NOAA Technical Memorandum NMFS NE 166
Report on the
Third Northwest Atlantic Herring
Acoustic Workshop,
University of Maine
Darling Marine Center,
Walpole, Maine,
March 13-14, 2001
by William L. Michaels1,3,
Editor and Coconvenor,
and Philp Yund2,4, Coconvenor
1National
Marine Fisheries Serv., Woods Hole Lab., 166 Water St., Woods Hole, MA 02543;
2National
Science Foundation, Biological Oceanography Program, 4201 Wilson Blvd., Arlington,
VA 22230
Print
publication date December 2001;
web version posted November 6, 2002
Citation: Michael WL, Yund P, editors/co-convenors. 2001. Report on the
Third Northwest Atlantic
Herring Acoustic Workshop,
University of Maine Darling Marine Center,
Walpole, Maine, March 13-14, 2001. NOAA Tech Memo NMFS NE 166; 18 p.
Download complete PDF/print version
BACKGROUND
Early cooperative efforts by U.S. and Soviet scientists to survey Atlantic
herring (Clupea harengus) by using hydroacoustics ("acoustics")
in the Northwest Atlantic began in the late 1960s. These research efforts
were abandoned by the early 1970s due to the severe decline of Atlantic
herring stocks on Georges Bank from overfishing, and to the limitations
of analog acoustic instrumentation. The use of fisheries acoustics for
research in U.S. waters of the Northwest Atlantic did not resume for
about two decades, while scientists in government, academia, and industry
working in other regions continued to advance fisheries acoustics. The
development during the last decade of scientific quality echosounders
with digital output and accurate calibration procedures has resulted
in the international acceptance of fisheries acoustics as a valuable
tool for surveying and assessing pelagic fish populations.
A number of independent research efforts by Canadian and U.S. scientists
implemented fisheries acoustics to survey and estimate the biomass of
Atlantic herring in various regions of the Northwest Atlantic during
the last decade. Some of these scientists from the National Marine Fisheries
Service's Northeast Fisheries Science Center (NEFSC), Canadian Department
of Fisheries and Oceans (DFO), Maine Department of Marine Resources (MDMR),
and Gulf of Maine Aquarium/University of Maine (GMA/UM) met in Woods
Hole, Massachusetts, in 1998 to discuss their acoustic research on Atlantic
herring in the Gulf of Maine, Scotian Shelf, and Georges Bank regions.
That meeting was the first in a series of Northwest Atlantic Herring
Acoustic Workshops. During the first workshop, scientists provided an
overview of their field operations and postprocessing procedures. Fisheries
acoustic research by the NEFSC included Simrad EK500, omnidirectional
sonar, midwater trawling, and underwater video operations in the Georges
Bank and Gulf of Maine regions. Scientists from the DFO's St. Andrews
Biological Station described their operations with Femto echosounders,
sidescan sonar, trawling, and seining from commercial and research vessels
on the Scotian Shelf and Georges Bank. The DFO had also begun collaborative
efforts with the University of New Brunswick to develop calibration and
postprocessing procedures for the recently developed Simrad SM2000 multibeam
system. The MDMR presented data based on the use of a Simrad EY500 aboard
charter vessels to survey Atlantic herring in coastal waters and nearshore
banks in the Gulf of Maine. The GMA/UM initiated a project to implement
automated acoustic data loggers aboard commercial Atlantic herring vessels.
An overview was also provided on Simrad BI500, SonarData, and Femto postprocessing
software. The need to coordinate field operations, compare procedures
and results, and develop cooperative Atlantic herring acoustic research
in the Northwest Atlantic was recognized at the first workshop.
The second Northwest Atlantic Herring Acoustic Workshop occurred at
UM's Darling Marine Center in Walpole, Maine, during January 18-19, 2000.
Overviews were provided by NEFSC, DFO, MDMR, and GMA/UM scientists of
their 1998 and 1999 research, and these scientists compared preliminary
biomass estimates. The results from an intervessel comparison between
the F/V Mary Ellen and FR/V Delaware II were also presented.
Scientists discussed the importance of accurate calibrations. Approaches
for deriving Atlantic herring abundance and biomass estimates by using
acoustic data were discussed and recommended. Representatives from the
Atlantic herring fishing industry attended the first day to provide suggestions
for improving survey operations. This workshop concluded with the planning
and coordination of 2000 field operations.
The Third Northwest Atlantic Herring Acoustic Workshop was held at
the Darling Marine Center during March 13-14, 2001. The workshop focused
on specific topics identified during the second workshop that had the
largest effect on the estimates and variances of the acoustic measurements.
The participation was also expanded to include additional Canadian scientists
who had analytical expertise in estimating the abundance of Atlantic
herring by using acoustic data from various regions in the Northwest
Atlantic. The goals of this workshop were to evaluate ongoing research
in, identify research requirements of, and improve cooperative operational
and scientific efforts in, fisheries acoustics to obtain more accurate,
cost-effective, and timely population estimates and variances for Atlantic
herring assessment. This document summarizes the presentations and discussions
of the workshop.
SUMMARY
OF PRESENTATIONS AND DISCUSSIONS
Automated Acoustic Data Logging aboard Commercial
Vessels
Philip Yund
Darling Marine Center
University of Maine
Walpole, Maine, USA
Presentation
The Gulf of Maine Aquarium's fisheries acoustic project has collected
acoustic data in both fishery-dependent (i.e., normal fishing
operations) and fishery-independent (i.e., systematic transects)
modes for 2 yr. Results from this year's (i.e., 2000) fishery-independent
survey of prespawning aggregations in the coastal waters of the Gulf
of Maine suggested a total spawning biomass of approximately 300,000
metric tons. A possible spawning aggregation on Jeffreys Ledge in early
November was not surveyed due to scheduling conflicts.
In the fishery-dependent portion of the project, 2 yr of data (i.e.,
1999 and 2000) from the F/V Providian showed consistently greater
biomass on Georges Bank than from coastal or nearshore bank waters (e.g.,
Jeffreys, Platts, and Fippennies Ledge) in the Gulf of Maine (Figure
1). This pattern was consistent with other acoustic results, and
stock assessments suggested that the Georges Bank component of the stock
is approximately an order-of-magnitude more abundant than the nearshore
Gulf of Maine stock.
Discussion
There is considerable U.S. interest for increasing the use of commercial
vessels to survey fish stocks, and of the automated acoustic data logging
system as a tool that can be readily implemented. Good progress has been
made during the last 3 yr in the Gulf of Maine and Georges Bank region
with the collection and processing of acoustic data from commercial Atlantic
herring vessels. Biomass estimates from these data seem reasonable. The
next step is to implement the use of the acoustic estimates into the
Atlantic herring management process. The Canadians have used acoustic
data from commercial vessels more extensively and provide many lessons.
Environmental
Influences on the Density-Distribution of Atlantic Herring
Philip Yund
Darling Marine Center
University of Maine
Walpole, Maine, USA
Presentation
We recently started to compare acoustic data to satellite and CTD (conductivity,
temperature, and depth) data on surface temperature and chlorophyll density
to assess possible oceanographic predictors of Atlantic herring distributions.
Preliminary analyses (of two strata on Schoodic Ridge) suggest that Atlantic
herring appear to be distributed along frontal boundaries between water
masses. Possible correlations with water depth and other variables will
be explored in future analyses.
Discussion
Atlantic herring are temperature sensitive, and their density distributions
can vary between years due to fluctuations in this variable. Environmental
influences are an important concern when determining the time and location
of an acoustic survey. Questions were asked as to whether Atlantic herring
feed during spawning; the NEFSC has observed Atlantic herring feeding
during the spawning period. A critical requirement of fisheries acoustic
research in support of management was recognized in that interannual
biomass indices should reflect trends in abundance rather than environmental
anomalies.
Fisheries-Independent
Estimates from Acoustic Survey Data
William J. Overholtz
Northeast Fisheries Science Center
National Marine Fisheries Service
Woods Hole, Massachusetts, USA
Presentation
Results from the autumn 2000 NEFSC Atlantic herring hydroacoustic surveys
were presented. Three different survey designs -- systematic evenly-spaced
parallel (Figure
2), stratified random parallel
(Figure
3), and systematic zigzag (Figure
4), were employed to survey the spawning concentrations of Atlantic
herring in the northern Georges Bank region during September and October
2000. Results suggest that Atlantic herring were abundant over the region,
especially in the Cultivator Shoals area. The general impression was
that the entire spawning stock was probably covered during each of the
three surveys. Survey estimates indicate that the spawning biomass in
the region is probably large.
Two surveys, each utilizing a smaller set of transects (i.e.,
seven), were conducted over a several day period in early October. These
surveys were designed to obtain information on the comparability of biomass
estimates from two replicates covering the same area. Biomass estimates
from these two surveys were similar, indicating that results are reproducible,
and that calibrations are reliable.
Future uses for the acoustics data were presented. These uses included
directly estimating abundance and biomass, tuning age-structured models,
and calibrating several types of biomass models, including surplus production
and delay-difference models.
Discussion
The NEFSC biomass estimates from acoustic surveys are similar to the
estimates from virtual population analysis, and the estimates within
and among acoustic survey designs are similar. The DFO applies the target
strength (TS)-length regression in a different manner than the NEFSC;
however, both approaches should produce the same result. A review of
these procedures and estimates will be conducted this year by the University
of Miami's Center for Independent Experts. Ongoing research is focused
on evaluating survey design and estimation using classical statistics
and geostatistical procedures.
In-situ
Acoustic Measurements for Northwest Atlantic Herring
William L. Michaels
Northeast Fisheries Science Center
National Marine Fisheries Service
Woods Hole, Massachusetts, USA
Presentation
Acoustic surveys have been conducted during the last 4 yr to estimate
Atlantic herring biomass in the Gulf of Maine and Georges Bank regions.
During these cruises, individual TS measurements for Atlantic herring
have been collected to determine an appropriate TS-length equation. This
relationship is important because it directly affects the abundance and
biomass estimates when scaling the areal backscatter indices.
An in-situ TS experiment on Atlantic herring was conducted on Fippennies
Ledge in the Gulf of Maine during August 1997. Acoustic data were collected
using a Simrad EK500 echosounder operating three hull-mounted transducers
(i.e., 12-kHz single beam, and 38- and 120-kHz split beams).
High-speed midwater rope trawl and Methot trawling operations were conducted
to determine fish and macrozooplankton composition, while underwater
static video provided direct observations of the acoustic targets (Figure
5). The biological composition contributing to the acoustic data
was almost entirely Atlantic herring, euphausiids (Meganyctiphanes
norvegica), and ctenophores. The EK500 data were processed using
the BI500 postprocessing software. The compensated TSs, target depths,
and offset angles from the 38- and 120-kHz data were used to remove potential
false individual targets from slant range and angle discrepancies. This
multifrequency filter removed on average about 98-99% of the TS measurements
to reduce the multiple targets associated with tightly aggregated fish
such as Atlantic herring. The filter removed multiple targets, resulting
in lower TS distributions; however, the mean TS for Atlantic herring
in the Gulf of Maine region remained relatively high in comparison to
the literature (Figure
6 and Figure
7).
Further analyses are needed separate the high TS values that were most
likely from larger fish (e.g., Atlantic cod and haddock) and
sharks that occurred in the study area.
Atlantic herring TS measurements during the day were also significantly
higher than during the night and twilight periods. The EK500 omnidirectional
sonar, midwater trawling, and underwater video sampling operations indicated
that Atlantic herring exhibited vertical migration patterns from near
bottom during the day into the water column at night. Future efforts
will be devoted to determining the TS-length relationship for Atlantic
herring in the Northwest Atlantic, and variability in TS distributions
due to the species' diel behavioral patterns, orientation, enlarged gonads,
or morphology.
Discussion
These high TS measurements for Atlantic herring are similar to observations
from an earlier experiment conducted by John Wheeler and George Rose.
Workshop participants presently use Foote's equation (except for DFO
in St. John's, Newfoundland), but all agree that a new TS-length equation
should be developed for Atlantic herring in the Northwest Atlantic. The
higher 38-kHz TS measurements observed by the NEFSC suggested that Foote's
intercept of -71.9 dB might be increased to possibly -69 dB, resulting
in a decrease of the present biomass estimates by half, although further
analyses are needed. High TS values for Atlantic herring from a pen experiment
by the DFO in St. John's, Newfoundland, have resulted in the use of an
intercept of -66 dB for 120 kHz.
Ex-situ
(laboratory) Experiments to Refine Acoustic Measurements
J. Michael Jech
Northeast Fisheries Science Center
National Marine Fisheries Service
Woods Hole, Massachusetts, USA
Presentation
Fish are complicated scatterers of sound due to the presence or absence
of a swimbladder, the shape of the body and/or swimbladder, their behavior,
and life history changes in anatomy. Anatomical attributes coupled with
organism behavior complicate predictions of fish length from acoustic
echoes, and influence the precision and accuracy of abundance or biomass
estimates from acoustic data. Combining theoretical acoustic models with
in-situ and laboratory measurements will help to explain variability
in acoustic backscatter, provide relationships between length and acoustic
measures, increase the precision and accuracy of acoustic estimates,
and improve target recognition and discrimination among acoustic targets.
The Fisheries Acoustics Research Group at NEFSC has been working toward
integrating acoustic models with in-situ and laboratory measurements
to improve our biomass estimates of pelagic fish stocks in the Gulf of
Maine and Georges Bank regions.
Procedures to obtain digital representations of a fish's body and swimbladder,
predicted backscatter using a Kirchhoff ray-mode model (KRM), and comparisons
of predicted and laboratory measurements were presented at this workshop.
The KRM model uses the fish's body and swimbladder morphometry to predict
echo amplitudes as a function of fish length, acoustic frequency, and
angle of insonification. Radiographs in the dorsal and lateral planes
(Figure
8) are used to construct digital images
of a fish's body and swimbladder. The KRM model approximates the swimbladder
and fish body shapes as finite cylinders and then estimates acoustic
backscatter as a function of acoustic frequency or fish aspect. The KRM
model predicts a nonmonotonic and nonlinear relationship between reduced
backscattering amplitude and fish length or acoustic frequency. In collaboration
with scientists at the Woods Hole Oceanographic Institution, measurements
on live alewife (Alosa pseudoharengus) were obtained in two
planes (dorsal/ventral and lateral) of fish aspect angle (Figure
9). Comparisons of model predictions and laboratory measurements
suggest that the KRM model predictions are reliable and robust over a
wide range of fish aspect angles. We have expanded the KRM model to predict
backscattering amplitude for 3600 of tilt and roll to produce
a three-dimensional backscattering surface (backscattering ambit, Figure
10). The three-dimensional backscattering surface allows visualization
and quantification of the effects of behavior on echo amplitudes and
prediction of acoustic backscatter measurements obtained by SONAR or
multibeam technology.
Discussion
The laboratory measurements presented on alewife, which has a similar
morphology to that of the Atlantic herring, provide a detailed understanding
of the species' acoustic ambit which can help to explain the variability
observed with in-situ measurements. It was recommended that similar laboratory
measurements on Atlantic herring are needed to validate ongoing field
investigations to derive a TS-length relationship for Atlantic herring
in the Northwest Atlantic. It was also emphasized that the multifrequency
measurements from the laboratory will also provide important parameters
that allow intercalibration between existing Atlantic herring acoustic
surveys that implement different frequencies for estimating abundance.
The
Role of Fisheries Acoustics in Support of the Atlantic Herring Industry
Jeff Kaelin
Stinson Seafood, Inc.
Winterport, Maine, USA
Presentation
Prior to joining Stinson Seafood, Inc., as Governmental Affairs Coordinator
in April 2000, Mr. Kaelin was employed at the Executive Director of the
Maine Sardine Council from March 1986 until the council was dissolved
on March 31, 2000. The Maine Sardine Council was an industry-funded governmental
entity of the State of Maine, established in 1951, that supported research
on the Atlantic herring fishery and product quality control of benefit
to the Maine sardine industry and the State of Maine.
For many years, Maine's sardine industry, which utilizes Atlantic herring
for its products, has supported cooperative fisheries research projects
to help estimate the abundance of Atlantic herring available for its
operations. During the late 1970s, for example, the Maine Sardine Council
and the MDMR cooperated in the tagging of juvenile and adult Atlantic
herring along the Maine coast to help understand the migration of Atlantic
herring from coastal Maine spawning grounds. The information gathered
from this work serves as the basis for many of the assumptions of Atlantic
herring spawning stock behavior contained in the federal Atlantic Herring
Fishery Management Plan, approved by the Secretary of Commerce in December
2000.
In recent years, the Maine sardine industry has supported investigations
into the use of acoustic technology to better estimate Atlantic herring
abundance in the Gulf of Maine and Georges Bank region, and to assist
in establishing sustainable harvest levels for the resource. Industry
supported an acoustic Atlantic herring survey conducted by Dr. Richard
Nash and Dr. David Stevenson of the MDMR around 1989. In 1995, the industry
supported efforts by Dr. Stevenson to attempt to use acoustics to survey
Atlantic herring abundance in the vicinity of Jeffreys Ledge, using a
combination of state and federal funds. In 1997, working in cooperation
with the GMA and the NEFSC, the industry helped to identify the need
to establish a world class acoustic survey capability at the NEFSC, and
supported the use of acoustic technology on commercial vessels so that
surveys could be carried out where NEFSC vessels cannot operate due to
shallow water depths.
During Fiscal Years 1998, 1999, and 2000, the Maine sardine industry,
with the assistance of the Maine Congressional Delegation, was successful
in having $800,000 added to the NEFSC budget to establish its acoustic
survey program. A portion of these funds has been utilized to continue
the development of the commercial-vessel survey component. For Fiscal
Year 2003 (i.e., October 1, 2002, through September 30, 2003),
Stinson Seafood and the GMA are again asking Congress to increase the
NEFSC acoustic survey budget, from $200,000 to $500,000, to ensure the
continued development of the region's ability to survey Atlantic herring
spawning stock abundance at a time when acoustic surveys have begun to
be used to determine the abundance of other pelagic species such as Atlantic
mackerel and squids.
Discussion
Good progress has been made with implementing fisheries acoustics for
estimating Atlantic herring populations in the region. The acoustic estimates
are needed to help managers make important Atlantic herring management
decisions. Fishers have reported large aggregations of Atlantic herring
on northern Georges Bank during the fall (similar to the NEFSC results),
and the acoustic estimates will help in assessing the size of the offshore
resource.
Cooperative
Fisheries Acoustic Research in the Gulf of Maine
Donald Perkins
Gulf of Maine Aquarium Development Corporation
Portland, Maine, USA
Presentation
The fisheries acoustic research on Atlantic herring in the Gulf of
Maine region during recent years provides an excellent example of cooperative
efforts between government agencies and industry. The Gulf of Maine Aquarium
Development Corporation's web site (http://www.FishResearch.org/)
was introduced as a tool for increasing cooperative opportunities between
scientists and commercial fishers. There is considerable interest from
industry and funding agencies to foster cooperative research such as
the ongoing Atlantic herring acoustic research.
Discussion
Funding and collaborative research opportunities were discussed.
Atlantic
Herring Assessment and Management in the Gulf of Maine
David A. Libby
Maine Department of Marine Resources
West Boothbay, Maine, USA
Presentation
Atlantic herring assessment and management in the Gulf of Maine were
summarized. Atlantic herring catches were presented for management areas
by month for 1998-2000. Graphical presentations of catches were compiled
by 10-min square on a bimonthly basis from August through December 1998-2000
to correspond with inshore and offshore acoustic surveys. Preparations
are being made for a collaborative United States - Canada Transboundary
Resources Assessment Committee (TRAC) assessment meeting on Atlantic
herring in 2002.
Discussion
Commercial catch data which corresponds in time and space with Atlantic
herring acoustic survey data can improve the partitioning of acoustic
data by species. There was discussion regarding the spawning tolerance
used to manage fishing on spawning components in the Gulf of Maine. Alternative
management strategies used in other areas were also discussed. There
was also discussion of the plans to incorporate and review the acoustic
estimates of Atlantic herring biomass at the TRAC meeting scheduled for
2003.
Length-Stratified
Target Strength Calculations for Biomass Estimates
Claude Leblanc
Gulf Fisheries Center
Department of Fisheries and Oceans
Moncton, New Brunswick, Canada
Presentation
The Southern Gulf of St. Lawrence annual acoustic survey is held at
the end of September on Atlantic herring feeding aggregations in inshore
areas between 20- and 60-m depths. All major acoustic backscattering
detected during survey operations was verified using a midwater trawl
for species compositions and length-frequency distributions of Atlantic
herring schools. To apply the length-frequency distribution in calculating
a mean TS, the following procedure is used. From the detailed biological
samples collected per trawl sets, a length-weight regression is calculated
to obtain slope and intercept values for the relation: W = aLb.
We then use Foote's formula to calculate a TS for every length in the
frequency distribution, with the second part of the equation giving us
a TS value per kilogram:
TS = (20 log10Lcm - 71.9) - (10
log10 Wkg)
We then linearize the TS values per length by using:
TSlinearized = 10 (TS/10)
This linearized value of the TS is then multiplied by the number of
fish measured for each length in the length-frequency distribution. We
next sum the products of this multiplication for all lengths, and divide
the sum by the total number of fish in the length-frequency sample. This
value is a weighted-mean, linearized TS for the entire length-frequency
sample. Finally, we take the antilog value of this weighted-mean, linearized
TS to give us a weighted-mean TS in decibels per kilogram.
Discussion
The calculations presented above are the procedures used by the GMA/UM
and DFO participants in which an average value of -35.5 dB/kg was used.
The NEFSC scientists presently derive the biomass estimates in a slightly
different manner, which should produce the same results. The NEFSC approach
is to obtain a mean length from a series of transects which is converted
to TS using Foote's equation. The individual TS is used to derive the
cross-sectional backscatter coefficient of an individual Atlantic herring
which is used to divide the mean areal backscatter (Sa) estimates to
obtain the number of fish. This estimate is converted to biomass by multiplying
it by the mean weight per fish derived from the length-to-weight conversion
from the survey. The NEFSC scientists will check both approaches to ensure
there are no discrepancies in their biomass estimates.
A
Method for Beam-Width Calibration
Allen Clay
Femto Electronics, Ltd.
Sackville, Nova Scotia , Canada
Presentation
Allen Clay from Femto Electronics spoke on recent developments in the
firm's Hydroacoustics Data Processing System (HDPS). The beam-angle calibration
initiative within the HDPS is primarily intended to improve the confidence
in a transducer manufacturer's beam-angle specification, an important
parameter in the acoustic integration process. It has also been used
to determine the health of all transducer elements, as well as for reviewing
the position and amplitude of side lobes. On one occasion, it was instrumental
in determining which of two possible transducers was installed on an
Atlantic herring seiner. The developed software predicts the beam pattern
based on slight ball movements as a result of changes in the line lengths
of each of the three suspension lines (Figure
11).
The software also accounts for the stretching and shrinkage in the monofilament
line as the ball is moved. Upon completion of the procedure, the software
predicts the error associated with the estimate of biomass should the
manufacturer's specification be chosen over the actual values.
Discussion
It was recognized that most transducer calibrations conducted in the
field rely on the manufacturer's beam-angle specification. Although there
are calibration programs (i.e., Simrad Lobe program) that derive
beam offset parameters, these values are not truly independent from the
manufacturer's specifications.
Size
Discrimination Using a Dual-Frequency System
Allen Clay
Femto Electronics, Ltd.
Sackville, Nova Scotia , Canada
Presentation
The new Femto DE9320 Dual Frequency Digital Echosounder with the 40/120/9.2
Dual Concentric Transducer was presented as a new initiative for addressing
areas of interest in the Atlantic herring fishing community. Femto plans
to exploit this technology to assist in such concerns as mean target
size within a school, mean maturity stage, and Atlantic herring behavioral
separation. In the long term, Femto also hopes to provide a tool for
bottom typing studies. Preliminary study indicates that the TS of an
Atlantic herring of length L obtained by using a frequency F may follow
a relationship described by:
TS = 10 Log [B L2 Trig (C L F)]
where B and C are constants and Trig is a trigonometric function to
be determined. In reviewing the modeling studies done by Mike Jech of
the NEFSC and presented at the workshop, there are indications that this
trigonometric function may exhibit multiple maxima over the commercial
size range of Atlantic herring. Although it is not anticipated that this
work will quickly lead to a quantitative measure of mean target size
since many other factors will influence variability in the signal, it
is felt that we will be moving closer to a solution, and that in the
meantime, the technology may help fishers in their efforts to target
fish of a size needed for a particular market.
Discussion
The advantages with two identical 3-dB beam patterns of different frequencies
from a single transducer were discussed (Figure
12). This development will provide improved target classification,
and does not have the disadvantage of the varying beam patterns associated
with broadband systems.
Recent
Multibeam Developments for Fisheries Acoustics
Gary Melvin (presented by Michael Jech)
St. Andrews Biological Station
Department of Fisheries and Oceans
St. Andrews, New Brunswick, Canada
Presentation
An overview on the history and recent developments in the application
of multibeam sonar technology to fisheries research was presented. Collaborative
efforts between scientists from the University of New Brunswick, DFO's
St. Andrews Biological Station, and the University of New Hampshire have
made good progress in the development of calibration procedures and postprocessing
software for the Simrad SM2000 Multibeam System. The SM2000 is presently
the only multibeam system available that collects backscatter data from
the water column for fisheries acoustic research. Development of the
calibration procedures and postprocessing software is a critical prerequisite
before SM2000 survey operations can be routinely implemented.
Extensive progress was made over the past year on calibrating the sonar
in order that signal returns from the system can be quantified. The first
calibration coefficients obtained from an experiment in February 2001
are presently being incorporated into the display software for computation
of volume backscatter. Advances have also been made in the 3D visualization
of multibeam data. At present, it is possible to display up to 60 pings
on the screen in real time. This provides 3D form to objects as the vessel
passes over the bottom. A function is available within the software to
count echoes when individual targets are observed in the data. The first
attempt to implement multibeam sonar on survey operations for estimating
Atlantic herring biomass is scheduled for 2001.
Discussion
Implementation of the SM2000 Multibeam System for fisheries acoustic
surveys has been proposed in the Northwest Atlantic by a number of agencies
and funding programs. SM2000 installation has also been proposed on new
NOAA research vessels. Ongoing research and development with SM2000 calibration
and postprocessing software are critical requirements before the SM2000
can be routinely used during survey operations.
Spatial
Statistical Approaches for Fisheries Acoustics
Patrick J. Sullivan
Department of Natural Resources
Cornell University
Ithaca, New York, USA
Presentation
Mean biomass estimates and variances were compared from three survey
designs (incorporating stratified random parallel, systematic evenly-spaced
parallel, and systematic zig-zag transects) on northern Georges Bank
during fall 2000 (Figure
13). There appeared
to be insignificant differences between the estimates, suggesting that
biomass estimates from fisheries acoustic operations were robust. A geostatistical
estimator (e.g., kriging) provided a similar mean with reduced
variance (Figure
13).
Spatial
and Temporal Analysis of Fisheries Acoustic Data from Commercial
Vessel Operations
Ross Claytor
Bedford Institute of Oceanography
Dartmouth, Nova Scotia, Canada
Presentation
A method for measuring the spatial and temporal distribution of fish
school densities and their exploitation rates using fishery-collected
acoustic data and Voronoi - natural neighborhood analysis was described.
An Atlantic herring purse seiner fishing on nonspawning feeding aggregations,
and an Atlantic herring gillnetter fishing on smaller, highly dense spawning
aggregations, in the southern Gulf of St. Lawrence, Canada, collected
acoustic data for this study during their regular fishing activity. The
declining-catch-per-unit-of-effort estimator, which was associated with
the gillnet fishing and which was used to assess this stock, reached
asymptotic values at lower-than-expected levels, and was not useful for
tracking daily trends in school density.
Gillnet and purse seine catch per distance searched during fishing
operations was linearly related to school density, and likely is a suitable
abundance index for stock assessment estimates. An individual boat which
collected data in this manner was found to represent trends in the entire
fleet. There was a threshold density below which exploitation rates remained
low. This threshold provides managers with a method for identifying high
exploitation rates and preventing overfishing.
A simulation model calibrated with data from the Pictou 1997 inshore
gillnet fishery compared the properties of abundance indices derived
from fishery acoustic data to those derived from survey indices. The
indices were examined over five fish distribution types, ranging from
a single spike to a uniform flat distribution, four conditions of fishing
and fish movement, and 16 stock sizes for each of these distribution
and conditions. These acoustic data are suitable for deriving abundance
indices provided the searching covers the entire population.
Fishery acoustic abundance indices provide a basis for adopting a decision-rule
management paradigm, and for allowing the metapopulation structure of
Atlantic herring to become the basic management unit for this species.
These results represent an important alternative to the current F0.1 management
paradigm for Atlantic herring populations, and offer an opportunity to
develop a more transparent and responsive management system for the long-term
viability of Atlantic herring fisheries.
Discussion
Discussions focused on the advantages of estimating exploitation rate
from acoustic data collected from commercial vessels. Catch per search
distance is considered more sensitive to changes in biomass than catch
per net, and less sensitive to varying catch efficiencies among vessels.
Survey
Design for Atlantic Herring Acoustic Surveys in the Newfoundland
Region
John P. Wheeler
Northwest Atlantic Fisheries Center
Department of Fisheries and Oceans
St. John's, Newfoundland, Canada
Presentation
Acoustic surveys have been conducted on an annual basis since the early
1980s as part of the research program to assess Atlantic herring stocks
in the Newfoundland Region. Currently, four coastal stocks are surveyed
acoustically on an alternate year basis to estimate stock biomass. Surveys
are conducted either during the late fall or overwintering period, depending
upon stock area.
In all surveys, the survey area is defined as the area from the coastline
to the 120-m depth contour. The 120-m depth contour was selected as the
outer boundary as it has been shown that most Atlantic herring are distributed
within this depth range during the survey period. The survey areas are
divided into strata based upon geographical features and Atlantic herring
distribution patterns. Acoustic sampling intensity (i.e., total
transect length) is allocated to these strata on a proportional basis
based upon Atlantic herring distributional patterns observed in the commercial
fishery and previous acoustic surveys.
The survey design within strata has evolved during the time series.
Originally, zigzag transects were used; subsequently, random parallel
and systematic transects were used. Currently, a multistart systematic
design is employed (Figure
14). Each stratum
is subdivided into blocks with an equal number of parallel transects
per block. Placement of transects is randomly selected in the first block
of a stratum, but is defined by this placement in the remaining blocks.
As transects can be of varying lengths within blocks, weighted mean
densities are calculated for each stratum and extrapolated to the stratum
area to estimate fish biomass. Strata estimates are summed to calculate
a total biomass estimate for the survey area.
The multistart systematic design, currently employed, is deemed to
be advantageous as it allows for the distribution of acoustic sampling
intensity over the survey area in a semisystematic manner, while retaining
the ability to calculate a survey-based variance estimate.
Discussion
There was discussion regarding the survey design. Strata were oriented
perpendicular to the coastal baseline. This design strategy has the advantage
of randomness and even spacing of the transects. This design was decided
upon after a working group reviewed the advantages and disadvantages
of various designs used in fisheries acoustic surveys. Further insight
was also provided on how cruise sampling was allocated among stratum
based on densities from previous cruises. The remaining discussion involved
concern about whether Foote's equation was appropriate for Atlantic herring
in the Northwest Atlantic. The DFO in Newfoundland used a different TS-length
equation with an intercept at -66 dB (for 120 kHz) derived from an earlier
TS experiment, and this seems to support the high TS measurements observed
by the NEFSC.
CLOSING
DISCUSSIONS
Data Management and Availability
Closing discussions began with a summary of NEFSC data management and
availability. Efforts are underway to archive these data in the NEFSC
Oracle database which can be readily accessed by assessment scientists.
The NEFSC Atlantic herring acoustic information (i.e.,cruise
reports, acoustic estimates, and station/catch/biological data from midwater
trawl operation) from the 1998-2000 fall surveys were distributed to
workshop participates. There was discussion about the advantage of minimal
thresholding. Relational data management using Oracle is agreed to be
a good approach which can be readily incorporated into new tools (e.g.,
trawl/acoustical visualization software).
Some participants from DFO and MDMR/GMA manage their acoustic data
using Femto's HDPS data structure. The HDPS archives raw data. The NEFSC
does not archive raw data. The NEFSC data are logged to the BI500 postprocessor
which derives estimates with 2-sec, 0.5-m depth bins, and 0.5-nmi resolution.
Data to this resolution are equivalent to about 2 gigabytes per 12-day
cruise (i.e., about 4,000 nmi).
In regard to the TS data, it was recommended that the raw data be collected
from the EK500 serial port rather than using only binned measurements
from the BI500. It was agreed that raw data will be collected on future
NEFSC acoustic surveys on selected transects where single-species TS
data can be obtained.
The MDMR provided commercial Atlantic herring catch data that can be
overlaid with the 1998-2000 Atlantic herring acoustic surveys. Concern
was addressed regarding accessibility and confidentiality of commercial
catch data. Implementation of VMS vessel tracking in the Gulf of Maine
this year will provide more accurate information of commercial operations
in real time. Canadian scientists have found VMS tracking not only useful
for research and management, but the commercial fishers also have readily
accepted vessel tracking as a valuable tool. The commercial vessels knew
where other vessels were fishing anyway, and they would prefer to work
cooperatively.
Acoustic Estimates and Variances
The discussion began with two questions: what level of accuracy is
needed for estimates and variances from Atlantic herring acoustic surveys,
and how will the estimates be used? Most participants are presently using
the estimates to tune virtual population analysis (VPA) estimates, while
Gary Melvin is presently the only participant who is using acoustic estimates
as a direct biomass estimate. Gary Melvin is dealing with relatively
low F values, while VPA seems to work better with high F values. The
goal of the NEFSC is to use the acoustic estimates as both an indirect
(i.e., tuning VPA) and direct (i.e., biomass estimate
to anchor population models) approach for assessment of Atlantic herring.
Point estimates with confidence intervals are required. Low variance
of about 25% is desirable. Geostatistics and bootstrapping appear to
be useful approaches for approximating variance. Discrepancies between
observed TS measurements and Foote's equation for the Northwest Atlantic
herring are a major concern that can significantly affect the biomass
estimates; therefore, TS needs further investigation. Design and timing
of the survey are another critical element. For example, annual acoustic
surveys for Atlantic herring in the Gulf of St. Lawrence began in 1985;
however, a meaningful time series did not begin until 1994. This is because
distributional shifts of Atlantic herring made for difficult timing and
placement of the surveys. Timing and placement of the Atlantic herring
acoustic survey on northern Georges Bank do not appear to be a problem
given the high biomass, however surveying the nearshore banks and coastal
waters of the Gulf of Maine is a problem which needs to be addressed
to attain the goal of annual biomass estimates.
RECOMMENDATIONS
FOR ONGOING RESEARCH AND THE NEXT WORKSHOP
Agenda suggestions for the next Northwest Atlantic Herring Acoustic
Workshop include: 1) sources of error in biomass estimates, 2) TS-length
relationship for Atlantic herring, 3) geostatistics, 4) survey design,
5) shallow-water acoustics, and 6) new acoustic technologies. It was
recommended that the next workshop be held at the St. Andrews Biological
Station in New Brunswick during spring or early summer 2002.
APPENDIX
A
Agenda
of the Third Northwest Atlantic Herring Acoustic Workshop,
University
of Maine Darling Marine Center, Walpole, Maine, March 13-14, 2001
March 13
09:00 General Scientific Goals of the Workshop
09:05 Automated Acoustic Data Logging aboard Commercial Vessels (P.
Yund)
09:35 Environmental Influences on the Density Distribution of Atlantic
Herring (P. Yund)
09:50 Fisheries-Independent Estimates from Acoustic Survey Data (W.
Overholtz)
10:15 Break
10:20 Verifying in-situ Acoustic Measurements for Atlantic Herring
(W. Michaels)
11:05 Ex-situ (laboratory) Experiments to Refine Acoustic Measurements
(M. Jech)
11:50 The Role of Fisheries Acoustics in Support of the Atlantic Herring
Industry (J. Kaelin)
12:10 Lunch
13:00 Cooperative Fisheries Acoustic Research in the Gulf of Maine
(D. Perkins)
13:20 Atlantic Herring Assessment and Management in the Gulf of Maine
(D. Libby)
13:30 Length Stratified Acoustic Estimates (C. Leblanc)
14:45 Break
15:00 A Method for Beam-Width Calibration (A. Clay)
15:50 Size Discrimination Using a Dual Frequency System (A. Clay)
16:20 Recent Multibeam Developments for Fisheries Acoustics (M. Jech
for G. Melvin)
March 14
08:00 Spatial Statistical Approaches for Fisheries Acoustics (P. Sullivan)
09:15 Spatial and Temporal Analysis of Fisheries Acoustic Data from
Commercial Vessel Operations (R. Claytor)
10:00 Survey Design for Atlantic Herring Acoustic Surveys in the Newfoundland
Region (J. Wheeler)
11:00 Break
11:10 Discussion on Acoustic Data Management
11:25 Closing Discussions of Workshop, Ongoing Research, and Next
Year's Agenda
12:00 Lunch
13:00 Closing Discussions on Estimates and Variance from Atlantic
Herring Acoustic Surveys
APPENDIX
B
List
of Participants for the Third Northwest Atlantic Herring Acoustic
Workshop,
University of Maine Darling Marine Center, Walpole, Maine, March 13-14, 2001
Peter Chase
Northeast Fisheries Science Center
National Marine Fisheries Service
166 Water Street
Woods Hole, MA 02543
Voice: 508-495-2348
Fax: 508-495-2258
Peter.Chase@noaa.gov
Matt Cieri
Maine Department of Marine Resources
West Boothbay Harbor, ME 04575
Voice: 207-633-9500
Matthew.Cieri@state.me.us
Allen Clay
Femto Electronics, Ltd.
P.O. Box 690
Sackville, Nova Scotia B4C-3J1
Voice: 902-865-8565
Fax: 902-865-8558
aclay@sprint.ca
Ross Claytor
Bedford Institute of Oceanography
P.O. Box 1006
Dartmouth, Nova Scotia B2Y-4A2
Voice: 902-426-4721
Fax: 902-426-1862
ClaytorR@mar.dfo-mpo.gc.ca
Kara Dwyer
Northeast Fisheries Science Center
National Marine Fisheries Service
166 Water Street
Woods Hole, MA 02543
Voice: 508-495-2274
Fax: 508-495-2258
Kara.Dwyer@noaa.gov
J. Michael Jech
Northeast Fisheries Science Center
National Marine Fisheries Service
166 Water Street
Woods Hole, MA 02543
Voice: 508-495-2353
Fax: 508-495-2258
Michael.Jech@noaa.gov
Jeff Kaelin
Stinson Seafood, Inc.
P.O. Box 440
Winterport, ME 04496-0440
Voice: 207-223-9013
Fax: 207-223-9900
msardine@mint.net
Kohl Kanwit
Maine Department of Marine Resources
West Boothbay Harbor, ME 04575
Voice: 207-633-9535
Kohl.Kanwit@state.me.us
Andone Lavery
Woods Hole Oceanographic Institution
Woods Hole, MA 02543
Voice: 508-548-1400
alavery@whoi.edu
Claude Leblanc
Gulf Fisheries Center
Department of Fisheries and Oceans
P.O. Box 5030
Moncton, New Brunswick E1C-9B6
Voice: 506-851-3870
Fax: 506-851-2620
LeblancCH@dfo-mpo.gc.ca
David Libby
Maine Department of Marine Resources
West Boothbay Harbor, ME 04575
Voice: 207-633-9532
David.A.Libby@state.me.us
|
Gary Melvin
St. Andrews Biological Station
Department of Fisheries and Oceans
St. Andrews, New Brunswick E5B2L9
Voice: 506-529-8854
Fax: 506-529-5862
MelvinG@mar.dfo-mpo.gc.ca
William Michaels
Northeast Fisheries Science Center
National Marine Fisheries Service
166 Water Street
Woods Hole, MA 02543
Voice: 508-495-2259
Fax: 508-495-2258
William.Michaels@noaa.gov
William Overholtz
Northeast Fisheries Science Center
National Marine Fisheries Service
166 Water Street
Woods Hole, MA 02543
Voice: 508-495-2256
Fax: 508-495-2393
William.Overholtz@noaa.gov
Donald Perkins
Gulf of Maine Aquarium Development Corporation
P.O. Box 7549
Portland, ME 04112
Voice: 207-871-7804
Fax: 207-772-6855
donnyp@maine.rr.com
Shale Rosen
Darling Marine Center
University of Maine
193 Clark's Cove Road
Walpole, ME 04573
207-563-3146 x 331
ShaleRosen@gma.org
Patrick J. Sullivan
Department of Natural Resources
214 Fernow Hall
Cornell University
Ithaca, NY 14853-3001
Voice: 607-255-8213
Fax: 607-255-8837
Pjs31@cornell.edu
Ed Tooley
Little Bay Lobster Co.
415 Turnpike Drive
Camden, ME 04843
Voice: 207-763-4470
Fax: 207-763-4176
herring@midcoast.com
John P. Wheeler
Northwest Atlantic Fisheries Center
Department of Fisheries and Oceans
P.O. Box 5667
St. John's, Newfoundland A1C-5X1
Voice: 709-772-2005
Fax: 709-772-4188
wheeler@athena.nwafc.nf.ca
Philip Yund
Biological Oceanography Program
Division of Ocean Sciences
National Science Foundation
4201 Wilson Blvd.
Arlington, VA 22230
Voice: 703-292-8582
Fax: 703-292-9085
pyund@nsf.gov |
Acronyms |
DFO |
= |
(Canada) Department of Fisheries and Oceans |
GMA |
= |
Gulf of Maine Aquarium |
HDPS |
= |
(Femto Electronics, Ltd.'s) Hydroacoustics Data Processing
System |
KRM |
= |
Kirchoff ray-mode model |
MDMR |
= |
Maine Department of Marine Resources |
NEFSC |
= |
(National Marine Fisheries Service's) Northeast Fisheries
Science Center |
TRAC |
= |
(United States - Canada) Transboundary Resources
Assessment Committee |
TS |
= |
target strength |
UM |
= |
University of Maine |
VPA |
= |
virtual population analysis |
