The SOW is intended to support the Southern Estuaries (SE) module of the MAP and is directly linked to the 3.2.3.5 and 3.2.4.5 monitoring components. This SOW includes the objectives of the work effort, a general description of the scope citing the methodologies to be used, a detailed breakdown of all tasks to be performed and their associated deliverables, and timeframes, planning, coordination, data review, report preparation and submittal, equipment purchases, rental and ownership and project management.

MAP Activity 3.2.3.5 - Seagrass Fish, Pink Shrimp, and Invertebrate Assessment Network

Since 1983, a series of studies, using the 1-m² throw-trap, have been conducted by the National Oceanic and Atmospheric Administration, National Marine Fisheries Service (NOAA/NMFS), and the U.S. Geological Survey (USGS) in shallow seagrass dominated waters of south Florida for the purposes of quantifying abundance and species composition of seagrass-associated fish and invertebrates, including the pink shrimp, Farfantepenaeus duorarum, in relation to habitat (bank, basin, near-key; grass canopy) and salinity. Collectively, a 20-year baseline dataset has been accumulated that includes within its period-of-record: seagrass die-off between 1987-1990 and 1998-2000; hypersalinity associated with the drought of 1989-1991; and the advent of persistent algal blooms in Florida Bay beginning about 1992.

At present this dataset consists largely of data for Johnson Key Basin (JKB) in western Florida Bay. The period-of-record in JKB begins in 1983 and is continuous, on a 6-week interval, from October 1994 through the present. It currently is being funded by the Critical Ecosystems Studies Initiative of the National Park Service. Over time the JKB dataset has been augmented and/or complemented with other studies, providing a bay-wide perspective on faunal distributions. More limited data is available from Biscayne Bay where, since October of 2002, a study funded by the Florida State Legislature (through the South Florida Water Management District) has been underway in southern Biscayne Bay (Chicken Key to Turkey Point) to characterize the near-shore epibenthic fish and invertebrate communities, including spatial and temporal variation. The study employs three types of gear, including the throw trap. Previously, the throw trap has been used only briefly on the southwest mangrove coast. In 1984 the throw-trap was used to sample pink shrimp in southeastern Whitewater Bay at Midway Pass and in Coot Bay. Pink shrimp were sampled monthly in Ponce-de-Leon Bay and Whitewater Bay using the throw-trap in 1995-1996.

This MAP activity will build upon the established baseline dataset for seagrass-associated fish and invertebrates and will enhance the value of ongoing studies in Johnson Key Basin and southern Biscayne Bay by incorporating them into an expanded regional network of 20 sampling locations and 3 reference locations (one reference site at each location, Biscayne Bay, Florida Bay, and along the southwest mangrove coast). The network will be sampled twice annually: during April/May at the end of the dry season and September/October at the end of the wet season. Sampling protocols with the quantitative 1-m² throw-trap will be standardized across the network with sampling tightly coordinated with the concurrent seagrass monitoring South Florida Fish Habitat Assessment Program (FHAP-SF) (MAP activity 3.2.3.3).

MAP Activity 3.2.4.5 - Factors Controlling Epibenthic Communities of Near-shore Biscayne Bay

Adjustment of established sampling protocols and sampling effort to local habitat and faunal distribution characteristics in order to minimize sample variability locally and across the sampling network is a critical component in the development of this assessment network. Minimizing gear bias by employing a single method across the network is highly desirable. Previous work demonstrates that seagrass and algal species composition and cover and fish and invertebrate species composition and density are linked, and all vary greatly seasonally and spatially in south Florida. Understanding and quantifying the influence of habitat on faunal distributions and density is critical to discerning the influence of salinity.

The 1-m² throw-trap, a device first used to sample freshwater fishes in the Everglades, was adapted to sample seagrass associated fish and shrimp (prey for higher trophic levels) in shallow marine seagrass beds in Florida Bay, where it has now been used for over 20 years. The throw-trap has advantages over traditional gear types such as trawls and seines because it is usable across the full depth range present in Florida Bay, it is quantitative for fish and invertebrates present within the seagrass/algal canopy, and it is highly efficient because the area enclosed can be repeatedly swept until clear of target organisms. Also important is the small area sampled, allowing samples to be tightly linked to habitat.

Gear comparisons, to date, corroborate the original assumptions about the throw-trap. Sampling with the throw-trap, as used in Florida Bay, has been compared with a 0.5-m sled net and a 1-m roller-frame trawl while, in the current Biscayne Bay study, the throw-trap is being compared with the 14-foot commercial roller-frame trawl as well as with a small pull-trawl usable in shallow water. The throw-trap is the only sampling gear that can be used effectively in all depths to be sampled in this study. One strength of the throw-trap is that it samples habitat specifically; however, because the area covered by each sample is small, the sampling design (i.e., number of replicates, stratification) must be carefully fitted to the special distributional characteristics of each area, with special attention to areas of high habitat heterogeneity such as Biscayne Bay.

This MAP activity will enhance the ongoing gear comparison work in Biscayne Bay by complementing it with field experiments to determine sample number, replication, and habitat specific sampling for their effect on variance and the ability to detect change. Similar experiments conducted in Florida Bay and along the southwest mangrove coast will allow network optimization of sampling effort and maximized capability to detect change given available funding. The emerging sampling strategy will be tightly coordinated with the concurrent seagrass monitoring South Florida Fish Habitat Assessment Program (FHAP-SF), MAP activity 3.2.3.3.

2. Objectives.

a. MAP Activity 3.2.3.5

1). Quantify fish and invertebrate (caridean and penaeid shrimp, crabs) species abundance and community composition at 23 locations (20 sites, 3 reference sites) in Florida Bay (includes Johnson Key Basin), along the southwest mangrove coast including Whitewater Bay (USGS), and in Biscayne Bay and Card and Barnes Sounds (NMFS) in relation to habitat and salinity.

2). Directly link species abundance and community composition with measures of habitat (seagrass/algae, water depth, etc.) and salinity.

3). Integrate faunal sampling with the concurrent seagrass monitoring FHAP-SF, MAP activity 3.2.3.3, by coordinating the timing of sampling and sampling within the 30-cell FHAP grid.

b. MAP Activity 3.2.4.5

1). Evaluate the 1-m² throw-trap for use in monitoring seagrass-associated fish and invertebrates by comparison with roller-frame trawls used in deep water and pull-nets used in shallow water by reviewing and summarizing available data including that resulting from the current Biscayne Bay study.

2). Evaluate the effect of sample size, replication, and habitat specific sampling on sample variability and change detection in Biscayne Bay, Florida Bay, and along the southwest mangrove coast with the throw-trap.

3). Develop a sampling strategy optimizing sampling effort and maximizing change detection capability within funding constraints across 20 sampling locations in Florida Bay (includes Johnson Key Basin), along the southwest mangrove coast including Whitewater Bay, and in Biscayne Bay and Card and Barnes Sounds. The emerging sampling strategy will be tightly coordinated with the concurrent seagrass monitoring FHAP-SF, MAP activity 3.2.3.3.

3. Scope of Work

This SOW will quantify change and trends in the nearshore epibenthic fish and invertebrate (caridean and penaeid shrimp, crabs) communities in Florida Bay, Biscayne Bay including Card and Barnes Sounds, and the southwest mangrove coast including Whitewater Bay. It has two components: (a) monitoring epibenthic densities periodically (3.2.3.5) and (b) special-focus studies to adjust the sampling design to the distribution characteristics of the sampling environment (3.2.4.5). The initial sampling design and throw-trap sampling protocols are based on sampling experiences in the dense grass beds in Johnson Key Basin, western Florida Bay. Recent experience in Biscayne Bay indicates that distributional characteristics of habitat and fauna are sufficiently different than the JKB sampling design needs adjustment. The limited experience with the throw-trap in southwest coast estuaries indicates that adjustments to the JKB sampling design also will be needed there. The purpose of the special focused study is to make these improvements.

Sampling, analyses, and assessment will be conducted by the USGS and NMFS. The USGS will have primary responsibility for monitoring sites in the southwest mangrove coast and Johnson Key Basin and Rabbit Key Basin of western Florida Bay. NMFS will have primary responsibility for monitoring sites in Biscayne Bay and Eastern and Central Florida Bay.

a. MAP Activity 3.2.3.5

The initial plan is that 8 locations in Florida Bay, 6 locations along the southwest mangrove coast including Whitewater Bay, and 6 locations in Biscayne Bay including Card and Barnes Sounds (see Figure 1) will be sampled twice during each year at the end of the dry and wet seasons in south Florida, April/May and September/October, respectively. Seasonal data available from the long-term dataset for Johnson Key Basin and the 2-year dataset for southern Biscayne Bay will be used to assess the appropriateness of these sampling periods. Adjustments may be made following preliminary sampling. Reference sites will be established in each of the three areas. The reference site in Biscayne Bay (bay-side Elliot Key), established in the current SFWMD funded study, will be continued. Reference sites in Florida Bay and along the southwest coast may be established from among sites already proposed for sampling but which are relatively distant from where impacts of upstream CERP projects are expected, e.g. Lostmans River (SW01), Johnson Key Basin (FB20). All sampling sites will be coordinated with, and conform to, those proposed in the MAP South Florida Habitat Assessment Program (MAP activity 3.2.3.3).

A 1-m² throw-trap will be used to sample fish and invertebrates (caridean and penaeid shrimp and crabs). The throw-trap is an open-ended 1-m² aluminum box, 45 cm deep, with panels of nylon netting (0.16 mm stretch mesh DELTA netting) attached on parallel edges at the top. Each panel of netting is large enough to cover the top of the throw-trap when it is used in water deeper than 45 cm. The throw-trap is thrown into undisturbed water dropping to the bottom and immediately covered in water deeper than 45 cm. After the trap is in place, it is cleared of animals with separate passes of a 1-m wide framed sweep net of mesh size similar to the panels. In water deeper than 45 cm the throw-trap is covered between each pass of the sweep net. SCUBA equipment is used while clearing the trap in deep water. All material from the three separate sweeps are combined and washed over a 1 mm sieve in the field. Washed samples are maintained on ice in the field and preserved in 10% formalin at the dock. Samples are processed in the laboratory where all fish, shrimp and crabs are identified to species, counted and measured. Data are entered into the Everglades Marine Benthic Database from the laboratory via the Web. A type set of the species found is maintained in the laboratory and identification reviewed by subject area experts. Preserved specimens, not needed in the laboratory, are maintained in the Dan Beard Center, Everglades National Park.

With each throw-trap sample associated seagrass and algal cover will be quantified using the Braun-Blanquet cover estimation technique and appropriate physical and environmental parameters (at minimum: salinity, temperature, turbidity, water depth, sediment depth) will be measured. Braun-Blanquet cover estimation is the technique used to quantify seagrass and algal cover in FHAP-SF (MAP activity 3.2.3.3). A random sampling design will be implemented within the 30-cell FHAP-SF sampling grid in order to ensure close coupling of seagrass and algae monitoring and fish and invertebrate sampling. Permits will be required from Miami-Dade County, Biscayne National Park, Keys National Marine Sanctuary and Everglades National Park.

b. MAP Activity 3.2.4.5

Available data comparing the throw-trap with roller-frame trawls and devices pulled through seagrass in shallow water grass beds will be evaluated and used where appropriate to improve throw-trap sampling protocols. In addition, the number of sweeps required to capture >95% of fish and invertebrates enclosed by the throw-trap will be evaluated for two samples at each sampling location as well as reference sites (see Figure 1). Each throw-trap will be swept six times with the 1-m sweep net, keeping material collected separate. Cumulative catch curves will be evaluated to determine the number of sweeps required.

Field experiments will be conducted to evaluate the effect of sample size and habitat on sample variability with the 1-m² throw-trap. Experiments will be conducted at four of the proposed sampling locations (see Figure 1), one each in western and eastern Florida Bay, one in Biscayne Bay including Card and Barnes Sound, and along the southwest mangrove coast including Whitewater Bay; locations with seagrass habitat representative of the area will be selected for sampling. A random sampling design will be used within the 30-cell FHAP-SF sampling grid established at each sampling location. Sample variability with the 1-m² throw-trap under three sampling scenarios will be determined: 1) number of samples (maximum 30 samples, random habitat), 2) replication (maximum 30 sample pairs, random habitat) and 3) restricting sampling to seagrass habitats (maximum 30 sample pairs in seagrass). The basic experimental design is to collect paired samples randomly (without regard to seagrass habitat) in each grid cell and to augment this collection with additional randomly located sample pairs in seagrass habitat such that a total of 30 sample pairs from seagrass habitat are collected at each location. Cumulative sample variability with increasing sample effort will be compared for individual samples versus paired samples and for sampling random habitat vs. restricting sampling to seagrass habitat. For the final sampling design, sampling effort will be adjusted to minimize sample variability locally or among locations in order to maximize the capability of the network to detect change with available funds.

Figure 1. Points (open or solid squares and circles) indicate proposed South Florida Habitat Assessment Network sampling locations: solid squares and circles indicate ongoing sampling. Dashed boxes enclose the FHAP-SF sites proposed for fish and invertebrate sampling. [larger image]

c. Coordination of Study.

This project requires the coordination of field teams lead by both NMFS and USGS. The expert biologists from each agency will work together to accomplish the objectives of the project. The expert biologist from USGS will serve as the project coordinator to conduct the day-to-day operations of the project. The coordinator will ensure standardized sample collection and processing, quality control, data management, and close communication between teams from both agencies. The expert biologists from both USGS and NMFS will be active in planning details of the project, particularly the design, data collection, pilot testing, and analysis.

4. Work Breakdown Structure

a. Introduction. The results of the work performed under this SOW will be used to develop the cumulative finds of the AAT System Status Annual Reports. These annual reports will be used by the AAT to develop a RECOVER Technical Report at five-year intervals, as pursuant to the regulations [Section 385.31 (b)(4)]. This Technical Report presents an assessment of whether the goals and purposes of the CERP are being achieved. The Report will also include an assessment of whether the Interim Goals and Interim Targets are being achieved or likely to be achieved and evaluating whether corrective actions should be considered based on scientific findings of system-wide or regional ecological needs. The Principal Investigator(s) (PIs) will be required to work with the AAT Modules Chair to assist in the development of the AAT System Status Annual Report and asked to include their participation as a task in this work breakdown structure. Additionally, the following reporting guidance is offered by AAT to the principal investigator(s):

1) Evaluate Ability to Detect Change - PI Level
a) Describe the results of the power analysis for the sampling design.
b) Determine the minimum detectable difference of the power analysis, and its associated confidence and uncertainty.
c) Describe changes in the MAP sampling design and its implications for the power analysis and the minimum detectable difference.

2) Establish Reference Condition - PI Level
a) Describe the non-MAP data sources, if any, used in the assessment. If non-MAP data were used, did the data meet the guidance criteria? If the non-MAP data were used and did not meet the guidance criteria, provide a rationale to justify the inclusion of the data.
b) Describe how representative the data are in space and time.
c) Describe the approaches used to address measuring variability.
d) Prepare date for entry into the CERP-Zone and update Module Group.

3) Measure Change from Reference Condition - PI Level
a) Describe the methods used to estimate the direction and magnitude of change in performance measures from the reference state both annually and back-cast for multiple years.
b) Compare current status of the PM with its desired trend or target.
c) Evaluate consistency of monitoring results with MAP hypotheses.
d) Determine if there are indications of unanticipated events and describe how they are affecting the desired outcome.

4) Annual Integration of Performance Measures (PM) To Evaluate Module Hypotheses -Module Group Level
a) Annually integrate multiple PMs to provide an assessment of module level hypotheses.
b) Describe the direction and magnitude of change in the integrated performance measures, and determine if the changes are consistent with expected responses described in the CERP hypotheses.
c) If the trends do not correspond to expected responses, provide scientific explanation.
d) Evaluate progress toward achieving module-level Interim Goals and Interim Targets.

5) System-Wide Performance Evaluation - AAT Level
a) Synthesize findings across-modules and across years to provide a holistic description of the status of the system.
b) Evaluate the results in relationship to supporting system level hypotheses and achieving system-wide Interim Goals and Interim Targets.
c) Summarize those system-wide changes that are consistent with goals and hypotheses and those that are not.
d) Provide a scientific discussion of why the goals and hypotheses are not being achieved.

b. Task Descriptions.

Task 1 - Initial Site Visits and Coordination
Sampling locations will be identified in coordination with the MAP South Florida Habitat Assessment Program (FHAP-SF), MAP activity 3.2.3.3). Reference sites will be identified in Florida Bay and along the southwest mangrove coast, as well as, the currently used western shore of Elliot Key in Biscayne Bay. At each site the 30-cell FHAP-SF sampling grid will be established. The details of site selection and coordination between studies will be established following the initial site visit.

(1) Deliverables - A map locating sample locations and the sampling grids in Florida Bay, Biscayne Bay including Card and Barnes Sounds and along the southwest mangrove coast including Whitewater Bay and appropriate reference sites with an accompanying report describing in detail the sampling design and procedures used to position 1-m² throw-trap samples in coordination with the South Florida Fish Habitat Assessment Program's sampling.

(2) Timeframes - Within 3 months of the Notice to Proceed (NTP)

Task 2 - Preliminary Sampling and Throw-trap Sampling Protocol Development
The 1-m² throw-trap is cleared of animals using a sweep net. A field test will be conducted at each location to determine the number of sweeps required to sample >95% of the fish, shrimp and crabs enclosed by the throw-trap.

(1) Deliverable - A summary report for Florida Bay, the southwest mangrove coast including Whitewater Bay, and Biscayne Bay including Card and Barnes Sounds describing the results of the sweep net tests. Data files of all raw data (quality assured) will accompany the report.

(2) Timeframe - Within 6 months of NTP

Task 3 - Sample Design Development and Work Plan
Field experiments to evaluate the effect of sample size, replication and habitat specific sampling on sample variability will be conducted at 4 sampling locations: Florida Bay (2 locations), southwest mangrove coast (1 location), and Biscayne Bay (1 location). Sampling locations with representative seagrass habitat will be used. A detailed work plan for the south Florida epibenthic fish and invertebrate assessment network (Florida Bay, the southwest mangrove coast including Whitewater Bay and Biscayne Bay including Card and Barnes Sound) will be developed that includes a review of the relevant baseline/historical dataset, describes the random sampling design and procedures for interfacing with the FHAP-SF 30-cell grid, establishes sampling protocols for the 1-m² throw-trap, describes methods for estimating habitat, physical and environmental conditions associated with each throw-trap sample, and establishes a schedule for routine sampling. QA/QC procedures will be detailed. The work plan will detail proposed analyses for characterizing the fish and invertebrate communities sampled, for evaluating faunal relationships with habitat and environment and for evaluating the CERP premise that changes in salinity lead to increases in shoal grass enhancing productivity and resulting in increases in the relative and absolute abundance of mesohaline fish and invertebrates, including the pink shrimp in south Florida. The work plan will include provision for experts to review and assist in the identification of the fish and invertebrates captured. The work plan will detail database management procedures. Initially study data will be locally maintained in the Everglades National Park Marine Benthic Database. Over time the work plan will be modified to provide the study data to RECOVER in the format requested. The work plan will include a proposed deliverables schedule and required content.

(1) Deliverable - The completed work plan delivered with an appendix of results for field tests evaluating sample variability in relation to sampling design alternatives. Results will be discussed in relation to balancing sample size and the magnitude of change that can be detected across sampling locations (plus 3 reference sites (one in each major sampling region). Data files of all raw data (quality assured) will accompany the work plan and appendix.

(2) Timeframe - Within 9 months of NTP

Task 4 - Seasonal Sampling and Reporting
Following work plan approval, sampling will be initiated for Florida Bay, the southwest mangrove coast including Whitewater Bay and Biscayne Bay including Card and Barnes Sounds. Two collections, end of dry season (April/May) and end of the wet season (September/October), are planned for each year. Since Year 1 is a partial year and preliminary experiments are required to refine the sampling design for the first scheduled sampling event will be in April/May 2005. Samples will be processed in the laboratory and data (quality assured), and entered into the NPS Marine Benthic Database.

(1) Deliverable -Annual Reports. An annual summary report comparing the epibenthic fish, shrimp, and crab community observed at sampling locations in Florida Bay, the southwest mangrove coast including Whitewater Bay, Biscayne Bay including Card and Barnes Sounds, and the reference sites at the end of the dry season (April/May). A comparison with available baseline data (Johnson Key Basin, southern Biscayne Bay) will be included. Relationships with habitat and salinity will be evaluated and discussed. The assumptions on sample variability used in the development of the work plan will be evaluated and proposed adjustments to the sampling design, if necessary, detailed. Data files of all raw data (quality assured) will accompany the annual report.

(2) Timeframe - Within 12 months of NTP

c. Relationship of Tasks

Relationships of Tasks 1-6: All four tasks are integral to the main deliverable annual and final summary reports.

1. For data in the southwest mangrove coast, Johnson Key Basin and Rabbit Key Basin of Western Florida Bay, USGS will accomplish Tasks 1-4 by utilizing the Memorandum of Agreement between DA and USGS.
2. For data in Biscayne Bay and Eastern and Central Florida Bay, NMFS will accomplish tasks 1-4 by utilizing the Memorandum of Agreement between DA and NOAA/NMFS.

d. Review and Planning.

This SOW builds upon a well defined and reviewed dataset, and involves a partnership between NOAA/NMFS and USGS two federal agencies that have worked together on similar studies for several years. A planning meeting among NMFS, USGS and USACE will be held following NTP of the work order at some location in south Florida to ensure that field methods, data collection, and data management are uniform and standardized.

e. Equipment Purchase, Rental, and Ownership.

Equipment purchases for this SOW include: one boat and motor. Since the SOW is long term (3 years), it is more cost effective to purchase the boat and motor than to rent or lease these items. Ownership of all capital expense purchases under this SOW with Government funds will become the property of the Government at the conclusion of the work order.

5. Project Management

a. SOW Change Control. Changes in the SOW must be requested of the project manager in writing, with supporting justification. Any requested changes in the SOW will require, on part of the contracting entity, submission of an updated project work plan with supporting detail, updated scheduling and budget information. No changes in the SOW will occur without permission from the project manager. Any delays or changes in the scheduling and budget of the project will require the approval of the Adaptive Assessment Team (AAT). After any approved changes have occurred, the contractor will include documentation of these scope changes in the "lessons learned" section of the final project report.

In addition, in multi-year projects, where the results of each year's work can or will modify what happens in the subsequent years of the contract/project, the annual report can or will provide the results of the work gathered and proposed revisions to the future schedule of tasks/deliverables.

b. Data Management. Submission of all data is required for contract closeout. Data formatting, analysis, and delivery is to meet all CERP data management standards. Any data derived from the project will be provided to the AAT at predetermined intervals. All data and results derived from this project must be made publicly available or available to the AAT at the end of the project.

c. Quality Control and Assurance. The work plan will include quality assurance plan(s). This planning process determines which quality control and quality assurance procedures are appropriate for each project (e.g., QASR, FDEP standards). Methods used for each study should be selected based upon the following criteria: cost-benefit analysis, flowchart diagram of the system process, and determination of the best statistical experimental design (if appropriate). The burden of proof of compliance with standardized quality control and assurance procedures is the responsibility of the contractor. In the case where there are not standardized methods for quality control and assurance (e.g., many ecological sampling methods), the contractor must prove that the suggested methodologies are rigorous. Examples may be citing peer-reviewed literature of a method.

d. Status Reporting. Regular progress reports will be made to the project manager as deemed by the task list. Reports will be written, and verbal reports are not acceptable. Informal reports regarding status of permits needed for the project or timely progress of field work, or completion of specific tasks may be transmitted via email or fax. Reports that include any type of data analysis, datasets, and formal quarterly or interim reports will also be sent via electronic mail; however, signed hard copies with data attached in appropriate format must be mailed to the project manager.

e. Lessons Learned. The causes of variances in the statement of work, scheduling and budgeting, the reasoning behind any corrective actions, as well as, any other lessons learned will be documented in the final report. These lessons learned will become part of the historical database for this study and other RECOVER MAP studies.

Table 1.