FINAL REPORT

for

A Scientific Assessment of Nutrient Concentrations, Loads, and Biological
Response in the Northern Gulf of Mexico

EPA Contract No. 68-C-03-041
Work Assignment No. 1-01 Task 5

September 13, 2004

U.S. Environmental Protection Agency
Oceans and Coastal Protection Division

Prepared by

BATTELLE
397 Washington Street
Duxbury, MA 02332
(781) 934-0571


-------
Final Report	September 2004

Work Assignment 1-01 Task 7	Page i

TABLE OF CONTENTS

1.0 EXECUTIVE SUMMARY	1

2.0 PROJECT DESCRIPTION	1

2.1	Problem Definition/Background	1

2.2	Proj ect Obj ective	4

2.3	Project/Task Description	4

3.0 DATA COLLECTION, DATABASE DEVELOPMENT, FORMATTING, PROCESSING, AND
QUALITY REVIEW METHODS	4

3.1	Data Collection	4

3.2	Database Development	6

3.3	Data Loading	9

3.4	Data Screening in Database	9

3.5	Format Standardization	9

3.5.1	Parameter Names	9

3.5.2	Result Values and Result Units	12

3.5.3	Locations	12

3.6	Sampling Locations Linked to Spatial Segments	12

3.7	Evaluation for Duplicate Records	13

3.8	Flag and Qualifier Consideration	13

3.9	Quality Assurance and Quality Control of Data Processing	13

3.10	Statistical Data Quality Assessment	13

4.0 DATA SUMMARIES	14

4.1	Study Area	14

4.2	Alabama	18

4.3	Louisiana	21

4.4	Mississippi	24

5.0 TOWARD THE DEVELOPMENT OF NUTRIENT CRITERIA	27

5.1	Scale	27

5.2	Characterization/typology	27

5.3	Statistical Power	28

5.4	Nutrient Criteria Technical Guidance Manual: Estuarine and Coastal Marine Waters (EPA
2001)	28

5.5	NOAA National Estuarine Eutrophication Assessment	30

5.6	Louisiana Coast-wide Reference Monitoring System (CRMS)	31

5.7	Australian Waterway Classifications and Deviation Index	32

5.7.1 U.S EPA Aquatic Stressors Framework	33

5.8	Discussion (change to section 5.6)	34

6.0 RECOMMENDATIONS	36

6.1	Data Gaps Analysis	36

6.2	Identify Appropriate Classification Process	36

6.3	Further Develop Database and Information Access Program	36

7.0 REFERENCES CITED	37

Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page ii

LIST OF TABLES

Table 3-1. Data Required to Support Nutrient Characterization Assessment	7

Table 3-2. Priority Parameters and Standardized Names in the Nutrient Characterization Database	11

Table 4-1. Qualifiers and Flags within the Database and Actions Taken when Observed	15

Table 4-2. Total number of observed data associated with each primary parameter (n = 49) within the

entire study area	16

Table 4-3. Method of aggregation associated with parameter sampling station spatial distribution shown

in Attachments F and G	17

Table 4-4. Ranking of Alabama 12-digit and 10-digit coastal HUCs by total records of primary

causal/response parameters	19

Table 4-5. Ranking of Louisiana coastal HUCs by total records of primary causal/response

parameters	22

Table 4-6. Ranking of Mississippi coastal HUCs by total records of primary causal/response

parameters	25

LIST OF FIGURES

Figure 2-1. Study Area	3

Figure 3-1. Entity Relationship Diagram of the Nutrient Characterization Database	8

Figure 3-2. Locations of all sampling sites also indicating those outside of the project study area	10

Figure 5-1. Ternary classification of coastal systems divided into seven classes (after Dalrymple et al.

[1992] and taken from Ryan et al. [2003])	33

ATTACHMENTS

Attachment A: Current inventory of collected datasets from sources in Louisiana, Mississippi,

Alabama, and other regional sources.

Attachment B: Bibliography of publications and reports that are relevant to some datasets reviewed

for incorporation in the database
Attachment C: Status matrix of identified sources of relevant project data indicating individual names,

affiliations, and type of response to data request
Attachment D: Data Dictionary

Attachment E: Statistical summaries of 12 representative primary standardized parameters.
Attachment F: Maps showing discrete sampling locations for each primary parameter (aggregated

categories) within the study area.

Attachment G: Maps showing relative magnitudes of observed data for each primary parameter

(aggregated categories) within the study area.

Attachment H: Temporal distribution of observed data for each primary parameter (n = 49) for each

selected hydrologic unit (HUC) within the study area.

Attachment I: Total observation counts for all primary parameter (n = 49) in Alabama.

Attachment J: Total observation counts for all primary parameter (n = 49) in Louisiana.

Attachment K: Total observation counts for all primary parameter (n = 49) in Mississippi.

Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 1

1.0 EXECUTIVE SUMMARY

A Scientific Assessment of Nutrient Concentrations, Loads, and Biological Responses in the Northern
Gulf of Mexico is centered on compiling information to support nutrient criteria development for coastal
ecosystems of the northern Gulf of Mexico. This project was conducted to identify and access the best
available data for establishing nutrient criteria for the project study area: the estuaries and nearshore
waters associated with Lake Borgne, Chandeleur and Breton Sounds, Mississippi Sound, and Mobile Bay.
This project will support the ultimate goal of establishing nutrient criteria, identifying data gaps, and
providing recommendations to the Gulf of Mexico Program (GMP) related to future monitoring and
assessment activities.

Approximately 50 water quality parameters, physical parameters, and biological attributes necessary to
characterize the nutrient conditions of the study area were selected as guidance for data collection. These
parameters were provided to approximately 60 academic investigators, state and federal agency scientists
and managers, and private industry researchers to request whether they had or knew of existing datasets
that may contain information pertaining to the parameters. Several datasets were identified and 36
datasets were collected. These datasets contain a total of 2.1 million records from 3,124 sampling stations
within the study area. The temporal timeframe of the data spans the time period between 1905 and 2003.
When datasets were received, they were pre-screened for several criteria (e.g., spatially and temporally
referenced locations, within study area, contained parameters of interest). Those datasets that met the
criteria were loaded into the uniform relational database established for this project. Within the database
for this project, 678 unique parameters have been loaded, For the purpose of this report, several
parameters of interest were prioritized, standardized and summarized.

This report provides a characterization of the database. It summarizes key parameters that will be useful
in the development of nutrient criteria (i.e., total phosphorus, ammonia, chlorophyll a, etc). It provides a
look into the data gaps for key parameters on an individual HUC spatial scale and an annual temporal
scale. The report does not, however, generate nutrient criteria for the individual states. It includes an
assessment of potential approaches to the development of reference conditions that will support nutrient
criteria development. A series of approaches that are relevant to coastal system classification and
reference condition are reviewed, and compared, to provide guidance, additional knowledge and
increased efficiency. It is suggested that scientists and managers carefully review the data for the spatial
regions included in this database and build upon this information to design monitoring programs that
would address any data gaps that need more detailed information prior to development of nutrient criteria.

2.0	PROJECT DESCRIPTION

2.1	Problem Definition/Background

Section 304(a) of the Clean Water Act (CWA) directs the U.S. Environmental Protection Agency (EPA)
to develop and publish criteria guidance to assist states and authorized tribes in developing water quality
standards for all waterbodies (i.e., lakes, rivers, streams, estuaries) that are protective of the designated
uses of those waterbodies. Among these standards, nutrient criteria recommendations are intended to
protect the integrity of coastal ecosystems from the adverse effects of cultural eutrophication. Cultural
eutrophication manifests itself in a variety of interrelated ecosystem responses, both primary and
secondary, which include phytoplankton and macroalgal blooms, increased production and decay of
organic matter, decreased light transmission, and several additional responses that are considered
detrimental to ecosystem health. It has been determined that the primary causal stressor associated with
cultural eutrophication in coastal ecosystems is increased nutrient enrichment, primarily nitrogen,

Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 2

although phosphorus also contributes to the process. Therefore, water quality standards for nutrients
(nitrogen and phosphorus) must be established to support Section 304(a) guidance.

Estuaries and nearshore coastal waters naturally vary in the type, physical and biological characteristics,
as well as geographic coverage of the ecological communities at risk from nutrient overenrichment. The
factors associated with this variability are numerous, but include qualities such as water residence time,
depth, water column stratification, salinity, and temperature. Therefore, it is important to note that one
single criterion is not applicable to national or regional systems, and nutrient criteria must be developed in
a manner that is consistent with the uniqueness of system types.

The EPA guidance document, Nutrient Criteria Technical Guidance Manual: Estuarine and Coastal
Marine Waters (EPA-822-B-01-003) suggests a reference approach be applied to nutrient and algal
criteria development. Reference conditions are defined as the best existing estuarine or marine water
within a watershed or coastal area. In some cases, historical data can be compared to more recent
observations to assess ecological departures from a reference state. A thorough review of water quality
data and supporting information must occur in order to identify and characterize reference conditions.

This review should identify current and historical data that are necessary for the development of nutrient
criteria.

A Scientific Assessment of Nutrient Concentrations, Loads, and Biological Responses in the Northern
Gulf of Mexico, previously Task 7 of Work Assignment 0-01 and currently Task 5 of Work Assignment
1-01, is centered on the first of two phases of compiling information to support nutrient criteria
development for coastal ecosystems of the northern Gulf of Mexico. This project was conducted to
identify and access the best available data for establishing nutrient criteria for the project study area. The
project study area includes the estuaries and nearshore waters associated with Lake Borgne, Chandeleur
and Breton Sounds, Mississippi Sound, and Mobile Bay (Figure 1). As part of this project Battelle is
providing assistance to the states of Louisiana, Mississippi, and Alabama in an analysis of data compiled
during this current effort toward the development nutrient criteria throughout the study area. The
analyses associated with this phase of work is focused on illustrating the type of data acquired and their
spatial and temporal distribution within the study area. Therefore, the analyses are considered
preliminary in nature and do not result in proposed site-specific nutrient criteria. Rather, the database is
described through a series of spatial and temporal assessments, including descriptions of data quality and
database development. A review of nutrient criteria development methods and potential applications
within the study area is provided.

The remainder of this document summarizes the progress associated with the project tasks. Section 2.2
defines the objectives of the overall project. Section 2.3 describes each of the tasks conducted. Section
3.0 summarizes data collection and the activities associated with the database development, including
formatting, processing and quality review methods. Data summaries for each state are provided in
Section 4.0 for the primary causal and response parameters that have been chosen by the project Steering
Committee. In Section 5.0, a review of methods and activities associated with the development of
nutrient criteria is provided. Finally, Section 6.0 presents lessons learned, key findings and further
recommendations with respect to developing nutrient criteria within the project study area.

Battelle

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 3

MJ0,0,16 N.,0,0,0E

Figure 2-1. Study Area.

Batreiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 4

2.2 Project Objective

The objective of the overall "Scientific Assessment of Nutrient Concentrations, Loads, and Biological
Responses in the Northern Gulf of Mexico" project is to provide data and information that can be used by
states to develop nutrient criteria and associated management responses.

2.3 Project/Task Description

This project has focused on the identification, collection, and compilation of available water quality data
associated with the study area into a comprehensive project database. Analyses conducted will support
the ultimate goal of establishing nutrient criteria, identifying data gaps, and providing recommendations
to the Gulf of Mexico Program (GMP) related to future monitoring and assessment activities. Project
activities include:

1.	Developing a detailed Combined Work/Quality Assurance Project Plan (CW/QAPP) for the data
collection, QA/QC, compilation, and analysis, and gain work assignment manager (WAM)
approval;

2.	Selecting attributes that best characterize the nutrient condition of the study area, and describe
rationale behind these selections and rationale for omitting other attributes that were considered;

3.	Identifying and acquiring existing data sets from agencies and universities;

4.	Performing QA/QC for all data sets;

5.	Compiling data sets into a uniform, electronic relational database;

6.	Preparing a draft and final Interim Report detailing an analysis strategy for the data, with specific
information about each data set including:

•	Collection method

•	Data gaps

•	GIS layers representing temporal and spatial data density

•	Preliminary evaluation of waterbodies and corresponding nutrient data

•	Preliminary evaluation of reference conditions within the study area

•	Recommendations for procedures for analysis

3.0 DATA COLLECTION, DATABASE DEVELOPMENT, FORMATTING,
PROCESSING, AND QUALITY REVIEW METHODS

This section summarizes the methods used to acquire the historical data, develop the database, and
process the data (i.e., load, organize, format, standardize, and evaluate). Working with multiple formats,
different presentations, and unique nomenclature, data processing required frequent checks to ensure data
were loaded correctly and to compare records for duplication and correctness.

3.1 Data Collection

The identification and acquisition of data requires significant effort and efficiency. The GMP initiated the
process by developing an inventory of data sources and had also begun the process of acquiring data. To
minimize duplication of effort, Battelle initially searched existing federal and state government sources
(e.g., STORET, NEPs, DEP/DEQs) and developed a "living" database inventory (Attachment A). This

Battelle

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 5

inventory has been continually updated as additional sources were identified. The inventory was also
distributed along with a list of the targeted parameters (Table 3.1) to scientists for review and response.
This database inventory was critical for minimizing duplication of effort as many data sets being sought
from particular sources may already have been available from existing sources. For example, many of the
state monitoring programs upload their data into STORET. If the information was being sought from that
state, but was already included in the STORET acquisition, duplicate data would exist in the database.
Loading duplicate data was avoided if possible and checks were necessary once all data were loaded.

Battelle used a variety of means to identify potential sources of relevant data. These included monthly
conference calls, internet searches, literature searches, personal inquiries to scientists and contacts with
individuals from conferences and/or scientific meetings.

Monthly telephone conference calls. Monthly conference calls with the nutrient pilot study Steering
Committee members were held throughout the project period. Attendance on the calls varied, but
participants typically included Battelle scientists, GMP staff, and representatives from Louisiana,
Mississippi, and Alabama state agencies. Additional scientists and managers from EPA contributed to
some of the calls. Meeting agendas consisted of briefing steering committee members on progress and
problems associated with this task. The calls also provided Battelle and the GMP the opportunity to
inquire about specific data sources within the study area. Hence, several data sets and potential sources
(individual scientists and managers) were identified throughout this process. When data or sources were
identified during these calls, Battelle followed up with the appropriate points of contact to acquire this
data. Data were collected in various formats (e.g., Excel, ASCI, hardcopy), screened for appropriateness
and relevance, inventoried, and loaded into the Oracle database for this project.

Internet searches. Battelle conducted numerous data searches through a variety of internet search
methods. Several large data sets were available and documented online. Examples of internet-based data
sources include EPA's STORET, USGS's NWIS, and a variety of state agencies that had data search
engines on, and/or linked, to their respective homepages. Additional agencies such as NOAA and
programs such as the National Estuarine Research Reserve (NERR) also provided data search engines
from internet websites. Often these searches resulted in identification of technical documents, peer
reviewed literature or points of contacts for electronic data sets. Data were collected in various formats
(e.g., Excel, ASCI, hardcopy), screened for appropriateness and relevance, inventoried, and prepared for
loaded into the Oracle database for this project.

Literature searches. Scientific literature was searched through a variety of methods, including key word
searches using various search engines including Current Contents, Ingenta, and DIALOG. Many journal
articles were collected and analyzed for relevant data. In some cases the authors of relevant journal
articles were contacted directly in order to inquire about the availability of the data that were presented in
the paper. A bibliography of literature associated with this component of the project is listed in
Attachment B.

A contact list of research scientists and organizations was developed and maintained throughout this
project (Attachment C). This list identifies points of contact for potential information, whether contact
was attempted, and the status of inquiry (e.g., response). Those scientists that authored papers containing
potentially relevant data to the project are included in the list presented in Attachment B.

Personal inquiries (existing relationships). Several scientists with Gulf of Mexico research experience
are colleagues of Battelle staff. These individuals were contacted and asked if they knew of data sets that
may be available. Individuals contacted included faculty at Dauphin Island, Louisiana State University,
University of Southern Mississippi, LUMCON, and the University of New Orleans. Those researchers
that were contacted are also listed in Attachment C.

Battelle

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 6

Miscellaneous contacts (conferences, meetings). Battelle scientists are active members of many
professional and scientific societies and often attend meetings and conferences related to nutrient
eutrophication issues in coastal areas. Battelle staff attending recent conferences (e.g., Estuarine
Research Federation, Northwest Atlantic Indicators Workshop, etc) made contact with scientists from
other parts of the country that are engaged in similar efforts. Those contacts are also listed in
Attachment C. In some cases, these efforts were fruitful and some significant datasets were attained
(e.g., J. Pennock, Mobile Bay, AL).

Environmental data were screened prior to loading them into the database. In order for a dataset to be
considered for loading, it had to meet the following three screening criteria:

•	data were in the project study area

•	data were geo-referenced

•	contained parameters that would support the nutrient criteria development.

The datasets meeting these criteria were prioritized for loading into the relational database.

A Nutrient Characterization Database was developed to support the data review process for nutrient
criteria development. The database was developed in Oracle 8i and later upgraded to Oracle 9i. The
relational database was designed to reflect the requirements of the end users and data systems into which
project data may be exported. The Nutrient Characterization Database is a short-term repository for
specific data sets and is not meant to be a complete environmental data system. The database was
developed to generate the best available data not the necessarily the best interface to other databases.

The Nutrient Characterization Database allows for the recording of laboratory analyses and field
measurements. It also contains location information for all sampling and field measuring events.
Waterbody and Ecoregion tables support the linkages of locations to geospatial boundaries. Figure 3.1
depicts the entity relationship diagram (ERD) for the Nutrient Characterization Database. The ERD
shows the tables, data fields, and linkages of the database. The data dictionary for the Nutrient
Characterization Database (Attachment D) defines the use of the fields used.

Sample-based analyses and field data can be linked by location and time period to provide an overall
description of a site or region. Trend analysis can also be performed by parameter in similar fashion.

3.2 Database Development

Battelle

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 7

Table 3-1. Data Required to Support Nutrient Characterization Assessment.





I'liv sic:il

Alkalinity

Benthic Community

Air Temperature

Ammonia (NH3)

Biochemical Oxygen Demand (BOD)

Color (CDOM)

Ammonium (NH4)

Chlorophyll a, b, c

Conductivity

Benthic Flux Measurements (nutrients,

Dissolved Oxygen (DO)

Depth

DO, salinity)

Dissolved Oxygen (% Saturation)

Depth Distribution (Hypsography)

Chloride

Fecals (bateria, pathogens)

Flushing Time

Dissolved Inorganic Nitrogen (DIN)

Fish Community

Groundwater Discharge

Dissolved Inorganic Phosphorus (DIP)

Macroalgae Biomass and Production

Light Extinction Coefficient (Kd)

Dissolved Organic Nitrogen (DON)

Macro invertebrate and Fish Community

Meteorological Conditions

Hardness

Phytoplankton Production

Precipitation

Nutrients (all available)

Primary Production (02 or 14C)

River Discharge

Oil Sheen

Salt Marsh Distribution

Secchi Depth

PH

Submerged Aquatic Vegetation (SAV)

Surface Irradiance

Salinity

Total System Metabolism (TSM)

Tidal Amplitude

Solids

Zooplankton Biomass and Growth

Velocities

Sulfate



Water Residence Time

Total Kjeldahl Nitrogen (TKN)



Water Surface Elevation

Total Nitrogen (TN)



Water Temperature

Total Organic Carbon (TOC)



Wind Speed/Direction

Total Phosphorus (TP)





Total Organic Phosphorus





Total Suspended Sediments (TSS)





Turbidity





Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 8

t*K STUPY 10

STUDY NAME
STARTJ3ATE
END DATE
ORGANIZATION
CONTACT_NAME
T..PHONE
viCE..FlLE
uriN ID

PK STAT ON jD

STATION NAME
STATION TYPE
LOCATION TYPE
Kt> l-*m!UDE
GITUDE
TH

FFFLULNT RELEASE LATITUDE

EPF LI LNT RELEASE. LONGITUDE

LOGIN 10

F*l ENAME

LOAD. DATE

UPDATE SCRIPT

UPDATE DATE

bT^ND " h h h

stand' latitude
¦vrmc !nhi-«ruoE

Mi.JG
PASit

^YAIERBODY NAME
HUC. 8
HUC jo
HUG II
HUGH- NAME
JL" 1 NAME
HUC USE



¦ i ¦¦

PK

SAMP ID



FIELD QC CODE



PA TEMP MARKER



PLANNED ANALYSIS LIST



CLIENT SAMP ID

FK2

STUDY ID

FK1

STATION ID



SURVEY ID



ECTOR



NAVIGATION S 3TEM



DL



DL UNIT



COMMENTS



TEST TYPE



UPDATE SCRIPT



UPDATE DATE



DEPTH UNIT CODE

ft f





PK : K1

STATION ID



PARFfjT STATIC")



KM MION^HIP TYPE



LOCATION DESGR



xgfkcy n



I OCJ* T|r?j rspf











Fil bf.Ak'f-



I A/vn fUTF



UPDATE SCRIPT



update date

PK,FK1
PK

STATION ID
ECOREGION ID



WATERBODY ID#



EPA ECOREGION ID



ECOREGION DESCR



FILENAME



LOGIN ID



LOAD DATE



UPDATE SCRIPT



UPDATE DATE

PK.FK1 STATION ID

'_TYPE_COOE

[LAKE AREA SOURCE
I LAKE_OUTL£T.TYPE

| USEJMPAIRED
I'IPAIKN I N > CONDITION

I REFERENCE

.COMMENTS

\>h i ( Pf L NTS

I filename

I Oc-- N Id
I LOAD DATE

RlPT

TE

SAMP
PARAM CODE

HK>Ki iSAfgiUa





PK

STUDY ID

PK,FK1

STATION ID

PK

ME AS DATE TIME

PK

DEPTH

PK

PARAM CODE



SURVEY ID



DEPTH UNIT CODE



RESULT



UNIT



INSTR CODE



LOGIN ID



FILENAME



LOAD DATE



COMMENTS



UPDATE SCRIPT



UPDATE... DATE



b i -NL> PAKmM NAME



;RiPTION



LOGIN ID



I 'I i Ml L



LOAD DATE



UPDATE SCRIPT



UPDATE,DATE

LAB

* Th PUN ID
LAB SAMP To
I	£

I	'H

'H

SAMP., WGT
SAMP WGT

DL

DL UNIT !

3iN

t U t. I4IL
LOAD DATE
!	NT

!	STURE

;	D

REV NUM

PT

UPDATE DATE

PK,FK1 ANALYSIS IP
PK PARAM CODE

RESULT

LAB OUAL
I NJAw OUAL
I1W1

us uu i n. if

DETECT. LIM. CODE
ANALrbiS UAH
SDG

VALIDATION LEVEL
LOGIN ID
FILENAME
LOAD.DATE

DILUTION
f NMU». avu
h< ' NUM
UPDATE SCRIPT
UPDATE.DATE

i (AM Pm *!„ r mME
MANP ! Nh
«¦ i*ND F LM !i_I
REPORTABLE _YN

COMMENTS
>M I Lh'-,

REPORT RESULT

PK.F K1 AMLY§JJLJD

' SDG

cP%fTION
I RESULT
| LAB _QUAL
FINAL_QUAL.

I UNIT

j CE^Ei T LIMIT
| IFTCT LIT,' CODE
ANA L Y S i S_D AT E
| W.LH^Tuh LEVEL
LOGIN ID
FILENAME
jLOAD_DATE
I DILUTION

|minimum_level

| REV NUM
[ UPDATE.. SCRIPT
|UPDATE_DATE
| PARAM..NAME
i ST f D P \P 1..NAME
STAND_UNIT
I STAND RESULT

:



PK

siudvjd

PK

SURVEY ID

PK.FK1

STATION ID

PK

OBS DATE TIME

PK

PARAM CODE



OBS DESCRIPTION



LOGIN ID



FILENAME



LOAD DATE



UPDATE SCRIPT



UPDATE DATE

RESULT

LAB QUAL

UNIT

LOGIN ID

FILENAME

L0AD...DATE

COMMENTS

Figure 3-1. Entity Relationship Diagram of the Nutrient Characterization Database.

Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 9

3.3 Data Loading

A unique identifier or 'Login_ID' was given to each dataset (i.e., filename) loaded into the database.
Every record in every database table contains a 'LoginID', the original filename, the date it was loaded,
the filename of the last script used to process or update it, and the date the last update took place. If any
changes were made previously, the output script will show the date and script name of that update. This
combination of information allows for rapid and complete auditing of the data loading and processing.

Thirty six data sets met the screening criteria for loading and were loaded into the database. The loaded
data comprise over 2.1 million results from 3,124 stations over the temporal span of 1905 to 2003. Of the
3,124 locations included in the datasets, 2,441 are within the study area. Figure 3.2 illustrates the spatial
distribution of these stations. A total of 678 original parameter names are associated with the stations
within the study area. Because of time constraints and the multitude of parameter names that were
included in the database, parameters were prioritized as to whether they were key to nutrient criteria
development. Over 200 of these original parameter names were prioritized and standardized to reduce the
number of ways to indicate the same parameter. The Format Standardization section further describes the
process used to normalize the parameter names as well as other fields in the database.

The data for the priority parameters were screened in the database for reasonableness with respect to high
and low outliers. A database field was created called "reportable_yn" to allow values to be "turned off'
and not used in subsequent analysis - (reportable_yn = N) or left "on" for future use (reportable _yn = Y).
Any negative values observed in the database were turned "off' (i.e., reportable_yn='N') and the yn_flag
'NN' was assigned. Potential outliers on the upper end (i.e., very high values for a given parameter) were
not addressed at this time. It will be left to the discretion of the scientists and managers using this
information to determine if these high values are true outliers and whether to exclude them (reportable_yn
= N) from subsequent analysis.

Because data originated from multiple agencies and organizations, they arrived in disparate formats. To
be able to compare the data to check for duplicates and subsequent summarization, the data had to be
normalized for parameter name, location, result value, and result unit.

3.5.1 Parameter Names

Different sources often reported the same parameter differently. For example, the parameter dissolved
oxygen can be reported as DO, Dissolved O, Diss O, etc. To facilitate future queries of a specific
parameter, the priority parameter names were standardized in a new parameter name field (i.e.,
STAND_PARAM_NAME) of the database. This new parameter name field was populated with the
standardized name for the prioritized parameters and the original parameter name remained in a separate
(original) database field. Table 3.2 depicts the priority parameters (bold) and the standardized naming
convention used in the database. Source documentation was reviewed to establish the meaning and new
naming convention for original parameter names having the same meaning. Even what would seem to be
a simple parameter to express like 'Water Temperature' was provided under multiple naming
conventions: 'Bottom Water Temperature', 'Surface Water Temperature', 'TEMP', 'Temperature',
'Water Temp', 'Water temp', 'TEMP (C)', 'TEMPERATURE WATER (DEGREES CENTIGRADE)',

3.4 Data Screening in Database

3.5 Format Standardization

Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 10

Figure 3-2. Locations of all sampling sites also indicating those outside of the project study area.

Batreiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 11

Table 3-2. Priority Parameters and Standardized Names in the Nutrient Characterization

Database.

AIR TEMPERATURE

ORTHOPHOSPHATE AS P - FILTERED

AMMONIA NITROGEN AS N - FILTERED

ORTHOPHOSPHATE AS P - UNFILTERED

AMMONIA NITROGEN AS N - UNFILTERED

OXYGEN TEMPERATURE

AMMONIA NITROGEN BOTTOM DEPOSIT DRY WEIGHT

PARTICULATE NITROGEN AS N

CHLOROPHYLL A CHROMA-FLUORO

PERIPHYTON CHLOROPHYLL A CHROMA-FLUORO

CHLOROPHYLL A CHROMA-SPECTRO
CHLOROPHYLL A FLUORO

PERIPHYTON CHLOROPHYLL A FLUORO
PERIPHYTON CHLOROPHYLL A SPECTRO
UNCORRECTED

CHLOROPHYLL A HPLC

PH

CHLOROPHYLL A SPECTRO ACID CORRECTED

PHOSPHATE AS P - UNFILTERED

CHLOROPHYLL A SPECTRO UNCORRECTED

PHOSPHATE AS P BOTTOM DEPOSIT DRY WEIGHT

CHLOROPHYLL A TRICHROM UNCORRECTED

POLYPHOSPHATE AS P

DISSOLVED ORGANIC PHOSPHORUS AS P - FILTERED

POTENTIAL WATER TEMPERATURE

DISSOLVED OXYGEN CONCENTRATION

POTENTIAL WATER TEMPERATURE

DISSOLVED OXYGEN PERCENT SATURATION

PRODUCTIVITY AS C

DISSOLVED PHOSPHORUS AS P - FILTERED

SALINITY

INORGANIC NITROGEN AS N - FILTERED

SALINITY - DAILY MEAN

KJELDAHL NITROGEN AS N - FILTERED

SECCHIDEPTH

KJELDAHL NITROGEN AS N BOTTOM DEPOSIT DRY WEIGHT

SUSPENDED KJELDAHL NITROGEN AS N

MEAN ANNUAL AIR TEMPERATURE

TOTAL KJELDAHL NITROGEN AS N

MEAN MONTHLY AIR TEMPERATURE

TOTAL NITROGEN AS N - FILTERED

NITRATE + NITRITE NITROGEN AS N - FILTERED

TOTAL NITROGEN AS N - UNFILTERED

NITRATE + NITRITE NITROGEN AS N - UNFILTERED
NITRATE + NITRITE NITROGEN AS N BOTTOM DEPOSIT DRY
WEIGHT

TOTAL ORGANIC PHOSPHORUS AS P
TOTAL PHOSPHORUS AS P - UNFILTERED

NITRATE NITROGEN AS N - FILTERED

TOTAL PHOSPHORUS BOTTOM DEPOSIT DRY WEIGHT

NITRATE NITROGEN AS N - UNFILTERED

TURBIDITY - LAB

NITRITE NITROGEN AS N - FILTERED

TURBIDITY - UNFILTERED

NITRITE NITROGEN AS N - UNFILTERED

TURBIDITY AS SI02 - UNFILTERED

ORGANIC NITROGEN AS N - FILTERED

UNIONIZED AMMONIA AS N

ORGANIC NITROGEN AS N - UNFILTERED

WATER TEMPERATURE

Note - BOLD indicates the priority parameters that are further summarized in this report.

Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 12

'TEMPERATURE WATER (DEGREES FAHRENHEIT)', 'Temperature water', 'Temperature, water,
degrees Celsius', 'pOOOlO - Temperature, water, degrees Celsius', and 'pOOOl 1 - Temperature,
water, degrees Fahrenheit'. The differences in units were also normalized into a standard unit. See the
Result Values and Result Units section below.

3.5.2	Result Values and Result Units

Units can also prevent data from being directly compared. Syntax used to express units can also vary.
Examples of syntax variations seen for the unit milligrams for liter, included: 'MG/L', 'PPM', 'mg/L',
'mg/1', 'mg/1', 'mg/1 ', and 'ppm'. Some units appear to be the same but extra spaces in the unit make
the unit expressions different when comparing records. Additionally the unit was not always in a separate
field when received but instead part of the original parameter name. In these cases, the unit, as indicated
in the original parameter name, was copied to the unit field. The units for milligram per liter were
standardized in the database in a new result unit field (i.e., STAND_UNIT) as 'MG/L' and the original
unit syntax were maintained in the UNIT database field.

To achieve a single result unit for a given standardized parameter name, result values with different units
were converted. These converted or standardized values were placed in a new result value field (i.e.,
STAND RESULT) while maintaining the original value and original unit in the database (database fields
RESULT and UNIT, respectively). When original result units were equivalent (e.g., MG/L = PPM and
MG/M3 = UG/L), one unit was chosen as the standardized unit and the other was translated. In the cases
of equivalent units, the original result value was copied to the standardized result value field. When
parameters were expressed in different but related units, the corresponding conversion was used to
generate the standardized result and update the standardized unit field (e.g., G/L x 1000 = MG/L). By
maintaining the original values and units, translations or conversions may be made if another unit is
preferred.

3.5.3	Locations

Each station's latitude and longitude, expressed as decimal degrees were rounded to three decimal places
to generate standardized latitude (STAND LATITUDE) and standardized longitude
(STAND LONGITUDE) fields in the database. By rounding to three decimal places, stations within
approximately 100 feet of each other would have the same standardized latitude and standardized
longitude. Those stations with the same standardized latitude and standardized longitude were given a
unique identifier or standardized station name (STAND STATION). The new standardized location
identifier allowed records to be grouped according to location and later compared to identify duplicate
records. The 1,972 original stations in the project area were grouped according to the method described
above and converted into 1,553 standardized stations. For example, the three original station identifiers,
'MP_237E_01_N1', 'MP_237E_01_N2', and 'MP_237E _01', describe a similar location (i.e., are within
100 feet of each other) and were given the same standardized station name.

3.6 Sampling Locations Linked to Spatial Segments

The Gulf of Mexico project area covers portions of Alabama, Louisiana, and Mississippi. These states
are segmented into geospatial watersheds or basins. Each watershed has an eight digit hydrologic unit
code or HUC 8. These watersheds are further segmented into drainage basins, subsegments, or HUC 12
areas. The location data in the Nutrient Characterization Database are linked to these geospatial regions
to allow for queries according to state, project area, watershed (i.e., HUC 8), or drainage basin (i.e., HUC
12).

Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 13

3.7	Evaluation for Duplicate Records

Once the data were normalized by location identifier, parameter name, result value, and unit, records
could be compared to identify multiple occurrences of the same measurements in the database.

Identifying duplicate records prevents their use in future characterizations and analyses. Characterization
results may be skewed if duplicates remain unchecked.

Duplicate data were identified as measurements or results having the same standardized parameter name,
standardized station location, date/time of sample or measurement, sample depth, standardized result
(rounded to two decimal places), and standardized unit. Once identified the duplicate records that had
qualifying information (i.e., flags, qualifiers, method detection limit, or comments) were maintained as
reportable "yes" and the related record was turned "off (i.e., reportable_yn = 'N'). Those duplicate
records that were turned "off were also given a flag of 'ND' in the 'YN_FLAG' field. This flag
indicates that the record was turned "off because it was a duplicate record. Counts of record changes
were reviewed to ensure that the duplicate record evaluation procedure ran correctly.

Of the priority parameter records evaluated for duplication, 2,793 records were turned "off as duplicates.

3.8	Flag and Qualifier Consideration

Data qualifiers and flags provided with the original data sets were loaded into the database and taken into
consideration. Actions were taken to turn "off or adjust the result value based on particular flags or
qualifiers. Table 3.3 shows the original flags and qualifiers used by the data sources and the actions
taken in the database. The fatal qualifiers associated with the data led to 15,080 records being turned
"off and flagged as 'NQ' in the YN_FLAG field.

3.9 Quality Assurance and Quality Control of Data Processing

Quality assurance and quality control (QA/QC) was performed on a randomly sampled portion of the
database contents. Approximately 10% of the records in the database were pulled and reviewed by
Battelle's QA staff. This group checked that the data sets were correctly mapped to the database
structure, checked the loading scripts, and verified completeness and accuracy of the data in the database
compared to the data sets received for loading. The standardization process was reviewed and the
conversion factors checked. Findings were reported and corrections were made accordingly in the
database.

3.10 Statistical Data Quality Assessment

Following data loading, formatting, standardization, duplication and qualifier checks, the priority
parameters were queried. Summary statistics of these priority parameters were run using SAS Release
8.02. Each priority parameter was summarized in tabular form by state, HUC12 name, and year. The
statistical summaries include the number of observations, mean value, minimum value, maximum value,
standard deviation, standard error, coefficient of variation, median and various percentiles that show the
distribution of data by HUC12 and year. The statistical summaries are presented in detail in Attachment
E.

Battene

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 14

4.0 DATA SUMMARIES

This section summarizes the spatial and temporal distribution of data that have been identified to support
the development of nutrient criteria in each of the states within the study area. Two approaches to this
summary have been chosen. First, the full set of 49 causal and response variables (that have been selected
by the project Steering Committee to be of primary interest) are summarized within the study area and
each state. This summary describes the number of observations that exist for each parameter in each state
in tabular form. For effective visualization of the spatial distribution of these data, several inter-
associated constituents have been aggregated and mapped. This was done for chlorophyll, nitrogen and
phosphorus because each of these general parameters consists of either additional constituents (i.e., sub-
species of analytes) or varied analytical methods. For example, the response variable "chlorophyll" is
associated with 10 varieties of component variables which are based on the application of varying
analytical methods and sub-species (e.g., periphyton chlorophyll a). Further examples reside in the
nutrient data. Total nitrogen was identified as a primary causal variable of interest, however the
individual constituents of TN include inorganic (nitrate, nitrite, and ammonium) and organic (organic
nitrogen and total kjeldahl nitrogen) forms. Further, these nitrogen constituents were reported in different
ways due to analytical methods (e.g., filtered vs. unfiltered).

The following subsections focus on all 49 parameters/variables within the study area and each individual
state. In order to facilitate the understanding of how data are distributed throughout space and time, a
series of key questions that support the ultimate goal of identifying which regions, subregions and
waterbodies possess adequate data and which do not.

•	Which waterbodies are richest in causal (nutrient) and response (chl, DO, Secchi) data?

•	Which waterbodies have earliest records of causal and response data?

•	Which waterbodies have riches physical data (salinity, temperature)?

•	Which waterbodies possess most significant data gaps?

The spatial and temporal distribution of the 49 standardized parameters are summarized for the entire
study area (Figure 1) in this section. These parameters, and their total counts (data counts) are shown in
Table 4-1. The process by which these parameters were standardized within the database is discussed in
Section 3.5. In addition, the statistical analyses conducted on the most popularly sampled parameters
within the study area (shown as shaded rows in Table 4-1) are discussed in Section 3.10 and illustrated in
Attachment E.

4.1 Study Area

Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 15

Table 4-1. Qualifiers and Flags within the Database and Actions Taken when Observed.

ORGANIZATION

LAB QUAL

QUAL DESC

ACTION

Texas A and M

-9

not sampled or bad data

Turn record "off

Texas A and M

-1

not detected

Keep, use % the reported value

USEPA

$

Calculated by retrieval software. Numerical value was neither
measured nor reported to the database, but was calculated from
other data available during generation of the retrieval report

Keep

AL-DEM, Field
Ops

<

Value Is Less Than Method Detection Limit

Keep, use % the reported value

Smith and
Hitchcock

<

Value Is Less Than Method Detection Limit

Keep, use % the reported value

USGS

<

Value Is Less Than Method Detection Limit

Keep, use % the reported value

USGS

>

Actual value is known to be greater than the value shown

Keep

USEPA

A

Value reported is the mean of two or more determinations

Keep - leave as is

USEPA

B

Results based on colony counts outside acceptable range

Turn record "off

USEPA

C

Calculated. Value stored was not measured directly, but was
calculated from other data

Keep - leave as is

USEPA

D

Field measurement.

Keep, leave as is

USGS

E

Value Is Estimated, Daily Detection Limit > Value > Method
Detection Limit

Turn record "off

USEPA

G

Value reported is maximum of two or more determinations

Turn record "off

USEPA

I

Value reported is less than the practical quantification limit and
greater than or equal to the method detection limit

Keep, use % the reported value

USEPA

J

Estimated. Value shown is not a result of analytical
measurement

Turn record "off

USEPA

K

Off-scale low. Actual value not known, but known to be less
than value shown

Keep, use % the reported value

USEPA

L

Off-scale high. Actual value not known, but known to be greater
than value shown

Keep, use the reported value

USEPA

M

Presence of material verified, but not quantified. Indicates a
positive detection at a level too low to permit accurate
quantification. In the case of species, M indicates male sex

Keep, use % the reported value

USGS

M

Presence of material verified, but not quantified. Indicates a
positive detection at a level too low to permit accurate
quantification. In the case of species, M indicates male sex

Keep, use % the reported value

AL-DEM-BMP

nd

not detected

Keep, use % the reported value

USGS

nd

not detected

Keep, use % the reported value

USEPA

0

Sampled for, but analysis lost. Accompanying value is not
meaningful for analysis

Turn record "off

USEPA

P

Too numerous to count

Turn record "off

USEPA

Q

Sample held beyond normal holding time

Turn record "off

USEPA

R

Significant rain in the past 48 hours

Keep - leave as is

USEPA

U

Material was analyzed for but not detected. Value stored is the
limit of detection for the process in use. In the case of species,
U indicated undetermined sex

Keep, use % the reported value

USGS

U

Analyzed for, not detected

Keep, use % the reported value

USEPA

Z

Too many colonies to count

Turn record "off

* denotes fatal qualifier - record turned "off' and yn_flag = 'NQ'

Battene

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 16

Table 4-2. Total number of observed data associated with each primary parameter (n = 49) within

the entire study area.

I'lii'iimclcr

( ) III III

AMMONIA NITROGEN AS N - FILTERED

4,552

AMMONIA NITROGEN AS N - UNFILTERED

2,333

AMMONIA NITROGEN BOTTOM DEPOSIT DRY WEIGHT

12

CHLOROPHYLL A CHROMA-FLUORO

340

CHLOROPHYLL A CHROMA-SPECTRO

15

CHLOROPHYLL A FLUORO

3,560

CHLOROPHYLL A SPECTRO ACID CORRECTED

111

CHLOROPHYLL A SPECTRO UNCORRECTED

9

CHLOROPHYLL A TRICHROM UNCORRECTED

42

DISSOLVED ORGANIC PHOSPHORUS AS P - FILTERED

2

DISSOLVED OXYGEN CONCENTRATION

101,748

DISSOLVED OXYGEN PERCENT SATURATION

93,027

DISSOLVED PHOSPHORUS AS P - FILTERED

3,194

KJELDAHL NITROGEN AS N - FILTERED

937

KJELDAHL NITROGEN AS N BOTTOM DEPOSIT DRY WEIGHT

128

NITRATE + NITRITE NITROGEN AS N - FILTERED

7,222

NITRATE + NITRITE NITROGEN AS N - UNFILTERED

2,852

NITRATE NITROGEN AS N - FILTERED

1,582

NITRATE NITROGEN AS N - UNFILTERED

1,155

NITRITE NITROGEN AS N - FILTERED

4,476

NITRITE NITROGEN AS N - UNFILTERED

1,287

ORGANIC NITROGEN AS N - FILTERED

88

ORGANIC NITROGEN AS N - UNFILTERED

556

ORTHOPHOSPHATE AS P - FILTERED

561

ORTHOPHOSPHATE AS P - UNFILTERED

532

PARTICULATE NITROGEN AS N

1,921

PERIPHYTON CHLOROPHYLL A CHROMA-FLUORO

1

PERIPHYTON CHLOROPHYLL A FLUORO

2

PERIPHYTON CHLOROPHYLL A SPECTRO UNCORRECTED

1

PH

92,703

PHOSPHATE AS P - UNFILTERED

5,174

PHOSPHATE AS P BOTTOM DEPOSIT DRY WEIGHT

2

POLYPHOSPHATE AS P

2

PRODUCTIVITY AS C

1,152

SALINITY

94,183

SALINITY - DAILY MEAN

7,960

SECCHI DEPTH

3,234

SUSPENDED KJELDAHL NITROGEN AS N

94

TOTAL KJELDAHL NITROGEN AS N

3,556

TOTAL NITROGEN AS N - FILTERED

1,516

TOTAL NITROGEN AS N - UNFILTERED

1,016

TOTAL ORGANIC PHOSPHORUS AS P

8

TOTAL PHOSPHORUS AS P - UNFILTERED

7,998

TOTAL PHOSPHORUS BOTTOM DEPOSIT DRY WEIGHT

45

TURBIDITY - LAB

509

TURBIDITY - UNFILTERED

92,576

TURBIDITY AS SI02 - UNFILTERED

216

UNIONIZED AMMONIA AS N

760

WATER TEMPERATURE

107,834

Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 17

An analysis of the spatial distribution of these primary parameters was conducted by mapping individual
sampling locations and relative magnitudes of sampling sites within each 12-digit HUC. Nitrogen,
phosphorus and chlorophyll parameters were aggregated to provide an easier visual depiction of the
spatial distribution of these parameters. These aggregations are illustrated in Table 4-2.

Table 4-3. Method of aggregation associated with parameter sampling station spatial distribution

shown in Attachments F and G.

I'lii'iimclcr



NITROGEN

AMMONIA NITROGEN AS N - FILTERED

AMMONIA NITROGEN AS N - UNFILTERED

AMMONIA NITROGEN BOTTOM DEPOSIT DRY WEIGHT

KJELDAHL NITROGEN AS N - FILTERED

KJELDAHL NITROGEN AS N BOTTOM DEPOSIT DRY WEIGHT

NITRATE + NITRITE NITROGEN AS N - FILTERED

NITRATE + NITRITE NITROGEN AS N - UNFILTERED

NITRATE NITROGEN AS N - FILTERED

NITRATE NITROGEN AS N - UNFILTERED

NITRITE NITROGEN AS N - FILTERED

NITRITE NITROGEN AS N - UNFILTERED

ORGANIC NITROGEN AS N - FILTERED

ORGANIC NITROGEN AS N - UNFILTERED

SUSPENDED KJELDAHL NITROGEN AS N

TOTAL KJELDAHL NITROGEN AS N

TOTAL NITROGEN AS N - FILTERED

TOTAL NITROGEN AS N - UNFILTERED

UNIONIZED AMMONIA AS N

PARTICULATE NITROGEN AS N

PHOSPHORUS
CHLOROPHYLL

ORTHOPHOSPHATE AS P - FILTERED

ORTHOPHOSPHATE AS P - UNFILTERED

PHOSPHATE AS P - UNFILTERED

PHOSPHATE AS P BOTTOM DEPOSIT DRY WEIGHT

POLYPHOSPHATE AS P

TOTAL ORGANIC PHOSPHORUS AS P

TOTAL PHOSPHORUS AS P - UNFILTERED

TOTAL PHOSPHORUS BOTTOM DEPOSIT DRY WEIGHT

CHLOROPHYLL A CHROMA-FLUORO
CHLOROPHYLL A CHROMA-SPECTRO
CHLOROPHYLL A FLUORO
CHLOROPHYLL A SPECTRO ACID CORRECTED
CHLOROPHYLL A SPECTRO UNCORRECTED
CHLOROPHYLL A TRICHROM UNCORRECTED
DISSOLVED ORGANIC PHOSPHORUS AS P - FILTERED
DISSOLVED PHOSPHORUS AS P - FILTERED
PERIPHYTON CHLOROPHYLL A CHROMA-FLUORO
PERIPHYTON CHLOROPHYLL A FLUORO
PERIPHYTON CHLOROPHYLL A SPECTRO UNCORRECTED

DISSOLVED OXYGEN

DISSOLVED OXYGEN CONCENTRATION
DISSOLVED OXYGEN PERCENT SATURATION

PRODUCTIVITY

PRODUCTIVITY AS C

SALIINITY

SALINITY

SALINITY - DAILY MEAN

SECCHI

SECCHI DEPTH

TURBIDITY

TURBIDITY - LAB
TURBIDITY - UNFILTERED
TURBIDITY AS SI02 - UNFILTERED

TEMPERATURE

WATER TEMPERATURE

The spatial distributions of these standardized parameters are shown in Attachments F and G.

Battene

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 18

The following sections summarize data richness and distribution within Alabama, Louisiana, and
Mississippi.

4.2 Alabama

Data distribution within Alabama coastal waters is illustrated in several summaries. Table 4.3 (below)
contains an aggregated summary of total observations in Alabama waterbodies. Analytical species of
nitrogen, phosphorus and chlorophyll were aggregated as shown in Table 4.2. The total observations are
indicated for each Alabama HUC and these have been placed in descending order; from the greatest to
least total number of observations. Attachments F and G provide visual distribution of total stations
within the Alabama portion of the project study area (Attachment F) and relative magnitudes among the
coastal HUCs (magnitudinal centroids shown in Attachment G). Temporal distributions are indicated in
Attachment H where information on total annual observations are provided for each HUC. Attachment I
summarizes the total counts for all primary standardized parameters (non-aggregated).

Table 4.3 indicates a significantly high overall percentage of DO, salinity, temperature, and turbidity
observations. However, these are the result of a continuous monitoring program which utilized an in situ
data logger (e.g., CTD). This is a good example of how summaries of water quality data can be biased
based on the application of different monitoring practices. Clearly the thousands of continuously
monitored data are valuable, but not necessarily comparable to other historical and current discrete data in
terms of identifying data gaps. If we disregard this first row of Table 4.3, we see that there are
approximately 7 to 10 HUCs that possess relatively high total numbers of all primary parameters. The
remaining HUCs possess various combinations of observations that are shown in descending order.

One interesting aspect of the Alabama data set is that phytoplankton productivity was measured quite
rigorously. The Pennock et al. Mobile Bay and Weeks Bay cruises in the late 1980s and 1990s resulted in
a very good spatial and temporal survey of productivity, standing stock, DO, nutrients, and physical
factors. This type of surveying, which encompasses many key nutrient response characteristics, are not as
common in other regions of the study area.

Baiteiie

The Business of Innovation


-------
Table 4-4. Ranking of Alabama 12-digit and 10-digit coastal HUCs by total records of primary causal/response parameters. Also shown
are total records for all nitrogen (N), phosphorus (P) and chlorophyll (Chi) analytical values, and phytoplankton productivity (Prod),
dissolved oxygen concentration and percent saturation (DO), Secchi depth, turbidity, salinity and water temperature (Temp).

ST

HUC

HUC Name

Basin Name

Totals

N

p

Chi

Prod

DO

Secchi

Turbidity

Salinity

Temp

AL

3160205060

Magnolia River

Mobile

425,730

0

0

0

0

170,292

0

85,146

85,146

85,146

AL

031602050307

Lower Fish River

MOBILE BAY

15,514

5,416

2,645

1,362

424

1,767

1,121

32

1,344

1,403

AL

031602050104

Bon Secour Bay

MOBILE BAY

8,210

2,844

1,121

669

210

1,242

413

23

841

847

AL

031602050102

Middle Mobile Bay

MOBILE BAY

6,483

2,068

807

434

175

1,152

248

145

706

748

AL

031602050101

Upper Mobile Bay

MOBILE BAY

6,444

2,196

803

485

144

1,174

298

21

618

705

AL

031602050103

Lower Mobile Bay

MOBILE BAY

4,889

1,688

632

382

109

844

185

28

510

511

AL

031700090201

Mississippi Sound

MISSISSIPPI

2,715

807

269

142

40

624

99

62

329

343







COASTAL





















AL

031602040201

Mittlin Lake

MOBILE-TENSAW

2,084

17

15

0

0

1,349

0

14

0

689

AL

031602050309

Skunk Bayou

MOBILE BAY

1,984

691

352

168

50

241

142

0

170

170

AL

031602050310

Bon Secour River

MOBILE BAY

1,854

14

6

0

0

975

0

2

366

491

AL

031602040505

Tensaw River-

MOBILE-TENSAW

1,139

152

50

0

0

580

0

42

16

299





Apalachee River























AL

031602040303

Grand Bay

MOBILE-TENSAW

832

101

35

0

0

419

0

40

14

223

AL

031602040404

Lower Chasaw Creek

MOBILE-TENSAW

664

124

41

0

0

278

0

28

54

139

AL

031602050308

Magnolia River

MOBILE BAY

540

220

84

35

0

68

9

15

36

73

AL

031602050204

Lower Dog River

MOBILE BAY

513

2

1

1

0

145

1

115

117

131

AL

031602040105

Big Chippewa Lake

MOBILE-TENSAW

497

98

41

0

0

201

0

4

0

153

AL

031602050301

Yancey Branch

MOBILE BAY

484

0

0

0

0

121

0

121

121

121

AL

031602040402

Seabury Creek

MOBILE-TENSAW

477

66

3

0

0

87

0

0

0

321

AL

031602050304

Upper Fish River

MOBILE BAY

432

159

67

0

0

27

0

13

0

166

AL

031602040504

Three Mile Creek

MOBILE-TENSAW

297

65

20

0

0

128

0

20

0

64

AL

031401070205

Little Lagoon

PERDIDO BAY

249

35

20

0

0

37

0

5

115

37

AL

031602040501

Upper Bay Minette

MOBILE-TENSAW

236

120

50

0

0

28

0

14

0

24





Creek























AL

031502040705

Majors Creek

LOWER ALABAMA

189

101

45

0

0

13

0

14

0

16

AL

031602050302

Fly Creek

MOBILE BAY

174

14

6

0

0

40

0

38

36

40

AL

031602040101

Barrow Creek

MOBILE-TENSAW

159

0

0

0

0

106

0

0

0

53

AL

031602040203

The Basin

MOBILE-TENSAW

96

7

3

0

0

54

0

1

0

31

AL

031602050311

Oyster Bay

MOBILE BAY

86

0

0

0

0

1

0

0

84

1

AL

031602040103

Upper Cedar Creek

MOBILE-TENSAW

79

0

0

0

0

0

0

0

0

79

AL

3160205010

Mobile Bay

Mobile

77

0

0

0

0

22

11

0

22

22


-------
ST

HUC

HUC Name

Basin Name

Totals

AL

031602040104

Lower Cedar Creek

MOBILE-TENSAW

62

AL

031602050206

Fowl River

MOBILE BAY

60

AL

031602050201

Garrows Bend-

MOBILE BAY

60





Mobile Bay





AL

031602040401

Meekers Creek

MOBILE-TENSAW

60

AL

031602040302

Bayou Sara

MOBILE-TENSAW

51

AL

031602050205

Deer River

MOBILE BAY

44

AL

031700090103

Grand Bay Swamp

MISSISSIPPI

42







COASTAL



AL

031602050303

Gum Swamp

MOBILE BAY

40

AL

031602050202

Upper Dog River

MOBILE BAY

36

AL

3170009030

Mississippi Sound

Escatawpa

35

AL

031700080302

Powell Creek

Escatawpa

25

AL

031700090202

Dauphin Island

MISSISSIPPI

18







COASTAL



AL

031602050306

Middle Fish River

MOBILE BAY

15

AL

031602040202

Rains Creek

MOBILE-TENSAW

15

AL

3170009020

Dauphin Island

Escatawpa

14

AL

031700090203

Pelican Bay

MISSISSIPPI

13







COASTAL



AL

031602040102

Halls Creek

MOBILE-TENSAW

13

AL

3170009040

West Fowl River

Escatawpa

7

AL

3160205050

Fish River

Mobile

7

AL

3160204060

Three Mile Creek

Mobile

7

AL

3160204040

Lower Tensaw River

Mobile

7

AL

3140107040

Wolf Creek

Perdido-Escambia

7

AL

031602040403

Eightmile Creek

MOBILE-TENSAW

2

Total Counts	483,767

%

Chi

Prod

DO

Secchi Turbidity Salinity Temp

2
0
8

0

3

6
0

0
16
0

7
0

7
6
0

6

7
0
0
0
0
0
0

17
15
12

0
29
9
28

10
8
10

0
15
4

0

1

3
0

10

4
0
0

0

1
1
0

0

1
0
0
0
0
0
0

0
15
12

0
0
9
0

10
0
10
0
18

0
0
4
2

0
2
2
2
2
2
0

43
15
12

60
15
9
14

10
4
10
4

0

2

3

4
2

1

2
2
2
2
2
2

17,073 7,147 3,686 1,152 182,180 2,545 85,983 90,735 93,266
4%	1%	1%	0% 38% 1%	18% 19% 19%


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 21

4.3 Louisiana

Data distribution within Louisiana coastal waters is illustrated in several summaries. Table 4.4 (below)
contains an aggregated summary of total observations in Louisiana waterbodies. Analytical species of
nitrogen, phosphorus and chlorophyll were aggregated as shown in Table 4.2. The total observations are
indicated for each Louisiana HUC and these have been placed in descending order; from the greatest to
least total number of observations. Attachments F and G provide visual distribution of total stations
within the Louisiana portion of the project study area (Attachment F) and relative magnitudes among the
coastal HUCs (magnitudinal centroids shown in Attachment G). Temporal distributions are indicated in
Attachment H where information on total annual observations are provided for each HUC. Attachment J
summarizes the total counts for all primary standardized parameters (non-aggregated).

The Louisiana HUCs indicate the highest concentration of nutrient criteria data in nitrogen (27%), DO
(9%), turbidity (18%), salinity (26%) and temperature (18%). Currently the project database does not
posses any phosphorus or productivity data in Louisiana. Also, relatively few records associated with
chlorophyll (1%) are present.

Baiteiie

The Business of Innovation


-------
Table 4-5. Ranking of Louisiana coastal HUCs by total records of primary causal/response parameters. Also shown are total records for
all nitrogen (N), phosphorus (P) and chlorophyll (Chi) analytical values, and phytoplankton productivity (Prod), dissolved oxygen
concentration and percent saturation (DO), Secchi depth, turbidity, salinity and water temperature (Temp).

ST

HUC

HUC Name

Totals

N

P

Chi

Prod

DO

Secchi

Turbidity

Salinity

Temp

LA

LA070301

LOWER MISSISSIPPI-BATON ROUGE

11,699

4,836

0

20

0

1,239

0

1,452

18

4,134

LA

LA042202

LAKE PONTCHARTRAIN

3,982

906

0

0

0

0

0

2,050

653

373

LA

LA042001

LAKE PONTCHARTRAIN

2,984

893

0

0

0

464

22

1,295

22

288

LA

LA042202

PONTCHARTRAIN

2,697

0

0

0

0

23

14

0

2,607

53

LA

LA041702

LAKE PONTCHARTRAIN

2,177

696

0

70

0

514

212

291

167

227

LA

LA041701

PONTCHARTRAIN

2,168

0

0

0

0

2

1

0

2,159

6

LA

LA042208

PONTCHARTRAIN

2,057

0

0

0

0

2

1

0

2,025

29

LA

LA070401

MISSISSIPPI

1,348

680

0

100

0

189

11

182

24

162

LA

LA041702

PONTCHARTRAIN

960

0

0

0

0

0

0

0

953

7

LA

LA041601

LAKE PONTCHARTRAIN

839

469

0

65

0

123

0

126

1

55

LA

LA042208

LAKE PONTCHARTRAIN

671

151

0

35

0

34

0

215

158

78

LA

LA042003

LAKE PONTCHARTRAIN

523

224

0

0

0

0

0

299

0

0

LA

LA041805

LAKE PONTCHARTRAIN

489

70

0

0

0

246

0

28

0

145

LA

LA042101

PONTCHARTRAIN

488

0

0

0

0

0

0

0

238

250

LA

LA042102

LAKE PONTCHARTRAIN

444

7

0

0

0

0

0

167

169

101

LA

LA041401

LAKE PONTCHARTRAIN

335

159

0

0

0

8

0

167

0

1

LA

LA041901

LAKE PONTCHARTRAIN

278

171

0

38

0

35

0

34

0

0

LA

LA042207

LAKE PONTCHARTRAIN

264

160

0

33

0

36

0

35

0

0

LA

LA041601

PONTCHARTRAIN

244

0

0

0

0

0

0

0

242

2

LA

LA042201

PONTCHARTRAIN

219

0

0

0

0

50

27

0

50

92

LA

LA070301

MISSISSIPPI

181

0

0

0

0

46

23

0

45

67

LA

LA042001

PONTCHARTRAIN

156

0

0

0

0

30

15

0

30

81

LA

LA041501

LAKE PONTCHARTRAIN

142

106

0

3

0

14

0

17

0

2

LA

LA042201

LAKE PONTCHARTRAIN

121

83

0

0

0

18

0

19

0

1

LA

LA041701

LAKE PONTCHARTRAIN

94

51

0

1

0

18

0

14

0

10

LA

LA042206

LAKE PONTCHARTRAIN

73

59

0

0

0

9

0

5

0

0

LA

LA090202

LOWER PEARL

66

57

0

0

0

6

0

0

0

3

LA

LA070404

LOWER MISSISSIPPI-NEW ORLEANS

65

57

0

0

0

4

0

4

0

0

LA

LA042003

PONTCHARTRAIN

63

0

0

0

0

0

0

0

0

63

LA

LA090102

LOWER PEARL

46

0

0

0

0

28

0

4

0

14

LA

LA042102

PONTCHARTRAIN

44

0

0

0

0

0

0

0

0

44

LA

LA070404

MISSISSIPPI

39

0

0

0

0

0

0

0

0

39


-------
ST

HUC

HUC Name

Totals

LA

LA042203

PONTCHARTRAIN

35

LA

LA042205

PONTCHARTRAIN

31

LA

LA041703

PONTCHARTRAIN

30

LA

LA070403

MISSISSIPPI

30

LA

LA042207

PONTCHARTRAIN

27

LA

LA070402

MISSISSIPPI

26

LA

LA041801

LAKE PONTCHARTRAIN

21

LA

LA042203

LAKE PONTCHARTRAIN

21

LA

LA042206

PONTCHARTRAIN

20

LA

LA042104

PONTCHARTRAIN

18

LA

LA041704

PONTCHARTRAIN

15

LA

LA090208

PEARL

15

LA

LA042004

PONTCHARTRAIN

13

LA

LA040910

PONTCHARTRAIN

12

LA

LA042204

PONTCHARTRAIN

10

LA

LA090103

PEARL

10

LA

LA041901

PONTCHARTRAIN

9

LA

LA020802

BARATARIA

7

LA

LA041703

LAKE PONTCHARTRAIN

7

LA

LA041801

PONTCHARTRAIN

4

LA

LA070403

LOWER MISSISSIPPI-NEW ORLEANS

4

LA

LA042103

PONTCHARTRAIN

3

LA

LA090202

PEARL

3

LA

LA042004

LAKE PONTCHARTRAIN

2

LA

LA070601

MISSISSIPPI

2

LA

LA040910

LAKE PONTCHARTRAIN

1

LA

LA041401

PONTCHARTRAIN

1

LA

LA041704

LAKE PONTCHARTRAIN

1

LA

LA042101

LAKE PONTCHARTRAIN

1

0

0

0

0

0

0

9

0

0

0

0

0

0

0

0

0

0

0

7

0

0

0

0

2

0

1

0

1

1

Total

%

36,335

9,876
27%

P

Chi

Prod

DO

Secchi

Turbidity

Salinity

Temp

0

0

0

4

2

0

4

25

0

0

0

0

0

0

0

31

0

0

0

0

0

0

0

30

0

0

0

0

0

0

0

30

0

0

0

4

1

0

4

18

0

0

0

0

0

0

0

26

0

1

0

1

0

0

0

0

0

0

0

0

0

11

0

0

0

0

0

2

1

0

2

15

0

0

0

0

0

0

0

18

0

0

0

2

1

0

2

10

0

0

0

2

1

0

2

10

0

0

0

0

0

0

0

13

0

0

0

0

0

0

0

12

0

0

0

0

0

0

0

10

0

0

0

2

1

0

2

5

0

0

0

2

1

0

2

4

0

0

0

2

1

0

2

2

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

4

0

0

0

0

0

4

0

0

0

0

0

0

0

0

0

3

0

0

0

0

0

0

0

3

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

2

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

366

0

3,159

335

6,419

9,581

6,599

0%

1%

0%

9%

1%

18%

26%

18%


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 24

4.4 Mississippi

Data distribution within Mississippi coastal waters is illustrated in several summaries. Table 4.5 (below)
contains an aggregated summary of total observations in Mississippi waterbodies. Analytical species of
nitrogen, phosphorus and chlorophyll were aggregated as shown in Table 4.2. The total observations are
indicated for each Mississippi HUC and these have been placed in descending order; from the greatest to
least total number of observations. Attachments F and G provide visual distribution of total stations
within the Mississippi portion of the project study area (Attachment F) and relative magnitudes among the
coastal HUCs (magnitudinal centroids shown in Attachment G). Temporal distributions are indicated in
Attachment H where information on total annual observations are provided for each HUC. Attachment K
summarizes the total counts for all primary standardized parameters (non-aggregated).

The data distribution in the Mississippi HUCs is concentrated within nitrogen (26%), phosphorus (14%),
DO (28%) and temperature (23%). The other parameters are relatively underrepresented. However, there
is a relatively good coverage of the database, i.e., some nutrient and DO data available for many of the
HUCs. Biological response information such as chlorophyll and productivity are significantly low, in
addition to information on water column light transparency (Secchi and turbidity).

Baiteiie

The Business of Innovation


-------
Table 4-6. Ranking of Mississippi coastal HUCs by total records of primary causal/response parameters. Shown are total records for all
nitrogen (N), phosphorus (P) and chlorophyll (Chi) analytical values, and phytoplankton productivity (Prod), dissolved oxygen
concentration and percent saturation (DO), Secchi depth, turbidity, salinity and water temperature (Temp).

^ !3

o as'

& S.

ST

HUC

HUC Name

Basin Name

Totals

N

P

Chi

Prod

DO

Secchi

Turbitity

Salinity

Temp

MS

3170009

West Mississippi
Sound

MISSISSIPPI
COASTAL

6,148

2,074

1,539

0

0

1,067

43

71

531

823

MS

031700060303

N/A

PASCAGOULA

4,965

1,072

511

0

0

1,663

38

114

142

1,425

MS

031700090309

Bayou Talla- Jourdan
River

MISSISSIPPI
COASTAL

3,259

476

213

1

0

1,253

13

88

174

1,041

MS

031700090704

N/A

MISSISSIPPI
COASTAL

3,159

557

247

0

0

1,128

34

68

100

1,025

MS

031700090703

Flat Branch- Turkey
Creek

MISSISSIPPI
COASTAL

2,394

703

274

0

0

651

55

102

86

523

MS

031700099999

N/A

MISSISSIPPI
COASTAL

2,204

531

281

0

0

632

80

98

205

377

MS

031700060302

Big Swamp-
Pascagoula River

PASCAGOULA

1,712

136

218

0

0

625

0

0

101

632

MS

3170009

East Mississippi
Sound

MISSISSIPPI
COASTAL

1,659

455

340

1

0

444

1

41

100

277

MS

031700090203

Polar Branch-Wolf
River

MISSISSIPPI
COASTAL

1,348

764

275

0

0

126

0

62

0

121

MS

031700090403

Cutoff Bayou-
Jourdan River

MISSISSIPPI
COASTAL

1,168

267

91

0

0

387

19

38

63

303

MS

031700090702

Bayou Portage

MISSISSIPPI
COASTAL

1,166

472

148

0

0

262

19

14

40

211

MS

031700090706

Old Fort Bayou- Biloxi
Bay

MISSISSIPPI
COASTAL

905

420

165

0

0

138

0

35

17

130

MS

031700090801

Bayou Cumbest

MISSISSIPPI
COASTAL

784

247

112

0

0

219

0

30

26

150

MS

031700090605

Parker Creek-

Tchoutacabouffa

River

MISSISSIPPI
COASTAL

697

148

93

6

0

222

0

1

26

201

MS

031800040704

Mulatto Bayou-Pearl
River

LOWER PEARL

588

142

55

1

0

174

16

58

70

72

MS

031700090701

Flat Branch- Bayou
Bernard

MISSISSIPPI
COASTAL

529

187

50

1

0

146

1

2

0

142

MS

031700080706

Cunningham Branch-
Escatawpa River

ESCATAWPA

413

113

81

0

0

106

0

21

26

66

MS

031700090602

Bayou Costapia-
Tchoutacabouffa
River

MISSISSIPPI
COASTAL

339

115

41

15

0

96

0

24

0

48

MS

031700090506

Mill Creek- Biloxi
River

MISSISSIPPI
COASTAL

266

26

52

0

0

91

0

0

26

71

MS

031700090308

Rotten Bayou

MISSISSIPPI
COASTAL

181

6

2

0

0

86

0

2

4

81

MS

LA92LR14

East Mississippi

Mississippi Coastal

163

12

24

0

0

38

13

0

38

38


-------
ST

HUC

HUC Name

Basin Name

Totals

N

P

Chi

Prod

DO

Secchi

Turbitity

Salinity

Temp





Sound























MS

031700090402

Lower Bayou la Croix

MISSISSIPPI
COASTAL

120

24

4

0

0

46

0

4

6

36

MS

031700060301

Black Creek Swamp-
Pascagoula River

PASCAGOULA

114

30

6

0

0

34

0

5

0

39

MS

03170009

N/A

MISSISSIPPI
COASTAL

91

0

0

0

0

26

13

0

26

26

MS

031700090604

Choctaw Creek-
Tuxachanie Creek

MISSISSIPPI
COASTAL

80

32

8

4

0

16

4

0

8

8

MS

031800040703

Upper Mulatto Bayou

LOWER PEARL

57

0

0

0

0

38

0

0

0

19

MS

031800040702

West Pearl River-
Pearl River

LOWER PEARL

43

1

0

0

0

26

0

2

0

14

MS

031700090705

Old Fort Bayou

MISSISSIPPI
COASTAL

34

14

3

0

0

6

1

3

0

7

MS

031700090501

Horse Creek- Biloxi
River

MISSISSIPPI
COASTAL

30

0

0

0

0

19

0

1

0

10

MS

031700090505

Lower Little Biloxi
River

MISSISSIPPI
COASTAL

30

0

0

0

0

20

0

0

0

10

MS

031700090202

Bell Creek-Wolf River

MISSISSIPPI
COASTAL

26

20

0

0

0

0

0

0

0

6

MS

031700090307

Bayou La Terre

MISSISSIPPI
COASTAL

25

7

2

0

0

8

0

2

4

2

MS

031800040701

Mikes River

LOWER PEARL

22

0

0

0

0

12

0

4

0

6

MS

031700060107

Indian Creek-
Pascagoula River

PASCAGOULA

18

6

0

0

0

1

0

0

0

11

MS

031700090201

Crane Creek

MISSISSIPPI
COASTAL

18

8

2

0

0

4

0

2

0

2

MS

031700090302

Catahoula Creek

MISSISSIPPI
COASTAL

18

8

2

0

0

4

0

2

0

2

MS

031700090303

White Cypress Creek-
Hickory Creek

MISSISSIPPI
COASTAL

17

7

2

0

0

4

0

2

0

2

MS

031700060203

Fourmile Branch-
Moungers Creek

PASCAGOULA

16

7

1

0

0

4

0

2

0

2

MS

031700060204

Paige Bayou-
Moungers Creek

PASCAGOULA

15

7

1

0

0

4

0

1

0

2

MS

031700060303

West Pascagoula
River-East
Pascagoula River

PASCAGOULA

7

0

0

0

0

2

1

0

2

2

MS

031700090704

Back Bay of Biloxi

MISSISSIPPI
COASTAL

7

0

0

0

0

2

1

0

2

2

MS

031700099999

N/A

MISSISSIPPI
COASTAL

7

0

0

0

0

2

1

0

2

2



































Total Counts

34,842

9,094

4,843

29

0

9,832

353

899

1,825

7,967







%



26%

14%

0%

0%

28%

1%

3%

5%

23%

o

K> O
4--


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 27

5.0 TOWARD l lll DEVELOPMENT OF NUTRIENT CRITERIA

The ultimate goal of this project is to provide the best available means to support the states of Louisiana,
Mississippi and Alabama in the development of nutrient criteria for coastal waters. The Steering
Committee has determined that one essential step toward developing nutrient criteria is to focus on the
identification of reference conditions for estuarine and other coastal waters of interest within the study
area. Reference conditions support comparative analyses among groups of similarly functioning
ecosystems such that responses to agents of change (e.g., nutrient delivery) can be equated. Therefore,
relative ecological states can be assessed and management sectors can be well-informed of likely
responses to management decisions.

The term reference condition can have multiple meanings. EPA (2001) defines reference conditions to be
"...the comprehensive representation of data from several similar, minimally impacted, 'natural' sites on
a waterbody or from within a similar class of waterbodies..." The use of reference conditions in some
ecosystem restoration activities (e.g., salt marsh restoration) are often used to track relative changes in
ecosystem function from some baseline state which is not necessarily pristine or uninfluenced by human
activities. This is often referred to as functional equivalency. Reference sites can be paired, or multiple
sites can be combined to create functional equivalency assessments over large areas (Steyer et al. in
press). For estuarine and coastal waterbodies, the definition for reference conditions is focused on the
assessment of a desired, or target condition. These desired conditions may be based on existing
unimpacted (or pristine) sites, or some knowledge of historical conditions at individual sites. The
determination of reference conditions is key to the development of nutrient criteria. The process is
dependent upon the classification of estuarine and coastal waters into similar groups. The approach of
"one size fits all" has proven futile in most ecological characterizations. This is particularly true for
estuarine systems where biophysical variability is relatively high (Costanza et al. 1993). Therefore, a
series of steps aimed at normalizing whole systems to a set of controlling variables is required to enhance
comparability across space and time.

There are several important factors associated with establishing reference conditions in ecological studies.
These include (but are not limited to) scale, characterization (system type), and statistical power:

5.1 Scale

Temporal and spatial scales play a significant role in ecological comparative analysis. Temporal scale is
important in determining previous, or pristine, conditions, as well as forward-moving trends. Temporal
scale is also critical for understanding responses to management activities (i.e., nutrient reduction
practices). Spatial scale is extremely important in defining and understanding coastal ecosystem
processes. For example, local vs. regional conditions vary but can be strongly or weakly associated based
on their level of "communication."

5.2 Characterization/typology

Physical characteristics such as geomorphologic properties and sediment type are increasingly being used
to normalize ecosystem functional behavior in coastal systems (Boyd et al. 1992, Dalrymple et al. 1992,
EPA 2001, Bricker et al. 2003). For instance, shallow lagoon systems that exhibit strong benthic-pelagic
coupling behave dramatically different than deeper, river-dominated systems. Grouping similar types of
coastal systems is required to perform reasonable comparisons over space and time. There are several
efforts currently underway that are associated with the process of system characterization and typing that
are highlighted in Section 3.1.

Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 28

5.3 Statistical Power

Comparative analyses are only as good as their ability to interpret significant differences among causal
and responsive factors. In order to assess trends or differences to some quantifiable degree of certainty,
minimum numbers and distributions of observations are necessary. Historical data applied to reference
condition assessment may not always represent conditions in a statistically sound manner (i.e.,
inappropriate sampling frequency, replication in time and space, etc.). Future monitoring activities
should be planned such that their results fit into an appropriate sound statistical design.

Issues associated with scale, characterization/typology and statistical power are inherent in several
documents that provide guidance for developing reference conditions in coastal systems. The following
subsections highlight and summarize four examples of methodological review and guidance:

1.	The EPA Nutrient Criteria Technical Guidance Manual (EPA 2001);

2.	NOAA's National Estuarine Eutrophication Assessment (Bricker et al. 1999, Bricker et al. 2003,
Bricker et al. in press);

3.	The Louisiana Coast-wide Reference Monitoring System (CRMS) (Steyer et al. in press), and;

4.	Australia's National Estuaries Assessment and Management (NE) Project (Heap et al. 2001,

Ryan et al. 2003).

5.4 Nutrient Criteria Technical Guidance Manual: Estuarine and Coastal Marine

Waters (EPA 2001)

This project is consistent with much of the direction contained within the Nutrient Criteria Technical
Guidance Manual for Estuarine and Coastal Marine waters (EPA, 2001). There are several chapters
devoted to background and historical research, monitoring, and modeling of nutrient enrichment to
coastal water bodies. The key chapters associated with the development of reference conditions for
estuaries and coastal waters in Northern Gulf of Mexico region include Chapter 3 (Classification of
Estuarine and Coastal Waters), Chapter 6 (Determining the Reference Condition), and Chapter 7
(Nutrient and Algal Criteria Development).

Chapter 3 deals specifically with classification of estuarine and coastal waters. According to the manual,
the purpose of classification is stated:

"Where possible, similar estuaries, tributaries, or coastal reaches should be equated through physical
classification to reduce the magnitude of the criteria development problem and to enhance predictability
of management responses. To be useful, classification should reduce variability of ecosystem-related
measures (e.g., water quality factors) within identified classes and maximize interclass variability. This is
important because managers need to understand how different types of estuaries and coastal waters, as
well as important habitat differences within these systems, respond to nutrient over-enrichment in or to
plan effective management strategies."

Several classification schemes are reviewed in this manual. These classification schemes are introduced
in conjunction with a series of factors associated with determining the susceptibility of estuarine systems
to nutrient loading. These factors, modified from the NRC (2000), include the following:

1.	System dilution and water residence time, or flushing rate

2.	Ratio of nutrient load per unit area of estuary

Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 29

3.	Vertical mixing and stratification

4.	Algal biomass (phytoplankton and macroalgae)

5.	Wave exposure

6.	Depth distribution (hypsography)

7.	Ratio of side embayment volume to open estuary (local vs. regional communication)

With the exception of nutrient load (2) and algal biomass (4), all of these factors are physical influences.
These physical factors, in conjunction with nutrient loading rates, are significant determinants in how
primary (e.g., algal production) and secondary (e.g., consumers, system metabolism, light attenuation)
factors respond to nutrient loads. Other physical factors include temperature, surface irradiance, and
meteorological events. These are assumed to remain relatively consistent at the regional or subregional
scale.

Although the guidance manual does not endorse any single method of classification, it does offer several
examples including geomorphic classification, man-made estuaries, physical and/or hydrodynamic factor-
based classifications, and other considerations such as habitat type or theoretical issues. These
classification approaches vary from the relatively simple (geomorphologic) to complex (hydrodynamic).
Geomorphic classification provides a first-order estimate of physical characteristics. The manual
identifies the four main groups including: coastal plain estuaries, lagoons, fjords and fjord-like estuaries,
and tectonically caused estuaries (Pritchard 1955, 1967; Dyer 1973). Man-made estuaries are estuarine,
or near-estuarine systems such as drainage canals, boat basins, and other types of waterbodies resulting
directly from anthropogenic activities. They are noted because they have different physical evolutionary
histories and different response characteristics than natural coastal features. Physical and/or
hydrodynamic classifications are based on physical or hydrodynamic factors. Approaches for classifying
estuaries using this approach employs parameters such as stratification, mixing, circulation, water
residence time, river flow, tides, waves, and tidal amplitude. NOAA also has a scheme for determining
estuarine susceptibility which is summarized below (Section 3.2) and one other approach for classifying
estuaries based on physical/hydrodynamic parameters uses comparative systems empirical modeling.
Classification using other types of estuarine categories involves evaluations of habitat type or theoretical
considerations.

Since system response to nutrient enrichment varies considerably, the National Nutrient Criteria Program
(EPA 2001) suggests measuring the nutrient characteristics of minimally impacted sites as a reliable goal.
Chapter 6 (Determining the Reference Condition) of the guidance manual compares and summarizes the
identification of "pristine" vs. minimally impacted waters. The determination of reference conditions
requires some level of appropriate comparison; that is, a minimum number of similar systems to compare.
Five concentration-based approaches for establishing reference conditions are presented along with a
series of relative levels of estuarine degradation (see Table 6-1 in the manual). The guidance manual
describes potential approaches for determining reference conditions by applying historical data, paired or
multisystem comparisons, and dose-response models. In all cases, the point of developing the reference
condition in any system is analogous to establishing a target or minimum water quality metric for either
TN, TP, Chlorophyll-a, Secchi depth, or some combination of these parameters based on site-specific
characteristics.

Chapter 7 (Nutrient and Algal Criteria Development) of the guidance manual suggests a generalized
sequence of milestones in the development of nutrient criteria. Several aspects of this sequence may run
simultaneously; however, they should progress as follows:

Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 30

1.	Historical Information. The identification, collection and review of historical, or antecedent,
conditions may provide immediate and useful information related to historically-driven
reference conditions.

2.	Development of Reference Condition. Resource managers should identify, if possible, a set of
reference sites or historical reference conditions.

3.	Models. Estimates of dose-response relationships, historical nutrient loads, and the management
requirements associated with nutrient reduction can be made through the development of
models. Models may range in complexity from simple, empirical regressions to time-dependent,
spatially-discrete ecosystem response models.

4.	Antidegradation Policy and Allowable Downstream Effects. Nutrient criteria must not
change due to the degradation of reference conditions, and downstream (e.g., shelf) water
quality should be considered in the development of nutrient criteria for estuarine and near-shore
systems.

5.	Regional Technical Assistance Group (RTAG). A group of regional experts should be
assembled to work with the States in the review of water quality data, development of
monitoring plans, and the development of nutrient criteria.

5.5 NOAA National Estuarine Eutrophication Assessment

The National Estuarine Eutrophication Assessment (NEEA) is an approach that surveyed the
characteristics of 16 nutrient related water quality parameters within 82 U.S. estuaries (Bricker et al.
1999). The results of the assessment provided a series of qualitative indicators for these estuarine
systems, and these were compiled into regional condition assessments. NOAA sent a survey to
approximately 400 coastal scientists and managers in the U.S. This survey asked that the scientists assess
regional coastal conditions based on the set of water quality parameters. Based on the responses from the
survey, a series of qualitative indicators were developed that illustrated relative levels (e.g., high,
medium, low, no expression, improving, and insufficient data) of the following primary (Chlorophyll a,
epiphytes, and macroalgae) and secondary (low dissolved oxygen, SAV loss, and nuisance/toxic algal
blooms) symptoms of eutrophication. An overall eutrophic condition was then assessed based on the
average of the magnitudes of these symptoms.

This work also produced a eutrophication survey of the Gulf of Mexico (NOAA 1997). The
eutrophication survey contains summaries of water quality (e.g., nutrients, dissolved oxygen, algae),
physical processes, and physical characteristics (e.g., salinity zones, tidal zones) of the region. This
summary also expresses the relative degree of primary and secondary symptoms in a manner consistent
with the overall assessment report (Bricker et al. 1999).

NOAA is currently planning the next steps of this National Assessment (S. Bricker, personal
communication 1/04). These are currently in draft condition, and include the following:

1.	Identification of a Steering Group to guide further development of the monitoring and assessment
program.

2.	Determination of members for several groups including "Typology" and "Monitoring,
assessment and classification" (including reference conditions and classification issues, such as
variable thresholds), "Modeling and Management" (for development of operational framework
and analytical tools).

3.	Drafting of the scope-of-work for each of the groups, remembering that overlap among groups is
key.

Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 31

4.	Identification and contact of appropriate economists and social scientists for participation and
guidance on development of the economic impacts/society's uses component.

5.	Compilation of a summary of data in preparation for selection of candidate estuaries for a pilot
study (includes searches for models of susceptibility and input/response, such as that for Tampa
Bay).

6.	Begin designing pilot studies.

5.6 Louisiana Coast-wide Reference Monitoring System (CRMS)

The Louisiana Coast-wide Reference Monitoring System (CRMS) is currently a proposed project
focusing on an integrated, statistically-designed, 4-5 year coastal wetland monitoring program. There are
two primary objectives associated with this proposed work. The first objective is to determine,
collectively, the effectiveness of the projects carried out under the Breaux Act restoration plan. This
would be done by providing a network or "pool" of reference sites by which to evaluate project
effectiveness, thereby eliminating the need for one-on-one pairs of reference and project areas. The
second objective of CRMS is to determine the ecological condition of the coastal wetlands based on the
variables measured to ensure that the strategic coastal plan for Louisiana (Coast 2050) is effective in
recreating a sustainable coastal ecosystem (Steyer et al. in press).

A series of questions are related to these two objectives:

1.	Did the projects carried out under the Breaux Act contribute to reducing coastal wetland loss in
Louisiana?

2.	Which project types are the most effective in creating, restoring, protecting and enhancing
wetlands?

3.	Is the Breaux Act effective in reducing the major stressors on wetlands (i.e., flooding regime,
salinity, marsh elevation change)?

4.	Are Louisiana's coastal wetlands accumulating vertically at a sufficient rate to achieve
sustainability?

5.	Are estuarine salinity gradients being maintained to sustain a diversity of vegetation types in the
hydrologic basins?

6.	Has the coastal landscape been maintained with sufficient hydrologic linkages to support efficient
exchange and interface between the marshes and estuaries?

The strategy formed to answer these questions and to achieve the two primary objectives is based on a
state-wide system of reference sites. These would be characterized into ecologically similar groups and
statistically evaluated for changes in function over time in comparison to ongoing restoration activities
(restoration sites). Steyer et al. (in press) state that pristine conditions are rare or perhaps non-existent in
Louisiana and that there are very few sites, if any, that possess pre-project conditions (controls).
Therefore, the project proposes a design that is a composite of several wetland functional assessment
approaches including:

1.	Hydrogeomorphic (HGM)

2.	Functional Assessment of European Wetland Ecosystems (FAEWE)

3.	EPA Ecological Monitoring and Assessment Program (EMAP)

4.	Cumulative Impact Assessment (Lee and Gosselink 1988, Gosselink and Lee 1989)

Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 32

5.	Before-Afiter-Control-Impact (BACI)

6.	Other miscellaneous sources

The approach is based on classifying the coast into a discrete number of similarly functioning regions and
basins, and into project and non-project areas. Further classification is performed based on dominant
vegetation types within these areas.

The monitoring design for the project includes six variables to be measured within 9 basins (project and
non-project results in 18 strata). Based on the requirements implied by the Central Limit Theorem, 30
samples per stratum are being proposed; thus a total of 540 reference sites will be evaluated in the study.
Monitoring information will provide an instantaneous "snapshot" that places project sites within the
functional continuum of all reference sites (current conditions). Monitoring will also track the trajectory
of change of project sites through time with respect to changes in functionally equivalent reference sites
(accounting for regional/sub-regional natural variability). Finally, monitoring information will allow for
comparisons to a group of reference standards (relative distance from desired conditions).

A statistically sound, synoptic monitoring plan is being developed. Remaining design activities involve
identifying an indicator aimed at evaluating responses rapidly, and testing the design in terms of its power
to characterize temporal and spatial variability. Ecological functional models following HGM
methodologies also need to be developed. Due to the inherent degree of coupling between Louisiana's
coastal marshes and estuaries (Childers et al. 1999, Turner 2001), Steyer et al. (in press) suggest that this
cumulative effort of restoration projects and parallel evaluation will "collectively lead to the general
restoration of Louisiana's estuaries."

5.7 Australian Waterway Classifications and Deviation Index

The National Estuaries Assessment and Management (NE) Project of Australia has made significant
progress in the development of an inventory of estuarine (or waterbody) type (classification) and
condition. This assessment includes 974 discrete systems throughout Australia and Tasmania. Their
efforts are illustrated on their web site, the "OzEstuaries Database" (www.ozestuaries.org).

The Australian NE Project is based on a geomorphic classification scheme which was developed by
Harris et al. (2002). The conceptual model of this classification scheme (Figure 3) includes the general
geomorphic and sedimentary environments (or habitats) of seven classes:

1.	Embayments and Drowned River Valley (EMB)

2.	Wave-dominated Estuaries (WDE)

3.	Wave-dominated Deltas (WDD)

4.	Coastal Lagoons and Strandplain-associated Creeks (CL/SP)

5.	Tide-dominated Estuaries (TDE)

6.	Tide-dominated Deltas (TDD)

7.	Tidal Creeks (TC)

Battene

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 33

River

Wave

Tide

Figure 5-1. Ternary classification of coastal systems divided into seven classes (after Dalryinple et

al. [1992] and taken from Ryan et al. [2003]).

Conceptual models, examples of systems associated with these classifications, and a full classification of
all 974 water bodies are documented in Ryan et al. (2003).

A report on the quantitative assessment of the health of Australia's estuaries and coastal waterways
(Heap et al. 2001) includes a review of the evolution and geomorphologic types of Australian coastal
systems, conceptual models of dose-response of nutrients in four classes of coastal systems, an overview
of biogeochemical and physical factors affecting the nature of dose-response relationships, a series of
statistical analyses of existing data, and the development of a "deviation index'' which is used to
determine relative ecological states and to rank all systems for comparative analysis (reference
conditions). Technical descriptions of data collection and mapping activities are also included in this
assessment.

The approach employed by the Australian NE Project focuses on habitat type based on sediment qualities
(facies). Much emphasis is put on sediment organic matter and sediment nutrient concentrations and their
use as indicators of system denitrification rates and habitat quality. The deviation index is aimed at
developing a reference condition-based tool to assess target or desired conditions with the consideration
of natural variability as well as ranking coastal systems in terms of their level of relative degradation.

The Estuarine Geoscience Database (OzEstuaries) is an online warehouse of publications, estuarine GIS
mapping, estuarine conditions assessments, estuarine water quality data, conceptual models and
descriptive information, an indicators program, and other sources of information. The site allows the user
to create and view site maps and Landsat images, review condition assessment reports, and review water
quality data for most of the 974 coastal systems throughout Australia and Tasmania.

5.7.1 U.S EPA Aquatic Stressors Framework

The EPA is currently developing a framework by which to assess coastal ecosystem responses to multiple
stressors. The process is driven by the EPA Aquatic Stressors Framework (USEPA 2002) which has
formed a Diagnostics Committee that oversees research and development associated with this work. The

Batreiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 34

goal of the research is to develop tools and methods that will promote management decisions related to
major aquatic stressors. At this time there are four primary stressors under evaluation: (1) nutrients, (2)
suspended and bedded sediments, (3) toxics, and (4) altered habitat. A series of conceptual models are
being developed that are intended to provide guidance in how coastal ecosystems respond to these
multiple stressors. The ultimate goal of this work includes an improved ability to support State and Tribe
303(d) impaired waters assessments and the implementation of necessary management activities to
improve water and ecosystem quality.

The approach to this study involves the development of four distinct stressor-response conceptual models
that possess physical and modifying factors. A series of primary physical factors tend to control, or
govern, the extent of ecosystem response to agents of change. For example, water residence time, depth,
hypsography, salinity, and other physical controls influence the type of response an ecosystem has to
individual, or multiple, agents of change. Therefore, this framework is based on a coastal classification
system whereby classifying similar biogeophysical attributes of systems will increase the ability to assess
response characteristics. The classification framework (in development) is currently at the scale of 8-digit
HUCs which are modifications of NOAAs estuarine and coastal drainage area (EDA & CDA) and USGS
HUC classifications. Further refinement to smaller spatial units (e.g., 12-digit HUC scale) may increase
the ability to assess ecological response across many systems.

Therefore, this research is quite relevant to studies of nutrient criteria in coastal waters. The approaches
being developed are similar to the Australian classification and indices work, but are more relevant
because they are covering water bodies and contributing watersheds within the project study area and at
similar spatial and temporal scales. The efforts associated with this program will likely be valuable to
nutrient criteria development, and vice versa.

5.8 Discussion (change to section 5.6)

In a recent paper by R.E. Turner (Turner 2001) the fundamental question is posed: are Gulf of Mexico
estuaries unique? In attempting to answer this question, Turner (2001) produced a series of descriptive
comparisons between Gulf estuaries and others across the U.S. He also reviewed historical and existing
research on Gulf estuaries to synthesize information that would support an overall classification and
comparative functional assessment associated with estuarine systems that are susceptible to
eutrophication. His findings and conclusions can be summarized as follows:

1.	Although the microtidal regime of Gulf estuaries sets them apart from most other U.S. regions,
there is no apparent regional signature to the estuarine landscape as measured by geomorphic
criteria (e.g., per unit watershed drainage, estuarine surface area, freshwater turnover time, and
wetland area/water surface area).

2.	Wetland biogeochemical responses to watershed and open water coupling can determine
important sources, losses, and transformations of dissolved and particulate organic matter in
Gulf systems (see also Childers et al. 1999).

3.	It would be useful for estuarine managers to know if wetland restoration rates vary among
estuaries in order to plan, allocate resources, and effectively evaluate projects.

4.	One primary goal associated with the understanding of variability imposed by physics,
chemistry, geology, and biology is to develop an efficient mechanism that makes comparable
measurements of key parameters in many estuarine systems.

5.	A field-based, cross-system comparison of estuaries that examines different landscape scales,
ages, and vegetation types, etc., will further reveal the strength of these bio-geo-physico-
chemical interactions.

Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 35

6.	Important physical parameters to consider are wave energy, freshwater turnover times, size,
watershed morphology, and water column stratification.

7.	Important biological parameters include wetland:water ratios, landscape mosaics (fragmentation,
source materials, and composition), nutrient loading rates, light regime, and couplings with
upstream and continental shelf systems.

8.	In summary, a comparative analysis emphasizing forcing functions, rather than geographic
uniqueness, will lead to significant progress in understanding how all estuaries function, are
perturbed, and even how they can be restored.

Thus, Turner (2001) suggests that the estuarine systems within the study area are no less variable in
nature than others throughout the U.S. However, they are significantly influenced by extensive riverine
discharge (e.g., Mississippi River Delta and tributaries to Mobile Bay) and relatively high ratios of
wetland:estuarine surface area. He further suggests that a comparative analysis, based on focused field
monitoring, will contribute positively to the understanding of system behavior, and therefore ecological
responses to nutrient loads. This view is consistent with that offered by EPA (2001) and NOAA (Bricker
etal. 1999).

The five approaches and one review paper summarized in this section all share similar fundamental
properties; however, they differ in specific methods. A synthesis and comparison of particular methods,
results, and lessons learned in these and other examples would likely benefit this project.

The EPA guidance manual (EPA 2001) offers a broad overview of methods and approaches available to
state and tribal managers for the development of nutrient criteria. The emphasis, however, seems to be on
using nutrient concentrations and Secchi depth as the primary response indicators. Additional
investigations into the use of other supporting biogeochemical and biological response indicators may be
necessary, particularly in cases where water column nutrient concentrations are too variable or
appropriate data do not exist.

The NOAA NEEA (Bricker et al. 1999) provides a broad comparison of relatively large estuarine systems
in the Gulf (i.e., on the scale of Mississippi Sound, Mobile Bay) using a significant number of causal and
response variables. Although this assessment includes several Gulf of Mexico coastal systems, it may be
too general (large-scale) and qualitative to be directly applicable to many of the water bodies of interest
within this project's study area. However, according to NOAA, further refinement of this assessment is
planned and collaboration between EPA, NOAA, Louisiana, Mississippi, and Alabama would be prudent
in order to maximize efficiency. This is particularly true if the NOAA NEEA work will include a
national "typology" effort.

Although the methods for conducting the CRMS work in Louisiana proposed by Steyer et al. (in press)
are associated primarily with wetlands assessments, there is significant overlap in relevance to water
quality assessments in many of the water bodies within the study area. The close coupling between
coastal wetland and estuarine habitats and processes within the study area justify a close collaborative
effort between the CRMS work and nutrient criteria assessments. The approach for determining reference
conditions will require significant resources and effort; however, it will likely be one of the most
comprehensive comparative studies to date. Similar classification and referencing methods may be
considered for developing nutrient criteria in the estuarine systems of the northern Gulf of Mexico.

The Australian National Estuaries Assessment and Management (NE) Project applies a geomorphologic
and sedimentary type habitat classification (facies) approach to developing a cross-system comparison of
ecological status. The approach to classification and comparisons is well-documented (Heap et al. 2001,
Ryan et al. 2003). Condition assessments are being completed for each system based on available data.

Baiteiie

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 36

Water quality data are available online through a GIS-centered, user-friendly website. This overall
approach may provide a decent model for further development of the database and criteria development
associated with this project. In addition, important "lessons learned" may provide added value and
efficiency to further regional comparative eutrophication studies.

6.0 RECOMMENDATIONS

This section is a summary of recommendations that are associated with further progress toward
development of nutrient criteria in the coastal waters within the study area.

6.1 Data Gaps Analysis

A thorough data gap analysis should be performed on priority and representative water bodies within the
study area. A data gap analysis will result in the prioritization of candidate reference sites for
comparative studies and also can be used to evaluate the need to augment and/or revise current and future
monitoring plans. This will be based on at least two methods of analysis: (1) identify locations in
priority coastal waters where temporal monitoring has been ceased or interrupted and (2) identify
locations within priority coastal waters where monitoring should be implemented. An overall,
statistically robust plan would be beneficial to the desired outcome of creating nutrient-related criteria
(water quality or ecosystem functional targets) and reference sites used to quantify improvement or
degradation of particular systems over time. Future monitoring may overlap or coincide with parallel
efforts of other investigators and agencies. Collaboration in planning and implementation would reduce
unwanted redundancy and augment efficiency.

6.2 Identify Appropriate Classification Process

Coastal and estuarine waterbodies are physically and biologically complex systems. System classification
is necessary in order to limit the problems of interpretation caused by natural variability among system
types. There are several approaches being applied to system classification and most are directly
associated with physical or developmental characteristics. The project's steering committee should
review the various methods of classification and determine the most appropriate one(s). Section 5 of this
report summarizes some recently developed methods related to system classification, biogeochemical
states, and responses to stressors. They are all focused on determining reference conditions and methods
to quantify system change attributed to management activities and independent of natural variability.
Perhaps the most relevant, regionally-specific method is associated with the EPA Multiple Stressor
research. Battelle has been communicating with members of the Diagnostics Committee to find relevance
between projects to increase efficiency. This research has already accomplished a classification analysis
within the Northern Gulf study area, but at a spatial resolution that may need to be further optimized.
Their conceptual model on ecosystem responses to nutrient enrichment can be further developed and
modified in combination with further site-specific system classification. Ongoing research in Europe,
Australia, and other regions of the US should still be reviewed on a regular basis to capture innovative
approaches and discoveries.

6.3 Further Develop Database and Information Access Program

The database for this project is being developed with no specific application design, i.e., for a variety of
end-users. Plans to further develop a user-friendly web interface or front-end where specific information
can be easily reviewed and/or retrieved is strongly suggested. Examples of such a design exists in
individual state websites, STORET, the newly created EPA National Nutrient Criteria database and the
Australian NE Project's web site. An interactive database as such would provide a tool for assessing

Battelle

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 37

quantitative changes in ecosystem function, or state, over time. It would also be an important resource in
communicating project scope and results to critical stakeholders and similar programmatic entities.

Modification of this database to make it compatible with the EPA's National Nutrient Criteria database is
a high priority. Recent dialog between project steering committee members, Battelle, and EPA
Headquarters has resulted in an initial assessment of good compatibility between databases. Further work
on this is necessary to effectively transfer data from the Northern Gulf database to the emerging National
Nutrient Criteria database.

7.0 REFERENCES CITED

Boyd, R., Dalrymple, R., and Zaitlin, B. A., (1992) Classification of clastic coastal depositional
environments. Sedimentary Geology. 80:139-150.

Bricker, S.B., C.G. Clement, D.E. Pirhalla, S.P. Orlando, and D.R.G. Farrow. 1999. National Estuarine
Eutrophication Assessment: Effects of Nutrient Enrichment in the Nation's Estuaries. NOAA,
National Ocean Service, Special Projects Office and the National Centers for Coastal Ocean
Science. Silver Spring, MD: 71 pp.

Bricker, S.B., G. Matlock, J. Snider, and A. Mason. In Press. National Eutrophication Assessment
Update: Workshop Summary and Recommendations for Development of a Long-term
Monitoring and Assessment Program (Pre-Distribution Copy). September 4-5, 2002. Patuxent
National Wildlife Research Refuge, Laurel, MD.

Bricker, S.B., J.G. Ferreira, and T. Simas. 2003. An integrated methodology for assessment of estuarine
trophic status. Ecological Modelling. 169:39-60.

Childers, D.L., S.E. Davis III, R. Twilley, and V. Rivera-Monroy. 1999. Wetland-water column

interactions and the biogeochemistry of estuary-watershed coupling around the Gulf of Mexico,
p. 211-235. In T.S. Bianchi, J.R. Pennock, and R.W. Twilley (eds), Biogeochemistry of Gulf of
Mexico Estuaries. J. Wiley & Sons, Inc., New York.

Costanza, R., Kemp, W.M. & Boynton, W. 1993. Predictability, scale, and biodiversity in coastal and
estuarine ecosystems : implications for management. Ambio. 22(2):88-96.

Dalrymple, R. W., Zaitlin, B. A., and Boyd, R., (1992) Estuarine facies models; conceptual basis and
stratigraphic implications. Journal of Sedimentary Petrology. 62:1130-1146.

Dyer, K. 1973. Estuaries: A Physical Introduction. New York: Wiley-Interscience.

Gosselink, J.G. and L.C. Lee. 1989. Cumulative impact assessment in bottomland hardwood forests.
Wetlands. 9:83-174.

Harris, P.T., A.D. Heap, S.M. Bryce, R. Porter-Smith, D.A. Ryan, and D.T. Heggie. 2002. Classification
of Australian clastic coastal depositional environments based on a quantitative analysis of wave,
tide and river power. Journal of Sedimentary Research. 72(6):858-870.

Battelle

The Business of Innovation


-------
Final Report

Work Assignment 1-01 Task 7

September 2004
Page 38

Heap, A., S. Bryce, D. Ryan, L. Radke, C. Smith, R. Smith, P. Harris, and D. Heggie. 2001. Australian
Estuaries and Coastal Waterways: A Geoscience Perspective for Improved and Integrated
Resource Management. AGSO Record 2001/07.
http://www.ozestuaries.org/document/documentation.html

Lee, L.C. and J.G. Gosselink. 1988. Cumulative impacts in wetlands: linking scientific assessments and
regulatory alternatives. Environmental Management. 12:591-602

Monsen, N.E, J.E. Cloern, and L.V. Lucas. 2002. A comment on the use of flushing time, residence
time, and age as transport time scales. Limnology & Oceanography. 47(5): 1545-1553.

National Oceanographic and Atmospheric Administration (NOAA). 1997. NOAA's Estuarine

Eutrophication Survey, Volume 4: Gulf of Mexico Region. Silver Spring, MD: Office of Ocean
Resources Conservation and Assessment. 77 pp.

NRC (National Research Council). 2000. Clean Coastal Waters: Understanding and Reducing the
Effects of Nutrient Pollution. Washington, DC: National Academy Press.

Pritchard, D. 1955. Estuarine circulation patterns. Proc. Am. Soc. Civil Eng. 81:717-1 - 717-11.

Pritchard. D. 1967. What is an estuary: physical viewpoint. In: G. Lauff, ed. Estuaries. Washington,
DC: American Association for the Advancement of Science. Publication 83.

Ryan, D.A., A.D. Heap, L. Radke, and D.T. Heggie. 2003. Conceptual models of Australia's estuaries
and coastal waterways: application for coastal resource management. Geoscience Australia
Record 2003/09. http://www.ozestuaries .org/document/documentation.html

Steyer, G.D., C.E. Sasser, J.M. Visser, E.M. Swenson, J.A. Nyman, and R.C. Raynie. In Press. A

proposed coast-wide reference monitoring system for evaluating wetland restoration trajectories.

Turner, R.E. 2001. Of manatees, mangroves, and the Mississippi River: Is there an estuarine signature
for the Gulf of Mexico? Estuaries. 24(2): 139-150.

U.S. EPA. 2001. Nutrient Criteria Technical Guidance Manual: Estuarine and Coastal Marine Waters.
EPA-822-B-01-003, October 2001, U.S. Environmental Protection Agency, Office of Water,
Washington DC

USEPA. (2002). Aquatic Stressors: Framework and Implementation Plan for Effects Research (EPA

600/R-02/074). Research Triangle Park, NC: U. S. Environmental Protection Agency, Office of
Research and Development.

Baiteiie

The Business of Innovation


-------