EPA-600/R-08/004 | January 2008 | www.epa.gov/ord
United States
Environmental Protection
Agency
An Approach for Developing
Numeric Nutrient Criteria for
a Gulf Coast Estuary
Office of Research and Development
National Health and Environmental Effects Research Laboratory, Gulf Ecology Division
-------
EPA/600/R-08/004
January 2008
An Approach for Developing
Numeric Nutrient Criteria for a
Gulf Coast Estuary
Contributing Authors
James D. Hagy III, Ph.D.
Research Ecologist
U.S. Environmental Protection Agency
Office of Research and Development
National Health and Environmental Effects Research Laboratory
Gulf Ecology Division
Gulf Breeze, FL
Janis C. Kurtz, Ph.D.
Research Ecologist
U.S. Environmental Protection Agency
Office of Research and Development
National Health and Environmental Effects Research Laboratory
Gulf Ecology Division
Gulf Breeze, FL
Richard M. Greene, Ph.D.
Branch Chief, Ecological Dynamics and Effects Branch
U.S. Environmental Protection Agency
Office of Research and Development
National Health and Environmental Effects Research Laboratory
Gulf Ecology Division
Gulf Breeze, FL
U.S. Environmental Protection Agency
Office of Research and Development
National Health and Environmental Effects Research Laboratory
Gulf Ecology Division
1 Sabine Island Drive, Gulf Breeze, FL 32561
-------
Disclaimer
This report provides information to States, Indian tribes (hereafter tribes), and other authorized
jurisdictions for the purpose of assisting development of numeric nutrient criteria for adoption into
water quality standards under the Clean Water Act (CWA). Under the CWA, States and tribes are to
establish water quality criteria to protect designated uses. State and tribal decisionmakers retain the
discretion to adopt approaches on a case-by-case basis that differ from the approaches used in this
report when appropriate and scientifically defensible. Although this report provides hypothetical
numeric nutrient criteria that would serve to protect resource quality and aquatic life, it does not
substitute for the CWA or U.S. Environmental Protection Agency's (EPA's) regulations nor is it a
regulation itself. Thus, it cannot impose legally binding requirements on EPA, States, tribes, or the
regulated community and may not apply to a particular situation or circumstance.
Citation
Hagy, J. D., J. C. Kurtz, and R. M. Greene. 2008. An approach for developing numeric nutrient criteria
for a Gulf coast estuary. U.S. Environmental Protection Agency, Office of Research and Development,
National Health and Environmental Effects Research Laboratory, Research Triangle Park, NC. EPA
600R-08/004. 44 pp.
Cover photo by James Hagy
-------
Table of Contents
List of Tables iv
List of Figures v
Executive Summary vii
1. Introduction 1
1.1 Background 1
1.2 Study Site 3
2. Methods 7
2.1 Water Quality Data for Pensacola Bay 7
2.2 Ecoregional Water Quality Characterization 7
2.3 Aerial Submerged Aquatic Vegetation Surveys 8
2.4 Watershed Characteristics and Land Use 9
3. Results and Discussion 11
3.1 Ecoregional Comparisons and the Percentile Approach 11
3.2 Algal Biomass, Productivity, and Community Composition 14
3.3 Preventing Hypoxia 16
3.4 Protecting Seagrass Habitats 18
3.5 A Watershed Perspective 20
3.6 Nutrient Criteria 22
4. Summary and Conclusions 27
5. References 29
Appendix A: Observations in EPA's Level III Ecoregion 75 33
Appendix B: Observations in Florida's Level IV Ecoregion 75a 34
Appendix C: List of Acronyms and Abbreviations 35
-------
List of Tables
Table 1. Sources of water quality data used in this study 7
Table 2. Ecoregional water quality conditions computed for U.S. Environmental Protection
Agency's Level III ecoregion 75 and Florida's Level IV ecoregion 75a 14
Table 3. Seasonal median chlorophyll-a in Pensacola Bay surface waters and its
interquartile range 15
Table 4. Plankton community respiration rate and sediment oxygen demand in the Pensacola Bay
system 17
Table 5. Area of submerged aquatic vegetation beds in the Pensacola Bay system circa
1950 and in 1960, 1980, 1992, and 2003 19
Table 6. Suggested nutrient and nutrient-related water quality criteria for Pensacola Bay
resulting from the weight-of-evidence approach described herein 25
IV
-------
List of Figures
Figure 1. Locations of stations sampled during two major studies of water quality conditions
in the Pensacola Bay system and a 2003-to-2004 study of benthic and planktonic metabolism 3
Figure 2. A conceptual model illustrating historic, ecologic, and socioeconomic factors of
potential importance for developing and evaluating water quality management options for
Pensacola Bay 4
Figure 3. The locations of water quality monitoring stations visited under the Environmental
Monitoring and Assessment Program during 2001 and 2002 and the Florida Inshore Monitoring
and Assessment Program during 2000 through 2004 8
Figure 4. Procedure for computing statistical comparison values for estuaries 8
Figure 5. Distribution of summer estuary median surface water quality values by salinity zones
for estuaries in Level 111 ecoregion 75 and Level IV ecoregion 75a 13
Figure 6. Frequency histogram of surface chlorophyll-a concentrations in mesohaline waters
of the Pensacola Bay system from 1996 to 2001 15
Figure 7. Observations of bottom dissolved oxygen in the Pensacola Bay system during the
summers of 1996 through 1999 indicating the locations and frequency of observations less than
2.0 mgL'1 16
Figure 8. Coverage of submersed aquatic vegetation in Pensacola Bay circa 1950 and in 1992 18
Figure 9. Land cover in the Pensacola Bay watershed circa 2001 21
Figure 10. Flow diagram illustrating the logical path followed to determine an appropriate approach
for determining nutrient criteria for Pensacola Bay 23
-------
Acknowledgments
This manuscript would not have been possible without the vision and sustained efforts of a
large number of people at the Gulf Ecology Division, whose work over more than a decade provided
much of the data available to characterize water quality and ecological processes in Pensacola Bay.
Much of this work has been published separately and cited herein. We thank Mike Lewis and Richard
Devereux for sharing unpublished data and Linda Harwell for assistance in obtaining data from the
National Coastal Assessment and Florida's Inshore Monitoring and Assessment Program. Ed Decker
helped us better understand how regulatory mechanisms function under the Clean Water Act. Ed
Dettman, Walt Nelson, and Cheryl Brown provided helpful comments at several stages of manuscript
preparation. We appreciate the significant contributions of several individuals who edited the
document and developed its final layout. These include John Barton and Keith Tarpley of NHEERL's
National Information Management and Site Support Staff and Barbra Schwartz and others at the
National Center for Environmental Assessment.
This is contribution no. 1322 of the Gulf Ecology Division of the U.S. Environmental Protection
Agency's National Health and Environmental Effects Research Laboratory.
VI
-------
Executive Summary
Nutrient enrichment is currently a major
cause of impaired water quality and degraded
ecosystem health in coastal marine ecosystems
such as estuaries. The Clean Water Act (CWA)
legislation provides much of the legal framework
for water quality management. Effective
management of nutrient enrichment in estuaries
under the CWA presently is inhibited by the lack
of numeric criteria for nutrients and nutrient-
related water quality indicators. We describe and
demonstrate a manageable, science-based
approach for developing water quality criteria for
nutrients (nitrogen and phosphorus), chlorophyll-
a, dissolved oxygen (DO), and water clarity for a
single estuary, Pensacola Bay, FL, with which we
were very familiar. As a component of the
approach, we modified the U.S. Environmental
Protection Agency's (EPA's) percentile approach
to compare water quality indicators in three
salinity zones of Pensacola Bay with comparable
values from other estuaries. We also conducted a
broad-based quantitative assessment of both
nutrient-related conditions and processes in the
bay and nutrient inputs to the bay, which
contribute to a logical process leading to
determination of nutrient criteria.
To apply the percentile approach, it is
necessary to identify a class of water bodies,
estuaries in this case, that can be compared
reasonably. We defined a class of estuaries on the
basis of EPA's Level 111 ecoregions and Florida's
Level IV ecoregions. Within these schemes,
Pensacola Bay is located within ecoregion 75 and
ecoregion 75a, respectively. Estuaries were
defined according to the National Oceanic and
Atmospheric Administration's coastal assessment
framework, which identifies estuarine drainage
areas. We assembled median summer surface
water quality data for 28 estuaries in Level 111
ecoregion 75 and 7 estuaries in Level IV
ecoregion 75a. Three salinity zones (<5, 5 to 18,
and >18) also were defined so that water quality
in each salinity zone of Pensacola Bay could be
compared with water quality within
corresponding salinity zones in other estuaries.
Chlorophyll-a in Pensacola Bay was comparable
with the median for ecoregion 75 and was
somewhat higher than other systems in ecoregion
75a. Nutrients and water clarity in oligohaline
water were also similar to the median. In
contrast, water clarity and nutrient concentrations
in mesohaline and higher salinity waters were in
the best 25% of estuaries (>75th percentile for
water clarity and <25th percentile for nutrients).
Our analysis of major nutrient-related
water quality concerns in Pensacola Bay found
that seasonal hypoxia and seagrass loss were the
major issues. Seasonal hypoxia affected an
average of 25% of the bottom area in recent
studies. Seagrass loss is extensive in the system.
About 95% of seagrass habitat was lost between
1950 and 1980. Almost no recovery was
observed in surveys in 1992 and 2003. Yet, other
indicators of ecosystem function point to lower
nutrient-related impacts. Nitrogen and
phosphorus loading rates to the system are
moderate. Phytoplankton productivity was
moderate (320 g carbon m"2 year"1) as was
plankton and benthic metabolism. Nutrient
concentrations were generally low, and
phytoplankton were nutrient limited. Studies
indicate that hypoxia may result significantly
from a natural propensity toward oxygen
depletion, whereas the extent of seagrass loss
may be a legacy of formerly more degraded
water quality conditions.
Nutrient yields from Pensacola Bay's
largely forested watershed are low (262 kg
nitrogen km"2 year"1 and 21.8 kg phosphorus km"2
year" ). Concentrations of nitrogen in rainfall are
also relatively low (14 jiM). However, rapid
population growth in the coastal counties of the
watershed may cause increased nutrient loading
in the future. We conclude that water quality
criteria for nutrients and nutrient-related water
quality measures could be based reasonably on
VII
-------
currently observed conditions because evidence
that more stringent criteria are scientifically
defensible, necessary, or even achievable, is
lacking. Hypothetical criteria resulting from
application of this approach are presented.
The approach to criteria development that
we describe could provide a broad template for
analyses in other systems. It would be most
useful for systems that, like Pensacola Bay, do
not have clear evidence of water quality or water
quality-related ecological impacts resulting from
anthropogenic nutrient enrichment. Conversely,
systems with very high nutrient inputs resulting
from point sources, landscape changes, or other
anthropogenic causes may require substantial
nutrient reductions. Although elements of our
analytical approach may be useful as part of an
approach for determining criteria for these
systems, additional elements will be needed to
determine reduced levels for nutrient
concentrations and nutrient-related water quality
measures that are appropriate to the system and
protective of designated uses.
VIII
-------
1. Introduction
1.1 Background
The U.S. Congress responded to the
desire of the American people to aggressively
protect U.S. waterways from pollution with
passage of the 1972 Federal Water Pollution
Control Act and subsequent amendments, which,
since the 1977 amendments, have been known
collectively as the Clean Water Act (CWA; U.S.
House of Representatives 2000). The CWA put
into place a water quality standards process,
within which ambient water quality criteria are
centrally featured. Ambient water quality criteria
are concentration limits determined such that the
legally identified or "designated" human and
aquatic life uses of a water body are not impaired
by the target pollutant. Since passage of the
CWA, standard approaches have been developed
for establishing aquatic life criteria for toxic
chemicals, generally utilizing experimentally
determined acute and chronic laboratory toxicity
data for multiple species (Stephan et al. 1985).
Management actions enabled by the CWA have
successfully eliminated the most egregious toxic
impairments.
Nitrogen and phosphorus are major plant
nutrients that occur naturally in varying amounts
and are necessary for normal ecological function
of aquatic ecosystems. Nutrients become
pollutants, however, when concentrations or
loadings are increased excessively.
Anthropogenic nutrient enrichment often causes
eutrophication (Nixon 1995) and a related
spectrum of usually adverse environmental
impacts. Coastal nutrient enrichment has
increased during the past half century and has not
been curtailed effectively by water quality
management under the CWA (NRC 2000). In the
United States, water quality impairment is
tracked by the U.S. Environmental Protection
Agency (EPA) via a list of impaired waterways
known as the Section 303(d) list, referring to the
relevant section of CWA legislation. Nutrient
enrichment or nutrient-related effects (e.g.,
oxygen depletion, algal growth, harmful algal
blooms) combine as the leading cause for
inclusion on the Section 303(d) list (U.S.
EPA/OWOW, public communication). Global
trends in coastal nutrient enrichment are similar
in many regards to those in the United States
(GESAMP2001).
In 1998, EPA expressed its view (U.S.
EPA 1998) that establishing numeric criteria for
nutrients and related water quality indicators
(e.g., chlorophyll-a) was an important step
toward reducing nutrient impacts. Others have
expressed a similar perspective. A survey of
water quality managers for 18 U.S. estuaries
found that one of the major impediments to
effective management of eutrophication was a
lack of regulatory authority to require nutrient
reductions, including a lack of numeric nutrient
criteria (NRC 2000). In the absence of numeric
criteria, determination of water quality
impairment generally is based on narrative water
quality criteria. Commonly referred to as "free-
froms," narrative criteria describe in qualitative
terms what undesirable environmental conditions
are proscribed (e.g., water should be free from
excessive algal blooms caused by nutrient
enrichment). A problem with narrative standards
is that a procedure for interpreting the standard to
determine numeric targets must be developed
before regulatory actions can be imposed (U.S.
EPA 1999a). With the possibility of regulatory
action undermined by poorly defined standards,
many efforts to manage eutrophication rely
primarily on nonregulatory (i.e., voluntary,
educational) approaches (NRC 2000). For
example, the more than 20-year, high-profile
effort to restore the Chesapeake Bay has
emphasized voluntary rather than regulatory
approaches, with numeric criteria being proposed
only recently (U.S. EPA 2003). Adoption of
numeric standards for nutrients and nutrient-
related water quality indicators could simplify
both determination of nutrient-related water
-------
quality impairments and implementation of
regulatory actions to improve water quality.
Regulatory approaches ultimately may be needed
to reverse current trends.
There are several reasons why most water
bodies are presently protected under narrative
rather than numeric nutrient criteria:
• natural nutrient concentrations vary widely
among aquatic ecosystems, requiring that
numeric criteria be derived on a regional or
even site-specific basis;
• nutrient enrichment effects usually are
expressed in relatively complex ways that
preclude the use of experimental assays,
similar to toxicity tests, to derive nutrient
criteria;
• sensitivity to nutrient enrichment is expected to
vary significantly among systems; and
• ambient nutrient concentrations are usually not
as useful as indicators of biological effects as
are ambient concentrations of toxic chemicals
(e.g., high nutrient concentrations may reflect a
lack of nutrient limitation, whereas low
nutrient concentrations may have been caused
by an active phytoplankton bloom).
Despite the challenges of deriving
ambient water quality criteria for nutrients, EPA
adopted a national strategy to develop nutrient
criteria in 1998 (U.S. EPA 1998) and has since
published technical guidance and recommended
nutrient criteria for most U.S. rivers, streams,
lakes, and reservoirs (e.g., U.S. EPA
2000a,b,c,d). EPA published technical guidance
for establishing nutrient criteria for estuaries and
coastal marine waters (U.S. EPA 2001) but
stopped short of recommending specific
approaches. Determining nutrient criteria for
estuaries may be more complicated than for
many freshwater systems because of a number of
factors that include those below.
• Estuaries are influenced to varying degrees by
exchanges of matter and energy (e.g., resulting
from tides) at their seaward margins.
• Widely varying physical and chemical
conditions strongly influence biotic
communities in estuaries (Kennish 1986).
Feedback effects contribute to significant water
quality differences among systems.
• Estuaries often are subject to multiple stressors
influencing nutrient-related water quality.
Examples include fishing, dredging, and exotic
species invasions.
• Strong gradients in water quality are often
present within estuaries, complicating
specification and implementation of numeric
criteria.
In view of these observations, developing
a scientific basis for nutrient criteria could
require an understanding of nutrient inputs,
transport, fate, and the important biological
response mechanisms of the resident biotic
communities in each system (NRC 2000). The
failure in most cases of simple nutrient dose-
ecological response models has been noted as a
reason that complex simulation models may be
needed to directly model complex ecological
behavior related to nutrients in coastal waters
(Fitzpatrick and Meyers 2000, but see Boynton
and Kemp 2007 for examples of simple models
that illustrate predictable ecological responses).
Cloern (2001) outlined how the field of coastal
eutrophication research has progressed largely
from limnological concepts to the point where it
is grappling increasingly with the additional
complexity of coastal marine systems.
Recognizing the importance of
developing numeric nutrient criteria and also the
difficulty of doing so, this study was undertaken
to demonstrate an approach that could be used to
determine numeric nutrient criteria (and criteria
for nutrient-related response indicators) for a
single coastal system (Pensacola Bay, FL) with
which we were very familiar. Like many
estuaries, Pensacola Bay has not received intense
research and management focus. Thus, this study
demonstrates that conclusions can be reached
regarding criteria with reasonable but still
incomplete scientific information. To address our
objectives, we undertook the following:
• examined the data available to characterize
current and past water quality conditions in the
Bay,
• reviewed the scientific knowledge regarding
the environmental conditions and processes
influencing water quality in the bay that could
inform nutrient criteria setting,
-------
• compared water quality conditions in
Pensacola Bay with a group or class of
estuaries that we expected to share similar
ecological attributes, and
• synthesized the assembled information using a
"weight-of-evidence" approach to suggest
scientifically justifiable water quality criteria
for nutrient and nutrient-related water quality
indicators.
To place our demonstration in a larger
context, we discuss the potential for applying the
demonstrated procedure to other estuaries.
1.2 Study Site
The Pensacola Bay system is a complex
of estuarine bays arranged in two major arms that
combine to form Pensacola Bay proper. The
component bays include Escambia Bay,
Blackwater Bay, East Bay, and Pensacola Bay
(Figure 1). Santa Rosa
Sound is excluded from
this analysis even
though it adjoins the
bay, because in terms of
important ecological
processes, it is
relatively distinct. The
combined Pensacola
Bay system is medium-
sized (370 km2) and
shallow (mean depth =
3 m) and has been
characterized as a
partially stratified,
drowned river valley
estuary (Schroeder and
Wiseman 1999). Tides
are diurnal and have a
low amplitude, ranging
from 15 to 65 cm. The
basin includes three
major watersheds that
drain via the Escambia,
Blackwater, and Yellow
rivers. The Escambia
River discharges into
Escambia Bay, whereas
the Blackwater and
Yellow rivers discharge
into Blackwater Bay and East Bay, the eastern
branch of the system. Both branches join
Pensacola Bay, which connects to the Gulf of
Mexico through the narrow (800-m-wide)
Pensacola Pass (Figure 1).
The condition of the Pensacola Bay
system, particularly that of Escambia Bay,
became a matter of public concern as early as the
late 1960s (Olinger et al. 1975), earlier than in
many estuaries. Initial evidence for a pristine
condition was based on biological surveys of
stream fauna (Academy of Natural Sciences of
Philadelphia 1953). By the early 1960s, after the
establishment of significant industrial point-
source discharges (especially ammonium and
organic matter), similar biological surveys
indicated declining health in the river. Reports of
fish kills and declining fisheries outputs
ultimately led to actions that, by the mid 1970s,
Monthly Survey
Quarterly Survey
Benthic-Pelagic Stations
Pensacola Pass
Figure 1. Locations of stations sampled during two major studies of water quality
conditions in the Pensacola Bay system (Table 1) and a 2003-to-2004 study of benthic
and planktonic metabolism. Monthly survey stations were sampled approximately monthly
from March 2002 through November 2004. Quarterly survey stations were sampled
quarterly from 1996 to 2000. Station identifications for benthic-pelagic stations corre-
spond to rates reported in Table 4. Santa Rosa Sound is not considered in this study to
be part of the Pensacola Bay system.
-------
greatly reduced the point sources and, in some
cases, eliminated them completely. Olinger et al.
(1975) provides a remarkable early compilation
of ecological conditions in the bay, intended
principally to document the recovery of the
system following reductions in industrial waste
loads. No comparable data were collected after
1975 until bay-wide water quality surveys were
begun by EPA beginning in 1996 (U.S. EPA
2005). Whereas
differences in
survey methodology
and data reporting
mostly preclude
quantitative analysis
of ecological
changes, comparing
early data and the
conclusions of the
early investigators
with more recent
studies suggests that
neither the
ecological
conditions nor the
nature of the major
ecological concerns
have changed
dramatically in the
past 30 years. The
major ecological
concerns in 1975
were bottom water
hypoxia; loss of
seagrass habitats;
toxic contamination;
and degradation of
biotic communities,
including fisheries.
Bottom water
hypoxia and loss of
seagrass habitats,
both of which are
likely to be related
to nutrient
enrichment,
continue to be
concerns for
Pensacola Bay (U.S.
EPA 2005). U.S. EPA (1999b) reported that
sediment toxicity in the Pensacola Bay system is
confined mainly to the three urban bayous
(Bayou Grande, Bayou Chico, and Bayou Texar;
Figure 1). However, toxic contamination of
seafood, especially by mercury, remains a
regional issue in the Gulf of Mexico. Blue crab
(Callinectes sapidus) in Pensacola Bay have been
shown to be contaminated with polychlorinated
Low-amplitude (<0.5 m) diurnal tide
River inputs of freshwater, dissolved
organic carbon, and nutrients
Atmospheric nitrogen deposition
Nutrient inputs from municipal
wastewater
Industrial point sources mostly
eliminated, coal-fired power
generation emits to atmosphere.
Cross-pycnocline diffusive
mixing is low.
Two-layer estuarine circulation,
sometimes very weak
Moderate light attenuation
(upper estuary), high light
transmittance (lower estuary)
Macrobenthos degraded, most
likely by hypoxia.
f
£**
High and low dissolved oxygen
Medium and low dissolved
nutrients
Particulate C and N
sinking
Denitrification
Recreational uses, including
contact and noncontact uses
Commercial fishing, principally
shrimp trawling
Large fraction of watershed is
forested.
Development along estuarine
shorelines
Seagrasses present in lower
Bay, but largely lost elsewhere.
Figure 2. A conceptual model illustrating historic, ecologic, and socioeconomic factors of
potential importance for developing and evaluating water quality management options for
Pensacola Bay. Iconography from the University of Maryland Center for Environmental
Science Integration and Application Network, http://ian.umces.edu.
-------
biphenyls, likely acquired by bioaccumulation
from residual contamination of sediments
(Karouna-Renier et al. 2007).
Many of the important ecological
processes and environmental concerns, including
the human and aquatic life uses of the bay and
the major stressors of the ecosystem can be
summarized effectively and communicated in a
conceptual model (Figure 2). Developing such a
model, or at least acquiring the information
necessary to do so, is probably an important step
toward determining a scientifically sound
procedure for deriving nutrient criteria. By
comparing and contrasting conceptual models for
different systems, one can gain a greater
understanding of the similarities and differences
among those systems and how those differences
may impact criteria development. We include our
conceptual model without extensive discussion
here because many aspects of the model are
addressed in subsequent sections.
-------
-------
2. Methods
This study is based on "found" data from
several types of sources, including national
databases, water quality surveys, and aerial
remote sensing. The following describes the data
collected and either the methods or appropriate
references describing the methods.
2.1 Water Quality Data for
Pensacola Bay
Two major studies provided the majority
of the water quality data for Pensacola Bay used
in this study (Table 1). In the first study
(hereafter "quarterly survey"), 38 sites selected
using a probabilistic survey design (Summers
et al. 1995, U.S. EPA 2005) were sampled
quarterly (spring, summer, fall, and winter)
during surveys beginning in May 1996 and
concluding in February 2001. In the second study
(hereafter "monthly survey"), 29 water quality
surveys were conducted approximately monthly
from April 2002 through December 2003 and
from May 2004 through November 2004 (e.g.,
Hagy and Murrell 2007). Monthly surveys
included 15 sites oriented along two axial
transects of Pensacola Bay (Figure 1). Additional
water quality data were obtained from Florida's
Inshore Marine Monitoring and Assessment
Program (IMAP; McRae et al. 2004), which
collected a total of 64 observations on multiple
water quality variables in the Pensacola Bay
system between 2000 and 2004. Most
observations (53 of 64) were made during an
intensive survey in 2003.
2.2 Ecoregional Water Quality
Characterization
To compare water quality in Pensacola
Bay with other estuaries in the region, water
quality statistics were summarized for estuaries
in EPA's Level 111 ecoregion 75 (Omernik 1995),
encompassing the southeast coastal plain and
Florida's Level IV ecoregion 75a (Griffith et al.
1994), encompassing the west Florida panhandle
region eastward to the Big Bend region of Florida
(Figure 3). Because these ecoregions are defined
to encompass regions of land, not water, estuaries
were associated with an ecoregion when the land
surrounding the estuary was substantially within
ecoregion boundaries. Thus, Florida's ecoregion
75a includes all inshore marine waters from
Perdido Bay eastward to the Suwannee River
(Figure 3). Water quality data for the ecoregional
characterization were obtained from IMAP
(McRae et al. 2004) and EPA's National Coastal
Assessment (NCA; U.S. EPA 2004). Both
assessment studies utilized a probabilistic survey
Table 1. Sources of water quality data used in this study. Water quality data include physical data (water
temperature, salinity, and DO), water clarity (secchi depth), chlorophylls, and nutrients (NO3, NO2, NH/, PO4*~,
total nitrogen, total phosphorus).
Data Set
Source
Survey Design/Scope/Resolution
Pensacola Bay Quarterly
Survey (U.S. EPA 2005)
Pensacola Bay Monthly Survey
(Hagy and Murrell 2007)
Florida Inshore Marine
Monitoring and Assessment
Program (McRae et al. 2004)
National Coastal Assessment
(U.S. EPA 2004)
EPA/ORD
EPA/ORD
Florida Marine
Research Institute
EPA/ORD
Probabilistic survey design, May 1996 to February 2001,
Approximately 40 stations sampled quarterly. Surface and
bottom.
Transect surveys, 2002 to 2004,15 stations sampled monthly.
Surface and bottom, with 25-cm profile of physical variables.
Florida coastal waters, probabilistic survey design. 2000 to
2004, late summer. 30 stations per year statewide, with
rotating schedule of regional focus at 180 stations per year. 64
station visits in Pensacola Bay system, most in 2003.
National scope. 2001 to 2002, late summer. Approximately
375 stations in southeast coastal plain.
-------
45
km
90
Figure 3. The locations of water quality monitoring
stations visited under EMAP during 2001 and 2002 and
IMAP during 2000 to 2004. The distribution of sampling
stations within the southern coastal plain ecoregion
(region 75 of EPA Level III ecoregions) (upper panel) and
within the Gulf Coastal Flatwoods region (Florida's Level
IV ecoregion 75a) (lower panel). NCA stations were
omitted from the ecoregion 75a analysis. Estuarine
drainage areas in ecoregion 75a, as defined under
NOAA's coastal assessment framework, are identified as
follows: A = Perdido Bay, B = Pensacola Bay,
C = Choctawhatchee Bay, D = St. Andrew Bay, E =
Apalachicola Bay, F = Apalachee Bay, and G =
Suwannee River.
design with sampling conducted only during a
late summer index period (Table 1, Figure 3).
The procedure for water quality
characterization (Figure 4) involved computing,
for each variable of interest, the median of all
water quality observations in each of three
salinity zones in each estuary. We defined three
salinity zones based on surface salinity values:
(1) oligohaline (salinity < 5), (2) mesohaline
(salinity 5 to 18), and (3) polyhaline (salinity >
18). We defined each estuary according to the
boundaries of estuarine drainage areas (EDAs)
defined by the National Ocean and Atmospheric
Administration's Coastal Assessment Framework
(NOAA 1999a). Data from IMAP and NCA were
referenced to EDAs within the coastal assessment
framework using a geographic information
system (ArcGIS 9.2, ESRI, Inc.). Data not falling
within EDAs, such as those falling within
"coastal drainage areas," were excluded from the
analysis.
2.3 Aerial Submerged Aquatic
Vegetation Surveys
Data describing the extent of submerged
aquatic vegetation (SAV) in Pensacola Bay were
obtained from investigations that employed
interpretation of aerial photographs taken as early
as 1949 and as recently as 2003. Maps of SAV
coverage for circa 1950 were obtained from
Olinger et al. (1975), who interpreted
photographs obtained for highway construction.
For this study, maps from Olinger et al. (1975)
were scanned to TIFF files, geo-referenced
against corresponding modern shorelines, and
then manually digitized to a vector coverage at
the same resolution as the original map (modern
shoreline extracted from 1:100,000 scale digital
line graphs by the U.S. Geological Survey
(USGS) National Wetlands Research Center,
Gulf Breeze, FL, Project Office). Tabulated SAV
coverages for 1960, 1980, and 1992, were
obtained from Schwenning et al. (2007), which
also describes the photo-interpretation and
digitization protocols (also see Lores et al. 2000).
SAV coverage data for 1992 and tabulated SAV
All Watei Quality
Observations in the Class during
.^^^ Summer ^^^,
/ T" \
Oligohaline Zone
(Surface Salinity<5)
Mesohaline Zone Polyhaline Zone
(Surface Salinity 5-18) (Surface Salinity >18)
Figure 4. Procedure for computing statistical comparison
values for estuaries. The 25th percentile of estuary
medians was used for comparison values for water quality
indicators where lower values are better (e.g., nutrients,
chlorophyll-a). The 75th percentile was used for secchi
depth, where higher values indicate higher water clarity.
-------
coverage data for 2003 were obtained from the
USGS National Wetlands Research Center,
Lafayette, LA.
2.4 Watershed Characteristics and
Land Use
Data describing land use in the Pensacola
Bay watershed were obtained from the 2001
National Land Cover Database using the Multi-
Resolution Land Cover Consortium Viewer
(Homer et al. 2004; http://www.mrlc.gov/). The
database provides land use and land cover
classifications at 30-m resolution. Classifications
are derived from Landsat-7 Enhanced Thematic
Mapper imagery and additional supporting data.
Land cover data for the Pensacola Bay watershed
were extracted from the larger data set and
tabulated by classification using the ArcGIS 9.2
geographic information system indicated above.
-------
-------
3. Results and Discussion
3.1 Ecoregional Comparisons and the
Percentile Approach
Identifying appropriate benchmarks or
"reference conditions" against which to compare
water quality conditions is a key feature of EPA's
recommended approaches for developing nutrient
criteria (U.S. EPA 2000a,b). We modified EPA's
percentile approach for freshwaters to define
benchmark water quality values that have
potential applicability to Pensacola Bay.
A central feature of the approach, which we
applied here, is identification of a suitable class
of similar estuaries from which a reference
condition can be determined. In the context of
nutrients, similarity refers to a similarity in
ecological attributes that influence the response
of a system to nutrient inputs.
There are many classification schemes for
coastal systems, each serving different objectives
and emphasizing different ecological attributes
(Kurtz et al. 2006). Unfortunately, there is little
consensus regarding which classification
schemes would be most appropriate in terms of
predicting response to nutrient inputs. None have
been demonstrated to be useful for that purpose.
One concern is that, even though many important
attributes of estuaries have been considered (e.g.,
residence time, mean depth, ratio of watershed
area to estuary area; Kurtz et al. 2006), a much
broader set of factors (e.g., climate, geology,
character of offshore water quality) likely
contributes to the ecological character of
ecosystems.
As an alternative to the available
estuarine classifications, we utilized EPA's Level
III ecoregions and, simultaneously, the more
finely resolved Level IV ecoregions for Florida
to define a class of estuaries for Pensacola Bay.
Pensacola Bay lies within the "southern coastal
plain" ecoregion 75 in the Level III scheme
(Figure 3). The shores of Pensacola Bay include
several of Florida's Level IV ecoregions, but the
"Gulf coastal flatwoods region" (ecoregion 75a)
encompasses much of the coast surrounding
Pensacola Bay and a significant span of coastline
mostly to the east of the bay (Figure 3). Although
the definition of these ecoregions is based on
climatological, geological, and biological
attributes (Omernik 1987) and pertains to the
watersheds of estuaries rather than to the
estuaries themselves, we expect that the
ecological character of inshore coastal waters
must reflect to some degree the ecological
attributes of the surrounding land. Moreover,
because the ecoregions define geographically
contiguous regions of the coast, we expect
similarity in many attributes to arise simply from
proximity. Important differences can occur,
however, particularly when the ecoregion spans
oceanographically distinct regions. For example,
tides vary dramatically within ecoregion 75;
minimal diurnal tides occur on the northern Gulf
coast, whereas large semidiurnal tides occur
along the Atlantic coast of Georgia and South
Carolina. These differences alone contribute to
other large ecological differences, such as the
presence of extensive intertidal salt marshes in
Georgia (Dame et al. 2000), which practically
define the Georgia coast, versus the more limited
extent of salt marshes in west Florida. The
smaller spatial extent of Florida's Level IV
ecoregions eliminates the most glaring within-
class differences, but at the cost of a class that
includes fewer systems. Even though it is likely
that important differences remain within even
these smaller classes, further division leads
toward the perspective that every estuary is
completely unique, precluding any role for
comparative ecological analysis in development
of nutrient criteria.
U.S. EPA (2000a,b) define a variety of
options for determining reference conditions for
classes of freshwater systems. The preferred
approach is to derive water quality criteria on the
basis of conditions in relatively pristine water
bodies. When a set of sufficiently pristine water
11
-------
bodies can be identified, criteria have been based
on the 75th percentile of water body median
values for such pristine sites (U.S. EPA 2000a,b).
Given the pervasive human presence in many
coastal marine areas, we assumed that few, if
any, estuaries were sufficiently pristine to adapt
this approach for estuaries. Another approach is a
historical reference condition in which water
quality conditions observed prior to significant
anthropogenic impacts provide a basis for
comparison. For most estuaries, limited
availability of historical data precludes this
approach. The record of historical water quality
data for Pensacola Bay is unusually extensive,
but, as is common, early monitoring was
undertaken in response to obvious pollution
impacts (Olinger et al. 1975) and does not
quantify the unimpaired or pristine condition.
Thus, historical data cannot be used to
characterize pristine water quality conditions in
Pensacola Bay. Lacking either pristine sites or
adequate historical data, we derived levels we
call "comparison values" from the statistical
distribution of water quality conditions in all the
estuaries in the class (Figure 4), adapting the
percentile approach of U.S. EPA (2000a,b).
Because the class includes sites subject to a range
of nutrient enrichment, we utilized the 25th
percentile of estuary median water quality values
to derive comparison values. EPA (2000a,b) cite
empirical evidence from lakes to support this
choice when reference quality sites are not
available. We did not assume that there is any
relationship between the 25th percentile of
estuary median water quality and water quality in
a hypothetical set of pristine estuaries. We simply
chose to examine the 25th percentile as a
potentially useful reference point in the
distribution of estuary medians. For water clarity,
the same ideas apply, but higher values indicate
"better" water quality. Therefore, we used the
75th percentile as a comparison value. We
contrasted values based on two classes: (1) all
estuaries in ecoregion 75 (Appendix A) and
(2) the subset of those estuaries within ecoregion
75a (Appendix B).
Our application of the percentile approach
differs from that of U.S. EPA (2000a,b) in
several important ways. First, because the
available regional-scale data sets from NCA and
1MAP include only late summer data (Table 1),
our analysis is limited to summer (June through
September). We do not see this as a major
limitation because most water quality problems
related to nutrient enrichment, especially
hypoxia, are expressed primarily during summer.
Thus, we assumed that, if water quality supports
the designated uses during summer, uses outside
of summer are also likely to be met. The second
difference is that we divided observations into
salinity zones, based on the principle that water
quality parameters in estuaries commonly vary
along the salinity gradient. We selected three
zones (salinity < 5, 5-18, and > 18) with the
objective of resolving differences associated with
the salinity gradient, while ensuring that a
sufficient quantity of data was available in each
zone in most estuaries. Our objectives in defining
the zones likely could have been met with
different definitions (i.e., breakpoints); however,
an attempt to define values for a greater number
of salinity zones likely would have resulted in
excessive parsing of the available data. Finally,
we consider the values obtained via this approach
to be guidelines or comparison values, not values
that can be adopted defensibly as criteria without
further scientific support.
The computed median values show that
summer water quality in a significant portion of
estuaries in ecoregions 75 and 75a was
characterized by low nutrient and moderate
chlorophyll-a concentrations, with concentrations
generally lower in ecoregion 75a than in
ecoregion 75 as a whole (Figure 5, Table 2).
Concentrations of chlorophyll-a, nitrogen, and
phosphorus are commonly higher in Mid-Atlantic
estuaries (NOAA 1997). Nutrient concentrations
are commonly higher in Pacific coast estuaries
influenced by coastal upwelling (NOAA 1998).
Summer median nutrient concentrations and
water clarity in Pensacola Bay were comparable
to or lower than the 25th percentile (higher than
the 75th percentile for water clarity) for the
ecoregion, with a few exceptions.
12
-------
O)
3.
1 JU
100 -
50 -
25^
20
15
10
5 •
I z a
Chlorophyll-a |
I 2.0
pi T ; i
-
.
•
30
25
20
15
10
5 .
"I
,
-
*?
:>
M
t
r
!
J;
—
T
mn
I \J\J
80
60
40
20 -
n .
1
^^H
-L
_
, y '
liil^mmi
H
-
p-n
^
3 ,
r I
»
j^l
no data
L/
O
M
P
Ecoregion 75
S-'S
'.,_
^.
j_
'"'"""^
-*~^-l
»*f
O
1.5
E
1.0
0.5
n n
Secchi E3epth
*
i n
-j- ~ i, I "Vj
pia r4i Ju S rr*i '-'• •'•'
H UiJ Pj H fed
"i |~^ ^j"^
. *
c n
DIN
fen
i n i Bgm
5.0
4.0
1 3.0
2.0
1.0
n n
• ^
~r ' po
JT, p-L '
•Ik.
.'/' fi
j Ej ED Q
o c
TN
r"!
Liij
J.%J :
3.0
2.5
5 2.0
^ 1.5
1.0
0.5
n n
r— 7-5 TP
^ ,
no data
>'i „
'•' ' ' P""! t;' I
r. .' 1 B1.., 1 ' 1
Lj_J
MP OMPOMP
Ecoregion 75a Ecoregion 75 Ecoregion 75a
Figure 5. Distribution of summer estuary median surface water quality values by
salinity zones for estuaries in Level III ecoregion 75 (left three plots) and Level IV
ecoregion 75a (right three plots). Salinity zones are oligohaline (O; salinity < 5),
mesohaline (M; salinity 5-18)], polyhaline (P; salinity > 18). Red lines indicate summer
medians by salinity zone for Pensacola Bay. Identical values are shown in both the left
and right halves of each plot. DIN = NO2" + NO3" + NH/; TN = total nitrogen, TP = total
phosphorus.
Inorganic nitrogen and phosphorus
concentrations in oligohaline waters were higher
than the 25th percentile, but concentrations in
higher salinity water were comparable or lower
than the corresponding 25th percentile in those
salinity zones. Total phosphorus (TP)
concentrations were comparable to the 25th
percentile values across the salinity zones,
whereas total nitrogen (TN) concentrations were
comparatively lower in oligohaline waters and
higher in polyhaline waters (Figure 5, Table 2).
Median chlorophyll-a was higher in Pensacola
Bay than the 25th
percentile values in all
salinity zones for both
ecoregion 75 and
ecoregion 75a (Table 2).
Instead, median
chlorophyll-a in
Pensacola Bay was
similar to the median for
ecoregion 75 and closer
to the upper quartile for
ecoregion 75a (Figure 5).
The statistical
approach described here
is a repeatable,
quantitative scheme for
computing water quality
measures that can serve
as comparison values for
both causal variables
(e.g., nutrients) and
nutrient-related response
variables such as water
clarity (secchi depth) and
chlorophyll-a. In the
absence of water quality
thresholds clearly tied to
loss of ecological
integrity, these
comparison values are
useful for placing values
for any one estuary in
perspective. There are
several factors that limit
the range of
interpretations for these
comparison values. One
question is whether the 25th percentile (or 75th
percentile for water clarity) of estuary medians is
an appropriate comparison value, and whether it
approximates a minimum condition protective of
designated uses as suggested by EPA (2000a,b).
A broader comparison with the distribution of
values in the ecoregion (i.e., Figure 5) gives a
more complete comparative perspective but still
leaves open the question of whether a particular
water quality value supports use attainment.
Water quality values at the 25th percentile of the
ecoregional distribution could be either
13
-------
Table 2. Ecoregional water quality conditions for chlorophyll-a (Chl-a), secchi depth, dissolved inorganic
nitrogen (DIN = NC>2~ + NO3~ + NH4+), phosphate (PO43"), total nitrogen (TN), and total phosphorus (TP) computed
for EPA's Level III ecoregion 75 and Florida's Level IV ecoregion 75a. Values are the 25th percentile of summer
estuary medians, except for secchi depth, where the value is the 75th percentile of estuary medians. Summer
medians for Pensacola Bay are shown for comparison. Data are for surface water. Values in parentheses for
ecoregion 75 and 75a are the number of estuaries included for each variable. Values in parentheses for
Pensacola Bay are the number of observations included. TN and TP values were not computed for ecoregion 75
because data were only available for a portion of the ecoregion. Pensacola Bay medians were computed from a
combined data set that includes EPA's EMAP quarterly surveys (1996 to 2001, EPA's monthly surveys (2002 to
2004), and Florida's IMAP surveys. IMAP sampled Pensacola Bay intensively in 2003 and also visited two
stations in 2004.
Group
Oligohaline (Salinity < 5)
Ecoregion 75
Ecoregion 75a
Pensacola Bay
Mesohaline (Salinity 5-18)
Ecoregion 75
Ecoregion 75a
Pensacola Bay
Polyhaline (Salinity > 18)
Ecoregion 75
Ecoregion 75a
Pensacola Bay
Chl-a Secchi Depth
(M9 1"1) (m)
3.8(16)
2.1 (6)
5.6(103)
5.1 (21)
4.1 (7)
7.3 (233)
4.3 (28)
3.5(6)
5.8(107)
0.93(12)
1.6(4)
0.9 (93)
1.0(21)
1.0(7)
1.2(198)
1.3(28)
1.4(5)
2.0(100)
DIN
(MM)
2.9(15)
2.6(6)
8.7(102)
1.3(20)
0.81 (7)
0.97(210)
0.59 (27)
0.47 (6)
0.44(102)
PO43"
(MM)
0.11 (15)
0.023 (6)
0.18(86)
0.13(21)
0.065 (7)
0.11 (122)
0.11 (28)
0.052 (6)
0.048 (82)
TN
(MM)
nd
23.4 (6)
29.2 (49)
nd
26.5(7)
31.7(59)
nd
20.5 (6)
32.7 (57)
TP
(MM)
nd
0.67 (6)
0.88 (32)
nd
0.56 (7)
0.55 (47)
nd
0.73 (6)
0.43 (56)
nd = No data
nonprotective of uses or excessively restrictive
for a particular estuary. A second concern is the
assumption that the estuaries in either ecoregion
75 or ecoregion 75a are an appropriate "class,"
and that Pensacola Bay is a member of that class.
We suggest that the percentile approach is likely
to produce values in a reasonable range for
estuaries, but that, because the approach is
inherently independent from effects-based
considerations and is not certain to generate
appropriate values, the values should be
evaluated in the context of other information
available about the estuary in question before
they could be used to determine criteria. In the
following sections, we discuss ecological
conditions and processes in the Pensacola Bay
system to provide a thorough evaluation of the
science that may be pertinent to selection of
numeric nutrient criteria.
3.2 Algal Biomass, Productivity, and
Community Composition
Chlorophyll-a is an easily monitored and
conceptually appealing indicator of trophic status
and eutrophication in aquatic systems, obvious
reasons why it already is used widely in water
quality management related to nutrients. Other
simple descriptors of the phytoplankton
community, such as productivity; community
composition, including presence of harmful
algae; and growth dynamics, also can provide
useful insights. We assembled basic information
about the phytoplankton community in Pensacola
Bay to explore its implications, if any, for
managing nutrients in the bay.
Seasonal and salinity zone chlorophyll-a
medians varied between 1.6 and 7.4 jug L'1, with
the highest values during summer (5.6 to
7.3 ug L"1; Table 3). Both summer and non-
summer values were "low" to "medium"
according to standard ranges described by NOAA
(1997). "High" concentrations (>20 jig L"1)
occurred very infrequently (Table 3, Figure 6).
Of 1,390 chlorophyll-a measurements in the
Pensacola Bay system between 1996 and 2004,
only 36 (2.5%) were greater than 20 |ag L"1.
Murrell et al. (2007) measured phytoplankton
production rates in Escambia Bay using carbon
14
-------
Table 3. Seasonal median chlorophyll-a in Pensacola
Bay surface waters and its interquartile range (25th
to 75th percentile). Salinity zones were determined
from surface water salinity.
Median
Chlorophyll-a
(ug L-1)
Oligohaline (Salinity
Winter
Spring
Summer
Fall
Mesohaline (Salinity
Winter
Spring
Summer
Fall
Polyhaline (Salinity >
Winter
Spring
Summer
Fall
<5)
2.2
4.0
5.6
1.6
5-18)
3.5
5.8
7.3
5.0
18)
3.4
3.8
5.8
3.4
Interquartile
Range
(ugL-1)
(1.2
(2.6
(3.8
(1.2
(2.6
(3.7
(5.1
(3.1
(2.6
(2.1
(3.8
(2.5
-3.7)
-5.4)
-8.7)
-3.7)
-5.2)
-8.4)
- 10.7)
-7.2)
-4.7)
-4.8)
-8.1)
-6.0)
14 uptake methods and estimated that annual
phytoplankton productivity was 320 g carbon m"2
year"1 (0.88 g carbon m"2 day"'). Among estuaries,
where annual production commonly varies
between 100 and 500 g carbon m"2 year"1
(Boynton et al. 1982, Boynton and Kemp 2007),
the annual production rate for Pensacola Bay is
"mesotrophic" or moderate. The summer
phytoplankton assemblage is dominated by
cyanobacteria (Murrell and Lores 2004,
Murrell and Caffrey 2005), whose low
chlorophyll-a content relative to carbon
(phycoerythrin is the dominant light-gathering
pigment in most cyanobacteria; Maclntyre
et al. 2002) undoubtedly contributes to the
high ratio of annual productivity to
chlorophyll-a. The phytoplankton community
in Pensacola Bay is usually nutrient limited
and can be limited by either phosphorus or
nitrogen (Murrell et al. 2002, Juhl and Murrell
2005). Relief of nutrient limitation associated
with periods of increased freshwater and
nutrient inputs manifests as increased
chlorophyll-a concentration, increased
phytoplankton production (Murrell et al.
2007), and a greater relative abundance of
eukaryotic phytoplankton (Juhl and Murrell
2005) in Pensacola Bay.
The available information points to
several key conclusions regarding the
phytoplankton community in Pensacola Bay.
First, there is no clear evidence that the
phytoplankton community is stimulated strongly
by excess nutrients, or that it is unbalanced or
otherwise causing adverse ecological effects. If
chlorophyll-a concentrations were frequently in
NOAA's (1997) high (20 to 60 ug L"1) or
"hypereutrophic" (>60 jig L"') range, or, if
annual phytoplankton productivity was among
the highest found in estuaries, one might have
a priori cause for concern regarding excess
phytoplankton. Median summer chlorophyll-a
concentrations were generally higher than the
25th percentile of systems within the ecoregion,
especially for ecoregion 75a (Table 2). However,
the small differences are probably not
ecologically significant. We do not know of any
study directly implicating chlorophyll-a
concentrations in the low range observed in
Pensacola Bay with failure to support any human
or aquatic life uses. Experimental evidence
(Murrell et al. 2002, Juhl and Murrell 2005)
shows that phytoplankton biomass and
production in Pensacola Bay are limited by
nutrients. This provides a mechanistic
expectation that increased nutrient inputs will
Low
Medum
Percent of Observations
30 -
20-
10 -
n -
30-
20 -
10-
30-
20-
10 -
n -
30 -
20-
10 -
n -
I Spring
I (N=88)
I Summer
1 (N=132)
I Fall
I (N=78)
• I Winter
In (N=113)
10 15
Chlorophyll-a (ug L'1)
20
30
40
Figure 6. Frequency histogram of surface chlorophyll-a
concentrations in mesohaline (surface salinity = 5-18) waters of
the Pensacola Bay system from 1996 to 2001. "Low," "Medium,"
and "High" designations refer to descriptors used by NOAA
(1997). Dotted lines indicate the median concentration. Data are
from the EPA/ORD quarterly surveys of Pensacola Bay.
15
-------
1996
1997
<2o
1998
•V -
1999
Jan Feb Mar Apr May Jun Jul Aug Sep Od Nov Dec
Figure 7. Observations of bottom dissolved oxygen in the
Pensacola Bay system during the summers of 1996 through
1999 indicating the locations and frequency of observations
less than 2.0 mg L"1, a level commonly used as an
operational definition of hypoxia. Data are from EPA/ORD
quarterly surveys (upper panels). The computed extent of
hypoxia in Pensacola Bay by month from 2002 to 2004,
based on EPA/ORD monthly transect surveys (lower panel).
increase phytoplankton production and possibly
phytoplankton biomass. Unfortunately, simple
linear dose-response relationships fail to capture
the relationship between nutrient loads or
concentrations and chlorophyll-a in Pensacola
Bay (Murrell et al. 2007). Relationships that we
expect to observe on the basis of known causal
mechanisms often are hidden by the natural
complexity that characterizes estuarine
ecosystems (e.g., flushing, food web interactions)
and, therefore, are not easily seen in field
observations (Cloern 2001). Some empirical
relationships between nutrients and chlorophyll-a
have been found among systems and, through
time, within systems (Boynton et al. 1982);
however, such relationships usually are only
adequate to provide broad direction for
management (i.e., increased nutrients are
associated with higher chlorophyll-a).
Nonetheless, it is important and useful to
recognize that, on the basis of experimental
evidence documenting nutrient limitation of
phytoplankton, we expect that increased nutrient
inputs would increase phytoplankton production
and biomass in Pensacola Bay.
3.3 Preventing Hypoxia
Hypoxia, which we define here as
DO < 2 mg L"', is a common phenomenon in
bottom waters of Pensacola Bay (Hagy and
Murrell 2007). Where it occurs, hypoxia should
be a concern because of its direct impact on
marine life (Diaz and Rosenberg 1995). Globally,
hypoxia has increased in coastal waters in some
relation to nutrient overenrichment, commonly
causing negative effects on the health and
productivity of biological communities (Diaz
2001). Thus, hypoxia is a significant threat to
both human (e.g., fisheries) and aquatic life use
attainment in coastal systems.
Observations of hypoxia in the Pensacola
Bay system date back to Hopkins (1969), who
found oxygen concentrations as low as
0.48 mg L" in bottom waters of Escambia Bay
during a late-summer study. Olinger et al. (1975)
reported bottom water oxygen less than 1 mg L"1
in Escambia Bay but concluded that conditions
had improved since 1969. The recent monthly
survey data (Table 1) show that hypoxic bottom
waters occurred between April and November,
with the maximum extent in late summer
(Figure 7). Hypoxia currently affects an average
of 25% of the bay bottom during summer and
can, at times, affect a much larger fraction
(Figure 7).
Olinger et al. (1975) recognized long ago
that the bay often was strongly stratified, and that
weak circulation probably contributed to the risk
of hypoxia. Hagy and Murrell (2007) quantified
the stratification and circulation to better
characterize conditions that create a high
susceptibility to hypoxia. They found that
hypoxia developed in Pensacola Bay despite
moderate oxygen consumption rates in the lower
water column and sediments. Total (i.e.,
combined) oxygen consumption for the lower
water column and sediments was estimated to be
0 to 1.5 g oxygen m2 day"1 using a box model
(Hagy and Murrell 2007). The model estimates
were in reasonable agreement with average
integrated oxygen consumption rates (0.94 g
oxygen m"* day"1) measured during summer at
16
-------
several stations throughout the Pensacola Bay
system (Table 4). Oxygen consumption in
Pensacola Bay is low compared with other
estuarine and coastal marine systems (e.g.,
Hopkinson 1985, Cowan et al. 1996) and is much
lower than rates from a very eutrophic system
such as the Chesapeake Bay, where integrated
metabolism below the pycnocline can exceed
10 g oxygen m"2 day"1 (Kemp et al. 1997).
Table 4. Plankton community respiration rate (Rp)
integrated from the pycnocline to sediments (i.e.,
lower water column) and sediment oxygen demand
(SOD) at stations in the Pensacola Bay system
(Figure 1) with varying levels of bottom water (BW)
oxygen (mg L'1). SOD was measured using 5-hour
incubations of diver-collected cores. Plankton
respiration rates were measured in 24-hour
incubations in BOD bottles. All rates have units of g
oxygen m"2 day'1. Unpublished data provided by M.
Murrell.
Station
PCOLA28
PCOLA20
PCOLA29
BP04
PCOLA26
BP03
P05C
(6/28/04)
P05C
(7/27/04)
Average
BW
02
5.2
2.5
0.7
4.2
1.1
5.0
1.6
1.4
—
SOD
0.32
0.27
0.04
0.37
0.12
0.38
0.48
0.50
0.31
Rp
0.34
0.07
0.25
0.86
0.88
1.48
0.58
0.59
0.63
Total
0.66
0.34
0.29
1.23
1.00
1.86
1.06
1.09
0.94
Coastal systems can be naturally
susceptible to hypoxia. A well-known example
on the Gulf coast is the occurrence of "jubilee"
events, in which marine animals (especially
crabs) climb onto the beach to avoid low-oxygen
waters. Jubilees have occurred in Mobile Bay,
AL, for more than a century (May 1973,
Schroeder and Wiseman 1988). There is
presently no evidence that hypoxia did not occur
in Pensacola Bay prior to significant human
influence, nor are there adequate data to
document any trend in dissolved oxygen (DO)
during the past several decades. Nevertheless,
because relatively extensive hypoxia occurs
presently in Pensacola Bay, hypoxia should be a
focus for water quality management. We can
infer that hypoxia presently limits human and
aquatic life uses of Pensacola Bay because the
direct effects of hypoxia on many species of
marine life are known (Diaz and Rosenburg
1995, Campbell and Goodman 2004). Although
current science for Pensacola Bay cannot predict
the magnitude of increase in extent or severity of
hypoxia expected from a particular increase in
nutrient inputs or the extent to which the severity
of hypoxia might increase, both conceptual and
quantitative models indicate that increased
nutrients most likely would increase the extent
and severity of hypoxia, resulting in loss of
existing aquatic life uses.
Defining a broadly applicable approach
for determining DO criteria in estuaries where
hypoxia may occur naturally is a difficult
proposition. The ecoregional approach that we
applied for nutrients and chlorophyll-a may not
be adequate for DO because EPA's Level 111
ecoregions do not reflect the physical
oceanographic characteristics of estuaries that
influence the development of hypoxia. In fact,
none of the classification schemes that have been
devised for estuaries have been shown to be
effective for grouping systems according to their
intrinsic vulnerability to hypoxia. Moreover, the
data sets and statistical indicators (i.e., medians)
that we used are not suitable to characterize the
extent of hypoxia in most instances; the estuary
median generally provides little information
about the extent of bottom DO. For Pensacola
Bay, where an average of 25% of the bottom is
hypoxic during summer, the median
concentration is, by definition, higher than our
defined threshold (i.e., >2.0 mg L"'). Relatively
elaborate DO criteria have been developed for
Chesapeake Bay based on a tiered aquatic life use
approach and the oxygen levels known to be
minimally supportive of those uses (U.S. EPA
2003). The complex approach utilized for
Chesapeake Bay, however, may not be feasible
for many of the nation's estuaries because the
information required to delineate tiered-use areas
is not available. For Pensacola Bay, it may be
reasonable to define only two areas: (1) the area
that is susceptible to hypoxia, which would have
a criteria defined in terms of areal extent; and (2)
an area that is much less susceptible to hypoxia
(e.g., surface waters), for which criteria could be
defined in terms of the requirements of resident
17
-------
biota. Current areal extent of bottom water
hypoxia is well defined and could provide the
basis for a criterion defined as a multiyear
average extent (e.g., 3 years). A reasonable
implementation strategy must include an
averaging period of several years because the
extent of hypoxia varies strongly on an
interannual basis (Figure 7). Freshwater inflow
level, which varies from year to year, is an
important driver of such changes (Hagy and
Murrell 2007). For above-pycnocline waters and
nonstratified waters, DO concentrations should
be greater than 4.0 mg L"', a level sufficient to
prevent most impacts on marine life in Pensacola
Bay (assuming summer temperature -30 °C) and
that is also consistent with current Florida criteria
for Class III marine waters (FAC 2005). In the
monthly survey data, DO concentrations in the
bay exceeded this value almost all the time. The
Florida statute also requires a minimum of
5.0 mg L"1 as a 24-hour average. This is most
likely met much of the time as well, but, because
most of the available data in the estuary are
daytime point observations, good estimates of the
24-hour average are not available.
Design of compliance monitoring is an
important consideration for DO because
sampling must adequately quantify spatial extent
for hypoxia in bottom waters and minimum
concentration for other waters. The data sets for
Pensacola Bay illustrate the potential for
problems. Florida's IMAP measured bottom
water oxygen at 29 stations in Pensacola Bay
during an intensive probabilistic survey in 2003,
observing oxygen < 2 mg L"1 at 6 stations (21%),
which can be interpreted to mean that 21% of the
bottom was hypoxic. The transect-based surveys
obtained similar results. In 2004, however, IMAP
visited just two sites and observed no hypoxia,
even though hypoxia was much more extensive
in Pensacola Bay in 2004 than in 2003 (Figure 7).
Neither the transect surveys nor the probabilistic
IMAP surveys sampled at night, when minimum
concentrations may have occurred.
3.4 Protecting Seagrass Habitats
One of the most widespread and well-
recognized consequences of nutrient enrichment
in coastal systems is the loss of submerged rooted
vascular plants (commonly referred to as
submerged aquatic vegetation [SAV]), which are
well adapted to low-nutrient, high-light
environments (Short and Wyllie-Echevarria
1996). Excess nutrients have been cited
specifically as a major cause of SAV loss in the
northern Gulf of Mexico, including every major
estuary in northwest Florida (USGS 2004). Other
causes of SAV loss, such as prop scarring,
disease, and even food web shifts also have been
identified (Dawes et al. 2004).
Historically, SAV beds extended along a
large fraction of the shoreline of Pensacola Bay
(Figure 8). Olinger et al. (1975) delineated the
extent of SAV beds from aerial photographs
taken as early as 1949. Aerial photography was
conducted for the purpose of highway
construction and was repeated at sporadic
intervals, according to highway construction
Floridatown
ca. 1950
nd
N
nd
nd
nd
nd
nd
1992
Figure 8. Coverage of submersed aquatic vegetation in
Pensacola Bay circa 1950 and in 1992. Data for circa
1950 were digitized from coverage maps reported by
Olinger et al. (1975). Areas for which no coverage data
was available are indicated by "nd." SAV coverage for
1992 is from the USGS National Wetlands Research
Center, Lafayette, LA.
18
-------
needs. The result is an irregular series of SAV
coverage maps (Olinger et al. 1975, also
reproduced by Lores et al. 2000). Although
several areas of the estuary never were surveyed
during this time period, SAV beds delineated
within the available imagery covered 17.7 km"
from 1949 to 1951 (Table 5). By overlaying the
circa 1950 SAV coverage map on the bathymetry
of the bay (based on 1993 surveys [Divins and
Metzger, public communication]), one can
estimate that SAV attained a maximum
colonization depth of as much as 6 m in
Pensacola Bay, 2 to 3 m in East Bay, and 1 to 2
m in the presently more turbid areas of northern
Escambia Bay. Based on average light
requirements for seagrasses (Duarte 1991),
average secchi depth may have been nearly 5 m
in portions of Pensacola Bay. Water clarity is
currently much less than this (Figure 5), although
several observations approaching this value were
made during the monthly surveys.
Table 5. Area (km2) of submerged aquatic vegetation (SAV) beds in the Pensacola
Bay system. Data for 1949 through 1951 were computed from maps reported by
Olinger et al. (1975). The computed area is a lower limit because some areas were
not covered by the aerial photography, which was conducted for the purpose of
highway construction. Omitted areas include Blackwater Bay, the western shore of
East Bay, western Pensacola Bay, and Santa Rosa Sound. Data for 1960,1980,1992,
and 2003 are from the USGS National Wetlands Research Center, Lafayette, LA.
thereafter (Table 5). The freshwater SAV species
Vallisneria americana accounted for >80% of the
1992 coverage in Escambia Bay, mostly in the
Escambia River delta, with limited beds of
Ruppia maritima also present (Lores et al. 2000).
These beds suffered massive mortality because of
high salinity in 2000 (Lores and Sprecht 2001)
and recovered only partially by 2003. Most of the
SAV habitats remaining in the Pensacola Bay
system are in Santa Rosa Sound. These beds are
dominated by Thalassia testudinum, with
Halodule wrightii also present. Although still
significant in coverage, we have observed that
the vegetation appears stunted and sparse
compared to apparently healthier beds in
northwest Florida (e.g., St. Joseph's Bay).
The decline in SAV coverage between
1950 and 1980 has been attributed to poor water
quality resulting from industrial pollution and, to
a very limited extent, to dredging for harbor
construction (Olinger et al. 1975). The causal
relationships between
current water quality
and the present
Basin
Escambia Bay
East Bay
Pensacola Bay
TOTAL
Santa Rosa Sound
1949-1951
4.82
8.21
4.70
17.73
nd
1960
1.05
4.76
3.71
9.52
25.00
1980
0.24
0.20
0.55
1.00
14.47
1992
1.78
0.69
1.14
3.62
11.17
2003
0.45
0.11
1.51
2.07
12.27
nd = No data
SAV surveys in Escambia Bay show that
significant SAV loss occurred within a few years
of the earliest surveys (Olinger et al. 1975), and
that SAV virtually was eliminated in much of the
Pensacola Bay system by 1980 (Schwenning
et al. 2007). Particularly rapid and complete SAV
loss occurred in the Floridatown area of northeast
Escambia Bay (Figure 8) following initiation of
industrial point source discharges in the area.
Approximately 50% of SAV was lost bay-wide
by 1960, and all but a few percent by 1980
(Table 5). SAV coverage in nearby Santa Rosa
Sound also declined about 50% between 1960
and 1980 and remained relatively constant
distribution of SAV are
not clear. The
underwater light
environment has been
cited as the most
important predictor of
SAV survival.
Thalassia requires 20%
to 25% of surface
irradiance at its
maximum colonization
depth (Duarte 1991, Dawes et al. 2004).
Assuming a conservative 25%, the light field in
Pensacola Bay is currently adequate to support
Thalassia to a depth of approximately 2 m in the
polyhaline reaches (where salinity is appropriate
for Thalassia), consistent with its present
maximum colonization depth.
SAV is absent from many places in
Pensacola Bay where light appears to be
adequate, an observation for which there are
many possible explanations (Koch 2001).
Metabolic stress associated with high sulfide in
sediment pore waters has been implicated as a
possible stressor in Pensacola Bay, where sulfide
19
-------
concentrations as high as 5 mM have been
measured (Devereux, unpublished data). Sulflde
is especially harmful to plants when water
temperature and salinity are high (Koch and
Erskine 2001), as is common during summer in
Gulf coast estuaries. Absence of Thalassia also
may result from poor propagation. Flowering and
seed production by Thalassia is reduced or
absent in northwest Florida, which is at the
northern limit of the more tropical range for that
species (Dawes et al. 2004). In the absence of
seed production, revegetation occurs slowly by
rhizome extension, which has been observed in
Pensacola Bay (Lores et al. 2000). H. wrightii
appears to be within both temperature and
salinity tolerances; the reasons for its absence are
not clear.
Based on the global and regional trends
relating nutrient enrichment and SAV loss (Short
and Wyllie-Echevarria 1996), one may infer that
SAV loss in Pensacola Bay is also a consequence
of nutrient enrichment. Because SAV beds create
important habitat for a variety of estuarine biota
(Dawes et al. 2004), their protection and eventual
restoration is critical to supporting aquatic life
uses of Pensacola Bay. Thus, the water quality
requirements for SAV growth and survival
should be a significant consideration for
determining nutrient criteria for Pensacola Bay.
Although much is known about seagrass ecology,
effective decision-support tools based on that
knowledge are not readily available for many
applications, including Pensacola Bay. SAV
growth models (e.g., Eldridge et al. 2004) have
the ability to integrate the effects of many factors
to predict SAV growth. Such models have been
implemented for many SAV species and have
been integrated in some cases with fully coupled
hydrodynamic and water quality models (Cerco
and Moore 2001). Although promising, the
results of these models cannot be transferred
across systems without adaptation to local
conditions, calibration, and validation. EPA is
presently adapting a SAV model developed for
Thalassia in Laguna Madre, TX (Eldridge et al.
2004), to growth conditions in Pensacola Bay, a
promising first step. However, approaches to
extrapolating from local SAV growth conditions
to appropriate water column nutrient
concentrations for the bay have not been
developed, and quantitative relationships
between nutrients and SAV are unclear.
As in the case of hypoxia, some policy
alternatives can be evaluated on the basis of our
knowledge of SAV ecology in Pensacola Bay.
First, consistent with Gulf-wide trends, SAV
habitats in Pensacola Bay are degraded presently
relative to their past condition. Second, the best
available science links losses of SAV to nutrient
enrichment both globally and in the Gulf of
Mexico region. In the case of Pensacola Bay, it
remains unclear to what extent, if any, current
nutrient concentrations or water clarity are an
impediment to SAV growth, coverage, or
restoration success. The weight of evidence
suggests that an increase in nutrients likely would
pose a risk of further degradation of these
important habitats and should be prevented
through appropriate numeric criteria. Meanwhile,
continued monitoring of SAV habitats should be
pursued at a regular interval, along with
advancements in predictive models relating water
quality conditions to growth and survival of SAV
species. An important need is relating SAV
requirements in the littoral habitats where they
occur to nutrient conditions at the larger scale to
which nutrient criteria likely would apply (e.g.,
Cerco and Moore 2001).
3.5 A Watershed Perspective
An increase in nitrogen and phosphorus
export from watersheds accompanies conversion
of primarily forested lands to cropland and
developed land uses (Reckhow et al. 1980,
Jordan et al. 1997). Therefore, information on the
impacts of human activity in a watershed
provides important insight into the extent to
which nutrient inputs to coastal waters may have
increased and the potential for management
actions to reduce nutrient enrichment and its
adverse impacts.
As of 2001, the Pensacola Bay watershed
included approximately 7% developed land and
6% cropland. The remaining land was forest
(including silviculture), shrub land, or pasture
and grassland (Figure 9). Nearly all of the
developed land is in the immediate vicinity of
Pensacola Bay, whereas the cropland is
concentrated in the lower Escambia River
20
-------
watershed (Figure 9). The relatively small extent
of development in the fluvial watershed
corresponds to the low human population
density, which has remained nearly constant at
about 10 km'2 from 1970 through 2000 (NOAA,
public communication). Population for the
watershed as a whole was 29 km"2 in 2000
(NOAA, public communication), very similar to
the 2000 median for Gulf of Mexico estuaries
(31 km"2), which, overall, are populated much
less densely than those in the Mid-Atlantic region
(182 km"2). As in many coastal areas, the
population in the Pensacola Bay watershed is
growing rapidly; population in the watershed
increased 42% between 1970 and 2000 (NOAA,
public communication). Developed land in Santa
Figure 9. Land cover in the Pensacola Bay watershed (data from the 2001
National Land Cover Database).
Rosa County, FL, one of two coastal counties in
the watershed is projected to increase 350% by
2060 (Zwick and Carr, 2006).
Nutrient loading from the watershed
remains relatively low, most likely because only
a small fraction of the watershed has been
developed or put into row-crop production.
Cherry and Hagy (2006) estimated the input from
rivers of dissolved inorganic nitrogen (DIN) and
phosphate (PO/") into Pensacola Bay. Based on
the ratio of average TN to average DIN for fresh
water (TN/DIN = 2), the estimated average input
of TN was 1.2 x 104 kg nitrogen day"1. When
scaled by the fluvial watershed area (17,069
km'), the average TN yield delivered from the
watershed to the bay was 262 kg nitrogen km"2
year"1. Using the same approach
for TP, the estimated average TP
load and average watershed TP
yield was 1.0 x 103 kg
phosphorus day"' and 21.8 kg
phosphorus km"2 year"1,
respectively. Both nitrogen and
phosphorus yields are consistent
with reported values for entirely
forested watersheds (Reckhow
et al. 1980), indicating that
current nutrient loadings from
upland areas of the Pensacola
Bay watershed are low. The
SPARROW model (Smith et al.
1997), which predicts nutrient
yields of watersheds at the
national scale, provides a point
of comparison for our estimates,
with a few caveats. The
SPARROW model estimate of
average nitrogen and
phosphorus yield for the
hydrologic units comprising the
Pensacola Bay watershed are
548 kg nitrogen km"2 year"1 and
53 kg phosphorus km"2 year"1.
Relative to the overall range of
values from SPARROW for the
conterminous United States,
these values are very similar to
our estimates for Pensacola Bay,
even though they are
BB Open Water
^] Developed, Open
| | Developed, Low
Developed, Med
Developed. High
Barren
Forest. Deciduous
Forest, Evergreen
Forest Mixed
Shrubland
Grassland
Pasture
Cropland
Wetland, Forested
Wetland, Herbaceous
21
-------
approximately twofold higher for both nitrogen
and phosphorus. As the developers of the
SPARROW model note, the estimates are most
reliable for comparing regional attributes of
watersheds but are less reliable at smaller scales
(Smith etal. 1997).
When nutrient export from the Pensacola
Bay watershed is expressed per unit of estuary
area, rather than per watershed area, the loadings
appear somewhat higher. Scaled in this way, the
TN and TP loading rates from the watershed to
Pensacola Bay are 12 g nitrogen m'2 year"1 and
1 g phosphorus m"" year" , respectively, well
within a moderate range of nutrient loading rates
for estuaries (Boynton et al. 1995). The contrast
between the relatively low yield rates and the
moderate loadings per unit of estuarine area
reflects the fact that the Pensacola Bay watershed
is nearly 50 times the size of the estuary,
approximately 2 times the median ratio for
estuaries (NOAA 1999b).
Atmospheric nitrogen deposition is a
possible source of anthropogenic nitrogen
enrichment that could impact Pensacola Bay
independent of land use change. Using local
measurements, Cherry and Hagy (2006)
estimated that wet nitrogen deposition in the
vicinity of the bay was 0.3 g nitrogen m2 year"1.
The 1999 to 2004 average for the watershed,
computed from national maps (NADP 2005; data
obtained electronically from source identified
therein), was comparable at 0.36 g nitrogen m2
year"1. Assuming that dry nitrogen deposition is
55% of wet deposition (ratio based on local data;
Hagy and Cherry 2006), atmospheric nitrogen
deposition is approximately 0.5 g nitrogen m"~
year"1 (= 500 kg nitrogen km"2 year"1). As a direct
input to the bay, this amounts to a small (4%)
fraction of TN inputs. In contrast, it is a
significant input to the watershed, about twofold
greater than nitrogen export via rivers. Assuming
that nitrogen exports via stream flow are 20% to
40% of the total of all nitrogen inputs to the
landscape (Boyer et al. 2002), atmospheric-
deposition may account for two-thirds of that
input. Given its likely quantitative significance, it
would be helpful to know if atmospheric nitrogen
deposition in the Pensacola Bay watershed has
increased substantially over time. Several facts
22
suggest that any increase was probably fairly
small. First, nitrogen concentrations in rainfall
remain among the lowest for U.S. coastal
watersheds (14 uM; NADP 2005). Only U.S.
Pacific coast estuaries have significantly lower
nitrogen concentrations in rainfall (NADP 2005).
Because U.S. nitrogen oxide emissions in 1990
were 2.3-fold higher than in 1900 (U.S. EPA
1995), and the largest increases probably
occurred where deposition presently is elevated,
the proportional increase in the Pensacola Bay
watershed was likely less.
The above observations regarding the
Pensacola Bay watershed suggest that significant
reductions in nutrient loading from the fluvial
portion of the watershed may not be possible,
even if slight improvements could be achieved by
implementing best management practices on
existing agricultural land, managed forests,
developed land, and point sources. Because the
landscape in the immediate vicinity of the bay is
much more intensely developed and populated,
loadings from that segment of the watershed may
be higher and, therefore, potentially could be
reduced through improved nutrient management.
For the largest portion of the watershed,
however, the risk of a significant increase in
nutrient loading because of conversion of forest
to urban and suburban development (Wickham
et al. 2002) is much greater than is the potential
for nutrient reductions.
3.6 Nutrient Criteria
Below we examine the information that
we used and the logic that we followed to define
an approach for deriving nutrient criteria for
Pensacola Bay. We then apply the approach to
derive hypothetical criteria. We also examine
how the scientific information about the bay
influenced the approach that was followed.
In our evaluation of Pensacola Bay water
quality, we examined the environmental history
of the bay, current attributes of the phytoplankton
community, the prevalence of low DO (i.e.,
hypoxia), the status of seagrass (SAV) habitats,
and the status of land use and nutrient exports
from the watershed. We related these elements
via a conceptual model of key ecosystem
processes related to nutrients (Figure 2). We also
evaluated water quality in Pensacola Bay in a
-------
comparative context by applying a modified
percentile approach in an ecoregional analysis of
water quality (Figure 5, Table 2).
We evaluated these data by posing a
series of questions for which the answers, as well
as the uncertainty associated with them,
influenced the logical progression. It can be
informative to consider the questions (below)
separately, but they also may be considered part
of a logic path (Figure 10).
• Has water quality declined relative to a
documented historical condition?
• Is the estuary degraded in comparison to
similar estuaries (i.e., estuaries in the same
class?)
• Are there environmental measures or indicators
of conditions widely associated with nutrient
overenrichment?
• Are nutrient loads significantly enriched
because of anthropogenic causes?
At the end of the process, we determined
that a nutrient criteria for Pensacola Bay
reasonably could be defined by the current water
quality conditions (Figure 10). Below, we explain
the process, as outlined in Figure 10, in greater
detail.
A Is water quality
degraded relative to a
documented pristine
historical condition?
W No,
Historical reference values
may inform nutrient criteria
development
or inconclusive
D
Is water quality
degraded compared
to similar estuaries?
w No,
Additional
monitoring and
condition
assessment
needed.
incon-
clusive
or inconclusive
C
Are uses impaired by
conditions commonly
associated with nutrient
overenrichment?
/*
1
Define criteria
based on current
water quality.
)
G Values derived from
percentile approach may be
useful for criteria
development
Yes
D
Are nutrient loads
significantly enriched due to
anthropogenic causes?
/No Yes
E Define criteria based
on current water
quality. Identify
causes of impairment
if possible.
<
H Nutrient-response
relationships or related
models needed to set
criteria to resolve degraded
condition.
Figure 10. Flow diagram illustrating the logical path followed to determine an
appropriate approach for determining nutrient criteria for Pensacola Bay (boxes A
through E) and a subset of possible alternative paths (boxes F through J) that may
have been followed had the scientific findings been different.
23
Our analysis began with the historical
data. We concluded that there were no data to
indicate clearly that water quality in Pensacola
Bay had declined because of nutrients (box A in
Figure 10). This effectively eliminated the
possibility of deriving criteria based, in large
part, on a historical reference condition approach
(box F). Had the available historical data shown a
substantial decline in water quality, we may have
considered a greater role for historical data in
deriving nutrient criteria (box F). Even if this had
been the case, additional supporting information
still would be useful to further evaluate and
support the values.
Comparison of water quality in Pensacola
Bay with other estuaries (box B in Figure 10) in
its ecoregion-based class (Table 2, Figure 5)
showed that water quality was often among the
best in the class, and that it was not degraded
relative to the class. Because current values for
many water quality indicators were better than
the comparison values obtained from the
percentile approach, it did not seem reasonable to
use these values as criteria. If, on the other hand,
the comparison values determined via the
percentile approach were better than current
values, one might reasonably
begin with the comparison
values as criteria, then
provide additional
supporting evidence as
possible (box G).
Despite finding
comparatively good water
quality in terms of nutrient
concentrations, water clarity,
and chlorophyll-a, Pensacola
Bay has both significant
bottom water hypoxia and
extensive loss of seagrass
habitat, conditions widely
associated with nutrient
enrichment in estuaries (box
C in Figure 10). Because
hypoxia and seagrass loss
have well-known negative
impacts on estuarine
systems, we expect that
human and aquatic life uses
-------
are limited by these conditions. Moreover,
seagrass loss and hypoxia often increase with
anthropogenic nutrient inputs. The juxtaposition
of seagrass loss and hypoxia with otherwise good
water quality presents a relatively complex
picture for nutrient criteria in Pensacola Bay. It
appears that much of the loss occurred during the
1950s and 1960s in association with industrial
nutrient and organic inputs that were eliminated
by the early 1970s. Thus, these causes are not
present today. The factors that impede recovery,
as well as the prospects and best approach for
successful restoration could be understood better
through additional research. The significant
extent of hypoxia in the bay appears to reflect a
natural susceptibility of the bay to hypoxia. It is
not clear that realistic reductions in nutrients
could eliminate hypoxia. Scientific uncertainty
surrounding this conclusion also could be
reduced through additional research, particularly
improved modeling of the coupled physical-
chemical dynamics of the ecosystem. Both SAV
loss and hypoxia commonly are associated in
estuarine and coastal waters with significant
phytoplankton blooms, sometimes including
harmful algal species. However, phytoplankton
production and biomass in Pensacola Bay are
relatively modest compared to other systems.
High-biomass phytoplankton blooms occur
infrequently; harmful algal blooms are not known
to develop in Pensacola Bay but, on occasion,
have been transported into the bay from the Gulf
of Mexico (Tester and Steidinger, 1997). It
appears that Pensacola Bay is prone to adverse
impacts resulting from nutrient enrichment and
already has been impacted in the past by
nutrients, but it may not be demonstrating clear
symptoms of nutrient enrichment at present.
To better evaluate this hypothesis, we
examined nutrient inputs to the bay (box D in
Figure 10). EPA did not include evaluating
nutrient sources as part of a recommended
criteria development process for freshwater
systems (U.S. EPA 2000a,b). Rather, this step
usually is undertaken as part of the total daily
maximum load process. However, this analysis
was a useful diagnostic step for Pensacola Bay
and could prove to be important for criteria
development for other estuaries. Although there
were no historical baseline data with which
current loading data could be compared, our
analysis of current nutrient inputs from rivers,
land use in the watershed, and atmospheric
nitrogen deposition suggested that it was unlikely
that nutrient inputs have increased dramatically.
Evaluating water quality in the major rivers
entering the bay showed that nutrient export from
the watershed was similar to pristine or nearly
pristine watersheds. This observation was
supported by the absence in the watershed of
major causes of increased nutrient export (e.g.,
low population, minimal row-crop agriculture,
low atmospheric nitrogen deposition). One could
conclude that reasonable management actions
probably could not effect significant reductions
in nutrient inputs to the bay via the major
tributary rivers. Further research could reduce the
scientific uncertainty regarding nutrient loads
from the more developed portions of the
watershed immediately surrounding the bay,
which remain poorly quantified.
Ultimately, our approach to developing criteria
(box E in Figure 10) was the same as if no
impaired conditions were identified (box 1). In
each situation, we suggest that it is scientifically
justified to define water quality criteria as the
current water quality conditions because there is
little justification for alternative, more stringent
criteria. Table 6 outlines hypothetical criteria that
are based on summer medians, by salinity zone,
for the 1996 to 2004 timeframe. We computed
criteria for chlorophyll-a, secchi depth, DIN,
PCV", TN, TP, and DO. For concentration-based
water quality indicators (e.g., chlorophyll-a), the
criteria include a 10% buffer and a suitable
averaging period (e.g., 3 years), such that small
climate-driven variations surrounding current
water quality would not trigger a determination
of impairment. Because we concluded that a
portion of the bay may be subject to hypoxia
even in the absence of anthropogenic nutrient
enrichment, DO criteria for bottom water must
accommodate this feature of the ecosystem in
some way. Our evaluation of oxygen dynamics in
the bay indicates that a sensible approach would
define hypoxia and limit the acceptable extent in
24
-------
Table 6. Suggested nutrient and nutrient-related water quality criteria for
Pensacola Bay resulting from the weight-of-evidence approach described
herein. All values are 3-year moving averages of summer medians. The
quantities are based on observed values (Table 2) that have been adjusted to
provide 10% allowance (i.e., 10% higher for concentrations, 10% lower for
secchi depth).
Parameter
Criteria Recommendation
Chlorophyll-a Summer median concentration <8 |jg L" (all salinity zones)
Secchi Depth Summer median >0.8 m (salinity 0-5), >1.1 m (salinity 5-18),
>1.8 m (salinity >18)
DO Bottom waters below a pycnocline must have oxygen >2.0 mg L"
except for in an area that may average up to 25% of the mean
low water surface area of the bay.
Surface waters or unstratified waters: oxygen >4.0 mg L"1 at all
times. 24-hour average oxygen >5.0 mg L"1
DIN Summer median <10 pM (salinity <5), <1.0 pM (salinity 5-18),
<0.5 MM (salinity >18)
PO43" Summer median <0.2 |jM (salinity <5), <0.12 pM (salinity 5-18),
<0.05 pM (salinity >18)
TN Summer median <35 pM (all salinity zones)
TP Summer median <1 pM (salinity 0-5), <0.6 pM (salinity 5-18),
<0.5 pM (salinity >18)
near-bottom waters (e.g., 0.5 m above the
bottom). An approach that defines either a
DIN = NO2" + NO3" + NH4* concentration minimum for
all bottom waters of the bay or even a lower limit
for the median concentration in the bay provides
less precise regulation of the extent of hypoxia.
Scientifically defensible definitions of hypoxia
for Pensacola Bay could arguably vary from 2 to
as high as 4 mg L"1, provided the limits of extent
are defined accordingly. In our example, we
defined hypoxia as <2.0 mg L"', the level at
which relatively hypoxia-tolerant species begin
to exhibit significant mortality (Diaz and
Rosenberg 1995). The corresponding limit for
average extent of hypoxia (over several years) is
approximately 25% of the surface area of the
bay, or 90 km2 (Table 6). Concerns related to
natural extent of hypoxia do not apply for surface
waters, where Florida's existing statute is both
consistent with existing water quality conditions
and with expectations of aquatic life uses
(Table 6).
Given the complex ecology and
ecological history of Pensacola Bay, it would not
be surprising if other estuaries were similarly
complex. However, we expect that criteria
development for many estuaries could follow a
more typical model. For example, one might find
that water quality in an estuary is degraded
relative to some limited
historical data, that values
also are degraded relative
to ecoregional reference
sites, that commonly
observed nutrient-related
problems are present, and
that nutrient inputs are
elevated because of
obvious anthropogenic
causes. Here, a coherent
picture emerges and
would be supported by
several lines of evidence,
each of which would
provide support and
insight and final criteria
determination. In an
estuary where reasonable
criteria would stipulate
reduced nutrients and
improved water quality, some additional
elements may be helpful or even necessary (box
H). For example, models describing relationships
among water quality variables may be helpful to
ensure that criteria for causal variables (i.e.,
nutrients) are reasonably consistent with
expected values for response variables (DO and
chlorophyll-a). Criteria other than those proposed
here, including biocriteria, could be useful where
the indicators that we used fail to capture the
ecological impacts resulting from nutrient
overenrichment. For example, a harmful
proliferation of macroalgae occurs in some
estuaries subject to nutrient enrichment and is an
important measure of the overall response to
nutrients. Development of additional case studies
would be a useful way to demonstrate approaches
for deriving criteria in situations that were not
encountered in the case of Pensacola Bay. Such
case studies, like this one, could provide a
template that could be adapted as desired as
States develop criteria for their estuaries.
25
-------
-------
4. Summary and Conclusions
We developed and applied an approach
for developing numerical water quality criteria
for nutrients and nutrient-related water quality
indicators in an estuary. The approach is best
defined as a weight-of-evidence approach whose
important elements include those below.
• Developing a conceptual model to organize
and present the important ecological science
and other information related to nutrients in the
system.
• Comparing water quality and other ecosystem
attributes in the system with values from other
estuarine ecosystems in a reasonable
comparison group.
• Evaluating the incidence of actual or potential
nutrient-related water quality impacts in the
ecosystem. The most common examples
include hypoxia, SAV loss, and high-biomass
or harmful phytoplankton or macroalgal
blooms.
• Evaluating the extent to which impacts have
been caused by anthropogenic nutrient
enrichment using historical observations,
analysis of ecosystem processes, ecological
models, or other approaches.
• Evaluating the extent to which nutrient inputs
to the system are likely to have been increased
by human activities.
An additional desirable element of an
analysis supporting criteria development is an
assessment of the risks for future environmental
changes, including the principal drivers of such
changes. Such an assessment would help inform
future programmatic reassessments of water
quality criteria. Our analysis of the Pensacola
Bay ecosystem indicates that increased human
development associated with current population
trends poses a significant risk for increased
nutrient inputs. The most sensitive and easily
monitored indicators of enrichment are likely to
be nutrient inputs themselves (especially in
freshwaters), phytoplankton biomass, and extent
of hypoxia. SAV extent is presently too
degraded; is too sensitive to other impacts,
particularly salinity; and appears too slow to
respond to improved conditions to serve as a
reliable indicator of enrichment for most of
Pensacola Bay.
We believe that the weight-of-evidence
approach we outlined here and applied to
Pensacola Bay would be applicable for many
estuarine systems. The approach is most
applicable where enrichment impacts are
relatively modest and where a suitable
comparison group provides an adequate context
that might be used to justify criteria requiring
some reduction in nutrients. The approach may
be a useful point of departure for systems that
have been impacted seriously by nutrients.
However, in these cases, additional tools likely
will be needed to support criteria development.
Better methods would be needed to determine
what level of aquatic life uses are achievable
(including what uses were supported historically)
and to identify the water quality requirements for
those uses and, particularly, the relationships
among different water quality indicators (e.g.,
nutrients, chlorophyll-a, hypoxia).
27
-------
-------
5. References
Academy of Natural Sciences of Philadelphia. 1953.
Escambia River, Florida, vicinity of the Chemstrand
Corporation Plant, Fall 1952 - Spring 1953. Stream
survey report for the E. 1. du Pont de Nemours and
Company. Department of Limnology, Academy of
Natural Sciences of Philadelphia, Philadelphia, PA.
December 1953. 165pp.
Boyer, E. W., C. L. Goodale, N. A. Jaworski, and R. W.
Howarth. 2002. Anthropogenic nitrogen sources and
relationships to riverine export in the northeastern
U.S.A. Biogeochemistry 57/58:137-169.
Boynton, W. R. and W. M Kemp. 2007. Estuaries. In
Capone, D. G., D. A. Bronk, M. R. Mulholland, and
E. J. Carpenter (eds.). Nitrogen in the Marine
Environment. Second Edition, Academic Press, New
York. In press.
Boynton, W. R., W. M. Kemp, and C. W. Kcefe. 1982.
A comparative analysis of nutrients and other factors
influencing estuarine phytoplanklon production.
pp. 69-90. In V. S. Kennedy, editor. Estuarine
Comparisons. Academic Press, New York.
Boynton, W. R., J. H. Garber, R. Summers, and W. M.
Kemp. 1995. Inputs, transformations, and transport of
nitrogen and phosphorus in Chesapeake Bay and
selected tributaries. Estuaries 18(1B):285-314.
Campbell, J. G. and L. R. Goodman. 2004. Acute
sensitivity of juvenile shortnose sturgeon to low
dissolved oxygen concentrations. Transactions of the
American Fisheries Society 133:772-776.
Cerco, C. F. and K. Moore. 2001. System-wide submerged
aquatic vegetation model for Chesapeake Bay.
Estuaries 24(4):522-534.
Cherry, J. S. II and J. D. Hagy III. 2006. Pensacola Bay,
Florida USA. Land-Ocean Interactions in the Coastal
Zone (LOICZ) Biogeochemical Budgets.
Electronically published by the Land-Ocean
Interactions in the Coastal Zone (LOICZ) programme.
(http://www.loicz.org).
Cloern, J. E. 2001. Our evolving conceptual model of the
coastal eutrophieation problem. Marine Ecology
Progress Series 210:223-253.
Cowan, J. L., J. R. Pcnnock, and W. R. Boynton. 1996.
Seasonal intcrannual patterns of sediment-water
nutrient and oxygen fluxes in Mobile Bay, Alabama
(USA): regulating factors and ecological significant.
Marine Ecology Progress Series 141:229-245.
Dame, R.. M. Alber. D. Allen, M. Mallin, C. Montague,
A. Lewitus, A. Chalmers, R. Gardner, C. Oilman,
B. Kjerfve, J. Pinckney, and N. Smith. 2000. Estuaries
of the south Atlantic coast of North America: Their
geographical signatures. Estuaries 23(6):793-819.
Dawes, C. J., R. C. Phillips, and G. Morrison. 2004.
Seagrass Communities of the Gulf Coast of Florida:
Status and Ecology. Florida Fish and Wildlife
Conservation Commission, Fish and Wildlife Research
Institute and the Tampa Bay Estuary Program.
St. Petersburg, FL. 74 pp.
Devereax, R., personal communication. U.S.
Environmental Protection Agency, National Health
and Environmental Effects Research Laboratory, Gulf
Breeze, FL. December 2006.
Diaz, R. J. 2001. Overview of hypoxia around the world.
Journal of Environmental Quality 30(2):275-281.
Diaz, R. J. and R. Rosenberg. 1995. Marine benthic
hypoxia: A review of its ecological effects and the
behavioral responses of benthic macrofauna.
Oceanography and Marine Biologv Annual Review
33:245-303.
Divins, D. L. and D. Metzger, public communication,
NGDC Coastal Relief Model. National Geophysical
Data Center. National Oceanic and Atmospheric
Administration, Boulder, CO. Accessed November
2006
(http://www.ngdc.noaa.gov/mgg/coastal/coastal.html).
Duarte, C. M. 1991. Seagrass depth limits. Aquatic Botanv
40:363-377.
Eldridgc, P. M., J. E. Kaldy, and A. B. Burd. 2004. Stress
response model for the tropical seagrass Thalassia
tesiudinum: The interactions of light, temperature,
sedimentation and geochemistry. Estuaries 27(6):923-
937.
FAC 2005. Section 62-302.530(31), Florida Administrative
Code. Table: Criteria for Surface Water Quality
Classifications. Effective July 27, 2005. Accessed at
U.S. EPA Internet site
(www.epa.gov/waterscience/standards/wqslibrary).
Fitzpatrick, J. J. and M. B. Meyers. 2000. A critical look at
criteria. A review of methodologies thai determine
nutrient criteria for coastal and estuarine systems.
Water Environment and Technology 12(2):34-39.
29
-------
GESAMP (IMO/FAO/UNESCO-
IOC/WMOAVHO/IAEA/UN/UNEP Joint Group of
Experts on the Scientific Aspects of Marine
Environmental Protection) and Advisory Committee
on Protection of the Sea. 2001. Protecting the oceans
from land-based activities - Land-based sources and
activities affecting the quality and uses of the marine,
coastal and associated freshwater environment.
GESAMP Reports and Studies Number 71. 162 pp.
Griffith, G. E., J. M. Omernik, C. W. Rohm, and S.M.
Pierson. 1994. Florida regionalizalion project. U.S.
Environmental Protection Agency, National Health
and Environmental Effects Research Laboratory,
Western Ecology Division, Corvallis, OR.
EPA/600/Q-95/002. 83 pp.
Hagy, J. D., Ill and M. C. Murrell. 2007. Susceptibility of a
northern Gulf of Mexico estuary to hypoxia: an
analysis using box models. Estuarine, Coastal and
Shelf Science 74:239-253.
Homer, C., C. Huang, L. Yang, B. Wylie, and M. Coan.
2004. Development of a 2001 national land cover
database for the United States. Photogrammetric
Engineering ana1 Remote Sensing 70(7):829-840.
Hopkins, T. S. 1969. The Escambia River and Escambia
Bay during summer 1969. University of West Florida,
Pcnsacola, FL (unpublished manuscript). 54 pp.
Hopkinson, C. S., Jr. 1985. Shallow-water benlhic and
pelagic metabolism: evidence of heterotrophy in the
nearshore Georgia Bight. Marine Biologv 87:19-32.
Jordan, T. E., D. L. Correll, and D. E. Wellcr. 1997.
Relating nutrient discharges from watersheds to land
use and streamflow variability. Water Resources
Research 33( 11 ):2579-2590.
Juhl, A. R. and M. C. Murrell. 2005. Interactions between
nutrients, phytoplankton growth, and
microzooplankton grazing in a Gulf of Mexico estuary.
Aquatic Microhial Ecology 38:147-156
Karouna-Renier, N. K., R. A. Snyder, J. G. Allison, M. G.
Wagner, and K. R. Rao. 2007. Accumulation of
organic and inorganic contaminants in shellfish
collected in estuarine waters near Pensacola, Florida:
Contamination profiles and risks to human consumers.
Environmental Pollution 145:474-488.
Kemp, W. M., E. M. Smith, M. Marvin-DiPasquale, and
W.R. Boynton. 1997. Organic carbon balance and net
ecosystem metabolism in Chesapeake Bay. Marine
Ecology Progress Series 150:229-248.
Kennish, M. J. 1986. Ecology of Estuaries, Vol. I: Physical
and Chemical Aspects. CRC Press, Boca Raton. FL.
254 pp.
Koch, E. W. 2001. Beyond light: physical, geological, and
geochemical parameters as possible submersed aquatic
vegetation habitat requirements. Estuaries 24( 1 ):1-17.
Koch, M. S and J. M. Erskmc. 2001. Sulfidc as a
phytotoxin to the tropical seagrass Thalassia
lestudinum: interactions with light, salinity and
temperature. Journal of Experimental Marine Bio/ogr
and Ecology 266:81-95.
Kurtz, J. C., N. D. Detenbeck, V. D. Engle, K. Ho, L. M.
Smith, S. J. Jordan, and D. Campbell. 2006.
Classifying coastal waters: Current necessity and
historical perspective. Estuaries and Coasts 29(1): 107-
123.
Lores, E. M. and D. T. Sprecht. 2001. Drought-induced
decline of submerged aquatic vegetation in Escambia
Bay, Florida. Gulf of Mexico Science 19(2):16M64.
Lores, E. M., E. Pasko, J. Patrick, R. L. Quarles,
J. Campbell, and J. Macauley. 2000. Mapping and
monitoring of submerged aquatic vegetation in
Escambia - Pensacola Bay System, Florida. Gulf of
Mexico Science 18:1-14.
Maclntyre, H. L., T. M. Kana, T. Aiming, and R. J. Geider.
2002. Photoacclimation of photosynthesis irradiance
response curves and photosynthetic pigments in
microalgae and cyanobacteria. Journal of Phvcologv
38:17-38.
May, E. B. 1973. Extensive oxygen depletion in Mobile
Bay, Alabama. Limnologv and Oceanographv
18(3):353-366.
McRae, G., M. Julian, and H. Brown. 2004. Florida's
Inshore Marine Monitoring and Assessment Program
(IMAP). Annual Report, Year Five. Florida Fish and
Wildlife Conservation Commission, Fish and Wildlife
Research Institute, St. Petersburg, FL. 97 pp.
Murrell, M., personal communication. U.S. Environmental
Protection Agency, NHEERL/Gulf Ecology Division.
1 Sabine Island Drive, Gulf Breeze, FL 32563 (data
provided November 2006).
Murrell, M. C. and J. M. Caffrey. 2005. High
eyanobacterial abundance in three northeastern Gulf of
Mexico estuaries. Gulf and Caribbean Research
17:95-106.
Murrell, M.C. and E. M. Lores. 2004. Phytoplankton and
zooplankton seasonal dynamics in a subtropical
estuary: importance of cyanobacteria. Journal of
Plankton Research 26(3):371-382.
Murrell, M. C., R. S. Stanley, E. M. Lores, G. T. DiDonato,
L. M. Smith, and D. A. Flemer. 2002. Evidence that
phosphorus limits phytoplankton growth in a Gulf of
Mexico estuary: Pensacola Bay, Florida, USA. Bulletin
of Marine Science 70( 1): 155-167.
Murrell. M. C., E. M. Lores, J. D. Hagy III, R. M. Greene.
2007. Phytoplankton production in relation to nutrient
distributions in a sub-tropical estuary: importance of
freshwater flow. Estuaries and Coasts 30(3):390-402.
30
-------
NADP. 2005. National Atmospheric Deposition Program.
2004. Annual Summary. NADP Data Report, 2005-01.
Illinois State Water Survey, Champaign, IL. 16 pp.
Nixon, S. W. 1995. Coastal marine eutrophication:
A definition, social causes, and future concerns.
Ophelia 41:199-219.
NOAA. 1997. NOAA's Estuarine Eutrophication Survey,
Volume 4: Gulf of Mexico Region. National Oceanic
and Atmospheric Administration, Office of Ocean
Resources Conservation and Assessment, Silver
Spring, MD. 77 pp.
NOAA. 1998. NOAA's Estuarine Eutrophication Survey,
Volume 5: Pacific Coast Region. National Oceanic and
Atmospheric Administration, Office of Ocean
Resources Conservation and Assessment, Silver
Spring, MD. 75 pp.
NOAA 1999a. Coastal Assessment Framework. Special
Projects Office, National Ocean Service, National
Oceanic and Atmospheric Administration, Silver
Spring, MD.
NOAA 1999b. Physical and Hydrologic Characteristics of
Coastal Watersheds. Coastal Assessment and Data
Synthesis System, 1999. National Coastal Assessments
Branch, Special Projects Office, National Ocean
Service, National Oceanic and Atmospheric
Administration, Silver Spring, MD. 6 pp.
NOAA, public communication. Spatial Trends in Coastal
Socioeconomics, Coastal Ocean and Resource
Economics Program, National Ocean Service, National
Oceanic and Atmospheric Administration, Silver
Spring, MD. Accessed February 2007.
NRC. 2000. Clean Coastal Waters: Understanding and
Reducing the Effects of Nutrient Pollution. Ocean
Studies Board and Water Science and Technology
Board, Commission on Geosciences, Environment.
and Resources, National Research Council.
Washington, DC.
Olinger, L. W., R. G. Rogers, P. L. Fore, R. L. Todd, B. L.
Mullins, F. T. Bisterfield, and L. A. Wise. 1975.
Environmental and recovery studies of Escambia Bay
and the Pensacola Bay system, Florida. U.S.
Environmental Protection Agency, Atlanta, GA.
904/9-76-016. 468pp.
Omernik, J. M. 1987. Ecoregions of the conterminous
United States. Map (scale 1:7,500,000). Annals of the
Association of American Geographers 77(1 ):118-125.
Omernik, J. M. 1995. Ecoregions: A spatial framework for
environmental management, pp. 49-62. In Davis, W. S.
and T. P. Simon (eds.). Biological Assessment and
Criteria: Tools for Water Resource Planning and
Decision Making. Lewis Publishers, Boca Raton. FL.
Reckhow, K. H.. M. N. Beaulac, and J. T. Simpson. 1980.
Modeling phosphorus loading and lake response under
uncertainty: a manual and compilation of export
coefficients. U.S. Environmental Protection Agency.
Washington. DC. EPA/440/5-80/011.
Schroeder, W. W. and W. J. Wiseman, Jr. 1988. The
Mobile Bay estuary: Stratification, oxygen depletion
and jubilees. In Kjerfve, B. (ed.) Hydrodynamics of
Estuaries, Volume II, Estuarine Case Studies. CRC
Press, Boca Raton, FL. pp. 41-529.
Schroeder, W.W. and W. J. Wiseman, Jr. 1999. Geology
and hydrodynamics of Gulf of Mexico estuaries, pp. 3-
28 In Bianchi, T. S., J. R. Pennock,, and R. R. Twilley
(eds.) Biogeochcmistry of Gulf of Mexico Estuaries.
John Wiley and Sons, Inc., New York.
Schwenning, L., T. Bruce, and L. W. Handley. 2007.
Pensacola Bay. pp. 129-142 in Handley, L., D.
Altsman, and R. DeMay (eds.). Seagrass Status and
Trends in the Northern Gulf of Mexico: 1940-2002.
U.S. Geological Survey Scientific Investigations
Report 2006-5287 and U.S. Environmental Protection
Agency. 855-R-04-003. 267 pp.
Short, F. T. and S. Wyllie-Echevarria. 1996. Natural and
human-induced disturbance of seagrasses.
Environmental Conservation 23( 1): 1 7-27.
Smith, R. A., G. E. Schwarz, and R. B. Alexander. 1997.
Regional interpretation of water-quality monitoring
data. Water Resources Research 33( 12):2781 -2798.
Stephan, C. E.. D. I. Mount, D. J. Hansen, J. H. Gentile,
G. A. Chapman, and W. A. Brungs. 1985. Guidelines
for deriving numerical national water quality criteria
for the protection of aquatic organisms and their uses.
National Technical Information Service, Springfield,
VA. PB85-227049.
Summers, J. K., J. F. Paul, and A. Robertson. 1995.
Monitoring the ecological condition of estuaries of the
United States. Toxicological and Environmental
Chemistry 49:93-108.
Tester, P. A. and K. A. Steidinger. 1997. Gymnodinium
breve red tide blooms: initiation, transport, and
consequences of surface circulation. Limnologv and
Oceanography 42(5): 1039-105 1.
U.S. EPA. 1995. National air pollutant emission trends,
1900-1994. U.S. Environmental Protection Agency,
Washington, DC. EPA-454/R-95-011. 219 pp.
U.S. EPA. 1998. National strategy for the development of
regional nutrient criteria. U.S. Environmental
Protection Agency, Washington, DC. EPA 922-R-98-
002. 47 pp.
31
-------
U.S. EPA. 1999a. Protocol for developing nutrient
TMDLs. First Edition. Office of Water (4503F).
United Stales Environmental Protection Agency.
Washington, DC. EPA 841-B-99-007. 135 pp.
U.S. EPA. 1999b. Ecological Condition of Estuaries in the
Gulf of Mexico. U.S. Environmental Protection
Agency, Office of Research and Development,
National Health and Environmental Effects Research
Laboratory, Gulf Ecology Division, Gulf Breeze, FL.
EPA 620-R-98-004.
U.S. EPA. 2000a. Nutrient criteria technical guidance
manual: Lakes and reservoirs. U.S. Environmental
Protection Agency, Washington, DC. EPA-822-BOO-
001.232pp.
U.S. EPA. 2000b. Nutrient criteria technical guidance
manual: River and Streams. U.S. Environmental
Protection Agency, Washington, DC. EPA-822-BOO-
002. 254 pp.
U.S. EPA. 2000c. Ambient water quality criteria
recommendations. Information supporting the
development of state and tribal nutrient criteria. Lakes
and reserviors in nutrient ecoregion XII. U.S.
Environmental Protection Agency, Washington, DC.
EPA822-B-00-013. 77pp.
U.S. EPA. 2000d. Ambient water quality criteria
recommendations. Information supporting the
development of state and tribal nutrient criteria. Rivers
and streams in nutrient ecoregion XII. U.S.
Environmental Protection Agency, Washington, DC.
EPA822-B-00-021.80pp.
U.S. EPA. 2001. Nutrient Criteria Technical Guidance
Manual. Estuarine and Coastal Marine Waters. U.S.
Environmental Protection Agency, Washington. DC.
EPA-822-B-01-003.
U.S. EPA. 2003. Ambient Water Quality Criteria for
Dissolved Oxygen, Water Clarity, and Chlorophyll a
for Chesapeake Bay and its Tidal Tributaries. U.S.
Environmental Protection Agency, Region III
Chesapeake Bay Program Office, Office of Water.
Washington, DC. EPA 903-R-03-002.
U.S. EPA. 2004. National Coastal Condition Report II.
U.S. Environmental Protection Agency, Office of
Research and Development, Office of Water,
Washington, DC. EPA-620/R-03/002.
U.S. EPA. 2005. The Ecological Condition of the
Pcnsacola Bay System, Northwest Florida (1994-
2001). U.S. Environmental Protection Agency, Office
of Research and Development, National Health and
Environmental Effects Research Laboratory, Gulf
Ecology Division, Gulf Breeze, FL. EPA/620/R-
05/002.
U.S. EPA/OWOW, public communication. U.S.
Environmental Protection Agency, Office of Wetlands,
Oceans, and Watersheds, Total Maximum Daily Load
Program Web site. Accessed December 2006
(http://www.epa.gov/owow/tmdl).
USGS. 2004. Seagrass habitat in the northern Gulf of
Mexico: Degradation, conservation and restoration of a
valuable resource. United States Geological Survey,
Gulf of Mexico Program Office, Stennis Space Center,
MS. 855-R-04-001.26pp.
U.S. House of Representatives. 2000. United Slates Code.
2000 Edition. Office of the Law Revision Counsel of
the United States House of Representatives, U.S.
Government Printing Office, Washington, DC. Title
33, beginning with section 1251 accessed at
http://www.gpoaccess.gov/uscode.
Wickham, J. D., R. B. O'Neill, K. H.Ritters, E. R. Smith,
T. G. Wade, and K. B. Jones. 2002. Geographic
targeting of increases in nutrient export due to future
urbanization. Ecological Applications 12(1 ):93-106.
Zwick, P. D. and M. H. Carr. 2006. Florida 2060, A
population distribution scenario for the State of
Florida. Geoplan Center, University of Florida,
Gainesville, FL. 29 pp.
32
-------
Appendix A
Observations in EPA's Level III Ecoregion 75
Estuaries included in EPA's Level III ecoregion 75, defined as the "Southeast Coastal Plain,"
and the number of surface water quality observations used to compute median water quality values for
surface water. The estuaries are located between West Mississippi Sound, MS, and Winyah Bay, SC.
Number of
Estuarine Drainage Area Observations
Altamaha River, GA 8
Apalachee Bay, FL 19
Apalachicola Bay, FL 58
Breton/Chandeleur Sound, MS 4
Broad River, SC 27
Charleston Harbor, SC 14
Charlotte Harbor, FL 48
Choctawhatchee Bay, FL 37
East Mississippi Sound, MS/AL 32
Indian River, FL 139
Mobile Bay, AL 52
North/South Santee rivers, SC 3
Ossabaw Sound, GA 13
Pensacola Bay, FL 70
Perdido Bay, FL 18
Sarasota Bay, FL 58
Savannah River, SC 14
St. Andrew Bay, FL 62
St. Andrew/St. Simons sounds, GA 21
St. Catherines/Sapelo sounds, GA 25
St. Helena Sound, SC 28
St. Johns River, FL 65
St. Marys River/Cumberland Sound, FL/GA 16
Stono/North Edisto rivers, SC 13
Suwannee River, FL 57
Tampa Bay, FL 74
West Mississippi Sound, MS 61
Winyah Bay, SC 7
TOTAL 1,043
33
-------
Appendix B
Observations in Florida's Level IV Ecoregion 75a
Estuaries included in Florida's Level IV ecoregion 75a, defined as the "Gulf Coast Flatwoods
Region," and the number of surface water quality observations used to compute median water quality
values for surface water. The estuaries are located in the northern Florida Panhandle between Perdido
Bay, FL, and Suwannee River, FL. The data are from Florida's Inshore Marine Monitoring and
Assessment Program.
Number of
Estuarine Drainage Area Observations
Apalachee Bay, FL 18
Apalachicola Bay, FL 55
Choctawhatchee Bay, FL 36
Pensacola Bay, FL 67
Perdido Bay, FL 7
St. Andrew Bay, FL 60
Suwannee River, FL 56
TOTAL 299
34
-------
Appendix C
List of Acronyms and Abbreviations
BOD biochemical oxygen demand
BW bottom water
C carbon
Chl-a chlorophyll-a
CWA Clean Water Act
DIN dissolved inorganic nitrogen
DO dissolved oxygen
EDA estuarine drainage area
EMAP Environmental Monitoring and Assessment Program
EPA U.S. Environmental Protection Agency
IMAP Florida Inshore Monitoring and Assessment Program
uM micromolar
mM millimolar
N nitrogen
N2 molecular nitrogen
NCA National Coastal Assessment
NH
-------
oEPA
United States
Environmental Protection
Agency
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
Office of Research and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
Recycled/Recyclable Printed on paper that contains a minimum of
50% postconsumer fiber content processed chlorine free
------- |