changing climate, crops, soil type, and topography
affects the retention parameters for the various land use practices
in the watershed.
5. Evaluate how retention parameters for atmospheric nitrogen that
is deposited on the watershed may differ from those for other
human nitrogen inputs, such as fertilizers and animal waste.
6. Study the effects of nitrogen deposition and its delivery from the
watershed to the watershed outlet. We can account for the
contribution of atmospheric nitrogen deposition on the water
surface; however, there is a paucity in the information regarding
the amount of atmospheric nitrogen that is deposited on the
watershed and is finally delivered to the aquatic system of interest.
7. Investigate the possible impacts, if any, of biological nitrogen
fixation on retention parameters, especially for crops and forests.
1.4 FORMAT OF REPORT
Section 2 describes the EDF methodology in detail. The results obtained from the EDF analysis
along with the major assumptions used in their study are presented. The EDF procedures were
reviewed and refined to develop the project methodology. The project methodology is discussed
in Section 3. In Section 4 of this report we describe and analyze the results obtained from the
application of the project methodology to the Chesapeake Bay, Galveston Bay, and Tampa Bay
watersheds. Sensitivity analysis was performed on key nitrogen load assessment parameters; the
results obtained from the sensitivity analysis are also presented in Section 4. In Section 5, the
study conclusions and recommendations for future work are presented.
-------
SECTION 2.0
REVIEW OF EDF ANALYSIS
In this section we provide a detailed review and description of the procedures, findings, and the
assumptions of the Environmental Defense Fund's (Fisher et al., 1988) study of atmospheric
nitrogen deposition as it relates to the total nitrogen load to the Chesapeake Bay Watershed. In
Section 2.1, the EDF methodology and the results obtained from the application of the
methodology to the Chesapeake Bay watershed are presented. The conclusions derived from the
EDF study are presented in Section 2.2, and assumptions used in the EDF study are listed in
Section 2.3.
It has been reported that in the past century anthropogenic emissions of nitrogen oxides to the
atmosphere of North America have increased (Fisher et al., 1988). The changes in the
composition of the atmosphere have resulted in an increased nitrate content of precipitation and
of surface waters. The coastal waters of the United States are reported to be receiving large
inputs of nitrogen. Fisher et al., (1988) analyzed the contributions of nitrogen to the Chesapeake
Bay above the Fall Line by considering the various possible sources of nitrogen, and how these
sources relate to one another in relation to the total nitrogen delivery to the Bay. The Fall Line
represents the topographic transitional area between the coastal plain and piedmont physiographic
divisions. The above Fall Line watershed segments of the Chesapeake Bay drain through the
river/stream network into the Chesapeake Bay. The below Fall Line segments, which were not
analyzed in the EDF study, drain directly into the Chesapeake Bay.
The EDF report lists the possible nitrogen sources resulting from human activity along with their
contributions. The sources reported consist of atmospheric deposition, runoff from agricultural
land, the outflows from sewage treatment plants, and industrial discharges. The contributions
of nitrogen to the Chesapeake Bay are estimated for the year 1984. The Chesapeake Bay
watershed is depicted in Figure 2.1.
2.1 EDF PROCEDURES USED FOR THE CHESAPEAKE BAY WATERSHED
Fisher et al. (1988) reported that atmospheric nitrogen deposition, a component of acid rain, is
a major source of nitrogen to the Chesapeake Bay and other Atlantic coastal waters, and is a
major factor in the decline of coastal water quality. To derive the above conclusion, the authors
performed nitrogen load computations to the Chesapeake Bay using two alternative procedures,
referred to as the equal retention and differential retention methods.
In the equal retention procedure, the authors first calculated the total nitrogen inputs to the
Chesapeake Bay watershed from point and nonpoint sources. The nonpoint source inputs to the
-------
WEST BRANCH
SUSOUEHANNA
.JUNIATA
LOWER
6USQUEHANNA
UPPER POTOMAC
SHENANDOAH
EAST BRANCH
SUSOUEHANNA
PATUXENT
LOWER
POTOMAC
RAPPAHANNOCK
MATTAPONI
PAMUNKEY
Figure 2.1 Chesapeake Bay Watershed
-------
watershed were comprised of atmospheric deposition, fertilizers and animal waste. The authors
then calculated the percent contribution from individual nonpoint sources using the ratio of the
input from the nonpoint source of interest to total nonpoint inputs. The U.S. EPA (1983)
estimated the total nonpoint and point source loads to the Chesapeake Bay. For the equal
retention procedure the nonpoint load to the bay was differentiated based on the ratios of
individual nonpoint source loads calculated above to estimate the contribution from that source
to the Chesapeake Bay. Additional contributions to the Chesapeake Bay were calculated by
direct atmospheric deposition to the water surface of the Bay.
The differential retention procedure was a refinement of the equal retention procedure, as the
procedure considered the effects of alternative land use categories in the watershed. The land
uses evaluated consist of forest, cropland, urban, and pastures. The authors calculated nitrogen
inputs to the above listed land uses, and then using literature reported retention factors and a
50% loss of nitrogen due to riverine uptake the nitrogen loads to the Chesapeake Bay were
estimated. In the sections below, we describe the two procedures in greater detail and present
the results obtained from both methods.
The anthropogenic nitrogen budget for the entire watershed was calculated using data from U.S.
EPA, Chesapeake Bay Program, the National Atmospheric Deposition Program (NADP), State
Agricultural Agencies, and U.S. Bureau of the Census. The authors used the NADP observed
data for calculating atmospheric nitrogen loadings to the watershed land surface as well as on
the water surface of the Bay. The atmospheric deposition to the land surface of the Chesapeake
Bay watershed was computed by dividing the watershed into nine segments (North Branch
Susquehanna, West Branch Susquehanna, Susquehanna below Sunbury, West Chesapeake,
Patuxent, Eastern Shore, Potomac, York-Rappahannock, and James River; see Figure 2.1). All
other nitrogen load calculations were done for the entire watershed. The total fertilizer
application in the watershed was calculated using fertilizer sales data from the State Agricultural
Agencies, and the Agricultural Census data was used to calculate animal population in the
watershed. Data on point source nitrogen loadings were obtained from U.S. EPA, Chesapeake
Bay Program.
2.1.1 Equal Retention Procedure
The steps used in calculating nitrogen loads to the watershed using the equal retention procedure
are listed below:
1. The atmospheric nitrogen deposition to the Chesapeake Bay watershed was calculated
using the National Atmospheric Deposition Program's (NADP) wet atmospheric nitrogen
deposition data. Based on the atmospheric deposition trends in the watershed the authors
divided the watershed into nine basins — North Branch Susquehanna, West Branch
Susquehanna, Susquehanna below Sunbury, West Chesapeake, Patuxent, Eastern Shore,
Potomac, York-Rappahannock, and James River. The Chesapeake Bay water surface
area was used to calculate direct atmospheric deposition to the Bay water body. Total
8
-------
atmospheric deposition was assumed to be two times wet deposition to account for
unavailable dry deposition data.
2. The data on total nitrogen fertilizer applied in a county were obtained from state
agricultural reports. These data were used to calculate commercial fertilizer usage for
the states of Pennsylvania and Virginia. For counties which were partially in the
watershed the authors used a planimeter to calculate the area and the percent of the
county in the watershed. The total amount of fertilizer consumed in the county was then
multiplied by the percentage of the county in the watershed to estimate nitrogen
consumed in the part of the county which was in the watershed. For those states where
the 1984 fertilizer consumption was unavailable, the authors used the 1982 Census of
Agriculture data.
3. Using the Agricultural Census data, the authors calculated the type and animal population
in each county of the watershed. The total population of a given farm animal was then
multiplied by the average amount of nitrogen that the animal excretes each year to
calculate the total nitrogen produced by that species of animal in the county.
4. The authors obtained the point source total nitrogen input from the U.S. EPA,
Chesapeake Bay Program for the year 1985. The point sources included discharges from
sewage treatment plants as well as industrial sources. All STP's with flow rates over 0.3
million gallons per day (1136 m3 per day) were included in calculating nitrogen loads;
this accounted for 96% percent of the total flow from all STP's. For industrial sources,
observed nitrogen loads were used whenever data were available, or were estimated
based on a typical value for that industry type.
The total nitrogen loads calculated by EDF to the land and water surfaces of the Chesapeake
Bay Watershed using the steps listed above are presented in Table 2.1.
As mentioned above, in this procedure the authors assumed that the contribution of each
nonpoint source to the Bay is proportional to the loading of each nonpoint source to the
watershed. The nitrogen loadings to the Chesapeake Bay as estimated by the US EPA (1983)
are listed in Table 2.2. Table 2.3 below lists the results obtained from the equal retention
procedure. As shown in Table 2.3, the point source loadings and the sum of nonpoint
contributions is the same as estimated by EPA (1983) and as listed in Table 2.2. The total
nonpoint contribution to the watershed as listed in Table 2.2 was multiplied by the last column
in Table 2.1 to calculate loadings from individual sources of nonpoint pollution. The values of
direct atmospheric deposition to the Bay were kept the same as in Table 2.1, as the EPA
estimates of N loads to the Bay in Table 2.2 do not account for direct atmospheric deposition
to the Bay.
-------
Table 2.1 Calculated Nitrogen Loadings to the land and water surfaces of the Chesapeake
Bay Watershed in 1984
Source 106 kg N yr1 % of Total % of NFS
Precipitation-Land
Nitrate 143 23 25
Ammonium 79 13 14
Animal Waste 195 31 34
Fertilizer 158 25 27
NFS subtotal 575 -- 100
Point Sources 41 7
Precipitation-Bay
Nitrate 8.3 1.3
Ammonium 4.6 0.7
Total 629 100
2.1.2 Differential Retention Procedure
In the Differential Retention procedure, the authors calculated the nitrogen loadings from each
N source to the various land types found in the watershed as shown in Figure 2.2. The authors
then estimated the amount of nitrogen that will be lost in runoff based on existing nitrogen
retention studies. In the following paragraphs we describe the literature sources used by the
authors to estimate retention parameters, i.e. the fraction of the nitrogen loading that remains,
from the various land uses in the watershed.
2.1.2.1 Cropland
Total fertilizer application to the cropland acreages of the Chesapeake Bay Watershed was
calculated as 158 x 106 kg N yr1. The product of average atmospheric deposition (the authors
did not use the watershed delineation as used for the equal retention procedure) rate with the
cropland area was used to calculate atmospheric N deposition on croplands. The authors
reported that 25.6 x 106 kg N yr1 was deposited as nitrate deposition and 14.1 x 106 kg N yr1
as ammonium. The total N load on cropland as reported by the authors is the sum of fertilizer
and atmospheric deposition, and is equal to 197.7 x 106 kg N yr1, with 13% from atmospheric
nitrate deposition, and 6% from atmospheric ammonium.
10
-------
Table 2.2 Nitrogen Loadings to the Chesapeake Bay (US EPA, 1983)
At Fall Line Average Year*
Point Sources 6.3
Non-Point Sources 50.6
Sub Total 56.9
Below Fall Line
Point Sources 26.6
Non-Point Sources 16.2
Sub Total 42.8
Watershed Total
Point Sources 32.9
Non-Point Sources 66.8
Total 99.7
The EPA estimated the loadings for the period May-October, EDF multiplied the
EPA (1983) values by 1.5 to obtain yearly nitrogen loads to the Bay.
Table 2.3 Equal Retention Estimates of Nitrogen Sources to the Bay
Source 106 kg N yr1
Point Source 32.9
Non Point Sources:
Atmospheric Nitrate (N) 16.6
Atmospheric Ammonium (N) 9.2
Fertilizer (N) 18.4
Animal Waste 22.6
NPS Contribution 66.8
Subtotal, Excluding Direct Deposition 99.7
Deposition to Surface of Bay:
Atmospheric Nitrate (N) 8.3
Atmospheric Ammonium (N) 4.6
Subtotal, Direct Deposition 12.9
TOTAL 112.6
Note: Point Source and NPS Contributions are from EPA Analysis (1983).
The EPA estimates are modified to account for direct deposition
to the Bay.
11
-------
The authors reported that the EPA estimates of total nitrogen loadings to the Bay originating
from cropland was 59.8 x 106 kg N yr1. It was assumed that 13% of runoff from cropland can
be attributed to atmospheric nitrate and 6% to atmospheric ammonium. Based on this
assumption, the authors divided the EPA cropland loading to differentiate the source of nitrogen
in cropland runoff.
Runoff N loadings from cropland, based on a series of field measurements were reported to
range from 0.8 to 20 kg N ha'1 yr1 (Smullen et al., 1982). The authors used 20 kg N ha'1 yr1
as the runoff loading rate; this rate when multiplied by the cropland area (2952855 ha) resulted
in 59.1 x 10* kg N yr1 load from cropland. The authors calculated a 0.70 cropland retention
factor which also included river uptake. This retention factor can be obtained by the following
equation:
Bay Loading (59.1 x 106 kg N yr1)
Retention Factor = 1 = 0.70 (2.1)
Watershed Loading (197.7 x 106 kg N yr1)
2.1.2.2 Forests
Nitrogen retention factors as reported in the literature for forested watersheds range from 0.52
to 0.99 (Kelly and Meagher, 1986; Likens et al., 1977; Weller et al., 1986; Schreiber et al.,
1976; and Linthurst, 1988). In the EDF study the authors used a retention parameter of 0.8 for
forested watersheds. The forest runoff was further reduced by 50% to account for riverine
uptake thus resulting in a overall retention factor of 0.90. i.e. only 10% of the N input to forests
reached the Bay.
2.1.2.3 Pastures
The two sources of nitrogen inputs to pasture land consist of atmospheric deposition and the
application of animal waste. The atmospheric nitrogen deposition on the pastures was calculated
using the area occupied by pastures in the watershed. The loads were reported to be 23.6 x 106
kg N yr"1 from atmospheric nitrate and 13.0 x 106 kg N yr"1 from atmospheric ammonium. Data
from Robbins et al. (1972) cited in the EDF report indicates that only 1.5% to 5% of nitrogen
in animal waste applied to land reaches the stream via runoff. The authors assumed that 5% of
nitrogen in animal waste runs off the land, and this runoff load was further reduced by 50% to
account for riverine uptake, thus resulting in a overall retention factor of 0.975. The retention
factor for atmospheric nitrogen loads were assumed to be 0.70, which was reported to be the
same as the cropland retention factor.
12
-------
Figure 2.2 EDF's Differential Retention Procedure Flowchart
13
-------
Urban
The cities located in the Chesapeake Bay Watershed were assumed to drain directly into the Bay,
hence the river uptake losses were assumed to be zero. It was expected that the urban lands will
retain between 0% (complete washoff) to 70% (same as cropland) of nitrogen inputs. The study
assumed a 0.35 retention factor for urban areas which is an average of 0% retention and 70%
retention.
The retention factors used for the various N sources are listed in Table 2.4 along with calculated
nitrogen loadings to the Bay.
2.2 EDF STUDY CONCLUSIONS
Atmospheric nitrate deposition is reported to be the major factor in the decline of coastal waters.
From the estimates of various sources of nitrogen to the Chesapeake Bay, the individual
contributions under the differential retention procedure were quantified as 23% from point
sources, 39% from atmospheric deposition (includes nitrate and ammonium), 34% from
fertilizer, and 4% from animal wastes. It is reported by the authors that, in the absence of
further nitrogen oxide emission controls, emission in the Chesapeake Bay region will increase
by 44% by the year 2030.
The Chesapeake Bay Program is currently evaluating strategies for attaining a 40% reduction
in nutrient loads (nitrogen and phosphorus) reaching the Bay. The EDF study indicates that if
nitrogen loads from other sources are reduced by 40%, but atmospheric loadings are not
changed, then the overall nitrogen reduction to the Bay will be only 16% instead of 40%.
2.3 ASSUMPTIONS USED IN THE EDF STUDY
The following is a list of the assumptions that were used in the EDF study, as documented in
their report (Fisher et al., 1988):
1. Dry nitrogen deposition is equal to wet deposition.
2. N source from forest runoff is due to atmospheric N.
3. There is a 50% reduction in the nitrogen load reaching the bay due to riverine
uptake.
4. For the equal retention procedure it was assumed that the loading to the bay are
proportional to inputs of fertilizer, animal waste and atmospheric deposition on
the land surface of the watershed.
14
-------
Table 2.4 Retention Factors and N Loadings Using Differential Retention Procedure.
Source Fraction Nitrogen
Retained Loadings
(106 kg N yr1)
Point Source
Below Fall Line 0.00 26.6
Above Fall Line 0.50 6.3
Subtotal 32.9
Direct Atmospheric Deposition to Bay
Atmospheric Nitrate Nitrogen 0.00 8.3
Atmospheric Ammonium Nitrogen 0.00 4.6
Subtotal 12.9
Cropland
Fertilizer 0.70 47.7
Atmospheric Nitrate Nitrogen 0.70 7.8
Atmospheric Ammonium Nitrogen 0.70 4.3
Subtotal 59.8
Forest
Atmospheric Nitrate Nitrogen 0.90 9.0
Atmospheric Ammonium Nitrogen 0.90 5.0
Subtotal 14.0
Pastures
Animal Waste 0.975 4.9
Atmospheric Nitrate Nitrogen 0.70 7.1
Atmospheric Ammonium Nitrogen 0.70 3.9
Subtotal 15.9
Urban
Atmospheric Nitrate Nitrogen 0.35 2.8
Atmospheric Ammonium Nitrogen 0.35 1.6
Subtotal 4.4
TOTAL 140.0
Note: Retention factors from various land uses include a 50% river uptake.
15
-------
5. All fertilizer sold is applied to the crops i.e. applications are derived from
fertilizer sales.
6. Atmospheric deposition from individual monitoring sites represent larger areas.
7. Storage, handling, and volatilization losses of animal waste are ignored.
8. Other fluxes of N from the watershed are not altered by increased atmospheric
N deposition.
9. Assumed 70% nitrogen retention from all croplands.
10. Assumed 97.5% retention percentage for animal waste applied to cropland,
however, the retention of atmospheric deposition was assumed to be 70%.
11. Assumed 35 % nitrogen retention from urban areas.
In Section 3, we describe the current methodology which was developed by refining some of the
EDF assumptions that are listed above.
16
-------
SECTION 3.0
METHODOLOGY USED IN ESTIMATING NITROGEN LOADS TO AQUATIC
SYSTEMS
In Section 2.0 the EDF procedures were reviewed and discussed in detail. Based on the EDF
assumptions that needed to be reevaluated, the following tasks were included in this effort:
1. Investigate retention parameters for cropland, urban, and pasture
areas.
2. Investigate the assumption of 50% loss of nitrogen due to riverine
uptake.
3. Evaluate application amounts of fertilizer and animal waste in
relation to agronomic application rates by crops (i.e. based on crop
needs).
4. Divide watershed into subbasins to account for varying NFS
loadings from alternative land uses and riverine travel time to the
Bay (or watershed outlet).
5. Investigate relationship between wet and dry atmospheric
deposition.
In this section we describe the steps and procedures of the current methodology used in
estimating nitrogen loads to Chesapeake Bay, Galveston Bay and Tampa Bay. The following
is a list of the EDF assumptions that are modified using the current methodology:
1. The nitrogen loadings from the watershed are calculated by dividing the watershed
into smaller subbasins to better represent the variation in land use in the
watershed, rather than lumping all the land uses together, as was done in the EDF
analysis.
2. Fertilizer applications are calculated using crop nitrogen demand, rather than
fertilizer sales data.
3. Variable relationship between wet and dry atmospheric deposition is used.
4. Riverine losses are calculated using travel time and nitrogen losses during
travel to the watershed outlet.
3.1 CALCULATIONAL STEPS IN CURRENT METHODOLOGY
The steps used in estimating nitrogen loadings to aquatic systems using the current methodology
are presented below:
17
-------
STEP 1. Define Study Region. Drainage Areas. Subbasins and Drainage
networks, and Land Use Distribution - Define and calculate the
drainage areas of the three study sites, i.e. the area that directly
drains into Chesapeake Bay, Galveston Bay, and Tampa Bay. Once
the study regions are defined and their total drainage area
calculated, the next step was to map and list the stream/river
network that carries the loadings from the watershed segments to
the watershed outlet. The stream/river network aids in dividing
the watershed into subbasins. For each subbasin we then calculate
the land use distribution.
STEP 2. Identify Data Needs - The following data items and categories
were needed for estimating nonpoint and point source contributions
of nitrogen:
a. Fertilizer
b. Meteorological
c. Point source
d. Animal waste
e. Land use
f. Sewage plants discharge
g. Industrial discharges
h. N deposition (wet and dry)
i. Channel network and morphology
STEP 3. Calculate Loadings from Subbasins and Local Tributary Areas
The two sources of nitrogen loads at the subbasin outlet consist of
nonpoint sources and point sources. The nitrogen loads to the bay
or watershed outlet will be estimated by calculating individual
subbasin loads and then routing this load to the bay or watershed
outlet using travel time and riverine uptake parameters. In the
following subsections we describe the procedures that will be used
to calculate the loadings from nonpoint and point sources.
3.2 DEFINE NPS LOADS BY LAND USE
Nonpoint loads to the Bay or Watershed Outlet are a function of retention by the land use and
loss due to riverine uptake. Calculation of nitrogen loads from all the land uses was a two step
procedure: the first step was to estimate the total nitrogen inputs for all the land uses found in
each subbasin, and the second step then involved the calculation of the edge-of-field nitrogen
loads from each land use based on the retention factor of the land use. For each land use
practice, we calculated the percent contribution of each nitrogen source with respect to the total
nitrogen input to the land use. This percent contribution was used to calculate the source of
18
-------
nitrogen from the total nonpoint nitrogen load generated from the land use.
Nitrogen Inputs: Following is a list of typical land use categories for which
nitrogen inputs were calculated:
Forest
Croplands
Wetlands
Pasture/Idle
Urban
Nitrogen inputs primarily consist of atmospheric deposition.
Using the unit area observed wet and dry deposition data
in conjunction with the area occupied by the forest in the
subbasin, we calculated the total nitrogen input to forests.
The nitrogen inputs to croplands consist of fertilizer
application, manure application, biological fixation
(for legume crops), and atmospheric deposition. For the
major crop grown in each subbasin we calculated the typical
agronomic fertilizer application rate. It was assumed that
manure is only applied to croplands and pasture/idle lands.
Based on the ratio of cropland area to pasture/idle area, an
area-weighted manure application rate was calculated.
The atmospheric deposition of nitrogen was calculated using
observed unit area wet and dry atmospheric N deposition
data. The area occupied by the cropland was then used to
calculate the total nitrogen input to croplands.
Nitrogen inputs to wetlands consist of atmospheric
deposition.
The two primary sources of nitrogen inputs consist of
atmospheric deposition and manure application. Manure
application amounts were calculated using the area
weighted procedure described above.
Nitrogen inputs consist of atmospheric deposition only.
For each of the land uses listed above we calculated the ratio of each individual
nitrogen source to the total nitrogen input to the land use; these ratios were called
Input Source Ratios. An example snowing the calculation of the Input
Source Ratio is given below:
Total Nitrogen Input to Cropland (TN) =
Fertilizer (F) -I- Manure (M)
+ Biological Fixation (BF)
+ Atmospheric Deposition
(AD)
19
-------
Atmospheric Deposition Input = [Atmospheric Deposition (AD)]/[Total
Source Ratio for Cropland Nitrogen Input to Cropland (TN)]
These calculated Input Source Ratios were then used to differentiate the sources
of nitrogen from each land use, adding identical nitrogen sources resulted in the
nitrogen load from each source at the subbasin outlet.
Nitrogen Outputs: Literature reported nonpoint loading retention parameters (e.g.,
Kelly and Meagher, 1986; Liken et al., 1977; Weller et ah, 1986; Schreiber et
al., 1976 for forest land use; and Robbins et al., 1972 for pastures) for each of
the land use categories found in the subbasin were used to estimate the amount
of nitrogen loading from the land use of interest. If retention parameters for all
the land uses were unavailable, then literature studies (e.g., Donigian et al., 1994)
which report NFS edge-of-field loadings for the land use of interest were used.
These loading rates were then used to back-calculate the retention factor from
equation 2.1 in Section 2.1.2.
3.3 DEFINE POINT SOURCE LOADS
The point sources of interest consist of industrial and sewage treatment plant discharges. The
point sources discharge directly into the stream/river and the retention or loss of nitrogen from
point sources is due to riverine uptake only. The point sources that discharge directly into the
bay or watershed outlet were not subject to riverine losses.
3.4 SUBBASIN LOADS
The total nitrogen load at the subbasin outlet is the summation of nonpoint and point source
loads. However, we routed the nonpoint and point source loadings individually from each
subbasin in the watershed. We computed the total NFS load from the subbasin by multiplying
the edge-of-field export from each land use by its corresponding area.
The source of nitrogen resulting from nonpoint source loadings from each land use was
calculated by multiplying the input source ratios by the total land use nitrogen load. This
procedure was used for all the land uses in the subbasin. The total subbasin load was computed
by summing the loads from all the land uses. Summing up identical nitrogen sources from all
the land uses in the subbasin resulted in the total loads from individual nitrogen sources in each
of the subbasins.
After the total subbasin NFS nitrogen load was calculated, the next step was to calculate the ratio
of N from each source (fertilizer, biological fixation, manure, and atmospheric deposition) to
the total NFS nitrogen load at the subbasin outlet. This ratio was called the Output Source
Ratio. This ratio is used to differentiate the various sources of nitrogen at the bay or watershed
outlet.
20
-------
STEP 4. Calculate Stream Losses and Contributions to Bay or Outlet
The flow of nitrogen from the watershed subbasins to the bay or
watershed outlet is shown in Figure 3.1. The load from each
subbasin was routed to the bay or watershed outlet using Travel
Time from each subbasin. The nitrogen loadings to the bay or
watershed outlet were assumed to be dependent on riverine uptake
and the travel time of the stream/river to the outlet. The nitrogen
load to the bay or watershed outlet from all the subbasins in the
watershed were calculated using the following first-order decay, or
exponential relationship (Barnwell, 1993, personal
communications):
Subbasins
TN = V { Subbasin^ *exp [ - k *
a-l
where:
TN = Total Nitrogen load to the bay or watershed outlet
Subbasin; = Total Nitrogen load from subbasin i (includes NFS and
point sources)
Tj = Travel time from subbasin i
k = First-order riverine N loss constant (literature values of
riverine loss will be used)
Riverine nitrogen losses include settling (sedimentation) of
particulate nitrogen, denitrification, ammonia volatilization, and
some algal and plant uptake (where subsequent settling removes
the nitrogen).
The nonpoint and point sources were routed separately using the
above equation. The above equation was not used for subbasins or
local tributary areas that drain directly to the bay or watershed
outlet, as the loadings from these local subbasins or local
tributaries are not subject to riverine losses.
The individual nitrogen source in the nonpoint source loadings to
the bay or watershed outlet from each subbasin was calculated by
multiplying the total load (as calculated by Equation 3.1) by the
Output Source Ratios calculated for each subbasin. The procedure
showing the calculation of nonpoint nitrogen loads from a subbasin
is depicted in Figure 3.2.
21
-------
Figure 3.1 Methodology Flow Chart for Estimating Nonpoint and Point Source Loads
22
-------
Example of NFS Load Computations from each Subbasin
Landuae A
Landuae B
Landuae C
N-lnputa
Atm. Depo. « 10
Manure * 40
Total Input « 50
N-lnputa
Atm. Depo. - 10
Manure - 40
Fertilizer « 100
Total Input • 150
N-lnputa
Atm. Depo • 10
Total Input • 10
I Calculation of Input Sourca Ratio* [
Atm Input Ratio • 10/50 • 0.2 Atm. Input Ratio • 10/150 • 0.07 Atm. input Ratio • 10/10 • 1
Manure Input Ratio • 40/50 • 0.8 Manure Input Ratio • 40/150 • 0.27
Fert. Input Retio - 100/150 • 0.66
Assume 80% Retention Aaaume 70% Retention Assume 60% Retention
Unit Area Load • N-lnput - (N-lnput x Fraction Retained)
Unit Area Load • 5 Unit Area Load • 45 unit Area Load « 4
Distribution of Loads Distribution of Loads Distribution of Loads
Load from Each N Source • N Source Ratio x Unit Area Load
Atm Load • i ' 4 • 4
Atm. Load • 0.2 '5-1
Manure Load • 0 8 ' 5 • 4
Atm. Load • 0 07 * 45 • 3.15
Manure Load • 0 27 • 45 • 12.15
Pert Load • 0 66 • 45 • 29 70
Total Subbasin Load « Unit Area Load 1rnm Landuse A->B->C
Total Subbesm Load • 5 •* 45 » 4 • 54
Nitrogen Sources of Subbasin Load
Atm Load • Atm Load from Landuse (A-B-C) • 1-3 15*4-8.15
Manure Load • Manure Load (torn Lenduse (A-B) * 4*12 15-16 15
Fert Load • Pert. Load trom Landuse A • 29 70
Calculation of Output Source Ratios
Output Source Ratios in Subbasin Load • Source Load / Total Subbasin Load
Aim Subbasin Ratio • 8 15/54 • 0 15
Manure SuDbasm Ratio • 16 15/54 - 0 30
Fert Subbasin Ratio • 29 7/54 • 0 55
Assume 50% Subbasin Load Lost Due to Riverine Uptake
Loading to Bay or Watershed Outlet • 54/2 • 27
Source Contributions in Bay or Watershed Load from Subbasin
Atm Contribution • 0 15*27 • 4 05
Manure Contribution • 0 30'27 • 6 10
Fertilizer Contribution • 055*27 • 14.85
Figure 3.2 Example of NFS Load Computations from each Subbasin
23
-------
The total loadings to the bay or watershed outlet is the sum of
loads from all the subbasins in the watershed plus direct
atmospheric deposition on the water surface. The sum of all
identical nitrogen source loadings to the bay or watershed outlet
from all the subbasins in the watershed divided by the total Load
from all the subbasins was used in calculating the contribution of
individual nitrogen sources at the bay or watershed outlet.
STEP 5. Perform sensitivity analysis
After initial computations of nitrogen loadings from the three sites
of interest, additional analyses were performed to test the
sensitivity of the nitrogen output to the following parameters and
assumptions used in the methodology:
a. Vary retention parameters for all land uses
b. Vary river/stream loss (k in equation 3.1)
c. Evaluate range of literature reported loading rates for all land uses
d. Investigate relationship between wet and dry deposition.
e. Evaluate sensitivity to fertilizer application amounts, i.e. varying fertilizer
applications to 1.25 and 1.50 times the agronomic application rates.
The results from the sensitivity analysis are presented in Section 4.4.
24
-------
SECTION 4.0
ASSESSMENT OF NITROGEN LOAD FROM STUDY SITES
The methodology developed for this project was described in detail in Section 3. In this section
the results from the application of the methodology to the three study sites are presented. In
Section 4.1 we first describe the results obtained from the application of the project methodology
to the Chesapeake Bay watershed. These results are then compared to the EDF results obtained
using the differential retention procedure only. The results obtained from the application of the
methodology to the remaining two study sites — Galveston Bay, and Tampa Bay — are presented
in Sections 4.2, and 4.3 respectively. Sensitivity analysis was performed on the following
parameters:
1. Ratio of Wet to Dry Atmospheric Deposition;
2. First-order nitrogen loss rate in rivers; and
3. Fertilizer application amounts (varying agronomic recommended rates).
The results obtained from the sensitivity analysis are presented and discussed in Section 4.4.
4.1 APPLICATION OF THE METHODOLOGY TO CHESAPEAKE BAY WATERSHED
For the application of the methodology to the Chesapeake Bay watershed, we divided the
watershed into fourteen subbasins ~ East Branch Susquehanna, West Branch Susquehanna,
Juniata, Lower Susquehanna, Conowingo, Upper Potomac, Shenandoah, Lower Potomac,
Rappahannock, Mattaponi, Pamunkey, James, Appomattox, and Patuxent (see Figure4.1). The
input data (nonpoint and point sources) for the above listed subbasins were obtained from the
watershed modeling study conducted by Donigian et al. (1994) for the Chesapeake Bay
watershed. The nonpoint, point, and atmospheric deposition data of total nitrogen inputs used
in the Chesapeake Bay application along with the EDF nitrogen inputs are presented in Table
4.1.
The land uses from which nonpoint source loads were computed consisted of cropland, forest,
pasture, and urban land use. The nonpoint source inputs to the Lotus 123* spreadsheet consist
of atmospheric deposition from all land uses, manure application to cropland and pastures, and
fertilizer application to cropland. We used the average annual atmospheric nitrogen deposition
data as reported by Fisher et al. (1988). The manure and fertilizer application rates used for all
the subbasins in the watershed were calculated by Donigian et al. (1994), based on fertilizer and
manure application data provided by local and state agencies.
25
-------
WEST BRANCH
SUSQUEHANNA
JUNIATA
LOWER
6USOUEHANNA
UPPER POTOMAC
SHEKANDOAH
EAST BRANCH
SUSOUEHANNA
PATUXENT
LOWER
POTOMAC
RAPPAHANNOCK
MATTAPONI
PAMUNKEY
Figure A.I Chesapeake Bay Watershed with Subbasin Boundaries
26
-------
Table 4.1 Comparison of Current Nitrogen Inputs to EDF Nitrogen Inputs
Current Inputs EDF Inputs
106 kg/yr 106 kg/yr
Direct Bay Inputs
Point Sources Above Fall Line 12.7 12.6
Point Sources Below Fall Line 27.0 27.0
Atmospheric Deposition 13.0 13.0
Forest
Atmospheric Deposition 112.0 139.0
Cropland
Atmospheric Deposition 38.1 39.7
Fertilizer 171.0 158.0
Animal Waste 49.5 0.0
Pasture
Atmospheric Deposition 19.0 36.6
Animal Waste 32.8 195.0
Urban
Atmospheric Deposition 14.2 6.7
GRAND TOTAL 489.0 627.0
27
-------
Other inputs needed for the spreadsheet consist of edge-of-field loads to the streams and travel
time from each subbasin to the watershed outlet. Donigian et al. (1994) in their Hydrological
Simulation Program-Fortran (HSPF) (Johanson et al., 1984) model application study of the
Chesapeake Bay watershed computed the total nonpoint loads from all the subbasins in the
watershed; these loads from each land use for all the subbasins are listed below in Table 4.2.
The land use areas for all the subbasins are listed in Table 4.3 and the above Fall Line point
source loads are listed in Table 4.4. Travel time from each subbasin to the watershed outlet was
calculated by Bicknell (1993, personal communications). In Table 4.5 a sample input data set
to the Lotus 123 spreadsheet for the East Branch Susquehanna Subbasin is presented.
Nonpoint and point source data were input to the spreadsheet for all the subbasins in the
Chesapeake Bay watershed. The dry atmospheric deposition was assumed to equal wet
deposition; this was done to replicate the EDF assumption. As discussed in the methodology,
the nonpoint and point source loads were routed separately to the Chesapeake Bay. Input and
Output Source Ratios (as discussed in Section 3) were calculated for all the subbasins. Output
Source Ratios were used to compute the source of nitrogen from the subbasin load that was
delivered to the Chesapeake Bay. The first-order rate constant for nitrogen loss in streams, k,
was determined to be 0.08/day to replicate the 50% riverine loss rate used by EDF; this rate is
typical of nitrogen loss/decay rates reported in the literature (Barnwell, 1993, personal
communications). Identical sources of nitrogen resulting from all the subbasin loads to the
Chesapeake Bay were added together to obtain the total contributions resulting from each source
of nitrogen. The nitrogen loads from various land uses and nitrogen sources are listed in Table
4.6. The percent contributions to the Chesapeake Bay from the various nitrogen sources in the
watershed are presented in Table 4.7.
4.1.1 Comparison of New Methodology Results with EDF Results
The total nitrogen load computed using the current methodology is 115 million kilograms, which
compares favorably with the EDF estimate of 140 million kilograms, a difference of 25 million
kg of nitrogen or 18% of the EDF value. One of the reasons for this is the larger watershed
area used in the EDF analysis. The watershed area above the Fall Line as estimated by The
Chesapeake Bay Program U.S. EPA, as reported by Donigian et al. (1994) was 13 million
hectares as compared to 16 million hectares estimated by EDF, a difference of 20% of the EDF
value. The EDF study reported larger acreages under agricultural, pasture, and forest land uses
by 0.5, 1.3, and 2.2 million hectares respectively. However, EDF reported 0.5 million less
hectares under the urban land use. In the following sections we compare and discuss the
nitrogen loadings from individual land use practices in the Chesapeake Bay watershed as
estimated by the current methodology and the EDF analysis.
4.1.1.1 Cropland
As noted in Section 3.0, nitrogen loads resulting from the cropland area of the Chesapeake Bay
watershed were due to three sources of nitrogen, -- fertilizer, atmospheric deposition, and
manure. In the EDF analysis the authors assumed that nitrogen loadings from cropland were
28
-------
only due to fertilizer and atmospheric deposition.
The total fertilizer load to the Chesapeake Bay as estimated by the current methodology was 25
million kg less than that estimated by the EDF study. The reasons for this discrepancy in the
loading estimates are as follows:
1. The nitrogen loads were calculated for the current methodology by
dividing the watershed in fourteen subbasins; however, EDF
results are based on considering the entire watershed as one basin
with four different land uses.
2. The cropland area used by the EDF study was greater than the
current methodology area by 0.48 million hectares. The current
methodology area is 16% less than the EDF estimated area.
3. In the EDF analysis the authors calculated the nitrogen loads to the
Chesapeake Bay using a constant nitrogen retention parameter for
the cropland in the watershed; whereas, in the current
methodology we calculated fourteen different nitrogen retention
parameters which were based on the estimated nitrogen loads
calculated for the fourteen subbasins by Donigian et al. (1994).
The range of retention parameters used in the current methodology
range from 0.80 to 0.90; however, in the EDF study a constant
retention parameter of 0.70 was used.
4. In the EDF analysis a 50% loss of nitrogen load from the cropland
segment was assumed due to riverine uptake; however, in the
current analysis we routed the nitrogen loads from each of the
fourteen subbasins to the Bay using travel time and a decay
parameter calculated for average annual flow conditions. The
decay parameter of 0.08/day used in the current methodology
replicated the EDF assumption of 50% nitrogen loss due to
riverine uptake (Barnwell, 1993, personal communications);
however, in our analysis the total nitrogen loss in the rivers was
calculated as a product of the first-order loss parameter and the
estimated travel time in each subbasin.
The EDF study assumed no manure application to cropland, hence we cannot compare manure
loadings from cropland. The fertilizer and manure applications to cropland as reported by
Donigian et al., 1994, were derived from application rates supplied by state and county
personnel. The total nitrogen load to the Chesapeake Bay as a result of fertilizer and manure
application was estimated as 27.5 million kg, which is less than the 48 million kg fertilizer
nitrogen load estimated by the EDF study. Atmospheric loads to the Chesapeake Bay due to
deposition on cropland are 8.85 million kg less than the EDF estimates, and this can be
29
-------
Table 4.2 Estimated Mean Annual Total Nitrogen Loads from the Subbasins in the
Chesapeake Bay Watershed (from Donigian et al., 1994)
Cropland Forest Pasture Urban
106kg lO'kg 106kg 106kg
East Branch Susquehanna 9.71 1.87 9.83 3.93
West Branch Susquehanna 5.37 0.70 7.55 0.88
Juniata 2.73 0.49 5.07 0.51
Lower Susquehanna 8.93 1.11 5.36 2.03
Conowingo 2.71 0.76 0.65 0.96
Upper Potomac 1.16 1.50 3.25 0.52
Shenandoah 2.32 1.19 1.63 0.34
Lower Potomac 8.77 4.07 6.78 2.51
Rappahannock 1.15 0.77 0.94 0.15
Mattaponi 0.30 0.01 0.11 0.03
Pamunkey 0.74 0.07 0.23 0.09
James 2.96 1.53 1.89 0.53
Appomattox 0.98 0.15 0.27 0.13
Patuxent 0.21 0.03 0.05 0.33
explained by the reasons listed above, i.e. different retention parameters and different loss rates
by subbasins.
4.1.1.2 Forest
The nitrogen inputs to the forest land use were assumed to be from atmospheric nitrogen
deposition only. The current methodology nitrogen loads from forested land use are higher than
7.7 million kg as compared to the EDF estimate. This suggests that the EDF assumed a high
nitrogen retention parameter (0.80) for the forest land use. The retention parameters for the
fourteen subbasins calculated using load estimates from Donigian et al. (1994) ranged from 0.40
to 0.92, with an average retention value of 0.68. Fisher et al. (1988) reported that the literature
cited nitrogen retention values for forested land areas range from 0.52 to 0.99 (Kelly and
Meagher, 1986; Likens et al., 1977; Weller et al., 1986; Schreiber et al., 1976; and Linthurust,
1988).
4.1.1.3 Pasture
The two sources of nitrogen inputs to pastures consisted of manure and atmospheric deposition.
The nitrogen loads to the Bay from these two sources using the current methodology were
estimated to be less than the EDF estimates. The atmospheric deposition was estimated to be
less than 6.5 million kg, while the manure component was estimated to be lower by 1.11 million
kg, compared to the EDF values. The reason for lower loading to the Bay using the
30
-------
Table 4.3 Land Use Areas in the Chesapeake Bay Watershed (from Donigian et al., 1994)
East Branch Susquehanna
West Branch Susquehanna
Juniata
Lower Susquehanna
Conowingo
Upper Potomac
Shenandoah
Lower Potomac
Rappahannock
Mattaponi
Pamunkey
James
Appomattox
Patuxent
Total Area
Cropland
Itfha
602
209
173
386
140
81
134
350
74
19
42
190
54
15
Forest
lO'ha
258
99
56
83
34
147
167
164
91
4
19
224
34
8
Pasture
Itfha
1730
1430
594
492
54
778
443
453
233
121
195
1270
241
29
Urban
ICPha
332
68
37
114
40
48
37
173
16
8
20
63
18
34
2470
8070
1389
1008
Table 4.4 Above Fall Line Nitrogen Point Loads (from Donigian et al., 1994)
Point Loads
106kg
East Branch Susquehanna
West Branch Susquehanna
Juniata
Lower Susquehanna
Conowingo
Upper Potomac
Shenandoah
Lower Potomac
Pamunkey
James
Patuxent
Total
3.50
1.20
0.51
1.88
0.74
0.34
0.86
2.19
0.03
0.92
0.53
12.70
31
-------
current methodology can be attributed to the following:
1. Nitrogen inputs to the Chesapeake Bay watershed as estimated by
the EDF study are higher than the estimates derived from Donigian
et al., 1994 (Table 4.1). The reason for higher inputs is because
the land use area under pastures as calculated by EDF was greater
than the area estimated by the current methodology by 1.33 million
hectares, the current methodology area is 51% of the EDF area.
Fisher et al. (1988) estimated that the pasture area occupied 16.5%
of the total watershed area, while in the current methodology we
estimate that pastures occupy 10.6% of the watershed area.
2. The EDF study assumed that there were no losses during the
handling and storage of manure, and hence all the animal waste
produced in the county was applied to the watershed. However,
in the current methodology the manure application accounted for
manure storage, handling, and volatilization losses. The animal
waste application accounts for 16.8% of the total nitrogen input
under the current methodology; however, for the EDF study it
accounts for 31.1% of total nitrogen application.
3. The EDF study calculated an average rate of atmospheric
deposition (wet+dry) using wet deposition data from the four sites
located in the watershed; the atmospheric deposition data ranged
from 11.40 to 17.00 kg/ha. The average rate (15.06 kg/ha) of
atmospheric deposition was used to calculate the total nitrogen
input to the land use from atmospheric deposition. In the current
methodology we used data from each of the four sites, applied to
the adjacent watershed areas, to calculate the atmospheric nitrogen
input to each land use.
4.1.1.4 Urban
The nitrogen inputs to the urban land use were assumed to be from atmospheric nitrogen
deposition only. The current methodology estimated nitrogen loads from the urban land use are
higher by 2.76 million kg, or 63% greater than the EDF estimate. The higher loadings
calculated by the current methodology can be explained by:
1. Nitrogen inputs to the Chesapeake Bay watershed as estimated by
the current methodology are higher than the EDF estimates by 7.5
million kg; which is 210% of the EDF nitrogen input (Table 4.1).
The reason for higher inputs is because the urban land use area
calculated by the current methodology is greater than the area
estimated by the EDF study by 0.50 million hectares.
32
-------
Table 4 5 An Example of the Input Requirements for the Lotus 123 Spreadsheet
EAST BRANCH SUSQUEHANNA
NONPOINT SOURCES
FORESTS
Aim. Wet Deposition kg/ha
Aim. Dry Deposition kg/ha
CROPLANDS
Atm. Wet Deposition kg/ha
Aim. Dry Deposition kg/ha
Fertilizer kg/ha
Manure kg/ha
Biological Fixation kg/ha
WETLANDS
CO
OJ
Atm. Wet Deposition kg/ha
Atm. Dry Deposition kg/ha
PASTURES
URBAN
POINT SOURCES
Atm. Wet Deposition kg/ha
Atm. Dry Deposition kg/ha
Manure kg/ha
Atm. Wet Deposition kg/ha
Atm. Dry Deposition kg/ha
Sewage Treatment Plants kg
Industrial Discharge kg
Pert. Rates
Factors kg/ha
1.00
1.00
1.00 99
1.00
1.00
1.00
INPUTS
kg/ha
7.74
7.74
7.74
7.74
99.00
14.00
0.00
0.00
0.00
7.74
7.74
11.00
7.74
7.74
1748864
1748864
INPUTS INPUTS
Total
Inputs Input
Source
kg Ratios
15.48 0.50
0.50
128.48 0.06
0.06
0.77
0.11
0.00
0.00 0.00
0.00
26.48 0.29
0.29
0.42
15.48 0.50
0.50
3497727
INPUTS INPUTS ATM
NPS Total Load
Study Using
Loads Landuse NPS Study
Site Areas
kg/ha ha kg
5.67 1.73E+06 9.81 E+06
16.06 6.03E+05 9.68E+06
0.00 O.OOE+00 O.OOE+00
7.21 2.58E*05 1.86E+06
11.80 3.32E+05 3.92E+06
INPUTS INPUTS
Sources of Travel
NPS Time
Loads days
kg
4.90E+06 8.60
4.90E+06
5.83E+05 8.60
5.83E+05
7.46E+06
1.06E+06
O.OOE+00
O.OOE+00 8.60
O.OOE+00
5.4SE+05 8.60
5.45E+05
7.74E+05
1.96E+06 8.60
1.96E+06
INPUTS
River
Decay
/day
0.08
0.08
0.08
0.08
0.08
-------
Table 4.6 Comparison of Nitrogen Loads to the Chesapeake Bay
Total Load 106 kg/yr
Current EDF
Methodology Procedure
Source
Point Sources (AFL) 7.38 6.3
Point Sources (BFL) 26.60 26.6
Direct Atmospheric Deposition
to the Bay 12.90 12.9
Cropland
Fertilizer 23.00 48.0
Atmospheric Dep. 3.25 12.1
Manure 4.48 0.0
Forest
Atmospheric Dep. 21.70 14.0
Pasture . -
Manure 3.79 4.9
Atmospheric Dep. 4.50 11.0
Urban
Atmospheric Dep. 7.16 4.4
Total Load 114.76 140.2
34
-------
2. The EDF study assumed a constant nitrogen retention factor of
0.35 for urban land use; whereas, our estimates for the range of
nitrogen urban retention factors varied from 0.14 to 0.59 due to
the different travel times for the fourteen subbasins.
3. The same procedures as described above were used to calculate
atmospheric deposition on urban land, thus contributing to the
differences in the estimates.
4.1.2 Comparisons of Percent Loadings from Individual Nitrogen Sources
The percent contributions to the Chesapeake Bay from the various nitrogen sources in the
watershed are presented in Table 4.7. As can be seen from Table 4.7, the current methodology
indicates that 43% of the total nitrogen load to the Chesapeake Bay results from atmospheric
deposition. This estimate is in close agreement with the 39% estimate obtained by the EDF
study. It can also be concluded from Table 4.7 that the percentage contributions calculated from
point sources and manure loadings to the Chesapeake Bay using the current methodology and
the EDF estimates are also in close agreement. The current methodology estimated slightly
higher contributions of nitrogen from point sources as compared to the EDF analysis; this could
be due to the data that were used to calculate point source loads. In the current study we used
average (1984-1987) point source loadings as estimated by the U.S. EPA Chesapeake Bay
Program Office (Donigian et al., 1994), whereas the EDF analysis used point source data for
the year 1985 only. The percent contributions from manure was 3.2% higher than the EDF
estimate. In the EDF analysis manure was only applied to the pastures, and the authors used
a retention factor of 97.5 % which resulted in very low nitrogen loadings to the Bay as a result
of manure application. However, in our analysis, manure was applied to both cropland and
pastures, and as cropland has lower retention of nitrogen than pastures a slightly higher
percentage nitrogen load to the Bay was estimated using the current methodology.
Table 4.7 Comparison of Percent Nitrogen Contributions from Various Nitrogen
Sources to the Chesapeake Bay
Percent Using
Nitrogen Sources Current Methodology EDF Analysis
Point Sources 29.6 23.5
Atmospheric Deposition 43.1 38.9
Fertilizer Application 20.1 34.1
Animal Waste 7.2 3.5
35
-------
The percentage nitrogen loadings to the Chesapeake Bay from fertilizer applications as estimated
by the two procedures are not in agreement. In the EDF analysis the authors used the county
fertilizer sales data to estimate the total amount of fertilizer applied to the entire watershed and
also used a single constant retention factor for all the agricultural cropland in the Chesapeake
Bay watershed. However, in our analysis we used fertilizer application data estimated by state
and county extension officials (Donigian et al., 1994). In the present analysis the total fertilizer
application was 13 million kilograms greater than the amount applied in the EDF study:
however, using variable retention parameters (computed from unit area loads as reported by
Donigian et al., 1994) for all the fourteen subbasins resulted in lower fertilizer load to the
Chesapeake Bay as compared to the EDF estimated load.
4.1.3 Conclusions from Chesapeake Bay Comparisons
The application of the current methodology yielded favorably close results to the EDF study.
There were some minor discrepancies from loads estimated for individual land uses, which could
be attributed to the following reasons:
1. Total watershed area used in the EDF analysis was 3.41 million
hectares larger than the area used in the current methodology. The
greater area used by the EDF study represents a 26% differential
in watershed area as compared to the current methodology area.
2. EDF used a constant average atmospheric deposition rate for the
entire watershed, however, in our analysis we used actual field
data collected by NADP at four sites (Aurora, NY; Penn State,
PA; Leading Ridge, PA; and Wye, MD) in the watershed.
3. The EDF study used constant retention factors for each of the four
land uses in the watershed; whereas, in the current analysis the
retention parameters varied for the four land uses in the fourteen
subbasins.
4. A constant 50% nitrogen loss was assumed by the EDF study due
to riverine fate and transport. In our analysis the nitrogen loss in
streams was calculated as a function of decay rate and travel time
in individual subbasins.
5. EDF assumed no loss of nitrogen from animal manure due to
handling, storage, and application, whereas these losses were
accounted for in our analysis.
6. Fertilizer application in the current methodology was based on
fertilizer application information provided by the local and state
agricultural agencies, not fertilizer sales as was done in the EDF
analysis.
36
-------
The objective of this project was to assess the applicability of the EDF approach and associated
findings to the other parts of the country, in particular the Gulf Coast region. The approach
consisted of reviewing and refining the EDF methodology and then developing a Lotus 123
spreadsheet using the refined procedures. The first component of the study dealt with applying
the Lotus 123 spreadsheet to the Chesapeake Bay watershed and comparing the results with the
EDF analysis. From the results and discussions that are presented in Section 4.1, it can be
concluded that the first component of the study was completed successfully. The EDF
methodology was reviewed and refined and the new methodology yielded favorable results as
compared to EDF results.
The current methodology yielded a better estimate of nitrogen loadings to the Chesapeake Bay
watershed for the following reasons:
1. The nitrogen load computations from various land use practices in
the watershed were not lumped together, as in the case of the EDF
study.
2. We used varying retention parameters for the same land use
practice in different subbasins of the watershed. This allowed us
to better represent soil, climate, crop and topographic
characteristics of the watershed.
3. Riverine decay was calculated based on travel time and first-order
decay rate, rather than a fixed 50% loss of nitrogen as assumed by
the EDF analysis.
4.2 APPLICATION OF THE METHODOLOGY TO GALVESTON BAY WATERSHED
The Gulf of Mexico region ranks first among those of the U.S. in total estuarine drainage area
and in total water surface area (Rabalais, 1992). Some of the water quality problems affecting
the Gulf Coast estuaries result from man-made pollution both within and upstream of the
drainage area. As the Gulf of Mexico waters are a habitat for marine and biological life, which
is affected by the nitrogen concentration in waters, it is important to control the nitrogen loads
reaching the coastal waters. The Galveston Bay watershed is an integral part of the area that
drains into the Gulf of Mexico.
4.2.1 Available Data for Analysis
Newell et al. (1992) have reported that the watershed area that drains into the Galveston Bay
occupies 10.84 million hectares. The water surface area occupied by the Galveston Bay is 1.40
million hectares (NOAA, 1990). Newell et al. (1992) conducted a study for the Galveston Bay
National Estuary Program to estimate the nonpoint loadings to the Galveston Bay. The authors
divided the watershed into 21 subbasins (Figure 4.2), 3 of the 21 subbasins drained into an
37
-------
OQ
H
n>
CO
rt
o
3
w
to
ad
(U
rt
n
oo m
(X
en
C
0)
CO
o
3
o.
ES)
CD
C3
ED
a
ca
a
C3
HI
G3
so
CD
El
GJ
a
SCALE
LEGEND
AufMn/Daftrap Bayou
Addlcto RtMivoIr
Armand /Taylor Dayou
Buffalo Bayou
Baricer RvMrvotr
Bray* Bayou
Ctnr Crack
Cedar Bayou
Chocoble Bayou
DKUmon Bayou
EaKBay
Crwn* Bayou
North Bay
South Bay
San ladiMo Rlvar
Sim Bayou
Trinity Bay
Trinity River
WMBay
While Oak Bayou
NON-POINT SOURCES AND LOADINGS
CHARACTERIZATION PROJECT
C*rv«lon Bay National Estuary Program
SUBWATERSHED DELINEATION
Groundwater Services, Inc.
Houston, Texas
Rice University
Houston, Texas
CSljobNo: C-1220
Dale:
10/25/91
FIGURE E.2
-------
Table 4.8 Land Use Areas in the Various Subbasins of the Galveston Bay Watershed (modified from Newell et al., 1992)
U)
Subbasin
Name
Armand/Taylor
Bastrop/Austin*
Brays Bayou
Buffalo Bayou
Cedar Bayou
Chocolate Bayou
Clear Creek
Dickinson Bayou
East Bay*
Greens Bayou
North Bay*
San Jacinto River
Sims Bayou
South Bay*
Trinity Bay*
Trinity River
West Bay*
White Oak Bayou
Total Area
% of Watershed Area 20.1 %
Land Use Areas (ha)
Urban
6734
5180
20979
52836
6734
2849
9324
3626
10101
23310
2849
4144
10101
8029
6475
12432
13727
18648
218078
20.1%
Open/
Pasture
7252
15022
6734
31339
12950
8288
17353
11655
18648
14245
2590
4403
8806
5698
17871
34965
27195
6475
251489
23.2%
Cropland
2590
22792
4144
41699
20720
24605
11655
5439
19166
4662
259
2072
2849
1813
20461
37555
20461
2590
245532
22.6%
Wetlands
2331
10878
1036
10360
8029
6734
7252
4921
23051
3626
518
2072
2072
3108
17353
39109
24346
111
167573
15.5%
Fore
777
777
0
1554
5957
1295
111
259
2072
8029
259
3885
259
0
16317
158767
518
0
201502
18.6%
- Subbasins that drain directly into the Galveston Bay (Higgins, 1993, personal communications).
-------
adjacent subbasin (Addicks Reservoir, Barker Reservoir, and Ship Channel drain to Buffalo
Bayou, Higgins, 1993, personal communications). Thus in the analysis presented below we
divided the watershed into 18 subbasins. It was also reported that 6 of the 18 subbasins drain
directly into the Galveston Bay. The six subbasins that drain directly into the Galveston Bay are
Bastrop/Austin, East Bay, North Bay, South Bay, Trinity Bay, and West Bay (Higgins, 1993,
personal communications).
The nonpoint loading data used in our analysis are derived from Newell et al., 1992. The land
uses in the watershed are comprised of high density urban, residential urban, open/pasture,
agriculture, barren, wetlands, and forest. In our analysis we grouped the high density urban,
residential urban, and barren land use into the category of urban. The percent of the total
watershed area occupied by high density urban, residential urban, and barren land use categories
are 10%, 9%, and 1% respectively. The modified land use categories and the areas associated
with each land use are listed in Table 4.8.
The nonpoint nitrogen loading data available consisted of the total nitrogen load that was
generated from each subbasin. The procedure that was used for estimating the total nitrogen
load from the subbasin was not obvious from the results presented in the report. It was reported
that EMC (Event Mean Concentration) data for each land use in conjunction with respective
runoff rates from the land use of interest were used to calculate the total nitrogen load from the
subbasin.
Moreover, the runoff data that the authors used to calculate loadings from each land use were
not available. Thus one of the major tasks during the evaluation of the nonpoint load data
consisted of estimating what portion of the total nitrogen load resulted from the various land use
categories found in the subbasin. The steps used for estimating nitrogen loads for each land use
within a subbasin are described below:
1. Newell et al. (1992), provided the following EMCs for the various
land uses found in the watershed:
Land Use EMC (mg/1) Normalized EMC
to Forest
Urban* 2.35 to 3.18 mg/1 2.84 to 3.83
Open/Pasture 1.51 mg/1 1.82
Agriculture 1.56 mg/1 1.88
Wetlands 0.83 mg/1 1.00
Forest 0.83 mg/1 1.00
* - For the urban land use category we calculated area weighted
EMC values for all the eighteen subbasins, as the area occupied by
the three land uses of high density urban, residential urban, and
barren changed from one subbasin to another. We did not
40
-------
calculate area weighted values for the other land use categories, as
these categories were the same as those reported by Newell et al.
(1992).
We normalized the EMC data to the forest and wetlands EMCs
and the normalized ratios are listed above.
2. Using land use from Table 4.8 we calculated the fraction of the
area that was occupied by each land use for all the 18 subbasins in
the watershed.
3. For each land use in a subbasin, we then multiplied the total
nitrogen load that was estimated by Newell et al. (1992) (here after
referred as "observed load") by the normalized EMC ratios and
the fraction of each land use found in the subbasin. We thus
obtained nitrogen loads from each land use that were normalized
by the EMC ratios.
4. For each subbasin we then summed the load that was obtained
from each land use to obtain the "normalized" loads. A
normalized load ratio of the individual land use load to the
normalized load was then obtained for all the subbasins.
5. We then estimated nitrogen loads for each land use in all the
subbasins by multiplying the observed load for the subbasin by the
normalized load ratio for each land use in the subbasin. Thus the
sum of the nitrogen loads from each land use in the subbasin
equalled the total observed nitrogen load, and they reflected
both the relative EMCs and the land use distribution in the
subbasin. The calculated land use loads for all the subbasins are
listed in Table 4.9.
The annual unit area loads (kg/ha/yr) were calculated for all the land use categories using Table
4.8 and 4.9, and are listed in Table 4.10. The average unit area load from agriculture was
estimated as 6.31 kg/ha/yr. This value is significantly lower than the Chesapeake Bay estimates
derived by Donigian et al. (1994), which were reported to range from 16 to 34 kg/ha/yr.
However, these differences are not unexpected considering the differences in climate (i.e.
rainfall, runoff), soils, topography, etc. between the two regions.
41
-------
Table 4.9 Total Edge-of-Field Nitrogen Loads from the Subbasins of the Galveston Bay Watershed (modified from Newell et
al., 1992)
.e-
K)
Subbasin
Name
Armand/Taylor
Bastrop/Austin*
Brays Bayou
Buffalo Bayou
Cedar Bayou
Chocolate Bayou
Clear Creek
Dickinson Bayou
East Bay*
Greens Bayou
North Bay*
San Jacinto River
Sims Bayou
South Bay*
Trinity Bay*
Trinity River
West Bay*
White Oak Bayou
Land Use Total Nitrogen Loads (kg)
Urban
83717
34651
306631
732969
79193
24979
100167
33750
109274
316324
39505
57517
134666
79294
66760
109980
1 15038
284073
Open/
Pasture
51905
48882
57762
209797
74480
35443
103086
55751
99462
101058
20103
30712
67805
36048
89870
146929
127185
54710
Cropland
19151
76621
36723
288394
123113
108705
71530
26879
105610
34169
2077
14931
22663
11850
106301
163038
98860
22608
Wetlands
9171
19457
4885
38122
25382
15829
23680
12939
67580
14140
2210
7944
8769
10808
47967
90334
62586
3609
Forests
3057
1390
0
5718
18832
3044
2537
681
6075
31309
1105
14896
1096
0
45103
366720
1332
0
- Subbasins that drain directly into the Galveston Bay (Higgins, 1993, personal communications).
-------
The land use categories for which the nitrogen loads were estimated consisted of urban,
open/pastures, agriculture, wetlands, and forests. The following nitrogen inputs were assumed
for each land use:
Land Use Nitrogen Inputs on the Land Use
Urban Atmospheric deposition
Open/Pastures Atmospheric deposition
Agriculture Fertilizer application and atmospheric deposition
. Wetlands Atmospheric deposition
Forest Atmospheric deposition
Animal waste is usually applied to croplands and pastures. However, we were told that there
are no animal waste data available for the Galveston Bay watershed (Hudson, 1993, personal
communications). The areas occupied by various crops grown in the watershed were obtained
from the NOAA data base of Agricultural Pesticide Use in Coastal Areas (1992). According
to the NOAA data, the crops grown in the Galveston Bay watershed consisted of com, rice, rye
grass-hay, and soybeans. The percent area occupied by corn, rice, rye-grass, and soybean as
reported by the NOAA data base is 16%, 36%, 25%, and 24% respectively. McCown (1993,
personal communications) from the local extension service provided information on fertilizer
application amounts; he indicated an average of 120 kg/ha/yr of nitrogen fertilizer is applied on
the agricultural land use in the watershed.
The point source data were obtained from Armstrong (1993, personal communications) and are
listed in Table 4.11. There was only one NADP atmospheric deposition monitoring site in the
Galveston Bay watershed. The site is located at Attwater Prairie Chicken NWR. Observed
atmospheric deposition data for the period of 1983-1991 were available from this site. We used
an average deposition rate for our analysis, and the data used are listed in Table 4.11. There
were no dry atmospheric deposition data collected in the watershed, hence in our analysis we
assumed that dry deposition equals wet atmospheric deposition. The travel time data from each
subbasin listed in Table 4.11 were provided by Newell (1993, personal communications).
After obtaining all the necessary inputs required for our analysis, we estimated the nitrogen loads
that are delivered to the Galveston Bay from the various nitrogen sources in the watershed. The
results obtained from the Lotus 123 spreadsheet analysis are presented in the section below.
4.2.2 Results Obtained from the Galveston Bav
The results obtained from the application of the project methodology to the Galveston Bay
watershed are presented in Tables 4.12 and 4.13. In Table 4.12 we present the annual nitrogen
loads in million kilograms from all the land use practices and all the sources of nitrogen. The
percent of the total nitrogen load that is being contributed by the various sources of nitrogen is
presented in Table 4.13.
43
-------
Table 4.10 Unit Area Annual Nitrogen Loads from the Various Land Use Categories in the
Galveston Bay Watershed (kg/ha/yr)
Urban Open/ Cropland Wetlands Forest
Pasture
Armand/Taylor 12.43 7.16 7.39 3.93 3.93
Bastrop/Austin** 6.69 3.25 3.36 1.79 1.79
Brays Bayou 14.62 8.58 8.86 4.71
Buffalo Bayou 13.87 6.69 6.92 3.68 3.68
Cedar Bayou 11.76 5.75 5.94 3.16 3.16
Chocolate Bayou 8.77 4.28 4.42 2.35 2.35
Clear Creek 10.74 5.94 6.14 3.27 3.27
Dickinson Bayou 9.31 4.78 4.94 2.63 2.63
East Bay** 10.82 5.33 5.51 2.93 2.93
Greens Bayou 13.57 7.09 7.33 3.90 3.90
North Bay** 13.87 7.76 8.02 4.27 4.27
SanJacinto 13.88 6.98 7.21 3.83 3.83
Sims Bayou 13.33 7.70 7.95 4.23 4.23
South Bay** 9.88 6.33 6.54 3.48
Trinity Bay** 10.31 5.03 5.20 2.76 2.76
Trinity River 8.85 4.20 4.34 2.31 2.31
West Bay** 8.38 4.68 4.83 2.57 2.57
White Oak Bayou 15.23 8.45 8.73 4.64
Average 11.46 6.11 6.31 3.36 3.17
Standard Deviation 2.43 1.51 1.56 0.83 0.75
**
- Subbasins that drain directly into Galveston Bay (Higgins, 1993, personal communications)
44
-------
Table 4.11 Point Sources, Wet Atmospheric Deposition, and Travel Time for the Galveston
Bay Watershed Subbasins
Point Sources
kg/yr
0
632602
1708663
30647
845
152939
3811
218997
1526219
403261
8176472
330078
6741559
Subbasin
Armand/Taylor
Brays Bayou
Buffalo Bayou
Cedar Bayou
Chocolate Bayou
Clear Creek
Dickinson Bayou
Greens Bayou
San Jacinto River
Sims Bayou
Trinity River
White Oak Bayou
Direct Point
Source Loads to
the Galveston
Bay
Note: Point source data from N. Armstrong (1993, personal communications)
Travel time estimates provided by Newell (1993, personal communications)
Atmospheric*
Deposition
kg/ha/yr
12.89
12.89
12.89
12.89
12.89
12.89
12.89
12.89
12.89
12.89
12.89
12.89
Travel
Time
days
0.4
0.6
2.4
1.3
0.9
1.3
0.8
0.9
0.6
0.7
3.3
0.3
* _
Average of 1983-1991 annual wet atmospheric deposition at Attwater Prairie Chicken
NWR (Data from National Atmospheric Deposition Program)
The conclusions that can be drawn from Tables 4.11 and 4.12 are listed below:
1. A total of 36.20 million kilograms of nitrogen is delivered to the
Galveston Bay annually.
2. Point sources deliver 17.54 million kilograms per year, which is
a major portion of the total nitrogen load that reaches the Bay.
3. Direct atmospheric deposition on the water surface of the
Galveston Bay accounts for 14.06 million kilograms of nitrogen
per year.
45
-------
Table 4.12 Estimated Total Annual Nitrogen Loads to the Galveston Bay from Various
Nitrogen Sources
Total Load (106 kg/yr)
Point Sources (delivered through rivers) 10.80
Point Sources (direct loadings to Bay) 6.74
Direct Atmospheric Deposition to Bay 12.90
Direct Loadings to Bay from Watershed*
Fertilizer Application 0.31
Atmospheric Deposition 1.16
Cropland
Fertilizer Application 0.63
Atmospheric Deposition 0.14
Forest
Atmospheric Deposition 0.36
Open/Pastures
Atmospheric Deposition 0.88
Urban
Atmospheric Deposition 2.03
Wetland
Atmospheric Deposition 0.22
Total Load 36.20
* - The estimates are for the subbasins that drain directly into the Galveston Bay without
undergoing any riverine losses.
Table 4.13 Percent Loadings from Various Nitrogen Sources to the Galveston Bay
Source Percent Contribution of Total Load
Point Sources 48.5
Atmospheric Deposition to the Bay 38.9
Atmospheric Deposition from Watershed 10.0
Fertilizer Application 2.6
46
-------
4. The land use categories in the watershed contribute 5.76 million
kilograms of nitrogen annually to Galveston Bay.
5. The biggest contribution from the five land use categories is from
the urban land.
6. Point sources and atmospheric deposition each account for 49% of
the total nitrogen load that reaches the Galveston Bay.
To summarize the results from the application of the methodology to the Galveston Bay, it can
be concluded that atmospheric nitrogen deposition is a major contributor of the nitrogen loadings
that are delivered to the Galveston Bay.
4.3 APPLICATION OF THE METHODOLOGY TO TAMPA BAY WATERSHED
As stated above, the Gulf of Mexico region ranks first among those of the U.S. in total estuarine
drainage area and in total water surface area (Rabalais, 1992). The Tampa Bay watershed, our
second study site is also an integral part of the area that drains into the Gulf of Mexico.
4.3.1 Available Data for Analysis
The data used in the analysis of the Tampa Bay watershed was obtained initially from Greening
(1993, personal communications) and was subsequently published by Coastal Engineering Inc.,
(Zarbock et al, 1994) who was conducting a nutrient balance study for the Tampa Bay watershed
as part of the Tampa Bay National Estuary Program (Greening, 1993, personal communications).
The watershed area that drains into the Tampa Bay occupies 0.60 million hectares and is divided
into 10 subbasins (Figure 4.3). The water surface area occupied by the Tampa Bay is 0.1
million hectares (NOAA, 1990).
The nonpoint loading data from each of the land use categories in the ten subbasins was provided
by Greening (1993, personal communications) and later published by Zarbock et al. (1994). The
land uses in the watershed are comprised of urban, agriculture, wetlands, and forest.
The nonpoint nitrogen loading data available consisted of bar graphs of total nitrogen loads
produced from all the subbasins and associated land use categories. We used the bar graph to
estimate the nonpoint loadings from the ten subbasins; these loadings are presented in Table
4.14.
47
-------
MAJOR DRAINAGE BASINS OF TAMPA BAY
Modified from USGS
Map Prepaiol by Coaml Ecvtronmcnul, Inc.
Figure 4.3 Tampa Bay Watershed with Subbasin Boundaries
48
-------
Table 4.14 Total Edge-of-Field Nitrogen Loads from the Subbasins of the Tampa Bay
Watershed (Zarbock et ah, 1994)
Subbasin
Name
Total Nitrogen Loads (kg/yr)
Urban Cropland Wetlands Forests
Alafia River 56127 256581 16036 85527
Boca Ciega Bay 128290 0 5345 0
Coastal Hillsborough Bay 53454 66818 10691 16036
Coastal Lower Tampa Bay 10691 20313 2138 10691
Coastal Middle Tampa Bay 90872 9622 2138 5345
Coastal Old Tampa Bay 145396 20313 13898 42764
Hillsborough River 160363 235199 48109 203127
Little Manatee River 20847 198851 26727 69491
Manatee River 48109 225578 64145 85527
Terra Ciea Bay 5345 2673 2000 0
The land use categories for which the nitrogen loads were estimated consisted of urban,
agriculture, wetlands, and forests. The following nitrogen inputs were assumed for each land
use:
Land Use
Urban
Agriculture
Wetlands
Forest
Nitrogen Inputs on the Land Use
Atmospheric deposition
Fertilizer application and atmospheric deposition
Atmospheric deposition
Atmospheric deposition
Animal waste is usually applied to croplands and pastures. However, no animal waste
application data were available for the Tampa Bay watershed. The area occupied by various
crops grown in the watershed was obtained from the NOAA data base of Agricultural Pesticide
Use in Coastal Areas (1992). According to the NOAA data, the crops grown in the Tampa Bay
watershed consisted of rye grass-hay, citrus fruits, vegetables, and corn. The percent area
occupied by rye-grass, citrus fruits, vegetables, and corn as reported by the NOAA data base
is 88%, 7%, 4%, and less than 1% respectively. Huff (1993, personal communications) from
the local Soil Conservation Office provided information on fertilizer application amounts; she
indicated an average of 120 kg/ha of nitrogen fertilizer is applied annually on the agricultural
land use in the watershed.
49
-------
The point source data were provided by Greening (1993, personal communications) and are
listed in Table 4.15. There was only one NADP atmospheric deposition monitoring site in the
Tampa Bay watershed. The site is located at Vema Well Field. Observed atmospheric
deposition data for the period of 1984-1991 were available from this site. We used an average
deposition rate for our analysis, and the data used are listed in Table 4.15. There were no dry
atmospheric deposition data collected in the watershed; hence in our analysis we assumed that
dry deposition equals wet atmospheric deposition. The travel time data from each subbasin
listed in Table 4.15 was provided by Zarbock (1993, personal communications).
After obtaining all the necessary inputs required for our analysis, we estimated the nitrogen loads
that are delivered to the Tampa Bay from the various nitrogen sources in the watershed. The
procedure adapted for routing the loads in this analysis is slightly different than the procedures
used in the Chesapeake Bay or Galveston Bay analysis. Here, we did not calculate retention
parameters or individual land use total nitrogen loads based on unit area loads as we did in the
analysis of Chesapeake Bay and Galveston Bay, as information available on the land use areas
was not available to us. The input source ratios were calculated based on the nitrogen inputs
on each land use. These ratios were then used to calculate the source of nitrogen load from the
total nitrogen load that was provided to us by Greening, for all the land use categories in the ten
subbasins. The procedure of estimating output source ratios and routing the nitrogen loads
through the river/stream channel network to the bay using travel time and first-order decay rate
was the same as described in the project methodology. The results obtained from the Lotus 123
spreadsheet analysis are presented in the section below.
4.3.2 Results Obtained from the Tampa Bay
The results obtained from the application of the project methodology to the Tampa Bay
watershed are presented in Tables 4.16 and 4.17. In Table 4.16 we present the nitrogen loads
in million kilograms per year from all the land use practices and all the sources of nitrogen.
The percent of the total nitrogen load contributed by the various sources of nitrogen is presented
in Table 4.17.
The conclusions that can be drawn from Tables 4.16 and 4.17 are listed below:
1. A total of 4.48 million kilograms of nitrogen per year is delivered
to the Tampa Bay.
2. Point sources annually deliver 0.87 million kilograms of nitrogen
to the Tampa Bay.
3. Direct atmospheric deposition on the water surface of the Tampa
Bay accounts for 1.80 million kilograms of nitrogen per year,
which is a major portion of the total nitrogen load that reaches the
Bay.
50
-------
Table 4.15 Point Sources, Wet Atmospheric Deposition, and Travel Time for the Tampa
Bay Watershed Subbasins
Subbasin
Alafia River
Boca Ciega Bay
Coastal Hillsborough Bay"
Coastal Lower Tampa Bay"
Coastal Middle Tampa Bay
Coastal Old Tampa Bay
Hillsborough River
Little Manatee River
Manatee River
Terra Ciea Bay
Point Sources Atmospheric*
kg Deposition
kg/ha
93282
15909
484800
96960
95454
90909
106818
2273
23636
2727
10.04
10.04
10.04
10.04
10.04
10.04
10.04
10.04
10.04
10.04
Travel
Time
days
5.0
0.5
1.0
1.0
1.0
2.0
6.0
3.0
5.0
1.0
Note: Point source data from Greening (1993, personal communications)
Travel time estimates provided by Zarbock (1993, personal communications)
* - Average of 1984-1991 annual wet atmospheric deposition at Verna Well Field
(Data from National Atmospheric Deposition Program)
** - Includes Fugitive Emissions from Fertilizer Production Plants as Point Source Loadings
(Greening, 1993, personal communications)
51
-------
Table 4.16 Estimated Total Nitrogen Loads to the Tampa Bay from Various Nitrogen
Sources
Total Load (106 kg/yr)
Point Sources (delivered through rivers) 0.87
Direct Atmospheric Deposition to Bay 1.80
Cropland
Fertilizer Application 0.63
Atmospheric Deposition 0.11
Forest
Atmospheric Deposition 0.14
Urban
Atmospheric Deposition 0.58
Wetland
Atmospheric Deposition 0.36
Total Load - 4.48
Table 4.17 Percent Loadings from Various Nitrogen Sources to the Tampa Bay
Source Percent Contribution of Total Load
Point Sources 19.35
Atmospheric Deposition to the Bay 40.17
Atmospheric Deposition from Watershed 26.44
Fertilizer Application 14.04
52
-------
4. The land use categories in the watershed contribute 1.82 million
kilograms of nitrogen annually to Tampa Bay.
5. The biggest contribution from the five land use categories is from
cropland.
6. Point sources and atmospheric deposition, respectively, account for
19% and 67% of the total nitrogen load that reaches the Tampa
Bay each year.
To summarize the results from the application of the methodology to the Tampa Bay, it can be
concluded that atmospheric nitrogen deposition is the major contributor of nitrogen loadings that
are delivered to the Tampa Bay. However, the total bay loads are considerably less than the
loads delivered in the other two study sites. Also, the dominance of the atmospheric component
is likely due to the relatively large water surface area and relatively small watershed areas of the
Tampa Bay system, compared to the other study sites.
4.4 SENSITIVITY ANALYSIS
The results obtained from the application of the methodology to the study sites were presented
in the sections above. In this section we present results obtained from the sensitivity analysis
of three parameters that were used in the Lotus 123 spreadsheet to estimate nitrogen loadings
to aquatic systems. The sensitivity analysis was performed for the Chesapeake Bay watershed
on selected parameters used in the assessment of the initial results, which we will call the base
scenario presented in Section 4.1. The sensitivity analysis was performed by varying one
parameter at a time in all the fourteen subbasins of the Chesapeake Bay watershed, while holding
the other parameters constant. The three parameters analyzed were:
1. Ratio of Wet to Dry Atmospheric Deposition;
2. First-order decay rate in rivers; and
3. Fertilizer application amounts (varying agronomic recommended rates).
One of the major assumptions of the sensitivity analysis was that the nitrogen retention parameter
for the various land use practices in the fourteen subbasins will not be affected; these retention
parameters are listed in Table 4.18.
4.4.1 Sensitivity of Wet and Drv Atmospheric Deposition Ratios
As this project deals jwith estimating the percent of nitrogen contribution from atmospheric
deposition with respect to total nitrogen loadings to the bay or watershed outlet from all sources,
nonpoint and point, it was very important to study the effect of the ratio between wet and dry
atmospheric nitrogen deposition. The major reasons to study the effect of the ratio between wet
and dry atmospheric deposition are:
53
-------
1. Wet atmospheric deposition data are readily available for 200
stations in the contiguous United States. The wet deposition data
are collected by the staff of the National Atmospheric Deposition
Program (NADP).
2. The dry atmospheric deposition data are very scarce and are not
readily available. Thus, the ratio of wet to dry deposition is often
used to determine Total deposition.
3. In the Lotus 123 spreadsheet developed under the current
methodology, we estimate dry atmospheric deposition as a factor
times the wet atmospheric deposition.
For the base scenario, 40% of the total nitrogen input on the Chesapeake watershed and water
surface was derived from atmospheric deposition (Table 4.1). Fisher et al. (1988) cited
literature where the ratios of dry atmospheric deposition to total atmospheric deposition were
reported; these ratios and the literature source are listed below:
Literature Source Percent of Total Atm. Deposition as
Dry Deposition
Levy and Moxim, 1987 and Ro et al., 1988 20 to 85 %
Logan, 1983 38 to 55%
Meyers and Sisterton, 1989 30 to 45%
During the sensitivity analysis, the wet deposition data were not altered as these data represented
measured wet atmospheric deposition at the four sites in the Chesapeake Bay watershed. In our
analysis for the base case scenario, we assumed that dry deposition is 50% of the total
atmospheric deposition, which meant that dry deposition was equal to wet deposition and,
therefore, total deposition is two times wet deposition. The following scenarios were analyzed
during the sensitivity analysis:
1. Dry deposition is 20% of Total Deposition (this means dry
deposition is 25% of wet deposition);
2. Dry deposition is 33% of Total Deposition (this means dry
deposition is 50% of wet deposition); and
3. Dry deposition is 60% of Total Deposition (this means dry
deposition is 150% of wet deposition);
The results obtained from the sensitivity analysis are presented in Table 4.19 and Table 4.20.
In Table 4.19 we present the total nitrogen loads from all the land uses and nitrogen sources,
and in Table 4.20 the percent of individual nitrogen source loads as they relate to the total
nitrogen load to the Chesapeake Bay are presented. The conclusions drawn from the sensitivity
analysis are presented below:
54
-------
Table 4.18 Retention Factors for the Base Case for the Chesapeake Bay Watershed
Cropland Forest Pasture Urban
East Branch Susquehanna
West Branch Susquehanna
Juniata
Lower Susquehanna
Conowingo
Upper Potomac
Shenandoah
Lower Potomac
Rappahannock
Mattaponi
Pamunkey
James
Appomattox
Patuxent
0.88
0.80
0.87
0.83
0.87
0.90
0.88
0.90
0.90
0.85
0.85
0.89
0.85
0.90
0.63
0.68
0.47
0.64
0.30
0.63
0.68
0.64
0.65
0.92
0.90
0.87
0.90
0.85
0.73
0.79
0.70
0.54
0.34
0.75
0.94
0.58
0.74
0.70
0.68
0.40
0.62
0.84
0.24
0.21
0.16
0.20
0.23
0.12
0.18
0.15
0.14
0.50
0.59
0.26
0.37
0.16
3.
The total atmospheric nitrogen input to the Chesapeake Bay was
63%, 75%, and 125% of the base case scenario input when dry
deposition rates were assumed to be 20%, 33%, and 60% of the
total atmospheric deposition respectively.
The total nitrogen load to the Chesapeake Bay was 80%, 89%, and
120% of the base case scenario load when dry deposition rates
were assumed to be 20%, 33%, and 60% of total atmospheric
deposition, respectively. This indicates that estimating the percent
of total atmospheric deposition resulting from dry deposition is
very important as it is one of the major sources of nitrogen input
to the water surface, and is deposited on all land uses of the
watershed.
Under the base scenario, we estimated that 43% of the total
nitrogen load resulted from atmospheric deposition; however,
decreasing the percent of dry deposition to 20% and 33% of total
atmospheric deposition resulted in 32% and 36% of the total
nitrogen load resulting from atmospheric deposition. When we
increased the percent of dry deposition in the total atmospheric
deposition to 60%, we estimated that 49% of the total nitrogen
load to the Chesapeake Bay results from atmospheric deposition.
These results were obvious because decreasing the percent of dry
55
-------
deposition as a percent of total atmospheric deposition resulted in
a lower input of atmospheric deposition to the various land uses in
the watershed, and increasing the percent of dry deposition
resulted in higher inputs of atmospheric nitrogen.
4. The direct atmospheric deposition to the water surface of the
Chesapeake Bay was estimated to be 11 % of the total nitrogen
load. Decreasing the dry atmospheric deposition to 20% and 33%
of total atmospheric deposition resulted in 8 and 10% of direct
atmospheric deposition to the water surface. Increasing the
percent of dry deposition to 60% of total atmospheric deposition
resulted in 12% of direct atmospheric deposition to the Bay water
surface.
5. Under the base scenario, we estimated that 43% of total nitrogen
load to the Chesapeake Bay resulted from atmospheric deposition.
The sensitivity analysis indicates that changing the percent of dry
deposition in total atmospheric deposition to 20%, 33%, and 60%
results in 74%, 84%, and 113%, respectively, of the total
atmospheric nitrogen deposition relative to the base case scenario.
From the above analysis it can be concluded that either better estimates of atmospheric dry
deposition or the total atmospheric deposition are needed. This is important because it has a
significant impact when calculating the percent contribution of atmospheric deposition to the total
nitrogen load delivered to the watershed outlet.
4.4.2 Sensitivity to Riverine Decay Parameter
Another important parameter that controls the total amount of nitrogen that is delivered to the
Chesapeake Bay or any watershed outlet of interest is the first-order riverine decay rate used in
our analysis. Under the base case scenario, we assumed a decay rate of 0.08/day, this rate is
a typical rate of nitrogen loss/decay reported in the literature (Barnwell, 1993, personal
communications). During the sensitivity analysis the decay rate was varied as 50% (0.04/day),
75% (0.06/day), and 200% (0.16/day) of the base scenario rate. The results obtained from the
sensitivity analysis are presented in Tables 4.21 and 4.22, and are discussed below:
1. The total nitrogen load to the Chesapeake Bay was 116%, 107%,
and 78% of the base case scenario load when decay rates were
assumed to be 50%, 75%, and 200% of the base scenario rate
respectively. This indicates that decreasing the decay rate results
in higher nitrogen loads to be delivered to the Chesapeake Bay and
vice versa. A decrease of the decay rate by 50% of the base
scenario decay rate resulted in 16% increase in total nitrogen load
that was delivered to the Chesapeake Bay; however, an increase in
56
-------
Table 4.19 Sensitivity of Ratio of Dry to Wet Atmospheric Deposition Rates to
Nitrogen Loadings to Chesapeake Bay
Base Scenario
Source
Point Sources (AFL)
Point Sources (BFL)
to the Bay
Cropland
Fertilizer
Atmospheric Dep
Manure
Forest
Atmospheric Dep.
Pasture
Manure
Atmospheric Dep.
Urban
Atmospheric Dep.
Total Load
Table 4.20
Source
Point Sources
Atmospheric Deposition
Fertilizer
Manure
Dry Dep. Dry Depo.
50% of 20% of
Total Total
Deposition Deposition
106 kg/yr 106 kg/yr
7.38 7.38
26.60 26.60
>eposition
12.90 8.06
23.00 23.00
>ep. 3.25 2.03
4.48 4.48
)ep. ^21.70 13.50
3.79 3.79
)ep. 4.50 2.81
)ep. 7.16 4.48
115.00 96.20
Dry Depo.
33% of
Total
Deposition
106 kg/yr
7.38
26.60
9.68
23.00
2.44
4.48
16.30
3.79
3.38
5.37
102.00
Dry Depo.
60% of
Total
Deposition
106 kg/yr
7.38
26.60
16.10
23.00
4.07
4.48
27.10
3.79
5.63
8.95
138.00
Percent Contribution to the Chesapeake Bay with Varying Ratios of Wet
to Dry Atmospheric Deposition Rates
Percent to the Bay
Base 20% of 33% of 60% of
Scenario Total
Depo. Depo.
29.6 35.4
on 43.1 32.1
20.1 23.9
7.2 8.6
Total
Depo.
33.2
36.3
22.5
8.1
Total
Depo.
26.7
48.7
18.1
6.5
57
-------
the decay rate to 200% of base scenario rate resulted in 22% less
nitrogen loadings to the Bay. We assumed that the nitrogen
delivery to the Bay from the subbasin river network to follow an
exponential decay process; thus nitrogen loss was a function of the
product of the decay rate and travel time in the rivers. This
analysis indicates that estimating nitrogen losses in rivers and
streams is important as it controls the amount of nitrogen that will
be delivered to the bay or watershed outlet.
2. The change in riverine decay parameter only affected the total
nitrogen load to the Chesapeake Bay; the percent loadings from
individual sources of nitrogen as presented in Table 4.22 did not
change significantly, as the nitrogen loadings from all the sources
were influenced by the change in decay rate.
4.4.3 Sensitivity to Fertilizer Application Amounts
The percent of nitrogen input from fertilizer application under the base scenario was 35 % of the
total nitrogen input to the Chesapeake Bay watershed and water surface (Table 4.1). Nitrogen
is an important nutrient for crop growth and is supplied to the crops from fertilizer and manure
application. Each crop needs a certain amount of nitrogen for attaining maximum growth and
production. This nitrogen rate is normally determined by agricultural extension scientists based
on field experiments. The rate is normally referred to as the recommended agronomic
application rate. It has been reported that farmers do not apply fertilizers at recommended
agronomic rates; they often apply excess fertilizer than what is needed by the crops. This
excessive application results in greater amounts of nitrogen being delivered to the streams and
rivers. During the sensitivity analysis we varied the fertilizer application under the base case
scenario to 50%, 150%, and 200% of base scenario fertilizer input. The results obtained from
the sensitivity analysis are presented in Tables 4.23 and 4.24, and are discussed below:
1. The total nitrogen load to the Chesapeake Bay was 90%, 110%,
and 120% of the base case scenario load when fertilizer application
rates were assumed to be 50%, 150%, and 200% of the base
scenario rate, respectively. This indicates that decreasing the
fertilizer application amount by 50% resulted in a 10% decrease
in total load to the Chesapeake Bay as compared to the base case
scenario. Increasing fertilizer application rates to 150% and 200%
of the base scenario rate resulted in a 10% and 20% increase in
nitrogen loads to the Bay.
2. From Table 4.24, it can be seen that the percent contribution to
the Chesapeake Bay was 20% from fertilizer application under the
base case scenario. Changing the application rates to 50%, 150%,
and 200% of the base application rates resulted in 11%, 27%, and
33% of loadings to the Chesapeake Bay due to fertilizer
58
-------
Table 4.21 Sensitivity of Riverine Decay Rates to Nitrogen Loadings to Chesapeake Bay
Source
Point Sources (AFL)
Point Sources (BFL)
Direct Atmospheric Deposition
to the Bay
Cropland
Fertilizer
Atmospheric Dep.
Manure
Forest
Atmospheric Dep.
Pasture
Manure
Atmospheric Dep.
Urban
Atmospheric Dep.
Total Load
Table 4.22 Percent Contribution
Rates
Source
Point Sources
Atmospheric Deposition
Fertilizer
Manure
Base
Scenario
10* kg/yr
7.38
26.60
12.90
23.00
3.25
4.48
21.70
3.79
4.50
7.16
115.00
50% Base
Scenario
106 kg/yr
9.14
26.60
12.90
28.33
4.11
5.67
27.90
4.54
5.26
8.86
133.00
75% Base
Scenario
10* kg/yr
8.20
26.60
12.90
25.50
3.65
5.03
24.60
4.14
4.86
7.95
123.00
200% Base
Scenario
106 kg/yr
5.02
26.60
12.90
16.00
2.12
2.87
13.60
2.72
3.45
4.92
90.20
to the Chesapeake Bay with Varying Riverine Decay
Percent to the Bay
Base 50% Base 75% Base 200% Base
Scenario Scenario Scenario Scenario
29.6
43.1
20.1
7.2
26.8
44.3
21.2
7.7
28.2
43.7
20.7
7.4
35.0
41.0
17.7
6.2
59
-------
Table 4.23 Sensitivity of Fertilizer Application Rates to Nitrogen Loadings to Chesapeake
Bay
Source
Point Sources (AFL)
Point Sources (BFL)
Direct Atmospheric Deposition
to the Bay
Cropland
Fertilizer
Atmospheric Dep.
Manure
Forest
Atmospheric Dep.
Pasture
Manure
Atmospheric Dep.
Urban
Atmospheric Dep. 7.16 7.16 7.16 7.16
Total Load 115.00 103.00 126.00 138.0
Base
Scenario
106 kg/yr
7.38
26.60
12.90
23.00
3.25
4.48
21.70
3.79
4.50
50% Base
Scenario
106 kg/yr
7.38
26.60
12.90
11.50
3.25
4.48
21.70
3.79
4.50
150% Base
Scenario
106 kg/yr
7.38
26.60
12.90
34.60
3.25
4.48
21.70
3.79
4.50
200% Bas
Scenario
10" kg/yr
7.38
26.60
12.90
46.10
3.25
4.48
21.70
3.79
4.50
Table 4.24 Percent Contribution to the Chesapeake Bay with Varying Base Scenario Fertilizer
Rates
Percent to the Bay
Source Base 50% Base 150% Base 200% Base
Scenario Scenario Scenario Scenario
Point Sources 29.6 32.9 26.0 24.7
Atmospheric Deposition 43.1 47.9 39.2 35.9
Fertilizer 20.1 11.2 27.4 33.4
Manure 7.2 8.0 6.5 6.0
60
-------
application. Inputs from fertilizer application accounted for 35%
of the total nitrogen input to the Chesapeake Bay watershed. As
fertilizer application rates increase, they account for a larger
percent of the total nitrogen load, and other sources of nitrogen
becomes smaller percentages, and vice versa.
4.4.4 Conclusions from Sensitivity Analysis
The results obtained from the sensitivity analysis were presented in Section 4.4.1, 4.4.2, and
4.4.3 for three assessment parameters — dry deposition to wet deposition ratio, riverine decay
rate, and fertilizer application amounts. Of the three parameters analyzed, dry deposition to wet
deposition ratio and fertilizer application amounts directly affect the total amount of nitrogen that
is input onto the watershed, or the water body, which has a direct impact on the total amount
of nitrogen load that is delivered to the watershed outlet. A similar magnitude of change in
these two parameters and its corresponding effect on total nitrogen load relative to base scenario
load is presented in Table 4.25. It can be concluded from the table that the total nitrogen
loadings are more sensitive to atmospheric deposition than fertilizer application. The reason for
this sensitivity is atmospheric nitrogen deposition occurs on all of the land use categories found
in the watershed along with the water surface occupied by the Bay. The fertilizers are however,
only applied to the cropland in the watershed.
The nitrogen decay in the riverine systems affects the total amount of nitrogen that will be
delivered by the rivers/streams to the watershed outlet. A decrease in decay rate by 50% and
75% of the base scenario decay rate resulted in a 16% and 7% increase in nitrogen load that was
delivered to the Chesapeake Bay. An increase in the decay rate by 100% of the base scenario
rate resulted in a 22% decrease in nitrogen load, while a decrease by 50% resulted in a decrease
of total nitrogen load by 12%.
From the above analysis it can be concluded that the output from the spreadsheet analysis is
sensitive to all three parameters analyzed. However, the ratio of dry to wet deposition has a
greater impact on the total amount of nitrogen delivered to the Chesapeake Bay as it directly
affects the total nitrogen input to the various land uses and water surfaces of the Chesapeake
Bay. The nitrogen decay in streams is the second most sensitive parameter as it controls the
amount of nitrogen that will be delivered to the watershed outlet via the stream/river network.
Fertilizers are predominantly applied to cropland in the watershed. However, the typical
variation in fertilizer (and manure) application rates is much larger than the variation in the other
parameters, with many farmers applying two to three times agronomic rates. In a watershed
where the majority of the land use is under cropland, then the amount of fertilizer application
will be an important parameter, if not the most important parameter, in estimating nitrogen loads
to aquatic systems.
61
-------
Table 4.25 Effect of Changing Atmospheric Deposition and Fertilizer Application of Total
Nitrogen Load from the Chesapeake Bay Watershed
Decrease by 37.5% Increased by 50%
Atm. Depo1 Pert. Rate2 Atm. Depo1 Pert. Rate2
Percent Change in
Total Nitrogen
Load Relative -20% -8% +20% +10%
to Base Scenario
Load
Note: A positive/negative sign represents an increase/decrease relative to base scenario.
1 - Total atmospheric input was increased or decreased relative to the base
scenario input by changing the dry to wet deposition ratio
2 - Fertilizer rates were increased or decreased relative to base scenario
fertilizer application
62
-------
SECTION 5.0
CONCLUSIONS AND RECOMMENDATIONS
The coastal waters of the eastern United States are receiving large inputs of nutrients, in
particular nitrogen (Fisher et al., 1988). Large inputs of nitrogen cause excessive growth of
algae, loss of oxygen and light to the water and the long term decline of marine life. Fisher et
al. (1988) conducted a study to estimate the various sources of nitrogen that reach the
Chesapeake Bay. The authors estimated the contributions to the Chesapeake Bay from various
nitrogen sources as they relate to total nitrogen load. The reported percentages from various
nitrogen sources are: 24% from point sources, 39% from atmospheric deposition, 34% from
fertilizers, and 4% from animal waste.
The overall objective of this project was to assess the applicability of the EOF approach and
associated findings to other parts of the country, in particular the Gulf Coast region. The
approach consisted of reviewing and evaluating the EDF methodology, improving and refining
the EDF methodology and assumptions, and then implementing these refined procedures to
reproduce their results (in a Lotus 123 spreadsheet format) using commonly available databases.
The EDF methodology was reviewed and the following EDF assumptions were modified under
the current methodology:
1. The nitrogen loadings from the watershed were calculated by
dividing the watershed into smaller subbasins to better represent
the variation in land use in the watershed, rather than lumping all
the land uses together, as was done in the EDF analysis.
2. Fertilizer applications were based on recommended fertilizer
application rates as suggested by local and state government
agencies.
3. The Lotus 123 spreadsheet developed for the methodology allows
the user to vary the relationship between wet and dry atmospheric
deposition, retention parameters for various land uses, riverine
decay rates, and fertilizer application rates. This is in contrast to
the constant rates that were used in the EDF study.
4. Riverine losses are calculated using travel time and nitrogen losses
during travel to the watershed outlet, rather than a constant loss of
nitrogen from all the rivers/streams in the watershed.
63
-------
The Lotus 123 spreadsheet developed under the new methodology was applied to the
Chesapeake Bay, Galveston Bay, and Tampa Bay watersheds. The following conclusions were
drawn from the three study sites analyzed:
1. The methodology developed for this project improved and refined
the assumptions and procedures that were developed in the EDF
study.
2. Our study results indicate that atmospheric nitrogen deposition is
a major source of nitrogen that is delivered to the study sites,
accounting for over 40% of the total nitrogen load that is delivered
to the bays.
3. The majority of the atmospheric deposition load occurs as direct
deposition onto the water surfaces.
4. For many East Coast and Gulf Coast watersheds, a majority of the
nitrogen input may be derived from atmospheric deposition as it is
deposited on all the land use categories found in the watershed.
•
5. The Lotus 123 spreadsheet developed under the methodology is
sensitive to the nitrogen inputs received from various nitrogen
sources.
6. Application of the Lotus 123 spreadsheet provides reasonable
estimates of nitrogen loads to the watershed outlet for an average
climate year.
Based on the analysis that was performed there is a need to further investigate the assumptions
that were used in the current methodology. The following is a list of research issues and
recommendations that are derived from the current study:
1. We need to further analyze nitrogen loadings from various sources
under transient conditions.
2. We need to better characterize and understand total nitrogen
deposition, and the relationship between wet and dry atmospheric
deposition.
3. We need to investigate possible changes in the first-order nitrogen
loss/decay rate with the travel time associated with various
river/streams found in the watershed.
64
-------
4. We need to study further how changing climate, crops, soil type,
and topography affect the retention parameters for the various land
use practices in the watershed.
5. We need to better evaluate retention parameters for atmospheric
nitrogen that is deposited on the watershed as opposed to other
human nitrogen inputs, such as fertilizers and animal waste.
6. We need to investigate the possible impacts, if any, of biological
nitrogen fixation on retention parameters, especially for crops and
forests.
7. We need to focus efforts on the effects of nitrogen deposition and
its delivery from the watershed to the watershed outlet. We can
account for the contribution of atmospheric nitrogen deposition on
the water surface; however, there is limited information regarding
the amount of atmospheric nitrogen that is deposited on the
watershed and is finally delivered to the aquatic system of interest.
65
-------
SECTION 6.0
REFERENCES
Donigian, A.S. Jr., B.R. Bicknell, A.S. Patwardhan, L.C. Linker, C.H. Chang, and R.
Reynolds. 1994. Chesapeake Bay Program - Watershed Model Application to Calculate
Bay Nutrient Loadings: Final Findings and Recommendations (FINAL REPORT).
Prepared for U.S. EPA Chesapeake Bay Program, Annapolis, Maryland.
Fisher, D., J. Ceraso, T. Mathew, and M. Oppenheimer. 1988. Polluted Coastal Waters: The
Role of Acid Rain. Environmental Defense Fund, New York, New York 10010.
Johanson, R.C., J.C. Imhoff, J.L. Kittle, Jr., and A.S. Donigian, Jr. 1984. Hydrological
Simulation Program-FORTRAN (HSPF): User's Manual for the Release 8.0. EPA-600/3-
84-066. U.S. Environmental Protection Agency, Athens, GA.
Kelly, J.M and J.F. Meagher. 1986. Nitrogen Input/Output Relationships for Three Forested
Watersheds in Eastern Tennessee. In: Watershed Research Perspectives. Correll, D.L.
(ed.). Smithsonian Institution Press, Washington, D.C.
Levy, H, and W.J. Moxim. 1987. Fate of U.S. and Canadian Combustion Nitrogen Emissions.
Nature 328, 414-416.
Likens, G.E., F.H. Bormann, R.S. Pierce, J. Eaton, and N.M. Johnson. 1977. Biogeochemistrv
of a Forested Ecosystem. Springer-Verlag, NY, 146p.
Linthurst, R. 1988. Testimony before the United States Senate, Committee on Commerce,
Science, and Transportation, June 8, 1988.
Logan, J.A. 1983. Nitrogen Oxides in the Troposphere:Global and Regional Budgets. J.
Geophys. Res. 88, 10785-10807.
Meyers, T., and D.L. Sisterton. 1989. Measurements of Dry Deposition of Atmospheric
Pollutants. Section 6.3. In: Deposition Monitoring:Methods and Results: State-of-
Science/Technology. Report 6. National Acid Precipitation Assessment Program,
Washington, D.C. Draft for Public Review.
National Atmospheric Deposition Program (NRSP-3)/National Trends Network. 1993.
NADP/NTN Coordination Office, Natural Resource Ecology Laboratory, Colorado State
University, Fort Collins, CO 80523.
66
-------
National Oceanic and Atmospheric Administration. 1990. Estuaries of the United States Vital
Statistics of a National Resource Base. A Special NOAA 20th Anniversary Report.
National Oceanic and Atmospheric Administration, Rockville, MD.
National Oceanic and Atmospheric Administration. 1992. Data Base of Agricultural Pesticide
Use in Coastal Waters. Strategic Environmental Assessments Division. National Oceanic
and Atmospheric Administration, Rockville, MD.
Newell, C.J., H.S. Rifai, and P.B. Bedient. 1992. Characterization of Non-Point Sources and
Loadings to Galveston Bay. Published by The Galveston Bay National Estuary Program,
Publication GBNEP-15, March 1992.
Rabalais, N.N. 1992. An Updated Summary of Status and Trends in Indicators of Nutrient
Enrichment in the Gulf of Mexico. EPA/800-R-92-004. U.S. EPA, Office of Water, Gulf
of Mexico Program, Stennis Space Center, MS 39529.
Ro, C.U., A.J.S. Tang, W.H. Chan, R.W. Kirk, N.W. Reid, and M.A. Lusis. 1988. Wet and
Dry Deposition of Sulphur and Nitrogen Compounds in Ontario. Atmos. Environ. 22,
2763-2772.
Robbins, J.W.D., D.H. Howells, and G.J. Kriz. 1972. Stream Pollution from Animal
Production Units. J. Water Pollut. Contr. Fed. 44:1536-1544.
Schreiber, J.D., P.O. Duffy, and D.'C. McClurkin. 1976. Dissolved Nutrient Losses in Storm
Runoff from Five Southern Pine Watersheds. J. Environ. Qual. 5:201-205.
Smullen, J.T., J.L. Taft, and J. Macknis. 1982. Nutrient and Sediment Loads to the Tidal
Chesapeake Bay System. Section II, Chapter III. In: Chesapeake Bay Program Technical
Studies: A Synthesis. U.S. EPA, Chesapeake Bay Program, Annapolis, MD.
Tarrant, R.F. 1983. Nitrogen Fixation in North American Forestry: Research and Application.
In: Gordon, J.C, and C.T. Wheeler (eds.). Biological Nitrogen Fixation in a Forest
Ecosystems: Foundations and Applications, pp. 261-278. Martinius Nighoff/Dr. W. Junk
Publishers. The Hague.
U.S. E.P.A. Chesapeake Bay Program. 1983. Nutrients. Chapter 3. In: Chesapeake Bay: A
Framework for Action, Annapolis, MD.
Weller, D.E., W.T. Peterjohn, N.M. Goff, and D.L. Correll. 1986. Ion and Acid Budgets for
a Forested Atlantic Coastal Plain Watershed and their Implications for the Impacts of
Acid Deposition. In: Watershed Research Perspectives. Correll, D.L. (ed.). Smithsonian
Institution Press, Washington, D.C., p. 392-421.
67
-------
Zarbock, H., A. Janicki, D. Wade, D. Heimbuch, and H. Wilson. 1994. Estimates of Total
Nitrogen, Total Phosphorus, and Total Suspended Solids Loadings to Tamps Bay,
Florida. Prepared for Tampa Bay National Estuary Program. Prepared by coastal
Engineering, Inc. St. Petersburg, FL.
68
-------
APPENDIX A
DESCRIPTION OF LOTUS 123* SPREADSHEET
A.I INTRODUCTION
A.2 INPUTS TO LOTUS 123* SPREADSHEET
A.3 OUTPUTS FROM LOTUS 123* SPREADSHEET
A.4 NOTES ON USING LOTUS 123* SPREADSHEET
A-l
-------
DESCRIPTION OF LOTUS 123* SPREADSHEET
A.I INTRODUCTION
The project methodology was described in Section 3.0. A Lotus 123 spreadsheet was
developed using the project methodology to estimate nitrogen loadings to aquatic systems. The
inputs required for the Lotus 123 spreadsheet are explained below in Section A.2. In Section
A.3 we provide instructions for obtaining the outputs from the spreadsheet, and Section A.4 we
provide notes on using the spreadsheet.
A.2 INPUTS TO LOTUS 123 SPREADSHEET
The spreadsheet is developed for a watershed with a twelve subbasins. If the watershed of
concern has only six subbasins then the user can input data for the six subbasins in the
spreadsheet. This is because, the default nonpoint and point loads from all the twelve subbasins
are set to zero in the spreadsheet. For watersheds with more than twelve subbasins, the user
can copy the block of cells A9..AI42 which represents one subbasin, to cell location A440 and
A475, etc. for each additional subbasin. The total nitrogen load at the watershed outlet is the
sum of nitrogen loads from identical nitrogen sources from each land use category and point
sources from all the subbasins in the watershed. For watersheds with more than twelve
subbasins, the user has to add the nonpoint and point source contributions from any additional
subbasin to the total nitrogen load that is computed using only twelve subbasins. The formulas
in the following cells need to be modified to include the contributions from the additional
subbasins: Point Sources - Cell AN15; Cropland - Fertilizer (Cell AN22), Atm. Dep. (Cell
AN23), and Manure (Cell AN24); Forest - Atm. Dep. (Cell AN27), Pasture - Manure (Cell
AN30), Atm. Dep. (Cell AN31); Wetlands - Atm. Dep. (Cell AN34); and Urban - Amt. Dep.
(Cell AN37).
The inputs to the spreadsheet include both nonpoint and point sources. The nonpoint source
inputs are needed for the land use categories of forest, cropland, wetlands, pastures, and urban
areas. The point source inputs are from industrial waste discharges and sewage treatment plant
discharges. The various nitrogen sources that a user can input for the land use categories is
listed in Table A.I.
As can be seen from Table A.I, the user does not input dry atmospheric deposition because we
calculate dry atmospheric deposition as a factor of wet atmospheric deposition. The dry
deposition is computed in the spreadsheet by multiplying the factor times wet atmospheric
deposition. This factor is an input to the spreadsheet. If dry atmospheric deposition data is
available to the user, then he/she can compute the factor which when multiplied with wet
A-2
-------
atmospheric deposition will yield dry atmospheric deposition. A layout of the Lotus 123*
spreadsheet is shown in Table A.2.
Table A. 1 Land Use Categories and Nitrogen Sources
Nitroen Sources
Forests Wet atmospheric deposition (kg/ha)
Cropland Wet atmospheric deposition (kg/ha), Fertilizer
application (kg/ha), and Manure application (kg/ha)
Wetlands Wet atmospheric deposition (kg/ha)
Pastures Wet atmospheric deposition (kg/ha) and Manure
Application (kg/ha)
Urban Wet atmospheric deposition (kg/ha)
The necessary inputs along with the cell locations in the spreadsheet are listed in Tables A.3 and
A.4. The input cells are high lighted in the spreadsheet for the user, the user can only
input data in these cells as all the other cells are protected. The user does not have to input
any data elsewhere in the spreadsheet, except for water surface area of the bay (ha), direct
atmospheric deposition (kg/ha) and direct point source loadings (kg) to the bay or
watershed outlet.
The spreadsheet computes the unit area inputs for each land use category by summing all the
nitrogen inputs from the all the sources for the land use of interest. The total nonpoint source
unit area inputs (kg/ha) and total point source inputs (kg) are displayed in column P of the
spreadsheet. In Column R, in the spreadsheet, we compute the Input Source Ratios using the
procedures described in Section 3.2.
The total nitrogen load from each land use is computed as a product of the land use area and
either the retention factor or literature reported edge-of-field unit area load. If retention factors
are not available to the user, or if he/she intends to use edge-of-field unit area loads, then the
user should input 999 in the retention factor cells of the spreadsheet; this flag indicates that no
retention factor is used for that land use. Table A.4 lists the last set of inputs that are required
to compute total nitrogen load from each land use category. In Table A.5 we list a sample input
data set required for the spreadsheet.
After computing nitrogen loads from each land use category, we compute Output Source Ratios
in column AB of the spreadsheet, the procedure used in computing this are described in Section
3.4. Nonpoint loads for each land use and the point sources are routed using an exponential
nitrogen decay relationship which is based on travel time and first-order nitrogen decay rate (as
discussed in Section 3.4 of the report); these computations are done in column AE of the
A-3
-------
spreadsheet.
As can be seen from Table A.3 and A.4 there are no cells in the subbasin input section where
the user can input direct point source and atmospheric deposition to the bay or watershed outlet.
The direct loadings of point source and atmospheric deposition are not included in the subbasin
inputs as these loads are not routed using travel time and the first-order nitrogen decay rate. In
cell B7, the user inputs direct point source loadings to the bay or watershed outlet in
kilograms, in cell C7 the user inputs the bay water surface area in hectares, and in cell A7
the user inputs direct atmospheric deposition in kilograms/hectare, which is the sum of wet
and dry deposition in kilograms to the bay or watershed outlet.
A.3 OUTPUTS FROM LOTUS 123 SPREADSHEET
The routed load then is multiplied by the Output Source Ratios to determine the source of
nitrogen in the subbasin load; the results from these computations are presented in column AI.
An example output from a subbasin is listed in Table A.6.
In Table A.7 and Table A.8 we present other forms of output that can be obtained by the user.
The user can print these tabular outputs by printing cells AK9..AN39 and cells AP9..AR21,
respectively. The output from the spreadsheet can also be obtained as a pie chart for the entire
watershed. The pie chart is depicted in cells AT9..AX21, and is shown in Figure A.I. The
outputs listed in Table A.7 are for a watershed with twelve subbasins. For watersheds with
more than twelve subbasins the user Jias to calculate the nitrogen load at the watershed outlet
by summing nitrogen loads from identical nitrogen sources from each land use category and
point sources from all the subbasins in the watershed.
A.4 NOTES ON USING LOTUS 123* SPREADSHEET
The user should follow the steps listed below to use the spreadsheet:
1. This spreadsheet is prepared using WINDOWS* LOTUS 123*,
Version 4.0. The name of the spreadsheet is NITRLOAD.WK4.
2. The spreadsheet is set up for a watershed with up to twelve
subbasins.
3. The nonpoint and point source inputs are highlighted in the
spreadsheet for each subbasin. The cells in which formulas are
used for computations are protected.
4. The user should use Tables A.3 and A.4 as a guideline to input
nonpoint and point source data for each subbasin.
5. After computing nitrogen loads from each subbasin, the user can
A-4
-------
obtain tables (such as Table A.7 and A.8) listing the total nitrogen
load from all the subbasins in the watershed by summing nitrogen
loads from identical nitrogen sources for each land use category
and point sources from all the subbasins in the watershed.
6. After computing Table A.8, the spreadsheet will display the
percent loadings from various nitrogen sources to the bay or
watershed outlet besides Table A.8.
A-5
-------
Table A.2 Format of Lotus 123 Spreadsheet for a Watershed with One Subbasin
Cell Ranees Spreadsheet Calculations
Cell A1..N435
Cell F16..N37'
Cell H40..N41*
Cell F52..N73*
CellH76..N77*
Cell F87..N108*
Cell HI 11..Ml 12*
Cell F123..N144*
Cell H147..N148'
Cell F159..N180*
Cell HI83..Nl84*
Cell F195..N216*
Cell H219..N220*
Cell F230..N251*
Cell H254..N255*
CellF266..N287*
CellH290..N291*
Cell F320..N323*
Cell H326..N327'
CellF338..N359*
Cell H362..N363*
Cell F373..N394*
Cell H397..N398*
Cell F409..N430*
Cell H433..N434*
Nonpoint and Point Source Data Input Sections
Nonpoint Source Data Input Range for Subbasin 1
Point Source Load Input Range for Subbasin 1 (kg)
Nonpoint Source Data Input Range for Subbasin 2
Point Source Load Input Range for Subbasin 2 (kg)
Nonpoint Source Data Input Range for Subbasin 3
Point Source Load Input Range for Subbasin 3 (kg)
Nonpoint Source Data Input Range for Subbasin 4
Point Source Load Input Range for Subbasin 4 (kg)
Nonpoint Source Data Input Range for Subbasin 5
Point Source Load Input Range for Subbasin 5 (kg)
Nonpoint Source Data Input Range for Subbasin 6
Point Source Load Input Range for Subbasin 6 (kg)
Nonpoint Source Data Input Range for Subbasin 7
Point Source Load Input Range for Subbasin 7 (kg)
Nonpoint Source Data Input Range for Subbasin 8
Point Source Load Input Range for Subbasin 8 (kg)
Nonpoint Source Data Input Range for Subbasin 9
Point Source Load Input Range for Subbasin 9 (kg)
Nonpoint Source Data Input Range for Subbasin 10
Point Source Load Input Range for Subbasin 10 (kg)
Nonpoint Source Data Input Range for Subbasin 11
Point Source Load Input Range for Subbasin 11 (kg)
Nonpoint Source Data Input Range for Subbasin 12
Point Source Load Input Range for Subbasin 12 (kg)
A-6
-------
Table A.2 (contd.)
Cell Ranges
Cell A7*
Cell B7*
CellC?'
Cell O1..S435
Cell X1..AI435
Cell AK9..AN39
Cell AP9..AR21
Cell AT9..AX21
Spreadsheet Calculations
Input of Direct Atmospheric Deposition (wet+dry) to the Bay or
Watershed Outlet in kg/ha
Input of Direct Point Source Load to the Bay or Watershed Outlet
in kg
Bay Water Surface Area in ha
Calculation of Input Source Ratios and Nitrogen Loads from various
nitrogen sources for all the land use categories in the subbasin,
as described in Section 3.2 and 3.3 in the report.
Calculation of Output Source Ratios and Routing Nitrogen Load from
each land use to the watershed outlet, as described in Section 3.4 of
the report.
Summary of Nonpoint and Point Source Loads from the Watershed
Percent Contributions from Various Nitrogen Sources to Bay
Pie Chart depicting source of nitrogen loads to the Bay or Watershed
outlet
* - These are the only cells for the user to input data
A-7
-------
Table A.3 Cell Locations in the Spreadsheet for Nonpoint and Point Source Inputs
Fertilizer Manure Sewage Industrial
kg/ha kg/ha kg kg
Subbasin 1
Wet Factor*
Deposition
kg/ha
Nonpoint
Forest
Cropland
Wetland
Pastures
Urban
Point
Subbasin 2
Nonpoint
Forest
Cropland
Wetland
Pastures
Urban
Point
Subbasin 3
Nonpoint
Forest
Cropland
Wetland
Pastures
Urban
H16
H20
H26
H30
H35
—
H52
H56
H62
H66
H71
~
H87
H91
H97
H101
H106
F17
F21 G22
F27
F31
F36
..
F53
F57 G58
F63
F67
F72
_.
F88
F92 G93
F98
F102
F107
^_
H23
H32
—
—
..
H59
H68
~
—
....
H94
H103
—
H40
H41
H76
H77
Point
Hill
H112
A-8
-------
Table A.3 (contd.)
Subbasin 4
Nonooint
Nonooint
Point
Wet
Deposition
kg/ha
Factor"
Fertilizer Manure
kg/ha kg/ha
Sewage
Industrial
kg
Forest
Cropland
Wetland
Pastures
Urban
Point
Subbasin 5
H123
H127
H133
H137
H142
—
F124
F128
F134
F138
F143
—
G129
H130
H139
H147
H148
Forest
Cropland
Wetland
Pastures
Urban
Point
Subbasin 6
Nonpoint
Forest
Cropland
Wetland
Pastures
Urban
H159
H163
H169
H173
H178
~
H195
H199
H205
H209
H214
F160
F164
F170
F174
F179
—
F196
F200
F206
F210
F215
G165
H166
H175
H183
H184
G201
H202
H211
H219
H220
A-9
-------
Table A.3 (contd.)
Subbasin 7
Nonpoint
Forest
Cropland
Wetland
Pastures
Urban
Point
Subbasin 8
Nonpoint
Point
Wet
Deposition
kg/ha
Factor"
Fertilizer Manure
kg/ha kg/ha
Sewage
Industrial
kg
H230
H234
H240
H244
H249
F231
F235
F241
F245
F250
G236
H237
H246
H254
H254
Forest
Cropland
Wetland
Pastures
Urban
Point
Subbasin 9
Nonpoint
Forest
Cropland
Wetland
Pastures
Urban
H266
H270
H276
H280
H285
—
H302
H306
H312
H316
H321
F267
F271
F277
F281
F286
—
F303
F307
F313
F317
F322
G272
H273
H282
H290
H291
G308
H309
H318
H326
H327
A-10
-------
Table A.3 (contd.)
Wet Factor" Fertilizer Manure Sewage Industrial
Deposition
kg/ha kg/ha kg/ha kg kg
Subbasin 10
Nonpoint
Forest H338 F339
Cropland H342 F343 G344 H345
Wetland H348 F349
Pastures H352 F353 - H354
Urban H357 F358
Point -- -- -- » H362 H363
Subbasin 11
Nonooint
H373 F374
G379 H380
H389
Point - - - -- H397 H398
Forest
Cropland
Wetland
Pastures
Urban
H373
H377
H383
H387
H392
F374
F378
F384
F388
F393
A-ll
-------
Table A.3 (contd.)
Wet Factor* Fertilizer Manure Sewage Industrial
Deposition
kg/ha kg/ha kg/ha kg kg
Subbasin 12
Nonpoint
Forest H409 F410
Cropland H413 F414 G415 H416
Wetland H419 F420
Pastures H423 F424 - H425
Urban H428 F429
Point - -- -- -- H433 H434
NOTE: Total atmospheric deposition is the sum of wet and dry atmospheric
deposition. Observed dry atmospheric deposition is computed in cells H9, H13,
H19, H23, H28 for the land use categories of forest, cropland, wetland,
pastures, and urban areas, respectively.
* - The factor calculates dry atmospheric deposition as a function of wet deposition
(Dry atmospheric deposition = Wet Atmospheric Deposition x Factor)
A-12
-------
Table A.4
Input Cell Locations of Retention Factor, Edge-of-Field Load, Area, Travel Time,
and First-Order Nitrogen Decay Rate.
Land Use
Area
(ha)
Retention
Factor"
Literature
Edge-of-Field
Nitrogen Load
(fraction) (kg/ha)
Travel
Time
(days)
First-Order
Nitrogen
In Stream
Decay Rate
(/day)
Nonpoint Subbasin 1
Forest
Cropland
Wetland
Pastures
Urban
J16
J20
J26
J30
J35
K16
K20
K26
K30
K35
L16
L20
L26
L30
L35
Nonpoint Subbasin 2
Nomxrint Subbasin 3
Ml 6
N16"
Forest
Cropland
Wetland
Pastures
Urban
J52
J56
J62
J66
J71
K52
K56
K62
K66
K71
L52
L56
L62
L66
L71
M52"
N52"
Forest
Cropland
Wetland
Pastures
Urban
J87
J91
J97
J101
J106
K87
K91
K97
K101
K106
L87
L91
L97
L101
L106
Nonooint Subbasin 4
M87"
N87"
Forest
Cropland
Wetland
Pastures
Urban
J123
J127
J133
J137
J142
K123
K127
K133
K137
K142
L123
L127
L133
LI 37
L142
M123'
N123*
A-13
-------
Table A.4 (contd.)
Land Use Retention Literature Travel First-Order
Area Factor* Edge-of-Field Time Nitrogen
Nitrogen Load In Stream
Decay Rate
(ha) (fraction) (kg/ha) (days) (/day)
Nonooint Subbasin 5
M159" N159*
Forest
Cropland
Wetland
Pastures
Urban
J159
J163
J169
J173
J178
K159
K163
K169
K173
K178
L159
LI 63
L169
L173
LI 78
Nonpoint Subbasin 6
M195" N195'
Forest
Cropland
Wetland
Pastures
Urban
J195
J199
J205
J209
J214
K195
K199
K205
K209
K214
LI 95
L199
L205
L209
L214
Nonpoint Subbasin 7
M230" N230*
Forest
Cropland
Wetland
Pastures
Urban
J230
J234
J240
J244
J249
K230
K234
K240
K244
K249
L230
L234
L240
L244
L249
Nonpoint Subbasin 8
Forest J266 K266 L266 M266" N266*
Cropland J270 K270 L270
Wetland J276 K276 L276
Pastures J280 K280 L280
Urban J285 K285 L285
A-14
-------
Table A.4 (contd.)
Land Use Retention Literature Travel First-Order
Area Factor* Edge-of-Field Time Nitrogen
Nitrogen Load In Stream
Decay Rate
(ha) (fraction) (kg/ha) (days) (/day)
Nonpoint Subbasin 9
M302" N302"
Forest
Cropland
Wetland
Pastures
Urban
J302
J306
J312
J316
J321
K302
K306
K312
K316
K321
L302
L306
L312
L316
L321
Nonpoint Subbasin 10
M338" N338*
Forest
Cropland
Wetland
Pastures
Urban
J338
J342
J348
J352
J357
K338
K342
K348
K352
K357
L338
L342
L348
L352
L357
Nonpoint Subbasin 11
M373" N373*
Forest
Cropland
Wetland
Pastures
Urban
J373
J377
J383
J387
J392
K373
K377
K383
K387
K392
L373
L377
L383
L387
L392
Nonpoint Subbasin 12
M409" N409*
Forest
Cropland
Wetland
Pastures
Urban
J409
J413
J419
J423
J428
K409
K413
K419
K423
K428
L409
L413
L419
L423
L428
* - Input 999 if the user wants to use edge-of-field nitrogen loads to compute
total nitrogen load.
** - Travel Time and First-Order Nitrogen Instream Decay Rate is only input once, the
same travel time and first-order decay rate is used for all land use categories.
A-15
-------
Table A.5 Example Input from Lotus 123 Spreadsheet for a Subbasin
(Cell Range A9..NA1)
SUBBASIN NAME
NONPCXNT SOURCES
FORESTS
Aim. Wet Deposition kg/ha
Aim. Dry Deposition kg/Via
CROPLANDS
Aim. Wet Deposition kg/ha
Aim. Dry Deposition kg/ha
Fertilizer kgAta
Manure kg/ha
WETLANDS
Aim. Wet Deposition kg/ha
Aim. Dry Deposition kg/ha
PASTURES
URBAN
POINT SOURCES
Aim. Wet Deposition kg/ha
Aim. Dry Deposition kg/ha
Manure kg/ha
Aim. Wet Deposition kg/ha
Atm. Dry Deposition kg/ha
Sewage Treatment Plants kg
Industrial Discharge kg
INPUTS INPUTS INPUTS INPUTS INPUTS INPUTS
NPS
Unit Area Retention Loads Travel River
Landuse Factor Literature Time Decay
Pert. Rates Areas (fraction) Values days /day
Factors kg/ha kg/ha ha kg/ha
1.00
1.00
1.00
1.00
1.00
1.00
7.74 1.73E+06 999.00
7.74
5.67
8.60
0.08
7.74 6.03E+05 999.00 16.06 8.60 0.08
7.74
99 99.00
14.00
7.74 O.OOE+00 999.00 6.00 8.60 0.08
7.74
7.74 2.58E+05 999.00
7.74
11.00
7.21
8.60 0.08
7.74 3.32E+05 999.00 11.80 8.60 0.08
7.74
15000
20000
8.60
0.08
A-16
-------
Table A.6 Example Output from Louts 123 Spreadsheet for a Subbasin
(Cell Range AG9..AI41)
OUTPUTS
To Bay
from
Subbasin
FOREST
Atm. Loads
OUTPUTS
Subbasin
Nitrogen
Sources
kg
4.93E+06
CROPLAND
Atm. Loads
Fertilizer
Manure
WETLANDS
Atm. Loads
5.86E+05
3.75E+06
5.30E+05
O.OOE+00
PASTURES
Atm. Loads
Manure
URBAN
Atm. Loads
5.47E+05
3.89E+05
1.97E+06
POINT SOURCES
1.76E+04
A-17
-------
Table A.7 Example Total Nitrogen Load Output from Lotus 123 Spreadsheet for the
Watershed (Cell Range AK9..AN39)
Total Nitrogen Loads from Watershed
Point Sources-Watershed
Point Sources-Direct to Bay
Direct Atm. Depo. to Bay
Cropland
Fertilizer
Atmospheric Dep.
Manure
Forest
Atmospheric Dep.
Pasture
Atmospheric Dep.
Manure
Wetland
Atmospheric Dep.
Urban
Atmospheric Dep.
Total Load
Watershed
Results (kg)
1.76E+04
4.00E+05
1.29E+07
3.75E+06
5.86E+05
5.30E+05
4.93E+06
3.89E+05
5.47E+05
0.00+00
1.97E+06
2.60E+07
A-18
-------
Table A.8 Example Output for Percent Nitrogen Contributions from Various Nitrogen Sources
using Lotus 123* Spreadsheet for the Watershed (Cell Range AP9..AR21)
Percent
of Total
Load
Point Sources 1.60%
Atm. Bay 49.57%
Atm. Watershed 30.88%
Fertilizer 14.41%
Manure 3.53%
A-19
U.S. Environmental Protection Agency
'**"»
-------
(49.6%)
(30.9%)
(1.6%)
(3.5%)
(14.4%)
• Point Sources eg Atm. Bay
^Fertilizer • Manure
lAtm. Watershe
Figure A.I Percent Nitrogen Loadings from Various Nitrogen
Sources at Watershed Outlet (Cell Range AT9..AX21)
A-20
------- |