UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON D.C. 20460
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
July 19,2011
EPA-SAB-11-010
The Honorable Lisa P. Jackson
Administrator
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, N.W.
Washington, D.C. 20460
Subject: Review of EPA's draft Approaches for Deriving Numeric Nutrient Criteria for
Florida's Estuaries, Coastal Waters, and Southern Inland Flowing Waters
Dear Administrator Jackson:
Nitrogen and phosphorus inputs from urban and agricultural sources are known to impact water quality,
and nutrient pollution has been identified as a source of impairment for estuarine, marine and fresh
waters in Florida. In 2009, the EPA determined that numeric nutrient criteria were needed to protect
aquatic life in Florida, and initiated a process to develop such criteria for categories of state waters. As
part of that process, the Agency developed a draft document, Methods and Approaches for Deriving
Numeric Criteria for Nitrogen/Phosphorus Pollution in Florida's Estuaries, Coastal Waters, and
Southern Inland Flowing Waters. The EPA document proposes a conceptual model for relating nutrient
levels to biological endpoints, and describes data sources and possible approaches to define criteria for
each of the categories of waters: estuaries, coastal waters, inland flowing waters (including canals) in
South Florida, and South Florida marine waters. The Science Advisory Board (SAB) was asked to
review and provide advice on the draft EPA document and an ad hoc SAB panel, the Nutrient Criteria
Review Panel, was formed for this task.
The SAB acknowledges the substantial effort that already has been made by EPA to get the work to this
stage, much of it solid and thoughtful. However, much work remains to be done to develop nutrient
criteria for Florida waters (for example, to develop and validate models for over 20 Florida estuaries).
Given the extent of the task at hand, the SAB has attempted to prioritize its recommendations and to
emphasize a role for future refinements of criteria in the spirit of adaptive management.
To guide its development of nutrient criteria, the EPA proposes a conceptual model that links nitrogen
and phosphorus levels in Florida waters to biological endpoints to be protected using one or more
analytical approaches. Nitrogen and phosphorus may be limiting in different portions of the fresh-to-
marine continuum, and in some cases may be co-limiting. Thus, a dual nutrient (N and P) strategy is
warranted, and we agree with the Agency's decision to take this approach. Although the general
conceptual model provides a starting point for choosing numeric criteria, the SAB has numerous
concerns about how the driver variables—total nitrogen (TN) and total phosphorus (TP)—would be
linked to measurable biological endpoints. The SAB recommends that the EPA provide a more detailed
-------
conceptual model that includes additional endpoints and flows, and suggests that system-specific
diagrams be included for each of the four waterbody types.
The EPA is proposing three general approaches to relate nutrient levels to balanced natural populations:
(1) reference conditions; (2) predictive stressor-response relationships; and/or (3) numerical water
quality models. The SAB noted the uneven treatment of the three approaches (i.e., the emphasis on
water quality modeling), and encourages the EPA to continue to develop all three. The SAB agrees that
these approaches all have utility and recommends that a combination be used where data and models are
available. However, the Agency should provide more detail on the adequacy of the data for applying
each approach; how decisions would be made on which approaches to use; and how discrepencies in
targets derived from different approaches would be resolved. Although a complete uncertainty analysis
may not be feasible, the document should clearly indicate what is included in any uncertainty analysis
undertaken or contemplated. In particular, the EPA may need to specify some probabilistic goals for
meeting the specified nutrient criteria and then set thresholds for nitrogen and phosphorus loading
accordingly, to ensure that the criteria are met with a desired level of confidence.
The proposed biological endpoints (healthy seagrasses, balanced phytoplankton biomass, and balanced
faunal communities) are appropriate, but it is critical that the EPA define "balanced" for each of these
endpoints, preferably in quantitative terms. The SAB agrees with the Agency's broad delineation of
Florida coastal waters into four categories (estuaries, coastal waters, South Florida inland waters, and
South Florida estuarine and coastal waters) for purposes of criteria development, but suggests some
refinements to segmentation within the categories. The EPA also proposes to use downstream protection
value (DPV) criteria to ensure that upstream nutrient criteria will be set at levels that will protect
downstream estuaries. While agreeing with the goal of downstream protection from nutrient impacts, the
SAB has concern with the overlap between DPV and the Total Maximum Daily Load (TMDL) process.
In the enclosed report, the SAB provides numerous recommendations to strengthen the application of
the three approaches to develop numeric nutrient criteria for Florida waters. However, given the
Agency's time frame, the SAB offers the following priority recommendations:
• To provide greater confidence in the criteria, a combination of approaches should be used to
develop numeric nutrient criteria for each category of waters where data and models are
available.
• For estuaries, the SAB recommends that the EPA adopt additional measures of seagrass health
beyond the proposed use of chlorophyll-a, and encourages the use of direct measures of the
faunal communities to be protected, rather than relying on a dissolved oxygen criterion.
• For coastal waters, the SAB agrees that a criterion based on satellite-derived estimates of
chlorophyll may be the only feasible approach for this large, poorly sampled region. However,
the SAB recommends that the Agency expand the dataset to include waters farther than three
miles offshore and verify the strength of the relationship between pollutant loads from land and
observed chlorophyll-a concentrations using direct measurements of nutrients, where possible.
• For South Florida inland waters, the SAB was not convinced by the available data that nutrient
criteria based on instream protection values were meaningful for man-made and managed canals.
The canals do provide ecosystem services, but habitat quality and flows—rather than nutrients—
have the greatest influence on biological condition in these managed waterways. The SAB does
agree that nutrients in canal waters should be managed to ensure downstream, estuarine
designated uses.
• For South Florida coastal and estuarine waters, the SAB recommends that seagrass endpoints be
considered in addition to chlorophyll-a.
-------
• If the DPV approach is pursued, the SAB recommends that apportionment strategies not preclude
flexible nutrient allocation across tributaries to achieve the necessary estuarine load reductions.
In closing, the SAB encourages the EPA to continue efforts to develop numeric nutrient criteria for
Florida estuarine and coastal waters, using the best available scientific data and methods. Ongoing
changes in regional hydrology and climate, which will alter freshwater flows (and thereby, nutrient
concentrations) and ecological responses, make it important that these nutrient criteria be revisited
periodically in an adaptive management approach. We appreciate the opportunity to provide advice on
this important endeavor, and look forward to your response.
Sincerely,
/signed/
/signed/
Dr. Deborah L. Swackhamer, Chair
Science Advisory Board
Dr. Judith L. Meyer, Chair
SAB Nutrient Criteria Review Panel
Enclosure
-------
NOTICE
This report has been written as part of the activities of the EPA Science Advisory Board (SAB), a public
advisory group providing extramural scientific information and advice to the Administrator and other
officials of the Environmental Protection Agency. The SAB is structured to provide balanced, expert
assessment of scientific matters related to problems facing the Agency. This report has not been
reviewed for approval by the Agency and, hence, the contents of this report do not necessarily represent
the views and policies of the Environmental Protection Agency, nor of other agencies in the Executive
Branch of the Federal government, nor does mention of trade names of commercial products constitute a
recommendation for use. Reports of the SAB are posted on the EPA website at
http ://www. epa. gov/ sab.
1
-------
U.S. Environmental Protection Agency
Science Advisory Board
Nutrient Criteria Review Panel
CHAIR
Dr. Judith L. Meyer, Professor Emeritus, Odum School of Ecology, University of Georgia, Lopez
Island, WA
SAB MEMBERS
Dr. Walter Boynton, Professor, Chesapeake Biological Laboratory, Center for Environmental Science,
University of Maryland, Solomons, MD
Dr. Deborah Bronk, Professor, Physical Science Dept, The College of William and Mary, Virginia
Institute of Marine Science, Gloucester Point, VA
Dr. Piers Chapman, Professor and Head, Department of Oceanography, Texas A&M University,
College Station, TX
Dr. Robert Diaz, Professor, Department of Biological Sciences, Virginia Institute of Marine Science,
College of William and Mary, Gloucester Pt., VA
Dr. Anne Giblin, Senior Scientist, The Ecosystems Center, Marine Biological Laboratory, Woods Hole,
MA
Dr. Kenneth L. Heck, Jr, Chair, University Programs/Senior Marine Scientist Ill/Professor, Dauphin
Island Sea Lab, Dauphin Island, AL
Dr. Mark Noll, Associate Professor, Earth Sciences & Special Assistant to the Provost, Academic
Affairs, The College at Brockport, State University of New York, Brockport, NY
Dr. Peter Ortner, Research Professor, Marine Biology and Fisheries, Rosenstiel School of Marine and
Atmospheric Science, University of Miami, Miami, FL
Dr. Hans Paerl, Professor of Marine and Environmental Sciences, Institute of Marine Sciences,
University of North Carolina - Chapel Hill, Morehead City, NC
Dr. Kenneth Reckhow, Chief Scientist, Global Climate Change and Environmental Sciences, RTI
International, Research Triangle Park, NC
Dr. James Sanders, Director and Professor, Skidaway Institute of Oceanography, Savannah, GA
Dr. David C. Schneider, Professor, Ocean Sciences Centre, Memorial University, St. John's, NL,
Canada
11
-------
Dr. Andrew N. Sharpley, Research Soil Scientist, Department of Crop, Soil and Environmental
Sciences, University of Arkansas, Fayetteville, AR
Dr. Andrew Solow, Associate Scientist, and Center Director, Woods Hole Oceanographic Institution,
Woods Hole, MA
Dr. Alan Steinman, Director, Annis Water Resources Institute, Grand Valley State University,
Muskegon, MI
Dr. Joseph C. Zieman, Professor, Environmental Sciences, College and Graduate School for Arts and
Sciences, University of Virginia, Charlottesville, VA
SCIENCE ADVISORY BOARD STAFF
Ms. Stephanie Sanzone, Designated Federal Officer, U.S. Environmental Protection Agency, Science
Advisory Board (1400R), 1200 Pennsylvania Avenue, NW, Washington, DC, 20460
111
-------
U.S. Environmental Protection Agency
Science Advisory Board
CHAIR
Dr. Deborah L. Swackhamer, Professor and Charles M. Denny, Jr., Chair in Science, Technology and
Public Policy- Hubert H. Humphrey School of Public Affairs and Co-Director of the Water Resources
Center, University of Minnesota, St. Paul, MN
MEMBERS
Dr. David T. Allen, Professor, Department of Chemical Engineering, University of Texas, Austin, TX
Dr. Claudia Benitez-Nelson, Professor and Director of the Marine Science Program, Department of
Earth and Ocean Sciences , University of South Carolina, Columbia, SC
Dr. Timothy Buckley, Associate Professor and Chair, Division of Environmental Health Sciences,
College of Public Health, The Ohio State University, Columbus, OH
Dr. Patricia Buffler, Professor of Epidemiology and Dean Emerita, Department of Epidemiology,
School of Public Health, University of California, Berkeley, CA
Dr. Ingrid Burke, Director, Haub School and Ruckelshaus Institute of Environment and Natural
Resources, University of Wyoming, Laramie, WY
Dr. Thomas Burke, Professor, Department of Health Policy and Management, Johns Hopkins
Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD
Dr. Terry Daniel, Professor of Psychology and Natural Resources, Department of Psychology, School
of Natural Resources, University of Arizona, Tucson, AZ
Dr. George Daston, Victor Mills Society Research Fellow, Product Safety and Regulatory Affairs,
Procter & Gamble, Cincinnati, OH
Dr. Costel Denson, Managing Member, Costech Technologies, LLC, Newark, DE
Dr. Otto C. Doering III, Professor, Department of Agricultural Economics, Purdue University, W.
Lafayette, IN
Dr. David A. Dzombak, Walter J. Blenko Sr. Professor of Environmental Engineering , Department of
Civil and Environmental Engineering, College of Engineering, Carnegie Mellon University, Pittsburgh,
PA
Dr. T. Taylor Eighmy, Vice President for Research, Office of the Vice President for Research, Texas
Tech University, Lubbock, TX
Dr. Elaine Faustman, Professor and Director, Institute for Risk Analysis and Risk Communication,
School of Public Health, University of Washington, Seattle, WA
iv
-------
Dr. John P. Giesy, Professor and Canada Research Chair, Veterinary Biomedical Sciences and
Toxicology Centre, University of Saskatchewan, Saskatoon, Saskatchewan, Canada
Dr. Jeffrey K. Griffiths, Professor, Department of Public Health and Community Medicine, School of
Medicine, Tufts University, Boston, MA
Dr. James K. Hammitt, Professor, Center for Risk Analysis, Harvard University, Boston, MA
Dr. Bernd Kahn, Professor Emeritus and Associate Director, Environmental Radiation Center, Georgia
Institute of Technology, Atlanta, GA
Dr. Agnes Kane, Professor and Chair, Department of Pathology and Laboratory Medicine, Brown
University, Providence, RI
Dr. Madhu Khanna, Professor, Department of Agricultural and Consumer Economics, University of
Illinois at Urbana-Champaign, Urbana, IL
Dr. Nancy K. Kim, Senior Executive, Health Research, Inc., Troy, NY
Dr. Kai Lee, Program Officer, Conservation and Science Program, David & Lucile Packard Foundation,
Los Altos, CA
Dr. Cecil Lue-Hing, President, Cecil Lue-Hing & Assoc. Inc., Burr Ridge, IL
Dr. Floyd Malveaux, Executive Director, Merck Childhood Asthma Network, Inc., Washington, DC
Dr. Lee D. McMullen, Water Resources Practice Leader, Snyder & Associates, Inc., Ankeny, IA
Dr. Judith L. Meyer, Professor Emeritus, Odum School of Ecology, University of Georgia, Lopez
Island, WA
Dr. James R. Mihelcic, Professor, Civil and Environmental Engineering, University of South Florida,
Tampa, FL
Dr. Jana Milford, Professor, Department of Mechanical Engineering, University of Colorado, Boulder,
CO
Dr. Christine Moe, Eugene J. Gangarosa Professor, Hubert Department of Global Health, Rollins
School of Public Health, Emory University, Atlanta, GA
Dr. Horace Moo-Young, Dean and Professor, College of Engineering, Computer Science, and
Technology, California State University, Los Angeles, CA
Dr. Eileen Murphy, Grants Facilitator, Ernest Mario School of Pharmacy, Rutgers University,
Piscataway, NJ
Dr. Duncan Patten, Research Professor, Hydroecology Research Program , Department of Land
Resources and Environmental Sciences, Montana State University, Bozeman, MT
v
-------
Dr. Stephen Polasky, Fesler-Lampert Professor of Ecological/Environmental Economics, Department
of Applied Economics, University of Minnesota, St. Paul, MN
Dr. Arden Pope, Professor, Department of Economics, Brigham Young University, Provo, UT
Dr. Stephen M. Roberts, Professor, Department of Physiological Sciences, Director, Center for
Environmental and Human Toxicology, University of Florida, Gainesville, FL
Dr. Amanda Rodewald, Professor of Wildlife Ecology, School of Environment and Natural Resources,
The Ohio State University, Columbus, OH
Dr. Jonathan M. Samet, Professor and Flora L. Thornton Chair, Department of Preventive Medicine,
University of Southern California, Los Angeles, CA
Dr. James Sanders, Director and Professor, Skidaway Institute of Oceanography, Savannah, GA
Dr. Jerald Schnoor, Allen S. Henry Chair Professor, Department of Civil and Environmental
Engineering, Co-Director, Center for Global and Regional Environmental Research, University of Iowa,
Iowa City, IA
Dr. Kathleen Segerson, Philip E. Austin Professor of Economics , Department of Economics,
University of Connecticut, Storrs, CT
Dr. Herman Taylor, Director, Principal Investigator, Jackson Heart Study, University of Mississippi
Medical Center, Jackson, MS
Dr. Barton H. (Buzz) Thompson, Jr., Robert E. Paradise Professor of Natural Resources Law at the
Stanford Law School and Perry L. McCarty Director, Woods Institute for the Environment, Stanford
University, Stanford, CA
Dr. Paige Tolbert, Professor and Chair, Department of Environmental Health, Rollins School of Public
Health, Emory University, Atlanta, GA
Dr. John Vena, Professor and Department Head, Department of Epidemiology and Biostatistics,
College of Public Health, University of Georgia, Athens, GA
Dr. Thomas S. Wallsten, Professor and Chair, Department of Psychology, University of Maryland,
College Park, MD
Dr. Robert Watts, Professor of Mechanical Engineering Emeritus, Tulane University, Annapolis, MD
Dr. R. Thomas Zoeller, Professor, Department of Biology, University of Massachusetts, Amherst, MA
vi
-------
SCIENCE ADVISORY BOARD STAFF
Dr. Thomas Armitage, Designated Federal Officer, U.S. Environmental Protection Agency, Science
Advisory Board (1400R), 1200 Pennsylvania Avenue, NW, Washington, DC, 20460
Dr. Angela Nugent, Designated Federal Officer, U.S. Environmental Protection Agency, Science
Advisory Board (1400R), 1200 Pennsylvania Avenue, NW, Washington, DC, 20460
vii
-------
Table of Contents
1. EXECUTIVE SUMMARY 1
2. INTRODUCTION 7
2.1. Background 7
2.2. Charge to the SAB 7
3. RESPONSE TO CHARGE QUESTIONS 8
3.1. Conceptual Approach 8
3.2. Florida Estuaries 15
3.3. Florida Coastal Waters 20
3.4. South Florida Inland Flowing Waters 23
3.5. South Florida Marine Waters 32
3.6. Downstream Protection Values 35
REFERENCES R-l
APPENDIX A: Charge to the SAB A-l
APPENDIX B: Phosphorus Transport and Fate in Freshwater Systems B-l
APPENDIX C: Options for Distributing Loads C-l
viii
-------
1. EXECUTIVE SUMMARY
Nitrogen (N) and phosphorus (P) inputs from urban and agricultural sources are known to influence
water quality, and nutrient pollution has been identified as a source of impairment for estuarine, marine
and fresh waters in Florida. The state of Florida has a narrative criterion for nutrients, which states that
"in no case shall nutrient concentrations of a body of water be altered so as to cause an imbalance in
natural populations of flora or fauna." The criterion specifically refers to degradation from
anthropogenic loads of total nitrogen (TN) or total phosphorus (TP). In 2009, EPA determined that
numeric nutrient criteria were needed to protect aquatic life in Florida, and initiated a process to develop
such criteria for categories of state waters.
The SAB was asked to review and provide advice on the proposed approaches for estuarine, coastal and
South Florida waters, as described in the draft EPA document, Methods and Approaches for Deriving
Numeric Criteria for Nitrogen/Phosphorus Pollution in Florida's Estuaries, Coastal Waters, and
Southern Inland Flowing Waters. An ad hoc SAB panel, the Nutrient Criteria Review Panel, was formed
for this task. The Charge to the SAB included questions about the conceptual model used to select
assessment endpoints, data sources, and possible approaches to define criteria for each of the categories
of waters: estuaries, coastal waters, inland flowing waters (including canals) in South Florida, and South
Florida marine waters.
The SAB acknowledges the substantial effort that already has been made by EPA to get the work to this
stage, much of it solid and thoughtful. However, much work remains to be done. It is not clear what
approaches will be selected, if multiple approaches will be used, and which approaches will provide
useful information towards the goal of developing nutrient criteria. Under a court-ordered consent
decree, the Agency has committed to proposing nutrient criteria for estuaries, coastal waters, and South
Florida inland flowing and marine waters by November 14, 2011 and final criteria by August 14, 2012.
Given this time frame, and the extent of the task at hand, the SAB has attempted to prioritize its
recommendations and to emphasize a role for future refinements of criteria in the spirit of adaptive
management.
Different approaches are being considered for each of four major categories of Florida waters.
Consistent with the Charge to the SAB, the SAB provides review and advice on the proposed
approaches for developing nutrient criteria for each category of waters.
Conceptual Model (Charge Question la)
EPA proposes a conceptual model that links nitrogen and phosphorus levels in Florida waters to
biological endpoints to be protected, using one or more analytical approaches. Although the general
conceptual model provides a starting point for choosing numeric criteria, the SAB has numerous
concerns about how TN and TP would be linked to measurable biological endpoints. The SAB
recommends that the EPA provide a more detailed conceptual model that includes additional endpoints
and flows, and suggests that system-specific diagrams be included for each of the four waterbody types.
TN and TP should be considered driver (rather than causal) variables because nitrogen and phosphorus
are two factors that regulate primary production but they do not "cause" it. The conceptual model should
acknowledge that there are many factors other than nutrients that control water column chlorophyll (Chl-
a). TN and TP loadings are likely to be better predictors of Chl-a, hypoxia and seagrass loss than TN or
TP concentrations. TN and TP concentrations also can be considered response variables in that they
reflect long-term loading rates; the temporal and spatial scales over which they would be measured
should be clarified.
1
-------
The conceptual diagram is a representation of important linkages, but Biological Endpoints and
Objectives are not discussed in sufficient detail. The three biological endpoints (healthy seagrasses,
balanced phytoplankton biomass, and balanced faunal communities) are appropriate. However, these
endpoints need to be much better defined and, in some cases, more clearly connected to the explanatory
variables that would be the basis for setting numeric criteria. It is critical that EPA define "balanced" for
each of the three biological endpoints, preferably in quantitative terms. Simply relying on the state's
assessment of whether or not waters are "healthy" is not an adequate definition.
EPA proposes to identify levels of water quality variables that would protect the biological endpoints
using reference condition, stressor-response, and/or modeling approaches for a particular water body.
The SAB strongly recommends that multiple approaches be applied to each of the systems where data
and models are available. The EPA document should discuss how the results from multiple approaches
would be integrated to develop the final numeric criteria. The three analytical approaches are being
applied somewhat differently within the different categories of Florida waters, and each approach has
different data requirements and—more importantly—different assumptions, limitations and
uncertainties. EPA should describe the uncertainty associated with the various approaches to criteria
development and discuss how this uncertainty might influence the use and appropriateness of specific
numeric criteria.
The SAB agrees that Chi-a concentration in the water column is both sensitive to nutrient inputs and an
important measure of ecosystem health and therefore a reasonable endpoint in itself. However, the EPA
document should be explicit that Chi-a measures phytoplankton biomass and cannot be used to infer
anything about primary productivity or whether or not phytoplankton populations are "balanced" in
terms of species composition or relative abundance/dominance. Water column Chl-a also is linked to
seagrass health because it is an indirect measure of water clarity, which influences seagrass
photosynthesis. However, the SAB is concerned about relying upon Chl-a as the sole criterion to protect
seagrasses because, in some systems, macroalgae or epiphyte growth can significantly impact seagrass
communities even as water column Chl-a levels remain low. Thus, the SAB recommends that EPA use a
stressor-response approach to link nutrient loading with seagrass areal extent for protecting seagrass
communities.
The SAB is concerned that no direct measures of the faunal community are proposed to define whether a
balanced community is being protected and maintained. Instead, EPA proposes to rely on attainment of
the State of Florida dissolved oxygen (DO) standard as an indicator for the presence of balanced faunal
communities. EPA proposes to look for relationships between TN and/or TP and DO, and use those
relationships to determine numeric criteria for TN and TP that are protective. The SAB recommends that
faunal metrics be considered to evaluate the estuarine endpoint of balanced faunal communities.
General Delineation of Florida Waters (Charge Question lb)
In the document, EPA proposes an initial grouping of Florida waters into four categories: estuaries,
coastal waters, South Florida inland flowing waters, and South Florida marine waters. Separation of
estuarine and coastal waters is appropriate, and the separate consideration of South Florida also is
warranted. The SAB recommends that the term "marine waters" be replaced with "estuarine and coastal
waters." Further, South Florida inland flowing waters appears to be a default category without strong
scientific rationale for this classification; for purposes of criteria development, these waters should be
subdivided into managed canals and natural streams. Additional comments are provided on sub-
delineation and classification within the four broad categories of waters, below.
2
-------
Florida Estuaries
Delineation and Data (Charge Question 2a). The geographic delineations of estuaries seem
appropriate although it was not clear why a salinity of 2.7 psu rather than 0.5 psu was used to delineate
the upper reaches of these systems. A finer classification based on degree of impact may be useful (e.g.,
to separate Caloosahatchee and St. Lucie estuaries from the others given their unique hydrologic
relationship to Lake Okeechobee). EPA should consider adding tidal creeks as a separate ecosystem type
because they have different characteristics than the open estuaries and therefore may require different
nutrient criteria. The SAB has few issues with the data sets presented and has provided suggestions for
additional sources of data.
Assessment Endpoints (Charge Question 2b). Healthy seagrass communities are an appropriate
biological endpoint for Florida estuaries, given their widespread occurrence and ecological (and
economic) significance. To protect seagrass communities, EPA should consider a measure of epiphyte
abundance in addition to the proposed determination of Chl-a in the water column. Whether Chl-a is an
appropriate endpoint for assessing "balanced" phytoplankton communities, depends on how EPA
defines "balanced", which has not been done. Direct indicators of faunal community balance should be
considered in addition to DO. When hypoxic conditions are observed, impacts on the biota usually have
already occurred. Therefore, it is preferable to identify indicators that show stress on the faunal
community before such degraded conditions develop. Diel variability in DO needs to be considered
when establishing DO water quality targets. DO criteria may be better characterized by percent
saturation than by concentration.
Approaches (Charge Question 2c). N and P may be limiting in different portions of the fresh-to-
marine continuum, and in some cases may be co-limiting. Thus, a dual nutrient (N and P) strategy is
warranted and the SAB agrees with the EPA's decision to take this approach. The SAB acknowledges
the substantial effort made to date to collect data and evaluate possible approaches to criteria
development for these waters, but is concerned about the timetable for completion of this work. Previous
experience suggests that if the reference condition approach can be implemented, it might be the most
"time-efficient" pathway to developing nutrient criteria, although the stressor-response approach is
worthy of more attention than it has been given. The SAB is aware of a separate EPA report on stressor-
response approaches, but would have preferred additional discussion in the present document on how
such approaches might be applied in Florida's estuaries. The SAB urges caution in EPA's apparent
emphasis on water quality modeling approaches because development and validation of simulation
models for numerous estuaries to establish nutrient criteria would be a major undertaking, requiring
considerable time and money, and useful results are not guaranteed; for waters where this approach is
selected, a reasonable representation of internal nutrient cycling needs to be included.
In light of climate-related effects on hydrology and temperature regimes, as well as increases in
freshwater withdrawals for human use, the SAB recommends that the EPA consider the possibility that
thresholds could be crossed, fundamentally changing these systems. In a spirit of adaptive management,
the SAB recommends that future refinements to proposed nutrient criteria for Florida estuarine, coastal
and marine waters explicitly consider the impact of changing hydrology and climate on the numeric
criteria needed to ensure the protection of designated uses.
Florida Coastal Waters
Delineation and data sources (Charge Question 3a). EPA proposes to use remotely sensed
chlorophyll (ChlRs-a) to develop a reference criterion associated with a balanced phytoplankton
population in coastal waters. The approach is appropriate and sensible for this large, poorly sampled
3
-------
region, although the SAB recommends that direct measurements of nutrients, where available, be used
to verify the strength of the relationship between pollutant loads from land and observed Chl-a
concentrations in coastal waters. EPA proposes to use chlorophyll data from coastal waters out to three
miles. However, under the Clean Water Act, water quality criteria apply to state coastal waters, defined
for Florida as waters out to three nautical miles on the Atlantic coast and nine nautical miles on the Gulf
coast. In developing coastal criteria, the SAB recommends that EPA consider remote-sensed chlorophyll
in waters beyond the three-mile zone because some blooms observed in coastal waters may form further
offshore. In addition, data from the entire shelf should be used to improve the calibration of remotely
sensed and field-measured {in situ) chlorophyll. The SAB further notes that the proposed coastal
segments are a result of historical precedence, and EPA may wish to consider segments defined in terms
of bathymetry.
Assessment Endpoints (Charge Question 3b). While acknowledging that ChlRS-a is the most feasible
indicator of nutrient status for coastal waters, given available data and the lack of a better alternative, the
SAB has several concerns: (1) there is a causal relationship between ChlRS-a and nutrient loads but a
number of other factors also influence ChlRS-a levels; (2) few in situ Chi -a data are available for
calibration; (3) the ten-year remote-sensing dataset may not constitute an adequate baseline, given
decadal-scale variability; and (4) the future availability of ChlRS-a data is in doubt, particularly given the
loss of the SeaWiFS sensor.
Approaches (Charge Question 3c). The SAB agrees that the reference condition approach for coastal
waters using remotely sensed Chl-a is an appropriate application of remote sensing data, with certain
caveats. In order to relate remotely sensed chlorophyll to water column chlorophyll levels, it is
necessary to calibrate satellite sensor readings using field-measured chlorophyll data. The approach to
calibration has been thorough, and the SAB agrees with use of in situ calibration data taken within 3
hours of the satellite overpass. The ratio between the chlorophyll concentrations in the upper two meters
and the full euphotic zone needs to be established, and obvious antecedent bloom data points should be
removed from analyses. Because available satellite-based sensors will change over time, as some
platforms are retired and others are launched, the SAB recommends that EPA cross-calibrate data from
existing sensors with future sensors as they become available.
South Florida Inland Flowing Waters
Rationale for Criteria (Charge Question 4). For purposes of nutrient criteria development, the SAB
recommends that EPA subdivide South Florida inland flowing waters into human-managed canals vs.
natural flowing streams. Canals would be viewed as a source of nutrients to downstream, more
oligotrophic, systems and canal nutrient criteria would be in the form of DPV. Depending on data
availability, criteria could be developed for natural flowing streams as instream protection values (IPV).
Delineation and Data Sources (Charge Question 4a). EPA proposes to derive numeric criteria for
South Florida inland flowing waters, including canals, as instream protection values (IPV) for TN and
TP using the reference condition approach. Canals apparently account for approximately 90 percent of
waters in this category. This extensive network of canals provides ecosystem services but these services
are controlled primarily by hydrology and habitat quality rather than nutrient levels. The SAB is not
convinced from the material provided that IPV nutrient criteria are appropriate for these uniquely
artificial and highly managed ecosystems. The underlying problem is that the canals are classified as
Class III waters (with a designated use of recreation and balanced population of fish and wildlife),
although their primary purpose is flood control and management of water supply. The proposed
inventory of inland flowing waters that catalogues and distinguishes natural streams and canals should
4
-------
provide very useful information, and could provide the basis for subdividing this category of waters as
recommended above.
EPA is considering use of the Landscape Development Index (LDI) as an approach to identify areas
with reference conditions. The LDI is a surrogate for stressors, including TN and TP loads, associated
with various land uses. The 100-m stream/canal buffers proposed for use with the LDI may be too
limited, and may not capture impacts from land uses in the full watershed. The proposed classification
appears reasonable as it incorporates surface and subsurface flow regimes, soil types and land use;
however legacy N and P loads from previous land management practices also must be considered. The
SAB recommends that the LDI not be used for canals.
Assessment Endpoints (Charge Question 4b). If canal IPVs are to be developed, more information
needs to be presented on how balanced natural populations would be assessed for these unique aquatic
ecosystems. The SAB recommends further consideration and assessment of the response variables to be
used and the form of the nutrients (i.e., those with short-term versus long-term bioavailability) that are
most relevant. It was not clear that sufficient data are available to support either invertebrate or
phytoplankton endpoints for canals.
Approaches (Charge Question 4c). EPA proposes two approaches for determining numeric criteria for
South Florida inland flowing waters. The first is based on reference conditions and the second is based
on stressor-response relationships. As noted above, based on the available data, the SAB recommends
that nutrient criteria for canals focus on downstream protection, rather than IPV. If canal IPVs are
developed, either of these approaches could work, although it was not clear if the available data would
show interpretable patterns. Finding patterns in natural streams may be more likely than in the highly
managed canals. For the reference condition approach, data on historical annual values of TN and TP
from a set of least-disturbed sites (identified using LDI < 2) would be used to develop lognormal
distributions of TN and TP under least disturbance. Variability of nutrient levels at the least-disturbed
sites will reflect heterogeneity in hydrology, geology, etc. Failure to account for such heterogeneity,
which is also present in disturbed sites, may result in numeric criteria that are under- or over-protective.
The SAB also notes that selecting "least disturbed sites" using an LDI < 2 may not be feasible in this
region that has been subject to active management for many years.
The stressor-response approach also should work if a suitable relationship between Chi-a and nutrient
load can be demonstrated, but several of the same caveats apply here as for setting limits in coastal
waters. As with the approach based on reference conditions, the relationship between Chl-a and TN or
TP is likely to be modulated by the effects of hydrological, geological, and other covariates. Failure to
account for such factors may lead to criteria that are over- or under-protective. The document needs to
address how regression models will be used to determine numeric criteria, specifically, how they will
determine the level of Chi-a considered to be protective of balanced phytoplankton and faunal
communities.
South Florida Estuarine and Coastal Waters
Delineation and Data (Charge Question 5a). South Florida estuarine and coastal waters have a
different nutrient regime than other parts of the state, given their oligotrophic nature and susceptibility to
upstream water management (versus nutrient regulatory) decisions. The SAB agrees that these waters
should be considered separately for purposes of nutrient criteria development. The data identified for
South Florida estuarine and coastal waters seem appropriate. However, the proposed
subdivision/subclassification of these waters does not clearly relate to the oceanographic circulation and
5
-------
degree of connectivity in the region. EPA might consider an alternative segmentation scheme that
incorporates alongshore and cross-shelf circulation patterns.
Approaches (Charge Question 5b). EPA proposes to use a reference condition approach using least-
disturbed sites or a binomial test applied to a distribution of raw data. Both approaches have merit and
the SAB encourages application of both to provide a more robust evaluation of criteria. However, if
least-disturbed sites are those most distant from land-based sources, nutrient levels at these sites may
reflect dilution by oligotrophic ocean water rather than a significant difference in nutrient loading. In
addition, the coastal and estuarine waters of South Florida have experienced enormous changes over the
last century in freshwater inflows, salinity and residence times, with associated changes in nutrient
cycling and seagrass extent. These past alterations should be considered when defining reference
conditions. The SAB recommends that seagrass coverage and extent of epiphytic colonization be
considered as endpoints in addition to water column chlorophyll. The document should clarify which
coastal and estuarine areas will be under the jurisdiction of the forthcoming nutrient criteria, versus other
regulations (e.g., for federal and state protected waters).
Downstream Protection Values (Charge Question 6)
EPA proposes to use DPV criteria to ensure that upstream N and P water quality criteria will be set at
levels that will protect downstream estuarine designated uses. While agreeing with the goal of
downstream protection from nutrient impacts, the SAB has concern with the overlap between DPV and
the Total Maximum Daily Load (TMDL) process. For development of an estuarine TMDL, standard
practice is to model the estuary and watershed to determine the additional pollutant load reduction
needed, and then to allocate the load reduction based on input from state and local officials. This process
thus allows consideration of socio-economic factors, natural background levels of nutrients in different
tributaries, etc. in determining the load allocation. However, a significant drawback of relying on
TMDLs rather than DPVs is the fact that the TMDL process is not triggered until impairment has
occurred in a waterbody.
The EPA document includes an example of a DPV developed using an equal load reduction for each
tributary to an estuary. The SAB notes that this equal allocation approach ignores the possibility of
different background nutrient concentrations in upstream segments, and other characteristics of
segments that would make nutrient load reductions impractical. In response to a SAB request for further
information, the Agency described four options for distributing loads for calculation of DPV, each of
which presents some difficulties. If the DPV approach is pursued, the SAB urges that selected
apportionment strategies not preclude flexible nutrient allocation across tributaries to achieve the
necessary load reductions.
Whether developing a DPV or a TMDL, the modeling of load reduction apportionment for upstream
segments is a valid approach, but watershed characteristics such as predominant land-use (especially
urbanized area) should be considered. EPA should justify the choice of the LSPC watershed model to
estimate average flow for streams that discharge to an estuary, and explain why it is the most applicable
model for this case. The timeframe of the modeling should be linked to the response of biological
endpoints in the receiving waters; annual average values may grossly under-predict the impact of large
storm events. In addition, the document needs to discuss the impact on criteria of P cycling and
transformation.
6
-------
2. INTRODUCTION
2.1. Background
Nitrogen (N) and phosphorus (P) inputs from urban and agricultural sources are known to influence
water quality, and nutrient pollution has been identified as a source of impairment for estuarine, marine
and fresh waters in Florida. The state of Florida has a narrative criterion for nutrients, and is in the
process of developing numeric nutrient criteria for its estuaries and coastal waters. In 2009, EPA
determined that numeric nutrient criteria were needed to protect aquatic life in Florida, and initiated a
process to develop such criteria for categories of state waters. Criteria for total nitrogen (TN), total
phosphorus (TP) and chlorophyll a (Chi-a)—a measure of water column algal biomass—were finalized
for Florida lakes and inland flowing waters in 2010. Numeric nutrient criteria for estuarine and coastal
waters, and South Florida inland flowing waters, are being developed separately, using a variety of
approaches and ecological endpoints. The SAB was asked to provide review and advice on the proposed
approaches for estuarine, coastal and South Florida waters, as described in the draft EPA document,
Methods and Approaches for Deriving Numeric Criteria for Nitrogen/Phosphorus Pollution in Florida's
Estuaries, Coastal Waters, and Southern Inland Flowing Waters (November 17, 2010 draft; U.S. EPA
2010).
An ad hoc panel of the SAB, the Nutrient Criteria Review Panel, was formed for this task. The Panel
met on December 13-14, 2010, to hear EPA technical presentations and public comments, and to discuss
responses to the questions in the Charge to the SAB (Appendix A). A follow-up public teleconference of
the Panel was held on February 7, 2011, to discuss an initial Panel draft report. The chartered SAB
conducted a quality review of the Panel's report on May 17, 2011. Public comments were received and
considered throughout the advisory process.
2.2. Charge to the SAB
The Charge to the SAB included questions about the conceptual model used to select assessment
endpoints, data sources for the various categories of waters, and possible approaches to define criteria
for each of the categories of waters: estuaries, coastal waters, inland flowing waters (including canals) in
South Florida, and South Florida marine waters. Relevant charge questions are included at the beginning
of each section of the Panel's report, and the full Charge to the SAB is included as Appendix A.
7
-------
3. Response to Charge Questions
3.1. Conceptual Approach
3.1.1. Conceptual Model
Charge Question 1(a). EPA has introduced a general conceptual model in Chapter 2, including
the selection of assessment endpoint and indicator variables. What is your perspective of the
general conceptual model?
The State of Florida currently has a narrative criterion for nutrients that says, "in no case shall nutrient
concentrations of a body of water be altered so as to cause an imbalance in natural populations of flora
or fauna," with specific reference to degradation from "man-induced nutrient enrichment (total nitrogen
or total phosphorus)" (Florida Administrative Code 62-302.530). EPA has put forward a conceptual
model, describing the relationships between nutrient levels (total nitrogen [TN] and total phosphorus
[TP]) and biological responses, to guide its efforts to translate Florida's current narrative nutrient
criterion into numeric nutrient criteria. Although the general conceptual model (Figure 1, below)
provides a starting point for choosing numeric criteria, the SAB has numerous concerns about how the
causal variables will be linked to biological endpoints.
The conceptual model would translate Florida's objective of "balanced natural populations of aquatic
flora and fauna" into numeric criteria for three biological endpoints: seagrasses, phytoplankton, and
faunal communities. While agreeing that these endpoints are appropriate, the SAB strongly felt that
these endpoints need to be much better defined and, in some cases, connected to the explanatory
variables that would be the basis for setting numeric criteria. As detailed below, the SAB did not always
see a direct link between causal variables and the endpoints.
The term "balanced" is not defined in the document and is subject to a great range of interpretation. EPA
needs to provide a definition of "balanced" early in the document. Simply relying on the state's
assessment of whether or not waters are "healthy" is not an adequate definition. EPA also must define
how it will measure and interpret these three endpoints, preferably in quantitative terms. The Florida
DEP has defined a balanced community as one that may have experienced "modest" change, but which
still has: (1) reproducing populations of sensitive taxa; (2) a balanced distribution of all expected major
groups; and (3) an ecosystem with functions that are largely intact due to redundant system attributes
(FDEP 2010). If EPA chooses to use this definition, then the link between the water quality targets and
these goals should be documented.
More information is needed on the methodologies that will be used in order to determine if the general
conceptual approach is workable within the time constraints1. The SAB recognizes that details on
methods are to some extent specific to type of water body and appropriate for later chapters, but further
information on the methods is needed in this chapter as well. EPA is proposing three general approaches
to relate nutrient levels to balanced natural populations:
1 Under a consent decree, EPA is required to issue proposed criteria by November 14, 2011, and final criteria by August 14,
2012 (U.S. District Court, Northern District of Florida, Tallahassee Division, Case No. 4:08-cv-00324-RH-WCS).
8
-------
1. Identify reference conditions for a water body type based on available data or best
professional judgment;
2. Use predictive stressor-response relationships and nutrient/algal thresholds; and/or
3. Use numerical water quality models to predict nutrient loadings that would be protective of
system biology.
The SAB agrees that these approaches all have utility and recommends that a combination of approaches
be used where data and models are available. However, more detail is needed on: (1) the adequacy of the
data for applying each approach; (2) how decisions would be made on which approaches to use; and (3)
how discrepencies in targets derived from different approaches would be resolved. As noted below, each
of the three approaches has strengths and weaknesses.
The use of nutrient reference conditions implies that nutrient concentrations and loadings to a system
are known with enough certainty that target values protective of biological endpoints can be determined.
In cases where data specific to a system are not sufficient, best professional judgment could be used to
determine suitable target values. There has been extensive hydrologic modification of Florida's waters
and extreme weather events, a number of which have occurred in the last 15 years, which complicates
defining reference conditions in the context of spatial and temporal variability. Current reference
conditions may not represent historical conditions. EPA may need to state explicitly the general
hydrologic ranges over which these targets will be useful and have clearly stated goals in cases where
remediation is suggested. When using the reference condition approach, EPA also needs to evaluate
whether data sets used to set values are representative of long-term averages or (in the case of relatively
short-term data sets) are heavily influenced by recent hurricanes. Comments by Briceno et al. (2010)
may be useful in this regard.
The use of predictive stressor-response relationships and thresholds assumes that data on nutrient-
organism interactions from Florida waters and other regions, or countries, could be appropriately applied
to setting protective target values. The use of numerical water quality models assumes that models
would be a useful and realistic representation of nutrients and other water quality parameters. For
practical application of numerical models, there still remain questions as to the appropriateness of
selected models, availability of data, and level of detail required to adequately populate each model
approach. For example, the EPA document states that a watershed model will be run with all
anthropogenic sources removed to determine background TN and TP levels. More information and
justification is needed to provide assurances that the models being used can adequately accomplish this
with the stated degree of certainty. Most water quality models have been developed to assess and predict
fate and transport processes as a result of anthropogenic activities and not for determining pristine
conditions. Detailed validation of such uses is needed, which means calibration with non-impacted
watershed loads. However, there are few non-impacted watersheds with conditions that reflect baseline
concentrations in relation to determining water impairment. A key factor involved in using numerical
models would be their validation and some analysis of uncertainty for each of the systems where they
are applied. It may not be feasible to apply these models to a large number of estuaries in a short period
of time.
The three methodological approaches listed above are being applied somewhat differently within the
different categories of Florida waters. Since each approach has different data requirements (and more
importantly, different assumptions, limitations and uncertainties), it is critical that the reasoning behind
applying or not applying a given method be provided. The EPA document notes that EPA may use one,
two or all three of these approaches for a particular water body. There would be a greater confidence in
the criteria if multiple approaches were applied to each of the systems for which data and models are
9
-------
available. This would provide an ensemble approach and a range of values for setting numeric criteria.
However, this could result in more than one answer as to what numeric values would be protective. This
is understandable given the different conceptual bases for each approach, but the EPA document should
discuss how the results from multiple approaches would be integrated to develop the final numeric
criteria.
Specific suggestions on the conceptual model for different ecosystem types follow. Further discussion of
EPA's proposed approaches can be found in the responses to the charge questions for specific categories
of waters.
Conceptual Diagram
The conceptual diagram (Figure 1) gives a general overview of how the conceptual model will be linked
to three biological endpoints and how the approaches would be applied in each of the categories of
Florida waters. The upper three levels of the diagram (Causal Variable, Response Variable, and Water
Quality Targets) are dealt with at great length, but the bottom two levels (Biological Endpoints,
Objective) are not discussed in sufficient detail. The SAB recommends that the EPA provide a more
detailed conceptual model that includes additional endpoints and flows. In addition, it would be helpful
to include system-specific diagrams for each of the four waterbody types, perhaps at the beginning of
each chapter.
Terminology that implies TN and TP are causal variables should be dropped in favor of terms such as
driver variables because N and P are two factors that regulate primary production but they do not
"cause" it. While there is a cause/effect relationship between nutrients and Chi-a, the conceptual model
should acknowledge that there are many other factors that control Chi-a.
In Figure 1, Chi-a also is shown to be a water quality target that relates to balanced faunal communities.
There are mechanisms by which these two are linked, in addition to changes in DO (e.g., through
increased sediment loading), but the SAB did not find any discussion of mechanistic links in the EPA
document. Also, while low DO is closely linked with eutrophication, it is not the only mechanism of
nutrient impacts, which is what is implied in the diagram. The SAB recommends that the EPA alter the
diagram or include an explanation in the text on how numeric criteria for Chi-a will be linked to
balanced faunal communities. EPA also should provide more background and theory on the
relationships between biological endpoints and water quality targets. There are many factors that
regulate "balanced" ecosystem functions in addition to the few listed in Figure 1, including predation
and other food web interactions, harvest, salinity, substrate, species turnover, and N:P ratios.
10
-------
Causal
Variable
Response
Variable
Water
Quality
Targets
Biological
Endpoints
Water Clarity for
Maintenance of SAV
Habitat
I
Percent Surface
Light Goal
V
Protection and
Restoration of Healthy
Seagrass Communities
Proposed TN/TP Criteria
Chlorophyll o Criteria
Chlorophyll o
Concentration Prior to
Shift in Species
Dominance
Chlorophyll a at
Recommended
Trophic
Boundary
V
Balanced
Phytoplankton Biomass
and Production
Dissolved
Oxygen
Balanced Faunal
Communities
Objective
Balanced Natural Populations of Aquatic Flora and Fauna
Figure 1. EPA's Proposed Conceptual Diagram Relating TN/TP Criteria to Florida's Narrative Nutrient Criterion
(Source: U.S. EPA 2010, Fig. 2-1)
An additional concern is that the conceptual model does not include any discussion about how seasonal
and inter-annual variability will be addressed. Average annual values for parameters may not be
protective if there are wide seasonal changes. Seasonal variability (e.g., warm vs. cold, wet vs. dry)
should be evaluated to determine whether multiple criteria are required.
The goal of the criteria development effort is protection of biological populations, and so the SAB
begins its comments with the biological endpoints to be protected and works through the conceptual
diagram from the bottom up.
Protection and Restoration of Healthy Seagrass Communities
Chapter 3 of the EPA document describes in more detail how a healthy seagrass population might be
determined using historical data and colonization depth. This is a specific and quantifiable parameter
(see section 3.2.2). A brief explanation of this is needed in Chapter 2 to outline the approach. The SAB
did not find specific decision criteria in the document for determining when management objectives
have been met for impaired water bodies or what magnitude of changes would be considered a
significant change (i.e., what percent of historical seagrass coverage would be set as a target for
restoration?). Data tables provided in Appendix B of the EPA document (Estuarine and Marine
Assessment Endpoints and Water Quality Indicator Variables Literature Review) may be used as a
11
-------
starting point. A thorough presentation of key parameters and breakpoints needs to be included if
numeric criteria are to be applied.
The SAB is concerned about relying upon water column Chl-a as the sole criterion to protect seagrasses.
No numeric criteria directly related to macroalgae or epiphytes are being proposed. In systems where the
nutrients are largely taken up by the phytoplankton, Chi-a will reflect the major impact of nutrient
loading. However, there are systems where macroalgae proliferate in response to increased nutrient
loadings, but water column Chi-a remains low due to short water residence times. In these systems,
water column Chi-a is a poor measure of nutrient effects. Hauxwell et al. (2001, 2003) found that light
levels in benthic macroalgal mats prevented young eelgrass shoots from being established. Epiphytes
also can increase in response to nutrient loading in systems even though water column Chi-a levels
remain fairly low.
EPA could consider an approach linking nutrient loading with changes in seagrass areal extent for
protecting seagrass communities. This approach has been successful in Tampa Bay (Greening 2010). It
was also applied to a range of systems in New England (Latimer and Rego 2010), with data on eelgrass
loss for a number of estuaries being compared to calculated nutrient loadings. The Latimer and Rego
study found eelgrass loss began to occur at N loads above 50 Kg ha"1 y"1 and eelgrass disappeared at 100
Kg ha"1 y"1. It may be possible to develop a similar relationship for Florida seagrass systems. The SAB
recommends that EPA consider this approach and also ways to back-calculate concentrations from
specific loadings that impair seagrasses.
Phytoplankton Production and Biomass
The SAB agrees that Chi-a concentration is both sensitive to nutrient inputs and an important measure of
ecosystem health and therefore a reasonable endpoint in itself. However, the EPA document should be
clear that Chi-a measures biomass and cannot be used to infer whether or not populations are "balanced"
in terms of species composition or relative abundance/dominance. In testimony, EPA provided examples
where toxic blooms are known to occur at high Chl-a values. Although undesirable species may be more
prevalent in areas with higher nutrient loading (and higher algal biomass), low biomass does not ensure
that a toxic species will not occur or that species composition has not changed. Similarly, while Chi-a is
a measure of biomass (standing stock), it is not a measure of production (a rate) and cannot be used to
assess the biological endpoint of production.
Balanced Faunal Communities
The conceptual model does not include a direct metric for balanced faunal communities, but instead
proposes that healthy faunal communities rely upon sufficient concentrations of DO. While this is true
and the document cites studies where low DO (hypoxia and anoxia) causes mortality and impairment of
marine life, low DO usually is a secondary response to excess nutrients in a system. The primary
response of excess nutrients is eutrophication, the excess production and accumulation of organic matter
in a system. For Florida, in particular, there are also many swamp and backwater systems that feed into
tidal waters that are naturally low in DO and not related to TN or TP concentrations or loadings.
EPA proposes to use the State of Florida's DO standard to maintain the biological endpoint of balanced
faunal communities. Specifically, EPA proposes to look for relationships between TN and/or TP and
DO, and use those relationships to determine numeric criteria for TN and TP that are protective (i.e., that
are associated with attainment of the existing DO standard). How linkages between these variables and
faunal metrics will be assessed needs to be clarified. Chapter 3 (p. 49) of the EPA document implies that
the absence of hypoxia will be the indicator for the presence of balanced communities, so that numeric
12
-------
criteria would be based on ambient nutrient levels where hypoxia is absent. The SAB recommends that
faunal metrics be considered to evaluate the endpoint of balanced faunal communities (see also response
to charge question 2).
Chapter 3 also notes that DO can be computed in water quality models from TN and TP loading.
However, in determining if hypoxia is present in any particular system, seasonality needs to be
considered as hypoxia is a seasonal phenomenon and not present all year round
Dissolved Oxygen Targets
Additional concerns arose during the discussions and submitted testimony about using a single DO
standard. Some seagrass meadows routinely exhibit low oxygen conditions at night even in the absence
of any nutrient impairment. This diel cycling of oxygen—from supersaturated during daylight hours to
undersaturated, and at times hypoxic, at nighttime—has recently been found to be common in shallow
vegetated and unvegetated habitats (Moore et al. 2004; Verity et al. 2006; Tyler et al. 2009). Similar
conditions appear to occur in some Florida waters (see Boyer and Briceno 2010). Another issue is that
oxygen is less soluble under higher temperatures and higher salinity, conditions seen in many of
Florida's warm temperate and subtropical waters. Hence, low DO criteria may be better characterized by
percent saturation. Although the Florida DO numeric standard is not a subject of the current review, the
SAB raises these issues to emphasize the challenges in relying on a DO standard to protect healthy
biological communities. There are factors other than nutrients involved in development of hypoxia and
anoxia such as water column stratification and water residence time.
TN and TP Criteria
In the EPA document, TN and TP are listed as "causal variables" and defined (p. 39) as concentrations
(mg/L) of total (organic and inorganic) N and P. This may lead to confusion. As the table on page 36 of
the document points out, TP and TN loadings are normally considered to be the ultimate driver of
ecosystem changes while TP and TN water column concentrations are "associated with influent loading
over the long term". Hence this would make water column concentrations of both TP and TN
explanatory or response variables. The narrative (p. 53) also refers to loading as the causal variable and
water column concentrations as a response variable. This distinction is important when considering
using TP or TN to predict other parameters.
Many assessments have been based upon loading. Loading data, when available, would be expected to
be a better predictor of Chl-a, hypoxia and seagrass loss than concentration. Furthermore, TN and TP
concentrations may co-vary with Chl-a, since both are contained in phytoplankton, so they also are not
completely independent from Chl-a. This presumes TN and TP are measured on unfiltered samples; yet
the document does not clearly state whether that is what is proposed. Given the availability of data, there
may be excellent reasons to use TN and TP concentrations as numeric criteria but they should be
considered response variables. It also would be useful to characterize these variables in more detail,
including the temporal and spatial scales over which they would need to be measured (i.e., weekly or
monthly averages, surface values, depth-integrated samples, discrete depths).
The issue remains of whether TN and TP or "reactive N and P" (i.e., dissolved inorganic N [DIN] and P
[DIP]) are the most relevant variables to link nutrient enrichment to specific effects on biological
endpoints (i.e., primary production, biomass as Chl-a, and cascading effects such as food web alterations
and hypoxia). This issue has been the subject of considerable research, discussion and controversy for
decades. Much of the uncertainty regarding whether to use TN and TP or more "reactive" dissolved
forms of these nutrients revolves around the bioreactivity and roles of organic forms of these nutrients.
13
-------
Bioreactivity may be system-specific (or even system-component-specific), adding to the complexity
and uncertainty of measuring responses and impacts on water quality and habitat condition. It is
important for EPA to discuss this issue in the context of developing numeric nutrient criteria for
nutrient-sensitive waters, both in Florida and nationally.
Regardless of form, some consideration is needed for what portion of TN and TP loading in a system is
from natural sources versus anthropogenic sources. This is particularly important for open coastal waters
where conditions may be influenced by non-anthropogenic nutrient sources from outside the geographic
boundaries of the coastal zone. While we know that reducing nutrients is key to restoring ecosystems in
general, careful consideration needs to be given to all sources of N and P that in combination affect the
biological endpoints for a system.
Uncertainty
Throughout the document, uncertainty is mentioned only briefly, and that in the context of uncertainty
introduced because some environmental variables can covary with the explanatory variable of interest. It
would be easier to convey the concept of statistical uncertainty by presenting confidence intervals.
However, the level of uncertainty or size of confidence intervals associated with particular numeric
criteria should be described further. Predictions developed using stressor-response models should
explicitly state the level of confidence and be validated with site-specific data. The possibility that
multiple covarying factors are involved in the cause-effect relationship for a particular system should be
assessed. For example, the morphology of a system, habitat, and biological interactions within the water
body may modify the relationship between nutrient concentrations (both N and P) and observed
biological endpoints. These covarying factors need to be better documented, so that the uncertainty of
the relationship can be reduced. In other words, a relationship between nutrient levels and a biological
metric could be derived that is statistically significant, yet which accounts for only a small portion of the
total variability in the data. Inclusion of covarying factors would strengthen statistical relationships,
reduce uncertainty, and provide greater confidence in the final criterion.
3.1.2. Categories of Florida Waters
Charge Question 1(b). EPA has delineated the State of Florida into 4 general categories of
waters—Florida estuaries, Florida coastal waters, South Florida inlandflowing waters, and
South Florida marine waters—for purposes of considering approaches to numeric nutrient
criteria development. Are these categories appropriate and scientifically defensible?
Separation of estuarine and coastal waters is appropriate given the differences in natural populations of
aquatic flora and fauna between higher salinity coastal systems and lower salinity estuarine and inland
systems. Freshwater management in the region is complex and the separate consideration of South
Florida also is warranted, although the SAB recommends that the term "marine waters" be replaced
with "estuarine and coastal waters" for clarity and consistency. A finer classification based on degree of
impact may be useful; for example, to separate the Caloosahatchee and St. Lucie estuaries from the other
Florida estuaries, given their unique (i.e., strong human influence) hydrological relationship to Lake
Okeechobee. While nutrients clearly influence the biota in these systems, salinity levels play a stronger
role than is typically the case in other Florida estuarine systems (Kraemer et al. 1999; Doering et al.
1999; Steinman et al. 2002a).
The category of South Florida inland flowing waters seems to be a catch-all for waters that are primarily
man-made canals, but which also include a small percent of natural streams. These canals present a
unique challenge for setting numeric nutrient standards because of their morphology and management
14
-------
regime. It would be preferable to have a strong scientific rationale for this classification, and further
information on canal management could be helpful in this regard.
3.2. Florida Estuaries
3.2.1. Delineation and Data Sources
Charge Question 2(a). Are the data sources identified appropriate for use in deriving numeric
criteria in Florida's estuaries (as discussed in Sections 2.4 and 3.2)? Is the SAB aware of
additional available, reliable data that EPA should consider in delineating estuaries or deriving
criteria for estuarine waters? Please identify the additional data sources.
In general the geographic delineations of estuaries seem appropriate. The SAB was unclear why a
salinity of 2.7 psu was used to delineate the upper reaches of systems. Traditionally, such salinity would
denote an oligohaline region of an estuary. Why not use "head of tide" or salinity of less than 0.5 psu? In
any case, sediment nutrient dynamics change in this salinity transition zone (from approximately 0.0 to
5.0 psu). For example, at the toe of the salinity front, P-releases from sediments can increase sharply.
Wherever the upper boundary is fixed, such issues need to be considered.
EPA should consider adding another unit to the estuary delineation that would focus on tidal creeks. A
case was made that these systems are common but have different characteristics than the open estuaries
and therefore should have different nutrient criteria.
The SAB has few, if any, issues with the data sets presented. The summary tables in the EPA document
indicate a careful review of data sources, including attention to time-series data. We encourage
continued searching for appropriate data. In public comments to the Panel, one researcher (Dr. Tom
Frazer, University of Florida) indicated that additional data are available for some estuarine areas and
have yet to be utilized. It may be that County agencies have data sets not yet considered. This effort
could be especially useful in the Big Bend area, where offshore seagrass beds are extensive, satellite
data on Chl-a are not useful, and existing data sets from prior studies are rare. All data sets would need
to meet EPA requirements for QA/QC, but the SAB encourages EPA to continue consultations with
state and local agencies and researchers to access additional data and local knowledge where possible.
3.2.2. Assessment Endpoints
Charge Question 2(b). Are the assessment endpoints identified in Sections 2.3 and 3.2 (healthy
seagrass communities; balanced phytoplankton biomass and production; and balancedfaunal
communities) appropriate to translate Florida's narrative nutrient criterion into numeric criteria
for Florida's estuaries, given currently available data? Does the SAB suggest modification or
addition to these assessment endpoints?
Healthy Seagrass Communities
Florida seagrass beds are an extremely valuable natural resource, and the two largest contiguous
seagrass beds in the continental United States are found in the Florida Keys and Florida's Big Bend
region. Approximately 2.2 million acres of seagrass have been mapped in estuarine and near-shore
Florida waters by researchers at the Florida Fish and Wildlife Research Institute in St. Petersburg
(Carlson and Madley 2007). However, when seagrass beds growing in water too deep to easily map are
included, the total area of seagrasses within Florida waters and adjacent federal waters is likely over 3
million acres. Florida's seagrass beds improve water quality and reduce shoreline erosion, but their most
15
-------
important ecological role is to provide food and shelter for many economically important finfish and
shellfish species (Carlson and Madley 2007). Estimates of the ecological services provided by Florida
seagrass beds are approximately $20,500 per acre per year (Handley et al. 2007), and it is entirely
appropriate that EPA use healthy seagrass communities as one of its assessment endpoints.
There are, however, several issues relating to seagrasses that deserve further consideration. First, as
acknowledged in the draft EPA document, Chi-a usually explains a significant amount of variation in
water clarity, but frequently does not explain the majority of this variation. Water clarity is often greatly
influenced by colored dissolved organic matter (CDOM) and inorganic suspended material in the water
column. Of greater importance, seagrasses in the shallow waters of Florida are typically shaded more by
epiphytes growing on their leaves and by associated macroalgae (see Dixon 1999, and review by
Burkholder et al. 2007) than by Chi-a in the water column. Thus, EPA should consider a measure of
epiphyte abundance in addition to the proposed determination of Chi-a in the water column. Epiphyte
abundance is also often controlled to a significant extent by the animals that feed on these epiphytes
(Hughes et al. 2004; Burkepile and Hay 2006; Heck and Valentine 2007; Baden et al. 2010). This
control by consumers, often referred to as top-down control, is beyond the scope of the draft document,
but it is notable that both nutrients and herbivorous consumers affect the condition of seagrass meadows.
As pointed out in Section 3.4.3, similar consumer effects can also be important in controlling Chi-a in
the water column (Cloern 1982; Cohen et al. 1984; Thompson et al. 2008).
It might also be possible to develop direct seagrass targets. In comments to the Panel, Dr. Paul Carlson
suggested submerged aquatic vegetation (SAV), which includes seagrasses, as a management endpoint
in support of the EPA plan but went further, arguing that SAV occurrence (abundance, density, etc)
might be thought of as an integrative numeric criterion. In support of this suggestion, he states the
following: (a) measures of SAV occurrence over time and space can integrate water quality issues of
hypoxia, water clarity and nutrient conditions; (b) Florida has many areas dominated by SAV and thus
these plant communities are representative of many estuaries; and (c) some beds are surveyed every two
years, so there is a very good temporal record of their condition. The SAB suggests that serious
consideration be given to Dr. Carlson's suggestion.
Balanced Phytoplankton Biomass and Production
As noted earlier (3.1.1), EPA needs to provide a clear definition of "balanced". If EPA is not referring to
species composition and relative abundance, but rather the entire phytoplankton or benthic microalgal
communities, then Chl-a or other indicators of biomass (i.e., dry weight, particulate C, total cell counts)
will suffice. If EPA is referring to species diversity or some other index of biological diversity, then
more specific techniques will have to be employed, including microscopic species identification,
photopigment analyses, molecular analyses, etc. The SAB recommends community-level biomass
metrics, using Chi-a or other indicators of biomass, as this is best related to nutrient and C-flux,
hypoxia, and other drivers/indicators of impacts of and responses to nutrient inputs. Endpoints that
require taxonomic-level resolution (e.g., to characterize harmful algal blooms, HAB) will need more
specific suites of indicators to identify, quantify and characterize factors such as toxicity and food web
effects. Taxonomic identification of HABs can provide diagnostic value, but the SAB acknowledges that
such taxonomic analysis may not be possible with current monitoring programs in the systems and
regions of interest.
Balanced Faunal Communities
There is little discussion in Chapter 3 of how "balanced faunal communities" are defined, and this is a
concern for several reasons. First, given the generally shallow nature of Florida estuaries and our general
16
-------
impression that water clarity is (or was) high, it is likely that these systems are (or were) benthic-
dominated. If this is the case, a variety of benthic heterotrophs could provide strong metrics for estuarine
health, as a complement to seagrass community measures. Second, a strong shift from one common
benthic species to another (e.g., a pollution tolerant species) can provide a good indicator of benthic
habitat condition or deterioration, although the "species pair" might differ among estuaries. Has this
approach, using indicator species, been considered as a measure of the health of faunal communities for
Florida estuaries?
Chapter 3 indicates that hypoxia will be used as an indicator of compromised benthic habitat condition.
As a first pass, this will certainly tell us something about these habitats but not all that is needed.
Unfortunately, when hypoxic conditions are observed, impacts on the biota usually have already
occurred. It would be useful to have indicators of stress on the faunal community before such degraded
conditions develop. In addition, DO in Florida seagrass meadows during the early morning hours is
often below the levels considered to be hypoxic in unimpaired Florida waters, owing to the low amounts
of oxygen that can be dissolved in the high temperature and high salinity waters characteristic of Florida
and the high rates of nighttime respiration in the biomass-rich seagrass meadows. Thus, as suggested in
comments provided by Boyer and Briceno (2010), a percent saturation criterion may be more useful than
an absolute measure of oxygen concentration.
3.2.3. Approaches
Charge Question 2(c). EPA describes potential approaches in Section 3.3 (reference conditions,
stressor response relationships, and water quality simulation models) for deriving numeric
criteria in Florida's estuaries. Compare and contrast the ability of each approach to ensure the
attainment and maintenance of natural populations of aquatic flora andfauna for different types
of estuaries, given currently available data?
The EPA proposes to use three approaches to develop nutrient criteria for Florida estuaries: reference
conditions, stressor-response models, and water quality simulation models. The SAB comments on the
application of each of these approaches to estuaries in Florida, and then discusses the need to consider
the impacts on the criteria of future changes to hydrology and climate.
As noted previously, the SAB recommends that EPA provide a more quantitative description of the
concept of balanced phytoplankton and faunal communities, and remove the word "production" in the
description of phytoplankton unless measures of production are added. Nutrient criteria development
should take into account the natural diversity of Florida estuarine systems. For example, in some
systems having low N/P nutrient ratios, blue-green algae may be the normal dominant species.
Recognition of special system features will prevent systems from failing to meet criteria on the basis of
natural background conditions. In addition, the estuarine continuum, from freshwater to the sea, often
involves a transition from P- to N-limitation and possibly zones where co-limitation occurs. Thus, a dual
nutrient strategy is warranted and we agree with EPA's decision to take this approach. Similar strategies
have been adopted in the Chesapeake Bay, Neuse River Estuary and Baltic Sea (e.g., see Elmgren and
Larsson 2001; Conley et al. 2009; Paerl 2009).
The SAB has a general concern about the timetable for completion of this work. A substantial effort
already has been made to get the work to this stage, much of it solid and thoughtful. However, much
work remains to be done. In the case of Florida estuaries, it is not clear what approaches will be selected,
if multiple approaches will be used, and which approaches will provide useful information towards the
goal of developing nutrient criteria.
17
-------
The SAB emphasizes that there is no single approach that is ideal for developing nutrient criteria. This
being the case, we support using multiple approaches where possible. If results for two or more
approaches converge, then there is increased confidence in the results, and EPA needs to provide
guidance on how to use this information to develop a criterion. If different approaches yield conflicting
results, then EPA needs to have a defensible methodology to resolve the differences and move forward.
Reference Condition
Philosophically, the reference condition approach is the most satisfying, although making it operational
is often difficult because sufficient data are lacking to define "reference conditions", and because of the
problem of "shifting baselines" (Pauly 1995)—in other words, many ecosystems have been impacted by
human activities for some time and we run the risk of using degraded coastal environments as reference
conditions when the true (unimpacted) reference conditions have long since ceased to exist. The SAB is
aware of at least one other State (New Hampshire) using the reference approach for developing nutrient
criteria and that effort yielded some useful results. Experience suggests that this approach, if it can be
implemented, might be the most "time-efficient" pathway to developing nutrient criteria.
Stressor-Response Models
The SAB was disappointed that more attention was not given to the stressor-response approach.
Although a previous SAB SAB was critical of this approach as a stand-alone method for developing
nutrient criteria (U.S. EPA SAB 2010), the available data for Florida suggest a stronger relationship
between stressor (nutrients) and response (chlorophyll) than for the stressor-response relationships used
in the EPA document reviewed by that panel. One of the Nation's best estuarine restoration examples is
Tampa Bay, where a stressor-response approach was used to develop local criteria. The Tampa Bay
success suggests that this approach should be utilized more in the State-wide effort. The SAB is aware
of a separate EPA report on the stressor-response approach (U.S. EPA 2010), but would have preferred
to see more discussion in the current document of how this approach might be applied to development of
nutrient criteria in Florida. Limnologists have had great success with this approach. Recently, EPA staff
in New England published results of an analysis relating nutrient loads to seagrass health in a variety of
small coastal systems (Latimer and Rego 2010). These sorts of studies suggest that the stressor-response
approach needs to be more seriously considered. The EPA document simply refers to "regression
models," leaving many readers with the impression that EPA is considering only the simplest forms of
regression analysis. (In contrast, the discussion of simulation modeling packages is presented in
considerable detail.) Statistical models are correlative and the amount of variance explained by the
correlations can be less than that needed for criteria development. However, the SAB encourages EPA
to apply the stressor-response approach, in combination with other approaches, when developing criteria
for Florida waters.
Water Quality Simulation Models
The level of detail on simulation modeling in the EPA document suggests that EPA has decided to use
modeling as the primary tool for development of nutrient criteria. This may not be the case but the SAB
urges some caution here. The description of the model(s) sounds great, which can be quite seductive,
and some issues can only be addressed with simulation models (e.g., forecasting, understanding highly
interactive processes). At the December 2010 meeting of the Panel, the EPA staff noted that significant
staff resources are being devoted to develop simulation models for specific Florida waters. Nonetheless,
using simulation models to establish nutrient criteria is a major undertaking, requiring considerable time
and money, and useful results are not guaranteed. The SAB notes as a caution that the Chesapeake Bay
model has been under development for about 25 years, and although it has achieved a great deal, it still
18
-------
does not predict inter-annual hypoxic volumes well. There also are very considerable issues related to
data needed for calibration and verification of model results.
For waters where the simulation modeling approach is used, a reasonable representation of internal
nutrient cycling needs to be included. In the generally shallow Florida systems, benthic processes will be
especially important. In addition, these processes will interact with temperature and flow changes.
Ultimately, nutrient concentrations reflect the net effect of these biogeochemical processes, as well as
loadings.
Implications of Changing Hydrology and Climate
The SAB recognizes that the Agency is required to establish water quality criteria using the best science
currently available. However, Florida is experiencing increased demand for freshwater and changes in
climate that may directly impact the response of ecosystems to nutrient loadings. In a spirit of adaptive
management, the SAB recommends that future refinements to proposed nutrient criteria for Florida
estuarine, coastal and marine waters explicitly consider the impact of changing hydrology and climate
on the numeric criteria needed to ensure the protection of designated uses.
• Hydrologic Forcing. From a spatial perspective, the location of phytoplankton production and
biomass responses to nutrient inputs is strongly influenced by freshwater inflow and its impacts
on estuarine residence time. Under drought conditions, the biomass peak, or Chl-a maximum
(Cmax) will tend to be in the most upstream portion of estuaries (Valdes-Weaver et al. 2006).
Under moderate freshwater discharge and flow conditions, Cmax will form in mid-estuarine
locations, while under high flow conditions, Cmax will tend to predominate downstream (Valdes-
Weaver et al. 2006; Paerl et al. 2007). Under extreme hydrologic conditions resulting from
tropical cyclones, Cmax may not form at all, but rather the maximum phytoplankton biomass
response will be in the sounds and coastal waters (Paerl et al. 2006a, b). These conditions
represent a special challenge, because it may be difficult to evaluate and assign numeric criteria
for nutrient loads to estuaries, as the response may not occur in estuarine waters.
The earth is experiencing a sustained period of elevated tropical cyclone activity and intensity
(Webster et al. 2005). Florida is particularly impacted, because it experiences more tropical
cyclone strikes than any other state in the U.S. Therefore, this aspect of climate change needs to
be factored into the anticipated/predicted responses to nutrient inputs and the development of
nutrient criteria. Conversely, periods of extreme (and record) droughts require additional
attention and consideration in the context of the development of nutrient criteria, as the location
and amounts of phytoplankton biomass responses to nutrient inputs will be dramatically affected.
19
-------
• Groundwater and Surface Water Withdrawals. Florida is undergoing significant increases in
freshwater withdrawal (for drinking water and agricultural irrigation purposes) from its lakes and
rivers (Steinman et al. 2004). In addition, continued (and increasing) groundwater withdrawal
will reduce freshwater recharge from springs and other natural seepage sources. These
withdrawals will impact freshwater discharge to estuarine and coastal waters, which in turn will
affect the location and magnitude of phytoplankton (including HABs such as cyanobacteria and
dinoflagellates) responses to nutrient inputs, as well as the responses of benthic microalgae and
macrophytes. Growing human demand for freshwater will need to be factored into the
formulation of nutrient criteria as it will influence freshwater discharge, nutrient loads, nutrient
concentrations and microalgal responses in impacted estuaries. Changes in freshwater flows add
further uncertainty to derivation of nutrient criteria based on historical relationships and will
need to be considered in triennial reviews of nutrient standards.
• Climate and Temperature Changes. In addition to climate-related hydrologic effects, changes in
temperature need to be considered. Changes in the range of 1.5 °C have been noted in some
systems during the past 60-70 years. Temperature changes on this order can influence
phytoplankton community composition (i.e., "cyanobacteria like it hot"; Paerl and Huisman
2008), as well as key biogeochemical nutrient and organic matter transformation processes (e.g.,
nitrification, denitrification, and sediment oxygen demand). These changes may alter
relationships between nutrients and chlorophyll based on previous climatic conditions and will
need to be considered in triennial reviews of nutrient standards.
• Threshold Changes. When setting nutrient criteria, the SAB recommends that EPA recognize
that threshold changes could occur in these systems. We include here non-linear responses, lags
relative to input changes and general "state changes". These changes will in part be a result of
changing nutrient loading and freshwater discharge dynamics due to changing anthropogenic
activities in watersheds and airsheds. They will also be modulated by climatic changes, including
changes in rainfall (and conversely drought) intensities, frequencies and geographic patterns, as
well as temperature changes (i.e., warming, which will favor the growth and proliferation of
nuisance taxa such as cyanobacteria). These changes need to be considered during future
triennial water quality standards reviews.
3.3. Florida Coastal Waters
3.3.1. Delineation and Data Sources
Charge Question 3(a). Are the data sources identified in Sections 2.4, 4.1.1 and 4.2 appropriate
for use in deriving numeric criteria in Florida's coastal waters? Is the SAB aware of additional
available, reliable data that EPA should consider in delineating coastal waters or deriving
criteria for coastal waters? Please identify the additional data sources.
The EPA document defines the outer boundary of the coastal zone as extending 3 nautical miles from
shore. However, Florida's territorial waters extend out to three nautical miles on the Atlantic coast and
nine nautical miles on the Gulf coast. Thus, additional data points could be included in the analysis and
the SAB recommends that EPA also consider monitoring remotely sensed chlorophyll in waters further
from shore. Given the dynamic nature of algal blooms in the Gulf of Mexico in particular, it is possible,
and perhaps even likely, that blooms that form more than three miles offshore will migrate toward the
coastline, thus eventually "appearing" in the state's coastal waters. It will be important to understand the
20
-------
source of such patches of elevated chlorophyll, and to determine whether they are found in close
proximity to the shoreline because of land- and estuary-derived nutrients or formed offshore.
Restricting the offshore boundary to three nautical miles greatly reduced the number of calibration
samples compared to the available data. As there is no clear boundary in water types at three nautical
miles, it is appropriate to use data from the entire shelf. Extending the outer boundary to the shelf break
in this way will improve the quality of the dataset. According to EPA personnel, adding these additional
data increased both the correlation and the slope of the calibration graph (Fig. 4.5 in the EPA document)
considerably (i.e., r increased from 0.52 to ~ 0.8). EPA might consider using anomalies relative to
either seasonal or annual means, rather than absolute Chi-a concentrations, in their estimates (see
Stumpf et al. 2003, 2009; Tomlinson et al. 2004). This approach will mitigate problems inherent in
working close to the coast, as bottom backscatter reflectance, for example, will be constant and therefore
disappear from the equation.
The coastal segmentation scheme suggested in the EPA document apparently is a result of historical
precedence, rather than any underlying scientific rationale. Given the general alongshore flow that
creates anisotropy with strong gradients perpendicular to the coast and weaker gradients parallel to it,
EPA may wish to consider segments defined in terms of bathymetry.
Another recurrent topic in SAB discussions was the "missing kilometer" at the coast where ChlRS-a data
are not being used because the satellite chlorophyll estimate is corrupted by the presence of land within
the pixels and because of backscatter from shallow water. A potential solution may be to use turbidity
data to connect conditions in the estuary proper with the coastal system just offshore, thus bridging the
km gap. Another potential solution would be to collect airborne spectrographic imagery, but this would
require a new data collection scheme.
The SAB agrees that the use of remotely sensed data to develop a reference criterion for Chi-a is
appropriate and sensible for this large, poorly sampled region. The use of these data, however, requires
calibration with in situ chlorophyll samples, of which there are few. The SAB accepts that these sources
are limited, but felt that additional sampling, including opportunistic sampling (using ferries, fishing and
charter boats, etc.), where feasible, would improve the dataset. While the use of a reference criterion
(Chl-a) is reasonable, the SAB is concerned with the sole reliance on a surrogate (see below) with no
direct measurements of nutrients being made.
The question also arises as to what reference level is applicable in this region. Historic nutrient
concentrations were likely very different from today (although few data are available to provide
quantitative information), yet the document assumes that these areas are currently supporting a balanced
phytoplankton community. Although the SAB recognizes that a longer data record is not available, it is
not clear that the ten-year dataset available from satellite observations constitutes an adequate baseline,
given decadal-scale variability.
The SAB notes that reliance on satellite observations may not be as feasible in the future, as existing
sensors near the end of their lifetime. The Panel's concerns about the continued availability of SeaWiFS
data were born out by the announcement in February 2011 that after several months of attempts to
communicate with the spacecraft, GeoEye has determined that the SeaWiFS mission is no longer
recoverable (per a notice posted on NASA's Ocean Color website, at
http://www.nasa.gov/vision/earth/lookingatearth/seawifs_10th_feature.html).While VIIRS may be
launched in time to avoid disruption in the collection of ocean color data, there is also question about
that sensor's capability to produce high quality data for chlorophyll. Therefore, the SAB recommends
21
-------
that the EPA ensure that data from the existing U.S. and European satellites, as well as future sensors, be
cross-calibrated to ensure as complete a data record as possible.
3.3.2. Assessment Endpoints
Charge Question 3(b). Is the assessment endpoint identified in Section 4.2 (chlorophyll-a to
measure balanced phytoplankton biomass and production) appropriate to translate Florida's
narrative nutrient criteria (described above) into numeric criteria for Florida's coastal waters,
given currently available data? Does the SAB suggest modification or addition to this
assessment endpoint?
EPA's suggested reference-based approach with satellite remote sensing of Chi-a (ChlRS-a) to derive
numeric values is likely to be effective in Florida coastal waters, because they are optically amenable to
remote sensing of chlorophyll, color (CDOM) and turbidity (Hu et al. 2005; Muller-Karger et al. 2005;
Palandro et al. 2004). Remote sensing technology has evolved sufficiently to begin using calibrated
imagery for estimating chlorophyll.
The SAB acknowledges that ChlRS-a is the most feasible indicator of nutrient status for coastal waters,
given available data and the lack of a better alternative (i.e., direct monitoring of nitrogen and
phosphorus offshore). However, we caution that Chl-a levels in these waters also are influenced by
seasonal water temperatures, circulation and mixing, and influx of nutrient-rich waters from advection
or upwelling. Thus, there may be only a weak relationship between nutrient concentrations and
chlorophyll concentrations. There is certainly a potential relationship between nutrients and organic
carbon production, but the strength of this relationship can vary depending on factors such as season or
relative availability of N and P, as shown clearly in Fig. 4.4 of the EPA document. Walker and Rabalais
(2006), cited in the EPA document, found only about 40 percent of the variance in phytoplankton
production could be ascribed to nutrient concentration, and this was in an area of the northern Gulf of
Mexico known to be affected strongly by nutrient inputs from the Mississippi River. Also, the carbon-
to-chlorophyll ratio within phytoplankton can vary by an order of magnitude (Banse 1977), while
Trichodesmium blooms can arise in low N regimes because these organisms are nitrogen fixers. The
SAB notes that Chl-a will not be useful as an indicator of species composition, as has been discussed
earlier.
3.3.3. Approaches
Charge Question 3(c). Does the approach EPA describes in Section 4.2 appropriately apply
remote sensing data to ensure attainment and maintenance of balanced natural populations of
aquatic flora andfauna in Florida's coastal waters? If not, please provide an alternate
methodology utilizing available reliable data and tools, and describe the corresponding
advantages and disadvantages.
EPA is considering a reference condition approach for coastal waters using remotely sensed Chi-a and
the SAB agrees this is an appropriate application of remote sensing data, with certain caveats. First, the
SAB recognizes the problems with other approaches, but notes that the reference condition approach is
being undertaken with no linkage to nutrient input into the coastal zone. The SAB recommends that a
preliminary assessment of nutrient inputs be undertaken to better understand chlorophyll levels in the
coastal zone (i.e., to relate observed chlorophyll levels in coastal waters to TN/TP concentrations or
loadings from land). A first pass calculation might consist of an estimate of the total nitrogen and total
phosphorus released into the coastal zones from all sources.
22
-------
Second, the SAB notes the thorough approach to calibration, but recommends several refinements:
• According to the document, EPA used a 3 x 3 km pixel matrix and used only in situ calibration
data taken within 3 hours of the satellite overpass. This should be sufficient, as tidal current
velocities, particularly off the Florida panhandle and over the wide West Florida shelf, are
generally small (< 10 cm/s; He and Weisberg 2002). Largest values are about 20 cm/s in the
vicinity of the Big Bend and Florida Bay. Tidal ellipses here tend to be perpendicular to the
bathymetry except very close to the coast, where they tend to parallel it (Koblinsky 1981; He and
Weisberg 2002).
• The SAB recommends that satellite data within a larger, "coastal" context be used for the
calibration, i.e., including data from outside the 3-mile zone. Because the calibration presented
was not strong (r2 = 0.52), this inclusion of additional data should improve the predictive power
of the model. The SAB also recommends that EPA cross-calibrate its imagery from multiple
existing (and future) sensors so that ChlRS-a values can be compared over time.
• Another issue on calibration concerns the relation between remotely sensed chlorophyll and
water column measurements. EPA calibrated the satellite data to chlorophyll measured in the
uppermost two meters of the water column. The ratio between the chlorophyll concentrations in
the upper two meters and the full euphotic zone needs to be established for the subtropical waters
around Florida. In many cases, there is not a strong relationship between total water column
chlorophyll and near-surface values, particularly in areas with upwelling and other circulation
complexity.
• The SAB recommends that obvious antecedent bloom data points be removed from analyses as
these are likely not representative of desired "reference conditions" (p. 83, paragraph 1,
regarding Karenia blooms).
3.4. South Florida Inland Flowing Waters
3.4.1. Rationale for Criteria
The SAB recognizes the considerable time and effort that has been put into identifying data sources,
assessing endpoints, and developing two approaches for deriving nutrient criteria for inland flowing
waters of South Florida. However, waters included in this category are dominated by man-made canals
and the SAB is not convinced from the material provided that nutrient criteria are appropriate for these
uniquely artificial and highly managed ecosystems. We identify five specific concerns in this
introductory section before addressing the specific charge questions. We acknowledge that these
comments and questions may not have explicit answers; however, they deserve thought and
consideration.
• The category of South Florida inland flowing waters includes a variety of different water types,
including both human-created/human-controlled canals and natural streams (the latter of which
are largely confined to the southwest region of South Florida; Tom DeBusk, personal
communication; Hecker and Young 2011). A rough estimate from the SFWMD (Garth Redfield,
personal communication) indicated that approximately 90 percent of the waters in this region are
classified as canals, and the remaining 10 percent are natural streams. Although canals dominate
in number, the range of water types included in this category make it unlikely that there will be a
consistent response to nutrients.
23
-------
• South Florida's inland flowing waters have a long history of being highly manipulated and
managed, and in this regard they represent a special challenge to developing numeric nutrient
criteria. The underlying problem is that the canals are classified as Class III waters, although
their primary purpose is management of water quantity to alleviate flooding potential. The SAB
notes the difficulty in determining what constitutes an appropriate reference condition for these
systems, and the related issue of whether or not appropriate data are available to help define
reference conditions.
• Internal nutrient loading from sediment accumulated in canals may confound the relationship
between nutrient criteria and system response if sediments are a major source of nutrients (and
based on SFWMD [2010], sediment accumulation and P mass are quite variable).
• It is unclear how nutrient sources from past activities (i.e., legacy nutrients), including legacy
losses or inputs of N and P to water bodies, will be accounted for in the proposed approach. If no
water quality improvement or indicator biological response is seen after numeric nutrient criteria
are put in place, is this because (1) the nutrient criteria are not stringent enough, (2) legacy
nutrient inputs are an increasingly significant contributor, or (3) the monitoring interval is not
long enough to capture the response of dynamic ecosystems and watersheds? How will
continued legacy inputs of N and P be distinguished from decreased loadings related to
management changes? Internal recycling of nutrients can mask water quality improvements
brought about by nutrient reductions resulting from land management changes. Given the role of
legacy nutrients in influencing water quality in these systems, the EPA should employ an
adaptive management approach (e.g., as part of the triennial review of water quality standards) to
incorporate new monitoring data and revise criteria or loading targets as appropriate.
• South Florida inland flowing waters involve a spatially and temporally dynamic interaction
between surface and groundwater flows and as such, biological condition of these waters may be
more responsive to hydrology than to nutrients. For instance, N and P loadings at different times
of the year can influence biotic responses in different ways. In other words, it is not just how
much or in what concentration, but at what time these loadings occur. In dry years, ground water
will greatly influence surface water chemistry/quality compared with wet years. There also is
concern that cross watershed / ecoregion / system transfers of water and nutrients in ground
waters could confound the ability to relate ecological response to water column nutrient
concentrations or loadings.
24
-------
3.4.2. Delineation and Data Sources
Charge Question 4(a). Are the data sources identified in Section 2.4 and 5.4 appropriate for use
in deriving numeric criteria in South Florida's inlandflowing waters (as discussed in Chapters 2
and 5)? Is the SAB aware of additional available, reliable data that EPA should consider in
delineating or deriving criteria for South Florida's inlandflowing waters? Please identify the
additional data sources.
The SAB recommends that the category of South Florida inland flowing waters be subdivided into two
subclasses, each with separate approaches for nutrient criteria development: (1) human-managed canal
systems (or some equivalent term) and (2) natural flowing lotic systems. EPA should identify
characteristics to separate these two classes of water bodies, with one suggested characteristic being
whether their flow regime is natural or managed.
The network of human-managed canals provides ecosystem services, but these services are controlled
primarily by hydrology and habitat quality rather than nutrient levels. Thus, for purposes of nutrient
criteria, these canals should be viewed as connecting channels from inland water sources to downstream
estuaries. This would be consistent with the canal science summary document (SFWMD 2010), which
describes the aquatic life in the canals (macroinvertebrates, fish, alligators), but acknowledges that the
ecological value of the canals is secondary to their use for water conveyance. These canals serve as a
water conduit, but also transport considerable nutrient loads to downstream receiving areas. Hence, the
nutrient content in the canals can serve as a proxy for "potential impact" to the more natural wetlands
and water bodies adjacent to and downstream from canals and may be best addressed through
downstream protection values (DPVs) for estuaries, rather than by establishing canal nutrient IPVs.
(However, see section 3.6.2 on DPV).
The SAB recommends that EPA evaluate the appropriateness of developing nutrient criteria for South
Florida natural flowing lotic systems, given their greater potential for valued ecological attributes
relative to canal systems. Suggestions on approaches for nutrient criteria development are provided in
the next section.
There is considerable debate as to whether or not the data in Sections 2.4 and 5.4 of the EPA document
are sufficient to derive numeric criteria for South Florida's inland flowing waters. These data sources are
certainly the most logical beginning point. However, EPA should look into datasets potentially available
from the local water/drainage districts (not water management districts), such as Lake Worth; these local
regulatory authorities can control water regulation and were created for the purpose of reclaiming the
lands within their boundaries and to provide water control and water supply for settlement and
agriculture (see www.lwdd.net). In addition, agricultural interests that border the canals (e.g., U.S. Sugar
Corp.) may have useful data, although these data may be proprietary.
The proposed inventory of inland flowing waters that catalogues and distinguishes natural streams and
canals should provide very useful information. This inventory could provide the basis for dividing
inland flowing waters into two subcategories for purposes of criteria development, as recommended
above.
The EPA document explores the use of the Landscape Development Index (LDI) as a potential approach
(and data source) for determining reference conditions in inland flowing waters (reference conditions
where LDI < 2, p. 105). It is well established that surrounding land use can have substantial impacts on
receiving water bodies, so this approach has conceptual and intuitive appeal. However, insufficient
information is available in the EPA document to determine the appropriateness of the LDI approach for
25
-------
South Florida's inland flowing waters. The SAB does not recommend that the LDI be applied to South
Florida canals. Further concerns include the use of a 100-m buffer along canals; would not the canal's
water quality to be determined by the entire area that drains into it? The document cites a study by Fore
(2004) to justify this approach. However, Fore (2004) was based on streams throughout the state and not
just canals; there are considerable differences in hydrology and land-water interactions between canals
and natural stream channels. The 100-m buffers proposed for use with the LDI (p. 105-106) may be too
limited, particularly where stormwater pipes convey runoff from distances much further than 100
meters.
The condition of inland flowing waters is highly influenced by geology and anthropogenic activity. In
this regard, there is logic to subdividing these waters according to basin and sub-basin soil types and
land uses. An additional challenge is incorporating groundwater hydrologic/nutrient dynamics, which
also have been altered, but are likely to be very important in determining nutrient sources and impacts.
The proposed classification scheme appropriately incorporates surface and subsurface flow regimes and
flow lines, as well as soil types and human agricultural and urban impacts (i.e., land use). Inland water
regions should be classified according to soil order, color of water, or a combination of several such
characteristics.
As mentioned above, legacy N and P loads from past management and from natural sources also must be
considered. The EPA document appears to minimize the importance of legacy effects of past
management (e.g., the statement on the top of page 40 in reference to Huang and Hong 1999). However,
soil nutrient levels vary greatly as a function of past management, within and among fields and within
areas with uniform LDIs for which specific criteria concentrations will be set. Additionally, geologic
rock deposits vary within areas assumed to have similar nutrient concentrations. There is a wealth of
data on soil nutrient levels (particularly P) available from NRCS and land-grant university extension
offices. EPA should evaluate these, and other, data to better support the regional classification of South
Florida inland flowing waters.
3.4.3. Assessment Endpoints
Charge Question 4(b). Are the assessment endpoints identified in Section 5.4 (balancedfaunal
communities, i.e., aquatic macroinvertebrates, and balanced phytoplankton biomass and production)
appropriate to translate Florida's narrative nutrient criteria (described above) into numeric criteria
for South Florida's inlandflowing waters, given currently available data? Does the SAB suggest
modification or addition to these assessment endpoints?
Philosophically (but with practical implications), one can question whether any assessment endpoint is
appropriate for systems that have been artificially created. How does one establish an appropriate
reference condition for such systems, especially when they are heavily managed? There are no easy
answers for these questions, although this has certainly been done for reservoirs. Furthermore, nutrient
inputs are not the main drivers of habitat quality in the canals of South Florida.
South Florida canals have been constructed continuously over the last century, so it is not clear how
reference conditions can be assessed for these highly variable systems that are designed to be dynamic
and flashy during wet periods (to get water off the landscape quickly) but are managed during dry
periods to be slow or non-flowing. Least-disturbed sites tend to be in one region only and may not be
transferable to other identified regions. Because canals are unique aquatic ecosystems, more information
needs to be presented on how balanced natural populations are to be assessed. An initial inventory of
science for South Florida canals, provided by SFWMD (2010), summarizes data on water quality and
biological conditions in the canals. The closest analog to South Florida canals would be in The
26
-------
Netherlands, where much of the inland waters flow through canals (locally called ditches). There is
some literature on assessment endpoints from Netherland ditches (e.g., see Verdonschot 1987) that may
be of use in developing methods for assessing the status of flora and fauna in Florida canals.
If nutrient criteria are to be developed for South Florida inland flowing waters, the SAB recommends
further consideration and assessment of both the response variables and the form of the nutrients being
assessed. For example, distinction is needed among nutrient forms that are immediately availability for
biological uptake —i.e., short-term bioavailability and growth response, such as inorganic nitrogen (NO3
and NH4) and phosphorus (PO4)—compared with losses, such as particulate and organic forms of N and
P (i.e., long-term availability).
For assessment endpoints, both Chi-a and invertertebrate measures have conceptual appeal, but their
utility is not straightforward. Some freshwater and estuarine ecosystem studies have shown that Chl-a
can be a function of grazing pressures rather than, or in addition to, nutrient concentrations. For
example, increasing nutrient concentrations in inland flowing waters can increase the number of grazers,
which can lead to a lower Chi-a concentration; i.e., a top-down regulation of primary production (Cloern
1982; Cohen et al. 1984; Hilton et al. 2006; Thompson et al. 2008; Lewis and McCutchan 2010).
Aquatic macroinvertebrate community structure and/or traits have been shown to be reliable
bioindicators in other aquatic ecosystems. Hence, they are a reasonable starting point for South Florida's
inland flowing waters. However, these systems are poorly understood, highly managed, and heavily
modified. As a consequence, it is unclear at present if these proposed assessment endpoints can be
applied effectively.
The SAB identified areas of uncertainty that need further attention before a reasonable level of scientific
confidence can be applied to the use of balanced faunal communities and/or balanced phytoplankton
biomass and production. We elaborate on these below.
Faunal Communities
The macroinvertebrate index used by Snyder et al. (1998), provided as Figure 5-8 in the EPA document,
shows a good relationship between land use and macroinvertebrate community structure. However,
macroinvertebrate data provided in a presentation to the SAB reveal a much more tenuous stressor-
response relationship between total P concentrations and macroinvertebrate indices (DeBusk 2010). It is
important that EPA examine possible reasons for the lack of correspondence in these two data sets.
Possible explanations include the use of different measures of stressor (land use vs. total P), different
types of indices (e.g., emphasizing different taxa), or inclusion of inland flowing waters from different
parts of south Florida (e.g., that experience different pressures). For example, the relatively high Stream
Condition Index (SCI) score for the wetland sites shown in the Snyder et al. data may have more to do
with habitat quality than nutrients. The summary of canal science prepared by the SFWMD (2010) notes
that "additional research is needed to select sensitive (macroinvertebrate) metrics and a quality threshold
applicable to low gradient streams and canals within the peninsula and Everglades bioregions." The
SAB agrees with this statement. If the different macroinvertebrate patterns in these data sets can be
explained and related to nutrients, aquatic macroinvertebrates may be a very useful assessment endpoint
and one that the SAB recommends be given more attention. However, that relationship has not yet been
established.
Phytoplankton Biomass
There is a relative paucity of phytoplankton data (either as Chl-a, species composition, or productivity)
in these inland flowing waters. Data from SFWMD (2010) show geometric mean Chl-a concentrations
27
-------
ranging from 2 |ig/L (Lower East Coast) to 8.0 |ig/L (Everglades Agriculture Area) in canals within the
South Florida region. However, these concentrations are not paired with data on hydrologic conditions,
and it is impossible to assess if they represent actively growing algae populations (as might be expected
in a non-flowing canal) or algae being transported downstream (i.e., in a flowing canal) and therefore
not representative of local conditions. The hydrologic status of the canal (non-flowing, slow-flowing,
fast-flowing) has enormous implications for the plankton community, and this needs to be accounted for
in EPA's assessment. In addition, both present and antecedent hydrologic conditions need to be
considered in any analysis, as the flow regimes are quite variable (and can be artificially controlled) in
these systems. Hence, a canal that may be flowing when sampled may have been stagnant for weeks
prior to sampling, and vice versa. These kinds of lags make relationship interpretation (and accurate
modeling) complex. At this point, it is unclear if there are sufficient data to know what a "protective"
level of Chi-a should be for these systems; as a consequence, it is currently not possible to assess
whether or not phytoplankton can be used as an effective assessment endpoint.
The inventory of the inland flowing waters, and subsequent screening of water bodies, is an important
step and may help in the selection of appropriate assessment endpoints. The approach provided in the
technical document is a good starting point, but the SAB has identified some issues and suggestions with
respect to the classification procedure. The SAB identified several factors that EPA may want to
consider with respect to the chlorophyll endpoint:
• EPA proposes a classification of inland water regions according to soil order, land management
systems or color of water (see below). Some combination of these factors should be considered,
taking into account covariates.
• Currently, EPA does not appear (at least explicitly) to consider the potential influence of humic
soils in their classification of inland flowing water types, with respect to their role in
discoloration of waters; phytoplankton response will be very different in waters that are
naturally colored (i.e., influenced by humics) vs. those that are not.
Dissolved oxygen concentration reflects the relative amount of photosynthesis (DO production) and
respiration (DO consumption) in aquatic ecosystems, so DO might be an alternative endpoint in this
geographic domain; however, new studies would be needed to determine if DO levels are linked to
nutrient loads or concentrations, and not to other factors (such as light), and if groundwater influx (low
DO) confounds the use of this assessment endpoint.
3.4.4. Approaches
Charge Question 4(c). EPA describes two approaches in Section 5.4 (reference conditions and
stressor-response relationships) for deriving numeric criteria in South Florida inlandflowing
waters. Compare and contrast the ability of each approach to ensure attainment and maintenance of
balanced natural populations of aquatic flora andfauna in different types offlowing water or
geographical areas, given currently available data?
The two approaches that EPA is considering for determining numeric criteria for South Florida inland
flowing waters are discussed in Section 5.4 of the EPA document. The first is based on reference
conditions and the second is based on stressor-response relationships. Based on the available data, the
SAB recommends against the establishment of instream protection values for nutrients in the human-
controlled canal systems; these inland waters are more appropriately thought of as hydrologic conduits
designed to rapidly remove storm water from the landscape, while at the same time delivering nutrients
to downstream estuaries. The SAB does recommend that EPA evaluate both approaches for the natural
flowing lotic systems, but recognizes that data availability may severely limit their ability to develop
28
-------
rigorous criteria at present. As such, the SAB recommends that, at least for the present, South Florida
inland flowing waters be addressed through downstream protective values. We comment on the
problems with EPA's proposed approaches below.
Reference Conditions
Briefly, under the approach based on reference conditions, a set of least-disturbed sites would be
identified using the Land Development Intensity (LDI) index. The total LDI for each site would be
calculated as an area-weighted sum of the LDI coefficients for all land uses within an area of influence.
Sites with total LDI below 2.0 or another specified threshold would be classified as least-disturbed and
would form the reference set. The historical annual values of TN and TP for these sites would be used to
fit lognormal distributions of TN and TP under least disturbance and specified quantiles of these fitted
distributions - the EPA document mentions the 0.75 and 0.90 quantiles - would be used as the numeric
criteria.
On the bottom of page 107, the EPA document discusses the question of the frequency with which these
numeric criteria could be exceeded. This discussion is difficult to follow, but the general point appears
to be this: Consider that the estimated 0.75 quantile for one nutrient is exceeded k or more times in n
years. Commonly used values are 1 in 3 and 2 in 5. Under the assumption that values in different years
are independent and have the same distribution as the reference set (and ignoring any error in the
estimation of the 0.75 quantile), the probability of this event is given by:
p(k,n) = YJ
j=k
( ri\
KJJ
0.25; 0.75"
For fixed k and n, this formula essentially provides a one-sided significance level for testing the null
hypothesis that the nutrient distribution is the same as that of the least-disturbed sites. So, for example,
under this null hypothesis, the probability p{ 1, 3) of at least 1 exceedance of the 0.75-quantile in 3 years
is 0.58, while the probability p{2, 3) of at least 2 exceedances in 3 years is 0.16.
The SAB notes the following:
• The choice of quantile, k, and n can have a profound effect on the performance of criteria derived
in this way and some discussion is needed about how this choice will be made.
• The probability calculations sketched here pertain to exceedances of a single nutrient criterion. If
the same rule is applied to both nutrients (and assuming that nutrient levels are independent)
then, under the null hypothesis that both nutrient distributions are the same as those for least-
disturbed sites, the probability that either or both nutrients exceed their respective 0.75-quantiles
in at least k of n years is 1- (1 ~p(k,n))2. Thus for both nutrients, the probability of at least 1
exceedance of the 0.75-quantile of either or both nutrients in 3 years is 0.82, while the
probability of at least 2 exceedances in 3 years is 0.29, i.e., much greater than for each nutrient
considered individually
As noted, these calculations assume that the relevant quantile of the annual nutrient levels in least-
disturbed sites are estimated without error. An assessment of the impact of estimation error - including
non-normality of the log of annual nutrient levels - on the accuracy of these calculations is needed.
29
-------
Although these calculations provide information about the rate of Type I error (i.e., the exceedance of
the criterion when the underlying distribution is the same as that for least-disturbed sites), they provide
no information about the rate of Type II error (i.e., the non-exceedance of the criterion when the
underlying distribution is different from that for least-disturbed sites). In the jargon of hypothesis-
testing, this analysis provides no information about power. To gain such information, it is necessary to
consider also the distribution of nutrients in disturbed sites.
To some extent, variability of nutrient levels in the least-disturbed sites will reflect heterogeneity in
hydrology, geology, etc. Failure to account for such heterogeneity, which is also present in disturbed
sites, may result in numeric criteria that are under- or over-protective for some sites. It would, therefore,
seem preferable to develop criteria that account for such factors. The EPA document briefly notes (on p.
108) that EPA also is considering following the reference conditions approach using all sites as a
reference set (and not only least-disturbed sites as discussed above). With the exception of the
identification of least-disturbed sites, the mechanics of these approaches are the same. However, the
underlying logic seems rather different - loosely speaking, one approach aims to reproduce conditions in
least-disturbed sites and the other aims to maintain conditions within a specified quantile of the
distribution of all sites, whatever their level of disturbance - and this needs to be discussed.
Another Possible Approach
The SAB briefly discussed another statistical approach to determining numeric criteria for TN and TP.
The ultimate goal of this determination is to ensure that waterbodies meet their designated uses. If
information is available about whether or not waterbodies with the same designated use are, in fact,
meeting this use, then it should be possible to construct logistic or other binary regression models
relating the probability that designated use is met to levels of nitrogen and phosphorus (and possibly to
other regressors such as rainfall).
Consider that for years Florida DEP scientists examined water quality data on Florida surface waters and
then made the judgment as to whether or not a waterbody was in compliance with the narrative nutrient
criteria. In effect, these scientists created a binary response variable (compliance or noncompliance) for
each "data row" of water quality data for each Florida waterbody. These data rows for each waterbody
type (e.g., lakes, rivers, estuaries) could be the basis for development of logistic or other binary
regression models with the water quality variables as predictors and the DEP scientists' classifications as
the response.
For example, if initial analyses indicated that TN and TP were clearly the best predictors of compliance
response in estuaries, then a binary regression model could be fitted with TN and TP as the two
predictors of the compliance response. Then new TN and TP data could be inserted into this binary
regression model to determine the probability of compliance with the nutrient criteria. So, for example,
with a 50:50 cutoff for compliance as the transition point, if an estuary's TN and TP result in greater
than 50% probability of compliance with the nutrient criteria, the estuary would be classified as not
impaired.
This approach can be undertaken with little cost and time, and it makes direct use of the State's
interpretation of the existing narrative criteria. This approach is similar in spirit to the one based on
reference conditions. However, unlike the approach based on reference conditions, the statistical
analysis discussed here would use information about both unimpaired and impaired cases. Although the
results of this statistical analysis will be only as good as the underlying categorization of waters as
impaired or unimpaired, the approach may offer additional insight for interpreting proposed criteria.
30
-------
Stressor-Response Relationships
The second approach that EPA is considering for developing numeric criteria for South Florida inland
flowing waters is based on stressor-response relationships. This involves developing a statistical model
relating the level of Chi-a to TN or TP. The EPA document presents examples involving linear,
nonparametric, and quantile regressions of log Chi-a (as response) and log TN or log TP (as stressor).
The SAB notes the following:
• A fundamental question that the EPA document leaves unanswered is how such fitted regression
models will be used to determine numeric criteria; i.e., how they will determine the level of Chl-
a that will be considered protective of balanced phytoplankton and faunal communities. This is a
serious shortcoming that needs to be addressed.
• As the EPA document notes, it has not been possible to develop stressor-response relationships
in which a biological endpoint other than chlorophyll serves as the response. However, if there is
a clear relationship between Chi-a and TN, say, and a clear relationship between another
biological endpoint (e.g., a faunal community metric) and Chl-a, then there would ordinarily be a
clear relationship between that biological endpoint and TN. The fact that it is difficult to identify
this latter relationship may reflect limitations of the statistical models considered so far. For
example, the effect of TN or TP on a biological endpoint may be modulated by other factors.
This effect could be obscured by omitting these factors from the regression model.
• As with the approach based on reference conditions, the relationship between Chl-a and TN or
TP is likely to be modulated by the effects of hydrological, geological, and other covariates.
Failure to account for such factors may lead to criteria that are over- or under-protective at some
sites, and it would again seem preferable to include such covariates in developing numeric
criteria. Furthermore, should TN and TP be considered simultaneously (i.e., is a multiple or
simple regression most appropriate)?
• Some of the variability in the stressor-response relationship could be a result of season. This
should be investigated, and it may lead to the formulation of different criteria for different
seasons.
• A substantial amount of effort will be put into identifying and quantifying stressor-response
relationships in these waters using correlative/regression analysis. Considering the difficulty of
working across the surface-subsurface interfaces in deriving nutrient loading estimates, as well
as effects of these loads, the authors have done a good job of addressing these challenges. This
section could however benefit from closer process/response connections (including applying
modeling approaches) to receiving estuarine and coastal waters.
• Reference conditions established using a distribution approach (p. 108) will be sensitive to the
distribution of sites along the disturbance gradient. If a larger proportion of the samples are from
more disturbed sites, then using the lower percentile to set the criteria will result in a higher
number than if a larger proportion of the samples are from less disturbed sites. Some
requirements for the distribution of sites along the disturbance gradient should be specified.
31
-------
3.5.
South Florida Marine Waters
3.5.1. Delineation and Data Sources
Charge Question 5(a). Are the data sources identified in Section 2.4 and 5.5 appropriate for use
in deriving numeric criteria in South Florida's marine waters (as discussed in Chapters 2 and
5)? Is the SAB aware of additional available, reliable data that EPA should consider in
delineating or deriving criteria for South Florida's marine waters? Please identify the additional
data sources.
One general recommendation is that the waters currently termed "marine waters" in the EPA document
be changed to "South Florida coastal and estuarine waters" to be consistent with the use of the terms
throughout the rest of the document.
The southern part of Florida has a different nutrient regime than other parts of the state: it is highly
oligotrophic and conditions can be rapidly altered by upstream water management (versus nutrient
regulatory) decisions. In addition, embayments with calcareous sediments are highly phosphorus (not
nitrogen) limited. The SAB agrees that these waters should be considered separately for purposes of
nutrient criteria development. However, the proposed subdivision/subclassification of South Florida
estuarine and coastal waters does not clearly relate to the oceanographic circulation and degree of
connectivity in the region, particularly on the seaward side of the Keys and the southeast and southwest
interior coasts. Alongshore circulation and interconnection is much more vigorous than across shelf
except for highly restricted intertidal exchange passages (Lee and Mayer 1977; Lee et al. 2002).
The data identified in the report seem appropriate for use in this exercise. There also are water quality
data from NOAA's Atlantic Oceanographic Meteorological Laboratory (AOML) that have been
collected for Florida Bay, Biscayne Bay, the Florida Keys and SW Florida Shelf for more than a decade
as part of the NOAA South Florida Program (www.aoml.noaa.gov/sfp). There are some possibly
significant differences between these data and the Southeast Environmental Research Center (SERC)
data, which covers the same domains. For some periods in some subregions, the NOAA data were
temporally more dense (bimonthly versus quarterly) in larger domains and the nutrient methodologies
were more sensitive (long path length liquid wave guide) in accordance with oceanographic practice for
oligotrophic open ocean waters (as established in JGOFS, GLOBEC and other international programs).
3.5.2. Approaches
Charge Question 5(b). EPA describes two methods in Section 5.6 for using a reference condition
approach for deriving numeric criteria in South Florida marine waters (least-disturbed sites or
bionomial test). Compare and contrast the ability of each approach to ensure attainment and
maintenance of balanced natural populations of aquatic flora andfauna in South Florida marine
waters, given currently available data?
Two approaches to nutrient criteria are being considered for South Florida coastal and estuarine waters.
The first is to identify criteria that are inherently protective based on a statistical evaluation of data from
least-disturbed sites. In some of these zones, least-disturbed sites may be those most distant from land-
based sources; this is not a result of a significant different in nutrient loading but rather a result of
dilution with naturally highly oligotrophic waters. Hence, by including sites being diluted by highly
oligotrophic ocean waters, the criteria may be overly protective. Moreover at such sites it is not atypical
that the limiting nutrient can change. A second proposed approach is also based on a statistical
evaluation, but in this case raw data are analysed using a binomial test and two criteria are generated -
32
-------
an average concentration and an upper percentile concentration that is more sensitive to higher
concentrations. Both approaches have merit and the SAB encourages the application of both to provide a
more robust evaluation of criteria.
A third alternative has been suggested by Briceno et al. (2010), based on statistical analysis of long-term
monitoring data for Chi-a and nutrient concentrations in South Florida coastal and estuarine waters.
While that approach has some merit in subregions where sufficient time-series data are available, it is
not applicable in many other areas. Moreover, an inherent lagtime is introduced in that analysis which
may make it insensitive to ecologically significant deleterious changes if they occur too rapidly.
It is also critical to address how the statistical approaches would be applied. For example, if a baseline
(i.e., reference) condition is established using the median or geometric mean of a decade of data for the
undisturbed condition, there still remains the major issue of how concentrations that exceed the criteria
will be determined. Will each year be assessed against the baseline (the approach taken with the CERP
System Status Report and the South Florida Ecosystem Restoration Task Force, SFERTF, Scientific
Indicators) or will five years of data be required to determine if 2 (or 3) had "exceeded" the baseline?
How would the variability in the two data sets (baseline and evaluation) be incorporated? Given how
variable some of these numbers are, it is a lot less powerful (less chance of seeing a change) to ask if the
means of the two datasets (in the example above, 10 and 5 yrs) differ versus whether a particular year
was significantly above the baseline mean. Furthermore, there may be major ecological differences
between two successive years of concentrations that exceed the criteria versus two years separated in
time, and the document does not discuss this. The SAB recommends that more thought be given to these
implementation issues.
The SAB recommends a reconsideration of the rationale for doing both a principal component and
cluster analysis. EPA proposes to use a combination of principal component analysis (PCA) and cluster
analysis to define coastal regions based on multivariate measurements with sites. As the goal of cluster
analysis is precisely to identify groups of similar sites, it is unclear why PCA is also being proposed in
this context.
Past Alterations
The coastal and estuarine waters of South Florida have experienced enormous changes over the last
century. For example, the Florida Bay of the early 1900's was a true estuary with low and highly
variable salinities for most of the year. Following widespread damming for the Flagler railroad, salinities
were lowered throughout much of the Bay; the effect on salinity was relatively minor, but the effect on
residence time was significant. Then, after canal construction from the 1920s to mid-1960s, the vast
majority of the water flowing out of Lake Okeechobee is now shunted out to sea before reaching the
southern-most waters of the state (Steinman et al. 2002b). Although there were a number of animal
studies conducted, there were few nutrient or chlorophyll measurements made because the water was so
clear that light penetrated to the bottom. In the 1970s, Thalassia covered Florida Bay, believed to be a
result of the artificially high salinities resulting from the eastward and westward shunting of water that
used to flow south into Florida Bay. A major drought in the mid-1980s resulted in Florida Bay salinity
going as high as 70 ppt, which killed off Thalassia and other sea grasses. Although nutrients were not a
cause of the sea grass dieoff, the result was that enormous amounts of nitrogen and phosphorus that had
been sequestered as detritus in the sediment were no longer protected by the dense sea grasses. A
subsequent large storm event then mixed large amounts of sediment nutrients into the water column. The
result was eutrophication, yet the ultimate cause was a change in salinity that killed off sea grasses years
before. Based on this brief history, the SAB has the following recommendations:
33
-------
• When setting reference conditions, EPA should consider historical water management and
structural changes and regional climatic variability that have affected water delivery to South
Florida estuaries and coastal waters, recognizing that current conditions do not reflect historical
conditions.
• In particular, salinity should be considered a covariate because of its role in maintaining the
water quality necessary for seagrasses. We note, however, that salinity is relevant (and in fact
variable as a result of water management) only in a very restricted part of this domain.
• Seagrass coverage and the extent of epiphytic colonization should be considered as endpoints, in
addition to water column chlorophyll (see also 3.2.2).
Geographic Areas to be Included
South Florida contains a number of parks and marine protected areas. Management jurisdiction in the
region has been clarified to a large degree by the formation of the National Marine Sanctuary. The
document should clarify which coastal and estuarine areas will be under the jurisdiction of the EPA
document under review vs. other regulations. We note that the Florida Keys National Marine Sanctuary
(FKNMS) domain (and that of the three national parks: Biscayne, Dry Tortugas and Everglades - a.k.a.
Florida Bay) are not the only federally protected waters, and there are also state-protected waters of
various types. It is our understanding that what is set by EPA will constitute a "de minimus" standard for
these areas, which could receive additional protection. Similarly, the EPA document should clarify the
relationship of the South Florida coastal and estuarine nutrient criteria to the relevant indicators and
performance measures already established by the Comprehensive Everglades Restoration Plan (CERP)
through its Restoration, Coordination and Verification (RECOVER) program.
34
-------
3.6. Downstream Protection Values
3.6.1. DPVs for Estuaries
Charge Question 6(a). Are the methods EPA is considering for deriving downstream protection
values (DPVs) for estuaries (excluding marine waters in South Florida) as described in Section 6.1-
6.4 appropriate to ensure attainment and maintenance of downstream water quality standards, given
available data? Please describe additional approaches and their advantages and disadvantages that
EPA should consider when developing numeric criteria to protect these downstream estuarine
waters (excluding marine waters in South Florida), given available data?
Rationale for DPVs
The 1972 Clean Water Act (CWA) states that:
In designating uses of a water body and the appropriate criteria for those uses, the
State shall take into consideration the water quality standards of downstream waters
and shall ensure that its water quality standards provide for the attainment and
maintenance of the water quality standards of downstream waters.
This provision has been the basis for ensuring that water quality standards in one state provide for
attainment and maintenance of water quality standards of downstream states and Tribes. The recently
published nutrient criteria for Florida's lakes and flowing waters (75 FR75762-75807, December 6,
2010) explicitly included the concept of downstream protection values (DPV) as a concentration or
loading value in a stream at the point of entry into a lake, set at a value to ensure that lake nutrient
criteria are attained. The rule also notes that wasteload and/or load allocations from an approved total
maximum daily load (TMDL) may be used as the DPV.
In the present document, the concept of DPV is included as a means of ensuring that upstream N and P
water quality criteria will be set at levels that will protect downstream estuarine designated uses. This is
particularly important because freshwater and estuarine systems may be sensitive to different nutrients.
However, the SAB had concern with the overlap between DPV and the TMDL process. To illustrate
this, consider Figure 2 below. Water quality criteria (WQC) are required for all waterbodies in this
figure - the estuary and the streams. If streams Al, Bl, and CI meet their WQC, yet the estuary does
not, additional pollutant load reductions to the estuary are required. These reductions could come from
direct loading, the atmosphere, or the tributaries (Al, Bl, CI). Standard practice in the TMDL process is
to model the estuary and watershed to determine the additional pollutant load reduction needed, and then
to allocate the load reduction based on input from state and local officials. One possible result of
implementation of the required load reduction is that one or more streams may require WQC more
stringent than those initially established in order to attain WQC in the estuary. Regardless, this
regulatory-driven analysis achieves compliance with all WQC without the additional regulatory entity of
DPV criteria.
The drawback of relying on TMDLs, rather than the DPVs, is that the TMDL process is not triggered
until impairment has occurred in a waterbody. However, the TMDL process includes the opportunity to
consider socio-economic factors when selecting an allocation scheme for load reduction. In principle, in
an environment with no water quality impairment, the DPV concept could be captured within the
designated use for each tributary and correspondingly reflected in the water quality criteria for each
tributary. However, in situations where impairment exists, the DPV enforces rigidity that likely will not
35
-------
accommodate variations in natural and human factors affecting nutrient loading. Of particular note,
nutrient trading between tributary watersheds, which might be viewed favorably from a social and an
economic perspective, may be prohibited due to a priori nutrient load restrictions imposed by the
rigidity of the DPVs. Therefore, if the DPV approach is pursued, the SAB urges that selected
apportionment strategies not preclude flexible nutrient allocation across tributaries to achieve the
necessary load reductions.
The EPA document includes an example of DPV developed using an equal load reduction for each
tributary to a waterbody. The SAB notes that this equal allocation approach ignores the possibility of
different background nutrient concentrations in upstream segments, and other characteristics of
segments that would make nutrient load reductions impractical. In response to a SAB request for further
information, the Agency described four options for distributing loads for calculation of DPV (Appendix
C). Each of these options presents some issues, as discussed below.
• Option A distributes the load based on discharge, making the assumption that high nutrient loads
are associated with higher flow rates. This option could present several unresolvable issues. For
example, a stream with a low discharge but high nutrient load may be required to reduce its load
to unachievable values. This could occur if a stream drains an area with elevated natural
background, and would result in criteria that would require load reductions below background.
Conversely, a stream with a high discharge may not be required to reduce its load even if there
were some relatively simple approaches for reducing it load.
• Option B would distribute load based on watershed area drained. This would appear logical if
the export of nutrients from all land uses were nearly equivalent, which it is not. As with Option
A, this option might result in reduction targets that are unreasonable or unachievable in small
subwatersheds while missing opportunities for reductions in larger watersheds. An integrated
approach to load reduction should focus on watersheds, rather than on individual segments.
• Option C would distribute load based on natural or background loading. This approach seems
reasonable if it is possible to accurately arrive at the natural background nutrient load for a
segment or watershed. Essentially, an estimate of the natural background for each segment, given
a set of criteria (slope, soils, etc), would set the achievable "DPV" for the stream segment. Such
an approach is intellectually appealing, but may not be practical in the time frame given for the
project. It does, however, strongly suggest the need to have a GIS-based model such as SWAT
developed for the project and later implementation.
• Option D would distribute load based on average background, as described in Option C, or
natural background, but include anthropogenic loading using existing data. The difference
between natural and anthropogenic load would be the load that could be targeted for reduction.
In the description of this option, it is suggested that average load for a stream/segment could be
divided by the average discharge. This is the concentration, and makes one question the need to
calculate loads or DPVs if we can simply express a maximum concentration for a stream or
segment.
The SAB notes a few additional points to think about when considering the wisdom of DPVs:
• There are other sources, such as direct (e.g., groundwater and atmosphere) loadings to the
estuary. How might those be addressed in the determination of tributary DPVs?
36
-------
• Referring to Figure 2 (below), why should there not be DPVs for streams B2a and B2b in order
to protect stream Bl?
B3b
A1
B3a
B2b
Bl
B2a
CI
Estuary
Figure 2. Example stream and estuarine segments.
Approach to Setting DPVs
EPA's proposed assessment of DPVs is based on watershed modeling (to be undertaken using the LSPC
model) which results in an apportioned pollutant load reduction for each tributary to the waterbody (e.g.,
estuary) of interest. In an example in the review document, EPA proposes to apportion the pollutant load
reduction (required to achieve compliance with the waterbody water quality criterion) as an equal
fractional load reduction for each tributary to the waterbody. As noted above, this approach would
formalize, and unnecessarily restrict, the more flexible pollutant load allocation process that occurs for
TMDL pollutant load allocation when a water quality standard violation occurs. That said, the SAB has
the following suggestions for the modeling of load reduction apportionment for upstream segments:
1. The watershed segment approach is valid, but care should be taken in selecting segments to take
into account available data and other watershed characteristics such as predominant land-use.
Given a need to complete watershed modeling for the purpose of determining DPVs, the division of the
watershed into segments for the purpose of predicting loadings at the "pour point" into the estuary or
marine receiving waters should not be limited to simple hydrologic division of the watershed. This may
conflict with the premise of using a 12-digit HUC, but the segmentation process needs to take into
account predominant land uses for a segment, and those land uses that may be significantly different.
For example, urban areas, with high impervious surface cover and altered stream channels, are likely to
behave in a way that is distinctly different than less developed areas. Therefore, a simple model
37
-------
delineation of subwatersheds may not be suitable and some expert analysis and adjustment of the
segments would be more appropriate.
2. The impacts of urban environments should be considered.
Urbanized areas have a distinct influence on normal stream processes given their large areas of
impervious cover. In addition to changes in stream habitat, runoff from impervious surfaces as well as
municipal and industrial discharges may contribute to stream nutrient loads. For this reason, the SAB
recommends that large urbanized areas be given special consideration in any modeling approach that
might be used to generate DPVs.
3. Given that a complete uncertainty analysis cannot be accomplished, it is essential that, in all text
in the revised report where uncertainty is mentioned, readers are clearly told what is included
(excluded) in any uncertainty analysis undertaken or contemplated.
4. EPA should provide justification for the choice of the LSPC model and explain why it is the most
applicable model for this case.
The LSPC model is an updated version of the older HSPF program. While the model can be integrated
with a Geographic Information System (GIS), it is not a GIS-based approach. Numerous models exist
for watershed management, physical flow, and water quality modeling that may better utilize the
strengths of current GIS platforms. For example, the SWAT (Soil and Water Assessment Tool) model is
commonly used in the development of TMDLs and is run within a GIS. Given the complexity of
watershed modeling at the proposed scale and the complex nature of the problem being addressed, it
may be prudent to build watershed models that can take advantage of a wider array of GIS-based tools
and data for the current project and in future applications during implementation. This would not be a
significant increase in work load, as the development of the GIS and SWAT model would replace the
development of the LSPC model.
5. The time frame of modeling is important and should be linked to the response of the endpoints in
the receiving waters.
In the EPA presentation on development of DPVs, it was indicated that adjustments for seasonal effects
and flow levels are being considered. This is a very important consideration and the EPA is encouraged
to analyze available data in the context of seasonal changes in the watershed and for the differences
between baseflow and storm event conditions. Seasonal changes in the watershed may result from both
natural processes (e.g., biotic activity) and from anthropogenic factors (e.g., agricultural practices). The
differences in loadings seen during baseflow and storm events may be dramatic, with the majority of
loading of TN and TP coming during a few large storm events. This is particularly true for N and P
species associated with suspended sediment. Using an annual average value may grossly underpredict
the impact of large storm events. Therefore, EPA should evaluate the sensitivity of the selected
biological endpoints to the potential influences of shorter-term (e.g., days to weeks) events that may
result in high levels of TN and TP loading to determine if annual or seasonal averages are sufficient to
protect estuarine biota.
38
-------
6. In-stream/watershed P transformations should be considered in greater depth for streams, lakes
and canals.
Species/fractions of N and P are often a part of TMDL modeling. If DPVs are to be developed in
Florida, expressed as loads, and serve in a TMDL-like role, then DPVs might be expressed as nutrient
fractions (for a biotic estuarine water quality criterion). In the discussion of nutrients, EPA correctly
identifies the role of N species/fractions, but does not consider P species/fractions.
The dynamics of P in watersheds, lakes and canals is important to any effort to produce DPVs or similar
water quality criteria. Foremost is the need to recognize the mobility, reactivity and bioavailability of the
different P species: soluble reactive phosphorus (SRP); dissolved phosphorus (DP), which is the sum of
SRP and total hydrolysable P (THP); and total phosphorus (TP), which is the sum of DP and particulate
phosphorus (PP). These phases exist in natural waters in varying degrees that are dependent on
processes within the waterbody and on external inputs. Furthermore, transitions among forms occur
during transport within watersheds and within sediments in streams, lakes and canals. Although the
modeling effort will evaluate TP for DPVs, the transport and fate of P should not be oversimplified. A
brief overview of some of the extensive literature on P cycling and transformations within watersheds is
provided in Appendix B.
7. How are nutrients, especially P, from natural or geologic sources separated from anthropogenic
sources?
Further complicating apportionment and determination of DPVs is the issue of background values for
nutrients, especially P. Given that some areas of Florida have bedrock geology with high P
concentrations, understanding background is critical. In watersheds where high P loadings are the result
of natural factors, DPVs may not be applicable.
8. The continuum offresh to saline waters in going from watersheds to the receiving estuarine or
coastal marine waters must be considered in the process of determining DPVs.
In many instances, fresh water systems are P-limited with respect to nutrient balance and the potential
for the development of eutrophic conditions. The opposite is often the case for estuarine or marine
waters, where N and P can be co-limiting or N is the limiting nutrient. This raises the potential where the
application of upstream water quality standards that may be focused at reducing P inputs could be
protective of the watershed, but create a situation in the downstream brackish or saline receiving waters
that creates a nutrient imbalance (Elmgren and Larsson 2001; Paerl et al. 2004). This issue should be
recognized during development of DPVs and implementation of recent inland water criteria.
3.6.2 DPVs for South Florida Estuarine and Coastal Waters
Charge Question 6(b). Are the methods that EPA is considering for deriving downstream protection
values (DPVs) for marine waters in South Florida as described in Section 6.5 appropriate to ensure
attainment and maintenance of downstream water quality standards, given available data? Please
describe additional approaches and their advantages and disadvantages that EPA should consider
when developing numeric criteria to protect downstream marine waters in South Florida, given
available data?
Unlike for estuaries in other parts of the state, the EPA document is not proposing an upstream
apportionment of load reduction by stream segment because of the highly managed hydrology in South
Florida. Instead, the document proposes setting a protective load at the terminal reach of each tributary,
39
-------
i.e., at the point where the tributary empties into estuarine or coastal waters. The EPA document
discusses several schemes for allocating acceptable estimated nutrient loads among tributaries, including
allocation based on flow-weighted concentration, flow-only, or total load for each tributary. The Panel's
thoughts on various possible allocation schemes are presented in the previous section.
As discussed in section 3.4.1, the SAB recommends that nutrient criteria for canals be focused on
downstream, rather than instream, protection from nutrient impacts. The SAB has the following
additional comments:
1. Provide more information on how canals will be evaluated.
A number of primary canals empty directly into coastal waters, so it will be important to incorporate all
available data on N and P species for the terminal reach of these canals and to provide a more detailed
approach on how DPV criteria will be developed.
2. The time frame of modeling is important and should be linked to the response of the endpoints in
the receiving waters.
Given the wide variation in flow conditions for canals, the concentrations of nutrients in canal waters are
likely highly variable. Hence average nutrient concentrations in canal waters when released to estuarine
and coastal marine waters may not adequately represent the concentrations needed to protect receiving
waters. If additional information is not available for nutrient concentrations in the canals, discharge of
canal waters to the receiving waterbodies needs to take into consideration loading rates on a daily basis
that will ensure the receiving waterbodies meet their water quality standards.
3. In-stream/watershed P transformations should be considered in more depth for streams, lakes and
canals.
Although less is known about P transformations within the canals of South Florida than for streams, the
physical and chemical processes that control P transport within a watershed should be the same for
canals. Additional consideration, however, must be given to the special situations that result as a
function of the wide-ranging flow situations for the canal system. Furthermore, it is important to
understand the temporal parameters and their range of variability. These factors will determine, in part,
the mechanisms that are most important under different sets of flow conditions.
4. The continuum offresh to saline waters must be considered in the process of determining DPVs.
For canal waters discharging to estuarine and coastal marine waters, the issue of the continuum of fresh
to saline water is the same as discussed earlier in the Panel's response to charge question 6(a), above.
40
-------
REFERENCES
Baden S., C. Bostrom, S. Tobiasson, H. Arponen and P.-O. Moksnes. 2010. Relative importance of
trophic interactions and nutrient enrichment in seagrass ecosystems: A broad-scale experimental
assessment. Limnology and Oceanography 55:1435-1448.
Banse, K. 1977. Determining the carbon-to-chlorophyll ratio of natural phytoplankton. Marine Biology
41:199-212.
Bostrom, B., A.K. Pettersson and I. Ahlgren. 1989. Seasonal dynamics of a cyanobacterial dominated
microbial community in surface sediments of a shallow eutrophic lake. Aquat. Sci. 51:153-178.
Boyer, J. and H. O. Briceno. 2010. Written comments provided to the SAB Nutrient Criteria Review
Panel, December 2, 2010. Available at
http://yosemite.epa.gov/sab/sabproduct.nsf/C485AFF485895924852577F 100569C8F/$File/Boye
r+Letter+to+SAB+on+EPA+NNC+Approach+2010-12-02.pdf
Briceno, H., J. Boyer and P. Harlem. 2010. Proposed Methodology for the Assessment of Protective
Numeric Nutrient Criteria for South Florida Estuaries and Coastal Waters. Comments submitted
to the SAB Staff Office, December 6, 2010. Available at
http://yosemite.epa.gov/sab/sabproduct.nsf/3759224AAlDA4CAD852577F3005B533F/$File/Br
iceno+comments.pdf
Burkepile, D.E. and M.E. Hay. 2006. Herbivore vs. nutrient control of marine primary producers:
context-dependent effects. Ecology 87:3128-3139.
Burkholder, J.M., D.A. Tomasko and B.W. Touchette. 2007. Seagrasses and eutrophications. J. Exper.
Mar. BiolEcol. 350:46-72.
Carlson, P.R. and K. Madley. 2007. Statewide Summary for Florida. In: Handley. L., D. Altsman and R.
DeMay.(eds.). Seagrass status and trends in the northern Gulf of Mexico, pp. 99-114. USGS SIR 06-
5287. http://pubs.usgs.gov/sir/2006/5287.
Carrick, H.J. and A.D. Steinman. 2001. Variation in periphyton biomass and species composition in
Lake Okeechobee, Florida (USA). Archivfur Hydrobiologie 152:411-438.
Cloern, J.E. 1982. Does the benthos control phytoplankton biomass in south San Francisco Bay? Marine
Ecology Progress Series 9:191-202.
Cohen, R.H., P.V. Dresler, E.J. P. Phillips and R.L. Cory. 1984. The effect of the Asiatic clam,
Corbicula fluminea, on phytoplankton of the Potomac River, Maryland. Limnol. Oceanogr. 29:170-
180.
Conley, D.J., H. W. Paerl, R.W. Howarth, D.F. Boesch, S.P. Seitzinger, K.E. Havens, C. Lancelot and
G.E. Likens. 2009. Controlling Eutrophication: Nitrogen and Phosphorus. Science 323:1014-
1015.
R-l
-------
DeBusk, T.A. 2010. Comments on Numeric Nutrient Criteria for Florida's Southern Inland Flowing
Waters. Presentation to the SAB Nutrient Criteria Review Panel, December 13, 2010.
Dixon, K. 1999. Establishing light requirements for the seagrass Thalsasia testudinum: an example for
Tampa Bay, Florida. In: Bortone, SA (ed) Seagrasses: Monitoring Ecology Physiology and
Management, pp. 81-98. CRC Press, Boca Raton, FL.
Doering, P.H., R.H. Chamberlain, K.M. Donohue and A.D. Steinman. 1999. Effect of salinity on the
growth of Vallisneria americana Michx. from the Caloosahatchee Estuary, Florida. Florida
Scientist 62:89-105.
Elmgren, R. and U. Larsson. 2001. Nitrogen and the Baltic Sea: Managing nitrogen in relation to
phosphorus. The Scientific Word 1: 371-377.
Evans, D.J. and P.J. Johnes. 2004. Physico-chemical controls on phosphorus cycling in twolowland
streams. Part 1 — the water column. Sci. Total Environ. 329:145-163.
Evans, D.J., P.J. Johnes and D.S. Lawrence. 2004. Physico-chemical controls on phosphorus cycling in
two lowland streams. Part 2—the sediment phase. Sci. Total Environ. 329:165-182.
Fore, L.S. 2004. Development and Testing of Biomonitoring Tools for Macroinvertebrates in Florida
Streams. Prepared for Florida Department of Environmental Protection by Statistical Design,
Seattle, WA. 74p. Available at http://publicfiles.dep.state.fl.us/dear/labs/sas/sopdoc/sci_old.pdf
(accessed 01/15/11).
Golterman, H.L. 1995. The role of the ironhydroxide-phosphate-sulphide system in the phosphate
exchange between sediments and overlying water. Hydrobiol. 297:43-54.
Greening, H. 2010. Presentation on EPA's Proposed Approaches for Florida ENCs: Observations from
Tampa Bay. Presentation to the SAB Nutrient Criteria Review Panel, December 13, 2010.
Handley, L. D. Altsman and R. DeMay. 2007. Abstract. In: Handley. L., D. Altsman and R.
DeMay.(eds.). Seagrass status and trends in the northern Gulf of Mexico. pp.1-3. USGS SIR 06-
5287. http://pubs.usgs.gov/sir/2006/5287
Hauxwell, J., J. Cebrian, C. Furlong and I. Valiela, 2001. Macroalgal canopies contribute to eelgrass
(.Zostera marina) decline in temperate estuarine ecosystems. Ecology 82:1007-1022.
Hauxwell, J., J. Cebrian and I. Valiela. 2003. Eelgrass Zostera marina loss in temperate estuaries:
relationship to land-derived nitrogen loads and effect of light limitation imposed by algae. Mar.
Ecol. Prog. Ser. 247:59-73.
He, R. and R.H. Weisberg. 2002. Tides on the West Florida shelf. J. Phys. Oceanography 33: 465-477.
Heck, Jr., K.L. and J.F. Valentine. 2007. The primacy of top-down effects in shallow benthic
ecosystems. Estuaries and Coasts 30:371-381.
Hilton, J., M. O'Hare, M.J. Bowes and J.I Jones. 2006. How green is my river? A new paradigm of
eutrophication in rivers. Science of the Total Environment 365:66-83.
R-2
-------
House, W.A., 2003. Geochemical cycling of phosphorus in rivers. Appl. Geochem. 18:739-748.
Hu, C., Z. Chen, T.D. Clayton, P. Swarzenski, J.C. Brock and F.E. Muller-Karger. 2005. Assessment of
estuarine water-quality indicators using MODIS medium-resolution bands: Initial results from
Tampa Bay, FL. Remote Sensing of Environment 94:425-427.
Huang, B.Q. and H.S. Hong. 1999. Alkaline phosphatase activity and utilization of dissolved organic
phosphorus by algae in subtropical coastal waters. Marine Pollution Bulletin
39:205-211.
Hughes, A.R., K.J. Bando, L.F. Rodriguez and S.L. Williams. 2004. Relative effects of grazers and
nutrients on seagrasses: a meta-analysis approach. Mar EcolProgr Ser 282:87-99.
Jarvie, H.P., C. Neal, M.D. Jurgens, E.J. Sutton, M. Neal, H.D. Wickham, L.K. Hill, S.A. Harman, J.J.L.
Davies, J.J.L.,A. Warwick, C. Barrett, J. Griffiths, A. Binley, N. Swannack and N. Mclntyre.
2006. Within-river nutrient processing in Chalk streams: the Pang and Lambourn, UK. J. Hydrol.
330:101-125.
Jensen, H.S. and F.0. Andersen. 1992. Importance of temperature, nitrate, and pH for phosphate release
from aerobic sediments of four shallow, eutrophic lakes. Limnol.Oceanogr. 37:577-589.
Johnes, P.J. and R.A. Hodgkinson. 1998. Phosphorus loss from agricultural catchments:pathways and
implications for management. Soil Use Manage 14:175-185.
Koblinsky, C. J. 1981. TheA/2 tide on the West Florida Shelf. Deep-Sea Res. 28A: 1517-1532.
Kraemer, G.P., R.H. Chamberlain, P.H. Doering, A.D. Steinman and M.D. Hanisak. 1999. Physiological
Responses of Transplants of the Freshwater Angiosperm Vallisneria americana along a Salinity
Gradient in the Caloosahatchee Estuary (Southwestern Florida). Estuaries 22:138-148.
Latimer, J.S. and S.A. Rego. 2010. Empirical relationships between eelgrass extent and predicted
watershed-derived nitrogen loading for shallow New England estuaries. Estuarine, Coastal and
Shelf Science 90:231-240.
Lee, T.N. and D.A. Mayer. 1977. Low-frequency current variabilty and spin-off eddies on the shelf off
southeast Florida. J. Mar. Res. 35:193-220.
Lee, T.N., E. Williams, D. Wilson, E. Johns and N. Smith. 2002. Transport processes linking south
Florida coastal ecosystems, pp. 309-342, In: The Everglades, Florida Bay and Coral Reefs of the
Florida Keys: An Ecosystem Sourcebook.
Lewis, W.M. and J.H. McCutchan. 2010. Ecological responses to nutrients in streams and rivers of the
Colorado mountains and foothills. Freshwater Biology 55:1973-1983.
McCormick, P.V. and M.B. O'Dell. 1996. Quantifying periphyton responses to phosphorus enrichment
in the Florida Everglades: a synoptic-experimental approach. Journal of the North American
Benthological Society 15:450-468.
R-3
-------
McCormick, P.V., P.S. Rawlik, K. Lurding, E.P. Smith and F.H. Sklar. 1996. Periphyton-water quality
relationships along a nutrient gradient in the northern Everglades. Journal of the North American
Benthological Society 15:433-449.
McCormick, P.V., S. Newman, S. Miao, D.E. Gawlick, D. Marley, K.R. Reddy and T.D. Fontaine.
2002. Effects of anthropogenic phosphorus inputs on the Everglades. Pages 83-126 in: J. Porter
and K. Porter (editors). The Everglades, Florida Bay, and Coral Reefs of the Florida Keys. An
Ecosystem Handbook. CRC Press, Boca Raton, FL.
Melack, J.M.I995. Transport and transformations of P, fluvial and lacustrine ecosystems. In Phosphorus
in the global environment, ed. H. Tniessen, Chapter 15. West Sussex, England: John Wiley and
Sons Ltd.
Muller-Karger, F. E., C. Hu, S. Andrefouet and R. Varela. 2005. The Color of the Coastal Ocean and
applications in the solution of research and management problems. In: Remote Sensing of
Coastal Aquatic Environments: Technologies, Techniques and Application, R.L. Miller, C.E. Del
Castillo andB.A. McKee [Eds.], Springer, 101-127.
Owens, P.N. and D.E. Walling. 2002. The phosphorus content of fluvial sediment in ruraland
industrialized river basins. Water Res. 36:685-701.
Paerl, H.W. 2009. Controlling Eutrophication along the freshwater-Marine Continuum: Dual Nutrient
(N and P) Reductions are Essential. Estuaries and Coasts 32:593-601.
Paerl, H.W. and J. Huisman. 2008. Blooms like it hot. Science 320:57-58.
Paerl, H.W., L.M. Valdes, M.F. Piehler and M.E. Lebo. 2004. Solving problems resulting from
solutions: The evolution of a dual nutrient management strategy for the eutrophying Neuse River
Estuary, North Carolina, USA. Environmental Science & Technology 38: 3068-3073.
Paerl, H.W., L.M. Valdes, A.R. Joyner, B.L. Peierls, C P. Buzzelli, M. F. Piehler, S R. Riggs, RR.
Christian, J.S. Ramus, E.J. Clesceri, L.A. Eby, L.W. Crowder and R.A. Luettich . 2006a.
Ecological response to hurricane events in the Pamlico Sound System, NC and implications for
assessment and management in a regime of increased frequency. Estuaries and Coasts 29:1033-
1045.
Paerl, H.W., L.M. Valdes, J.E, Adolf, B.M. Peierls and L.W. Harding Jr. 2006b. Anthropogenic and
climatic influences on the eutrophi cation of large estuarine ecosystems. Limnology and
Oceanography 51:448-462.
Paerl, H.W., L.M. Valdes, A.R. Joyner and V. Winkelmann. 2007. Phytoplankton Indicators of
Ecological Change in the Nutrient and Climatically-Impacted Neuse River-Pamlico Sound
System, North Carolina. Ecological Applications 17:88-101.
Palandro, D., C. Hu, S. Andrefouet and F.E. Muller-Karger. 2004. Synoptic water clarity assessment in
the Florida Keys using diffuse attenuation coefficient estimated from Landsat imagery.
Hydrobiologia 530-531(l):489-493.
Pauly, D. 1995. Anecdotes and the shifting baseline syndrome of fisheries TREE 10:430.
R-4
-------
Ryan, R.J., A.I. Packman, and S.S. Kilham. 2007. Relating phosphorus uptake tochanges in transient
storage and streambed sediment characteristics inheadwater tributaries of Valley Creek, an
urbanizing watershed. J. Hydrol. 336:444-457.
Rydin, E. 2000. Potentially mobile phosphorus in Lake Erken sediments. Water Res. 34:2037-2042.
Rydin, E. And A.K. Brunberg. 1998. Seasonal dynamics of phosphorus in Lake Erken surface
sediments. Arch. Hydrobiol. Spec. Issues Advanc. Limnol. 51:157-167.
South Florida Water Management District (SFWMD). 2010. Canals in South Florida: A Technical
Support Document (April 28, 2010). SFWMD, West Palm Beach, FL.
Snyder, B., M. Barbour and E. Leppo. 1998. Development of a watershed-based approach for
biomonitoring of fresh surface waters in coastal Florida canal system. Prepared by TetraTech,
Inc., Owings Mills, MD under contract with Metro-Dade Environmental Resources
Management, 201 p.
Steinman, A., K. Havens, L. Hornung. 2002a. The Managed Recession of Lake Okeechobee, Florida:
Integrating Science and Natural Resource Management. Conservation Ecology 6:243-256.
Steinman, A.D., K.E. Havens, H.J. Carrick, and R. VanZee. 2002b. The past, present, and future
hydrology and ecology of Lake Okeechobee and its watersheds. Pages 19-37 in: J. Porter and K.
Porter (editors). The Everglades, Florida Bay, and Coral Reefs of the Florida Keys. An
Ecosystem Handbook. CRC Press, Boca Raton, FL.
Steinman, A.D., M. Luttenton and K.E. Havens. 2004. Sustainability of surface and subsurface water
resources: case studies from Florida and Michigan. Water Resources Update 127:100-107.
Stumpf, R.P., M.E. Culver, P.A. Tester, M. Tomlinson, G.J. Kirkpatrick, B.A. Pederson, E. Truby, V.
Ransibrahmanakul, and M. Soracco. 2003. Monitoring Karenia brevis blooms in the Gulf of
Mexico using satellite ocean color imagery and other data. Harmful Algae 2:147-160.
Stumpf, R.P., M.C. Tomlinson, J.A. Calkins, B. Kirkpatrick, K. Fisher, K. Nierenberg, R. Currier, and
T.T. Wynne. 2009. Skill assessment for an operational algal bloom forecast system. J. Mar. Syst.
76:151-161.
Thompson, J.K., J.R. Koseff, S.G. Monismith and L. V. Lucas. 2008, Shallow water processes govern
system-wide phytoplankton bloom dynamics: A field study. Journal of Marine Systems 74:153-
166.
Tomlinson, M.C., R.P. Stumpf, V. Ransibrahmanakul, E.W. Truby, G.J. Kirkpatrick, B.A. Pederson,
G.A. Vargo and C.A. Heil. 2004. Evaluation of the use of SeaWifs imagery for detecting
Karenia brevis harmful algal blooms in the eastern Gulf of Mexico. Remote Sensing of the
Environment 91:293-303.
U.S. EPA (Environmental Protection Agency) Science Advisory Board (SAB). 2010. SAB Review of
Empirical Approaches for Nutrient Criteria Derivation. EPA-SAB-10-006. Available at
http://www.epa.gov/sab.
R-5
-------
U.S. EPA (Environmental Protection Agency). 2010a. Methods and Approaches for Deriving Numeric
Criteria for Nitrogen/Phosphorus Pollution in Florida's Estuaries, Coastal Waters, and
Southern Inland Flowing Waters (November 17, 2010 Draft).
U.S. EPA (Environmental Protection Agency). 2010b. Using Stressor-response Relationships to Derive
Numeric Nutrient Criteria. EPA Office of Water, November 2010. EPA-820-S-10-001.
Valdes-Weaver, L.M., M.F. Piehler, J.L. Pinckney, K.E. Howe, K. Rosignol and H.W. Paerl. 2006.
Long-term temporal and spatial trends in phytoplankton biomass and class-level taxonomic
composition in the hydrologically variable Neuse-Pamlico estuarine continuum, NC, USA.
Limnology and Oceanography 51:1410-1420.
Verdonschot, P.F.M. 1987. Aquatic oligochaetes in ditches. Hydrobiologia 155:283-292.
Walker, N.D. and N.N. Rabalais. 2006. Relationships among satellite chlorophyll a, river inputs, and
hypoxia on the Louisiana continental shelf, Gulf of Mexico. Estuaries and Coasts 29:1081-1093.
Webster, P.J., G.J. Holland, J.A. Curry and H.R. Chang. 2005. Changes in tropical cyclone number,
duration, and intensity in a warming environment. Science 309:1844-1846.
R-6
-------
APPENDIX A. CHARGE TO THE PANEL
( har«e to the
Sci(-!K'<- \dvisory llmird
Nutrient < riteria Ke\k-\\ Panel
Methods and Approaches tor l)eri\ in;; Numeric ( riteria for ^irn>«eii,'Phosphorus
Pollution in Florida's Estuaries, Coastal Waters. anil Southern Inland Fhmin« V\ aU-rs
Background
In 2011. EPA will propose numeric criteria lor nitrogen'phosphorus pollution to protect
estuaries, coastal areas ami South Florida inland flowing waters thai have been designated Class
I, II and III1, as well as downstream protective values (DPVs)to protect estuarine and marine
waters, This is the second phase of a rulemaking effort as EPA has already established numeric
criteria for Florida's lakes, flowing waters2 and springs within the State of Florida, including
DPYs for lakes.
As part of this current effort to derive chlorophyll-a, total nitrogen (TN) and total phosphorus
(TP) criteria to protect estuarine and coastal water bodies. EPA must also develop additional
criteria to assure that upstream criteria w ill meet standards established for downstream estuarine
and marine waters in Florida, These DP Vs. will supplement the existing inland water 1"N and TP
stream criteria, if the applicable DP\ is more stringent.
()%eralf these numeric criteria are being developed to translate and implement Florida's existing
narrative nutrient criterion, to protect the designated use that Florida has previously set for these
waters, at Rule 62-302,530{47){b), F.A.C, which provides that
"In no ea.if .shall nutrient concentrations of a body of water he altered so as to cait.st an
imbalance in nutuial populations of aquatic flora or fauna. "
Under tlx Clean Water Act and EPA's implementing regulations, these numeric criteria must lie
based on sound scientific rationale and reflect the best available scientific knowledge.
EPA has previously published a series of peer review ed technical guidance documents' to
develop numeric criteria to address nitrogen'phosphorus pollution in different water body types.
EPA recognizes that available and reliable data sources for use in numeric criteria development
vary across cstuarine and coastal waters in Florida and flowing waters in South Florida, hi
addition, scientifically defensible approaches for numeric criteria development have different
requirements that must be taken into consideration in the context of the specific application and
' Class I Potable water supplies. Class II Shellfish propagation or harvesting, Class III Recreation, propagation, and
maintenance of a healthy. \\ ell-b«ilunecd population of fish and wildlife
1 In the November 2010 rulemaking. EPA did net establish numeric criteria tot inland flowing waters in South
Florida Far the purpose of* this effort, HP A has distinguished South Florida as those areas south of Lake
Okeechobee and the Culoosahatehee Rivet watershed to the west oi"Lake Okeechohec and the St Lucie watershed
to the east of Lake Okeechobee
!f S FPA 2000 \utrient Criteria 1 eehnieal Guidance Manual Rn ers
-------
APPENDIX A. CHARGE TO 'HIE PANEL
available information. This document describes the scientific approaches EPA is considering the
derivation of numeric criteria to address nitrogen phosphorus pollution in Florida estuarine and
coastal waters, and inland flowing waters in South Florida, given the available data currently
available.
Review Document; The SAB is asked to review the draft document. Methods ami Approaches
far Deriving Xumertc < 'ntena Jar Xitrogcn Phosphorus Pollution in Florida's Estuaries,
Coastal Haters, and Southern Inland Flowing Haters, and respond to the following charge
questions.
Charge Questions:
!, General Approach
a) EPA has introduced a general conceptual model in Chapter 2, including the selection of
assessment endpoinl ami indicator variables. What is your perspecti ve of the general conceptual
model?
b) EPA has delineated the State of Florida into 4 general categories of waters Florida estuaries,
Florida coastal waters. South Florida inland (lowing waters, and South Florida marine waters
for purposes of considering approaches to numeric nutrient criteria development. Are these
categories appropriate and scientifically defensible? (Note that the details of segmentation of
waters within these categories is addressed in subsequent charge questions.)
Florida Estuaries (Chapter 3)
KPA is considering three approaches; (1) reference conditions. (2) stressor.response models, and
(3) water quality simulation modeling that could he used independently or in combination to
develop numeric criteria for estuaries (exclusive of marine waters in South Florida that are
covered in Chapter ?}. Estuarine waters are defined as "a part of a river or stream or other body
of water that has an unimpaired connection wish the open sea and where the sea water is
measurably diluled with fresh water derived from land drainage"1.
2. Estuaries
a) /Yrc the data sources identified appropriate for use in deriving numeric criteria in
Florida's estuaries (as discussed in Sections 2,4 and 3.2)? Is the SAB aware of additional
available, reliable data that KPA should consider in delineating estuaries or deriving
criteria for estuarine w aters? Please identify the additional data sources.
b) Arc the assessment etidpoiirts identified in Sections 2.3 and 3,2 (healthy seagrass
communities; balanced ph\toplankton biomass and production; and balanced faunal
41. S FPA, 2tW Fstunrics ami C lean Wotcis Act at 3*X>, littp \uitci cp;i gov ,'tvnc oceh.''icp LIP.ci'ro i,;icces.eJ
! 1/1/2010)
A-2
Page 2 of 7
-------
APPENDIX A, CHARGE TO THE PANEL
communities) appropriate to translate Florida's narrative nutrient criterion (as cited
above) into numeric criteria lor Florida's estuaries, gi\en current!} av ailable data? Doe*
the SAB suggest modification or addition to these assessment endpomts? A literature
review of endpoints considered can be found in Appendix B
e) EPA describes potential approaches in Section 3.3 (reference conditions, stressor
response relationships, and watet quality simulation models) for deriving numeric criteria
in Florida's estuaries. Compare and contrast the ability of each approach to ensutc the
attainment and maintenance of natural populations of aquatic flora and fauna lor different
types of estuaries, given currently available data'.'
A-3
Page 3 of 7
-------
APPENDIX A, CHARGE TO THE PANEL
Florida Coastal Waters (Chapter 4)
KPA is considering ;t reference-based approach to derive numeric crileria for most of Florida's
coastal waters. Coastal waters are defined as marine waters up to three nautical miles* front
shoie^, Specifically, KPA is considering the use ofa remote sensing method to de\elop numeric
Chlk^-a crileria tor the Northwest Gull'Coast. West Gulf Coast, and Atlantic Constat areas, Due
to interference from colored dissolved organic matter and bottom reflectance on satellite
measurements. KPA is no! considering the derivation of Chlns-a criteria using remote sensing
data in coastal v\ aters from Apalaehieola Ray to S u\v annee Ri\ or (Big Bend) and South Florida,
3. ( cnlstill \\ aters
a) Are the data sources identified in Sections 2 4, 4.1.1 and 4,2 appropriate for use in
deriving numeric ci iteria in Florida's coastal waters? Is the SAB aware of additional
a\ailabk\ reliable data that KPA should consider in delineating coastal waters or derix ing
criteria for coastal waters'? Please identify the additional data sources,
b) Is the assessment endpoint identified in Section -4.2 (chtorophyll-a to measure balanced
phytophaiklon biomass and production) appropriate to translate Florida's narrative
nutrient criteria (described above) into numeric criteria for Florida's coastal waters, given
currently available data'? Does the SAB suggest moditlcation or addition to this
assessment endpoint?
c) Does the approach KPA describes in Section 4,2 appropriately apply remote sensing data
to ensure attainment and maintenance of balanced natural populations of aquatic flora and
fauna in Florida's coastal waters? If not. please provide an alternate methodology
utilizing available reliable data and toots, and describe the corresponding advantages and
disadvantages.
5 Based on the ('lean Water Act definition of "Waters of the United States"
Page 4 of ?
-------
APPENDIX A, CHARGE TO THE PANEL
,South Florida Inland Mowing Waters (("hauler 5)
EPA is considering a rcfcrenec-based approach to derive numeric criteria for South Florida
inland flowing waters using least-disturbed sites that support balanced natural populations of
aquatic flora and fauna. Alternative methods of criteria derivation lor inland flowing waters
include stressor-rcsponsc relationships between chlorophyll-a and TN and TP, and a
distributional approach using alt sites. South Florida inland flowing waters are defined for this
effort as free-How ing. predominantly fresh surface water in a tie lined channel and include,
streams, rivers, creeks, branches, canals, freshwater sloughs, and other similar water bodies
located in the South Florida nutrient watershed region''.
4, South Florida Inland Flowing Watets
A) Arc the data sources identified in Section 2.4 and 5,4 appropriate for use in deriv ing
numeric criteria in South Florida's inland Hewing waters (as discussed in Chapters 2 and
5)",* Is the SAB av\ are of additional available, reliable data that FPA should consider in
delineating or deriving criteria for South Florida's inland flow ing w at era? Please identify
the additional data sources,
b) Are the assessment endpoints identified in Section 5.4 (balanced faunal communities, i.e.,
aquatic niacroinveriebrates. and balanced phytoplankton biomass and production)
appropriate to translate Florida's narrative nutrient criteria (described above) into
numeric criteria for South Florida's inland flowing waters, given currently available data?
Does the SAB suggest modilication or addition to these assessment ondpoints?
c) EPA describes two approaches in Section 5.4 (reference conditions and stressor-response
relationships) for deriv ing numeric criteria in South Florida inland flowing waters.
Compare and contrast the ability of each approach to ensure attainment and maintenance
of balanced natural populations of aquatic flora and fauna in different types of flowing
water or geographical areas, given currently available data?
ft The South Fkmda nutr.ent watershed legion is the atea ,sr>uth of! ake I >keeehobec and the >.'ahn'sahatchee Riv er
vMtetshcJ to the west ot lake Okceehobei and the St Lucie watershed to the et of Lake i %eechobee E)l\ i\ not
deriving criteria that Mould apply to wakrs located on tin. Seminole Indian Reservation or the Nhecosukee Indian
Reservation. waters located in storm water treatment areas f STAsl. wetlands, or marches, or t 'lass IV canals FRY is
also net establishing new TP criteria for the Everglades Protection 'Irea iF.U'Ai m deference tu the Evcigtades
borever Act (Hi* AV
A-5
I\ige S of 7
-------
APPENDIX A, CHARGE TO THE PANEL
South Honda Marine Watt rs (C hapter 5S
EPA is considering a reference-based approach lo derive numeric criteria in South Florida
marine waters using least-disturbed sites that support balanced iKtUw.il populations of aquatic
flora and fauna. South Florida marine waters include estuarine and coaxial waters extending three
nautical miles offshore Fsluarine and coastal waters in South Florida are considered together
because the watershed based approach for delineating water bodies is Jess suited in South Florida
due to its highly managed inland flows and open-\\ ater dominated systems (i.e.- Florida Bay and
Keys)
5. South Florida Murine Wafers
a) Are the data sources identified in Section 2.4 and 5.5 appropriate for use in deriving
numeric criteria in South Florida's marine waters (as discussed in Chapters 2 and 5)'* Is
the SAB aware of additional available, reliable data that EPA should consider in
delineating or deriving criteria for South Florida's marine \\ aters? Please identify the
additional data sources,
b) FPA describes t\t o methods in Section 5.6 for using a reference condition approach for
deriving numeric criteria in South Florida marine waters (least-disturbed sites or
bionomial test). Compare and contrast the ability of"each approach to ensure
attainment and maintenance of balanced natural populations of aquatic flora and fauna
in South Florida marine waters. ghen currently available data?
A-6
Page 6 of?
-------
APPENDIX A, CHARGE TO THE PANEL
Downstream Protection Values for Florida Estuaries and South Florida Marine Waters
(( liapter 6)
The approach that EPA is considering for developing stream DPY criteria is to adjust upstream
limits on TN and TP loading rates thai arc needed to support balanced natural populations of
aquatic flora and fauna in downstream cstunrine waters. The loading limits will be determined as
part of the criteria development effort for estuarine and matine waters (as described in Chapters
3 and 5) and are scaled based on the average streaniilovv entering the estuary to determine
criteria for TN and TP concentrations in streams as thev discharge into estuaries or marine
waters.
DPVs can be determined for upstream reaches within watersheds by accounting for expected loss
or permanent retention of 'FN and TP within the stream network. Because of the complexities
associated with the managed Hows in South Florida inland flowing waters (hat are covered in
Chapter 5, the fraction of TN or TP from the upstream tributary reach that eventualh flows into
the marine waters in South Florida cannot be estimated or predicted. Therefore. EPA is
considering expressing DPYs at the terminal reach of the tributary into a South Florida estuary as
protective concentrations or. alternatively, protective loads.
6. Downstream Protection Values tor Florida Estuaries ami South Florida Murine Waters
a. Are the methods EPA is considering for deriving downstream protection values
(l)PVs) lor estuaries (excluding marine waters in South Florida) as described in
Section 6,1-6.4 appropriate to ensure attainment and maintenance of downstream
water quality standards, given available data? Please describe additional
approaches and their adv antages and disadvantages that FPA should consider
when developing numeric criteria to protect these downstream esutarine waters
(excluding marine waters in South Florida), given available data?
b. Ate the methods that FPA is considering for deriving downstream protection
values (DPYs) for marine waters in South Florida as described in Section. 6.5
appropriate to ensure attainment and maintenance of downstream water quality
standards, given available data? Please describe additional approaches and their
advantages and disadv antages that FPA should consider when developing
numeric criteria to protect downstream marine waters in South Florida, given
available data?
A-7
Page 7 of?
-------
APPENDIX B: Phosphorus Transport and Fate in Freshwater Systems
While streams are often viewed as simply a transfer mechanism for P_ recent woik has. investigated
processes that occur during transport. There are mechanisms that transform Is w ithin different physico-
chemical fractions within the stream channel (Melack, 1995; livuns and Johnes, 2004; If vans el al,.
2004). and the speciation of soluble P phases and fractionation of P are critical for any evaluation of
transport or retention within a watershed. Various processes transform P including sorption, co-
precipitation, and redox reactions (e.g.. House 2003), and SRP interacts with stream sediments Stream
sediments act as both sinks mid sources for SRP within the stream depending on the SRP concentration
in the stream water and may change both temporally and spatially w ithin a watershed (e.g.. Jarvie et at.
2006; Ryan et ah, 2007). This would suggest that TP A should evaluate existing data sets with regard to
SRP and TP concentrations.
hi comparing rural vcreus urbanized watersheds. Owens and Walling (2002) found that PP increased in
stream sediments receiving point source discharge high in SRP. and that PP (inorganic and organic) may
he the most significant mechanism for P transport. Up to 20% of the PP in stream .sediment is likely to
be easily bioavailable as inorganic P phases dominate. These mechanisms may also be active in take or
canal sediments. Given the short-term bioavailability of some fraction of the PP. it is important to
evaluate TP in the context of SRP. DP and PP with some evaluation of the immediacy of the impact of
each fraction.
Phosphorus retention within watersheds is typically dominated b> calcite co-precipitation within bed
sediment and physical trapping of sediment b\ reduction of flow velocity. I uke sediments may act as
both sinks and sources for P cycling, with a large fraction of the inorganic P in surface sediments in
equilibrium with the water column (Ooiterman 1995). The c\cling of P is most prevalent in stratified
lakes with anoxic hypoltmnion, but significant cycling of P also occurs from ovie sediments (Bostrom ct
al., 19X9; Jensen and Andersen, 1992; Rydin and Hranbcrg. 1998) found in nearshore environments,
stream sediments and likely in canals.
P mobilization occurs under both ovic and anoxic conditions, and exchangeable and Fe-hound P arc
generally mobile (Rvdm 2(XX)), Organic-associated P is about (Wo mobile, with greater mobility in
anoxic sediments. P associated with Al and Ca is immobile and may be considered permanently bound.
P release from aerobic sediments may deplete the Fe-bound P despite Fe remaining in the solid phase
(Jensen and Andersen 1992). The release process involves a complex relationship between nitrate
concentrations and microbial activity resulting in seasonal effect of increasing sediment P retention
during w inter w itli subsequent release during late summer and autumn. Biota also play a role in P
cycling in lake sediments (e.g., bioturbation. rooted maerophvtes that alter the sediment
biogeoehemistry). 'Hie likely lack of available data on the fractionation of P between the various
physico-chemical phases will limit a detailed evaluation; however, it is important that modelling of P
transport include some recognition of the biogeochcmieal processes involved in P e\ cling.
13-1
-------
APPENDIX C. OPTIONS FOR DISTRIBUTING LOADS
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
FEB 1 8 2011
OFFICE OF
MEMORANDUM water
SUBJECT: Science Advisory Board Nutrient Criteria Review Panel Request for Information
FROM: Elizabeth Behl, Director
Health and Ecological Criteria Division
Office of Science and Technology
TO: Stephanie Sanzone, Designated Federal Official
Science Advisory Board
This memorandum is in response to the request for information from the U.S.
Environmental Protection Agency Science Advisory Board (SAB) on the proposed methods and
approaches to derive numeric criteria protective of downstream estuaries in Florida. In the SAB
teleconference cail on February 7,2011, we indicated that there were several approaches that
could be used to allocate nutrient loads among stream reaches within an estuanne watershed tor
the purpose of computing downstream protective values (DPVs). We provided one of those
approaches to the Panel as an illustrative example at the SAB meeting in December (Option A
below). EPA also is considering several other options, described below, including two (Options
C and D) which take into consideration initial feedback EPA received from the SAB in
December 2010.
Options for Distributing Loads
Option A: Distribute Load in Proportion to Flow. This option would distribute the watershed
load among stream reaches according to the fraction of the total watershed discharge contributed
by that reach, thereby addressing the fact that higher nutrient loading is often associated with
higher freshwater flow. In developing this option, EPA also considered in-stream processing of
nutrients, computing the aggregate loss and/or retention of nutrients within the stream network.
Criteria derived using this approach could be higher for streams for which a significant quantity
of transported nutrients are lost or retained before reaching estuaries. EPA described this
approach in the Methods and Approaches document submitted to the EPA SAB tor review.
Option B: Distribute Load in Proportion to Area. This option would distribute the watershed
load among stream reaches according to the fraction of the total watershed area that is drained
via that reach. DPV concentrations could be computed by dividing the loading limit for each
reach by the average freshwater discharge from the respective reach. EPA does not intend to
consider this approach further, recognizing that it does not take into account differences in
freshwater yield (i.e., runoff per area), an important factor affecting nutrient transport from
watersheds. Freshwater yield varies significantly among watersheds due to both natural and
PRO^
Interne! Address (URL) • http://www.epa.gov
Recycled/Recyclable • Printed with Vegetable Oil Based Inks on 100% Postconsumer, Process Chlorine Free Recycled Paper
c-i
-------
APPENDIX C. OPTIONS FOR DISTRIBUTING LOADS
anthropogenic causes. For example, a watershed with high relief (i.e.. slope) or significant
impervious surfaces may have higher freshwater yield than a low-relief, forested watershed.
DPV criteria derived using this approach could be much lower for high-runoff watersheds, and in
some cases could be unreasonably low.
Option C: Distribute Load In Proportion to Natural or "Background" Loading. This
option distributes the load among terminal reaches in a watershed in proportion to an estimate of
the loading that would occur from each watershed in the absence of anthropogenic influence,
which we refer to as "background loading." These estimates could be derived using the Loading
Simulation Program in C++, or I.SPC. This option would enable EPA to consider a diverse
array of environmental information (such as slope, natural land cover type, soil types, local
rainfall) in deriving numeric criteria. As an example, if the watershed TN or TP loading rate for
an estuary is, in aggregate, 20% more than the estimate of background loading, the watershed
load distributed to each terminal stream reach would be 20% more than the respective
background load. DPV criteria could be computed by dividing the distributed loading by the
average freshwater discharge. Since the estimates of these loads consider hydrological and other
landscape factors, the DPV estimates will also reflect these factors. DPV criteria for upstream
reaches could be determined recursively using the same process. DPV criteria derived using this
approach limit loading to estuaries to the watershed load and DPVs criteria reflect a range of
natural factors that impact nutrient concentrations in watersheds.
Option D: Distribute Load in Proportion to Existing Loading. This option distributes the
watershed load among the terminal reaches in a watershed based on both the average background
loading (as described above) and average existing loading (e.g., 1997-2009), recognizing that the
difference between the estimate of existing loading and background loading is an estimate of
anthropogenic loading. For example, if the aggregate watershed loading to the estuary is 30%
less than the current loading rate. DPV criteria would be computed based on 100% of the
background loading plus a fraction of anthropogenic loading, such that the total loading is equal
to the watershed load. DPV criteria would be computed by dividing the divided load by the
average freshwater flow. DPVs for unmodified watersheds with low nutrient yields would be
relatively low, whereas DPVs for sub-watersheds with developed or agricultural lands would be
expected to be higher, but lower than existing concentrations. DPV criteria derived using this
approach limits loading to estuaries to the watershed load and DPVs reflect both natural and
anthropogenic factors that impact nutrient concentrations in watersheds.
C-2
------- |