Summary work statement

U. S. GLOBEC: Broad-scale patterns of the distribution of zooplankton and nekton in relation to micro- to coarse-scale physical structure in the Georges Bank Region.
Peter H. Wiebe, Timothy Stanton, Ray Schmitt, and Charles Greene+
Woods Hole Oceanographic Institution
Woods Hole, MA
+Cornell University
Ithaca, NY

Sumary

As part of the U.S. GLOBEC NW Atlantic/Georges Bank study, we will determine the broad-scale distribution of zooplankton and nekton populations on Georges Bank using either a towed body equipped with multiple frequency acoustical, bio-optical, and physical sensor systems or the system used in our previous acoustical studies of the Bank. Our primary focus is on making high resolution measurements of volume backscattering and target strength of plankton and nekton throughout the Georges Bank region complemented with optical and physical data to provide a multi-dimensional base of estimates of the spatial distribution of biomass and size frequency of acoustical targets which span the size range of the target species (cod, haddock, Calanus, and Pseudocalanus) and their predators. When used, the multi-frequency multi-beam system will provide high spatial resolution estimates of volume backscattering and target strength. The bio-optical system will provide high resolution video images of the plankton which will be used to determine the relative composition and size distribution of the taxa along the tracklines in the immediate vicinity of the towed body. These data together with theoretical models will be used to interpret the acoustical data in terms of meaningful biological properties, such as biomass and size distribution. The physical sensors will provide the information about fine to coarse-scale physical structure of the Bank environment. One of the towed systems will be deployed on as many of the broad-scale surveys of the Georges Bank region in each of the years 1997, and 1998 as the funding permits. These surveys are designed to cover the principle period of the pelagic stages of the gadid ichthyoplankton (Jan-June). The acoustic measurements can be directly related to the data generated by the optical and net collections of plankton and nekton. The acoustical spatial maps and time-series will also provide essential linkage between the physical oceanographic conditions on the Bank and the biological distributions of the species as determined from the net collections at the stations distributed throughout the Georges Bank region.

Specific Work Plan

The goal of this component of the program is to provide a basic description of the animal distributions over the Bank and surrounding waters based on acoustical data collected during the along-track steaming of the research vessel between the broad-scale stations. This work will focus on the use of a new multi-sensor towed platform which can acoustically image the entire water column remotely and instantaneously while acquiring optical and physical data which are essential to interpret of the acoustical images and understand biological/physical coupling processes. With the appropriate "decoders", the acoustically based information can be used to describe the patterns of distribution and abundance of the organisms at high resolution and over large regions. The acoustical estimates will be compared with the physical data to determine correlations of abundance and animal behavior with physical structure on various scales. This new system will be phased in; the bioacoustic system used during the first two years of the program will be continued in service.

The sampling program

We will make multi-sensor measurements along each of the tracklines between stations on the broad-scale survey cruises that we are able to participate on. With our new system (BIOMAPER II), the sensors include active acoustics (five transducers facing up and five facing down in range of 38 to 1000 kHz), bio-optics (including VPR at two magnifications), and physical properties at fine- to coarse-scales (temperature, conductivity, fluorometry, transmissometry, downwelling light).

The sampling instrumentation

To significantly advance our ability to conduct quantitative surveys of the spatial distribution of Georges Bank zooplankton and nekton, we are currently constructing BIOMAPER II which will consist of a multi-frequency sonar, video plankton recorder system (VPR), and environmental sensor package. Also included are an electro-optic tow cable; a winch with optical slip rings; and a van to hold the electronic equipment for real-time data processing and analysis. The system will be able to operate in a surface towed down-looking mode at speeds up to 10 kts, in a vertical oscillatory "towyo" mode, or in a sub-surface up/down looking horizontal mode at slower speeds. The latter arrangement will permit acoustical measurements close to the surface and over greater range of depths, especially at the higher frequencies. As an additional measure to ensure that we obtain sufficient data to evaluate the range dependent bias against observing small targets in dense layers or far from the transducer, when this system is used, the towed body can be lowered obliquely to within 10-15 meters of the seafloor to obtain a profile of target strength distributions as a function of depth. These data can be collected as the research vessel leaves each station and should prove essential to the interpretation of the along-track data.

When using BIOMAPER II, the tow cable consists of a strength member with optical fibers inside for high speed data transmission and communication between the underwater portion of the system and the ship-board data logging and processing computers. The computers will produce acoustical and optical "maps" of the animal distributions and maps of micro- to coarse-scale physical structure. The calibrated acoustic data will be presented both in raw form (i.e., volume scattering strength, with no assumptions) and in the form of size distributions of animals. The latter requires the use of scattering models, inverse techniques, and occasional samples of animals for use in the inversions. The optical/video data will be viewable in real-time, but currently, post processing is required to acquire the data for quantitative comparison with the acoustics. The system is designed for work in near shore coastal regions with a working depth of 300 meters.

Acoustics: The optimal frequencies for in situ studies will vary with application. On BIOMAPER II, the following frequencies will be available for continuous operation: 38 kHz, 120 kHz, 200 kHz, 420 kHz, and 1 MHZ. This combination of frequencies provides the ability to reach the entire water column at the lower frequencies and detection of the sub-millimeter sized animals at the higher frequencies. The principal output of the system will be volume scattering strength at each frequency over a range of depth bins. Also, split-beam data will be collected at 38, 120, 200, and 420 kHz for direct target strength measurements on individual macrozooplankton, micronekton, and nekton (for conditions under which they are acoustically resolved). Those frequencies with split-beam target-strength estimation (and target-tracking capabilities while on station) can be used to derive more information for acoustically characterizing larger isolated targets in the animal assemblage, including macrozooplankton, micronekton, and fish (Greene et al., 1992). The other system will provide echo integration data at 120 kHz and 420 kHz.

VPR: BIOMAPER II will also carry a Video Plankton Recorder (VPR) that is a high resolution camera video imaging system designed to unobtrusively quantify the distribution and abundance of zooplankton encountered along the tow path at 60 fields per second. The BIOMAPER II VPR will be equipped with one camera which images concentrically positioned volumes of 28 cm3. This image volume can be modified as needed. The other system will not carry a VPR.

Acoustical data analyses:

Acoustic data from the broad-scale survey cruises will be processed in real time and plots of volume backscattering as a function of depth will be available during the cruise. In-depth analysis will occur after each cruise at both WHOI and Cornell. We will use state-of-the-art techniques in volumetric geographical information systems to manage, statistically analyze, and visualize the acoustic data from each survey from four aspects.

Qualitative visualization of echograms and the relationship of echogram features to physical structure of water and bathymetry. This is the most direct form of presentation as it directly displays the various data sets without having been subject to interpretation or transformation.
Use of MOCNESS and VPR data to scale the acoustic data so that size distribution and biomass of taxa can be extrapolated to the entire water column. This analysis assumes that the relative proportions of taxa remain fixed at a given depth. This method will only be used if the net or VPR samples support the assumption.
Doing the forward problem to check the validity of (ensemble averaged) scattering models for each hydrographic and biographic region on the Bank. Here, comparisons will be made between measurements of acoustic scattering levels and predictions of acoustic reverberation based on VPR or net samples. If successful, then an inversion of the multifrequency data may be possible as discussed in 4.
If the results of number 3 are acceptable, use inversion algorithms to estimate size distributions of taxa. This analysis would make use of a multi-taxa inversion algorithm. We plan to modify the algorithm for use in this multi-taxa environment. The analysis will be used under conditions where the signal-to-noise ratio is high enough in each channel, the number of different taxa contributing to the echoes is not too great, and the Rayleigh/geometric scattering transition point is bracketed for each taxa.

VPR data analysis:

The VPR will be mounted on a rail projecting from the anterior end of BIOMAPER II so that disturbance of the imaged volume is minimized. Video data from the camera will flow through coaxial cable via a fiber optic multiplexor and up the fiber optic tow cable to a surface demultiplexor. Raw video data will be archived to SVHS tape and it is hoped that a PC based image processor will be available to extract regions of interest (ROIs) from the video field which contain in-focus targets. These ROIs will be written to disk for subsequent identification. The VPR data are processed in four stages. (1) The image processor examines each field of the videotape for targets which are in focus. (2) Subsections of each video field which contain an in-focus target are cropped and stored in files along with a video time code. The original tape is archived and can be re-run through the processor for additional ROI extraction as desired. Postprocessing is currently performed manually. (3) A MATLAB routine sequentially displays ROIs and allows an observer to identify and measure the dimensions of each target using a calibrated cursor. (4) Data consisting of VTC timecode, target identity, length, and any user comments are written to an ASCII file for subsequent analysis. Efforts are proceeding to develop an automated identification algorithm based on texture, length:width ratios, structural taxonomic characters and other factors, and the results to date have been extremely promising.

Sample And Data Needs From Broad-Scale Surveys

In addition to the VPR data collected when BIOMAPER II is deployed, in order to interpret the acoustic data, stratified net tows at a number of locations across the Bank are required. Out of 60 stations, at least 20 should have stratified information, 10 taken during the day and 10 taken at night. A subset of these zooplankton samples will be silhouette photographed and the silhouettes analyzed for the size frequency distribution of the taxa. An extensive set of length-to-wet-weight regression equations derived for each taxonomic group will be used to estimate the biomass of each individual. These data will provide a detailed description of the variations in the size spectra of the zooplankton at each station and enable interpretation biologically of the mobile acoustics data. The ADCP current data collected on the broad-scale surveys are also needed to interpret the bank-wide distribution patterns of the acoustic field. A significant problem in interpretation is that the fluid field that is being ensonified is moving. The ADCP data will permit the tracking of the water movements. With this information, the effects of water motion can be removed and the 3-D distribution of the organisms reconstructed.

Analysis and Synthesis

Some specific problems to be addressed:

Identification of the Well-Mixed versus Stratified Regions and the Tidal Mixing Fronts. The location of the tidal mixing front along the southern flank of Georges Bank is believed to be a critical region with regard to the retention of larval cod and haddock on the Bank. Much of the water flowing along the southern flank carrying the developing eggs and larvae exits the Bank via storm events, entrainments of water around warm-core rings, or flows southwest to the mid-Atlantic Bight. Some flow makes the turn at the southwestern end of the Bank in the vicinity of the Great South channel and re-circulates. In addition, some water and animals are thought to cross through the tidal mixing front and enter the well-mixed area where residence times are longer, especially during the late spring and summer periods. The seasonal development and position of the tidal mixing front is poorly known, as is its relationships to the distributions of biomass and zooplankton taxa. We should be able to map the occurrence and to some extent the structure of this frontal region, and should be able to relate this structure to the animal distributions and the physical factors affecting it.
Identification of large scattering patches associated with submarine canyons and other bathymetric features. The bank-wide surveys provide a base of high resolution data that can be used to identify the intensity and spatial dimensions of patches of the zooplankton on scales from meters to ten's of kilometers. Of particular interest in the early portion of the year are spawning cod and haddock aggregations. We believe that we have observed such aggregations on the flank of the Bank near Great South Channel and believe we will be able to see such aggregations on the northeast peak region of the Bank, a dominant spawning area. The acoustic data will need to be corroborated with collections from trawl surveys and we will have the opportunity to do this when the broad-scale surveys parallel the NMFS groundfish assessment cruises.
Confounding factors: The occurrence of sand particles in the water column of the well-mixed areas of the Bank represents a significant problem in the interpretation of the acoustic data. We will, however, be able to independently assess the amount of sand present in the water column from pump and MOCNESS collections and possibly from the VPR images when the system is deployed. We also expect a distinct signature from sand in the microstructure records. Documenting the presence or absence of sand in the water column and relating the occurrence to physical forces (e.g. wind events, tidal flows during spring and neap tidal periods in relation to topography) will be a first step in determining the impact of this material on the zooplankton populations.
Internal waves and solitons. Once stratification takes place on the edges of the Bank, internal waves and solitons can propagate freely on and off the Bank. These physical structures can strongly affect the vertical distribution of the plankton and may enhance the turbulence structure in localized portions of the water column. Internal waves propagating onto the Bank may also play a role in determining the position and structure of the tidal mixing front since the energy they carry must be dissipated when the waves meet the front. Little is known about this process, however, and we view the mapping of these structures as a first step in understanding the role they play in creating physical structure on the Bank.

Significance of Proposed Research and relationship to other components in the Program

This program of broad-scale along-track surveys of integrated acoustical/bio-optical/physical measurements will play two roles important to the U.S. GLOBEC Georges Bank Program's ultimate success as a climate change program. First, intensive, broad-scale surveys are essential to determine the pattern of the annual cycle and some of its inherent variability. Ultimately, these must be specified in sufficient detail to recognize longer term trends from natural, interannual variability. Once recognized, these longer term trends may be the signal of global climate change or they may correspond to natural, inter-decadal or longer time-scale variability. This also puts point samples into proper context and connects them.

Second, the products of the acoustical program should enable us to determine what level of effort is required to monitor fundamental changes in the ecosystem over time. Obviously, the current intensive level of effort for the survey program cannot be sustained indefinitely because of the high financial and manpower costs of extensive shipboard operations. One goal is to determine how to reduce the level of shipboard measurements and to replace them with cost effective remote-sensing technologies. Satellite remote sensing alone cannot provide us with sufficient information about the processes at work in the Georges Bank ecosystem. New acoustic technologies hold great promise for fulfilling many of the long-term measurement requirements. To realize this promise, however, these technologies must be integrated into the full survey program to determine what information they can and cannot provide. When the U.S. GLOBEC Georges Bank Program is completed and the shift is to long-term environmental measurement, we must ensure that appropriate, cost-effective acoustic methods for getting the job done are operational.