The CCAMLR krill synoptic survey

Report of Planning Meeting

British Antarctic Survey, 8-12 March 1999

1 A planning meeting for the CCAMLR-sponsored multi-national, multi-ship near-synoptic acoustic survey for krill biomass in Area 48 to be conducted in January 2000 (hereafter referred to as CCAMLR-2000) was convened by J. Watkins and held at the British Antarctic Survey during the week of 8 through 12 March 1999. Also in attendance were A. Brierley, I. Everson, C. Goss, A Murray, J. Priddle, K. Reid, P. Trathan, and P. Ward from the UK; T. Pauly from Australia; S. Kim from Korea, M. Naganobu and S. Kawaguchi from Japan; and D. Demer and R. Hewitt from the U.S.A.

2 S. Hedley, representing the International Whaling Commission (IWC) was also present and expressed her gratitude for the opportunity to explain the IWC’s broad objectives with regard to the study of cetaceans and their habitat and to present the IWC request to participate on CCAMLR-2000. She also expressed the hope for fruitful collaboration between IWC and CCAMLR scientists as well as a closer relationship between the two organizations.

Survey Design

3 The group reaffirmed that the principal participants conducting of the survey will be Japan, the U.K. and the U.S.A. The time period of the survey would be early January to mid-February with specific start and stop dates dictated by the necessities of national programs. Each country would contribute 30 days of ship time for the conduct of CCAMLR-2000. Specific ship schedules are listed under Itinerary.

4 S. Kim noted that the CCAMLR Sub-Group on International Coordination intends to encourage several countries who plan to have field programs in the vicinity of the South Shetland Islands during the austral summer of 1999/2000 to repeat the CCAMLR-2000 transects in this area. The close-spaced CCAMLR-2000 transects on the north side of the South Shetland Islands are likely to be surveyed four times (one by Korea in late-December, one by Japan in late-December, one by the CCAMLR-2000 survey vessel in late January/early February, and one by the U.S.A. in late February/early March).

5 It was understood that Brazil, Russia and the Ukraine are also interested in participating but that each of these countries is not in a position to make firm commitments at this time to CCAMLR-2000. It was further reported that the Ukraine will be conducting field work in the vicinity of the South Orkneys during the 1999/2000 austral summer and that their ship will be equipped with an echo sounder other than a Simrad EK500; that Russia may have a research vessel available during the survey period and that it will be equipped with a Simrad EK500 echo sounder; and that Brazil has a research vessel equipped with an EK500 but that the availability of this ship during the survey period is less certain. Accordingly, it was decided that the Ukrainians would be encouraged to conduct an acoustic survey with a calibrated system in the vicinity of the South Orkneys and that this information could be used to complement the planned survey coverage and as an aid in the interpretation of survey results. It was also decided that if Russia were able to participate that they would be encouraged to conduct a replicate of one of the three planned survey tracklines with a calibrated EK500 system.

6 It was recognized that the extent of sea ice may affect the degree to which the southward extent of planned transect lines may be conducted. It was agreed, therefore, to examine recent trends in the annual extent of sea ice and if a reasonable probability existed that the planned transect lines could not be completed, then the survey design would be adjusted so as to achieve a more efficient use of time.

7 During a discussion of the survey design it was noted that the proposed transects run along meridians were not parallel, converging as they approach the pole. Considerable discussion ensued weighing the advantages of design simplicity against the disadvantages of over-sampling the higher latitudes relative to the lower ones (transect spacing at the highest latitudes would be approximately 65% of the transect spacing at the lowest latitudes). Ultimately, it was decided to use transects that were parallel on the earth’s surface. In order to orient these transects as closely as possible along the prevailing topographic gradient the area was divided into two grids. The first grid includes subareas 48.2 and 48.3 and was aligned N-S along the 40° W meridian. The second was aligned along a bearing of 330° at 50° W in order to take account of the topography in subarea 48.1. These grids were used to describe the nominal survey design, which would yield the maximum survey coverage using the available ship time. A randomization scheme will now be applied to all possible parallel transects on this grid to achieve the final survey design. Every third transect will then be assigned to each ship and cruise tracks will be laid out. Each transect will have a unique number. In addition the nominal noon and midnight stations will be laid out for each transect and be assigned a unique number. A. Murray agreed to undertake these tasks with the understanding that his work is critical to the success of CCAMLR-2000 and should be thoroughly checked.

8 Weather contingencies were discussed and it was agreed that the following guidelines would be adopted by each cruise leader in the event that weather and/or equipment failure caused introduced delays such that the survey could not be completed within the allotted time. Noon and midnight stations will be placed along each transect (the actual time of the station will vary according to the net sampling rules laid out in paragraph 10 below and adjustment for local apparent time). The cruise leader on each survey vessel will check progress against the expected time at the station and make adjustments if necessary according to the following hierarchical scheme:

Lengthen daytime acoustic survey operations by beginning and ending acoustic transects at the local apparent time of civil twilight; else
Increase vessel speed without sacrificing quality of acoustic data (see Acoustic Sampling Protocol for guidance); else
Delete daytime net sampling and CTD cast.

In addition, the cruise leader will check progress against the expected time at the approximate mid-point of each major transect (7 for each ship) and make adjustments if necessary according to the following hierarchical scheme:

Break survey work on current line and redirect the survey vessel toward the beginning of the next transect; else
Break survey work on current line and redirect the survey vessel toward the most adjacent point on the next transect; else
Delete an entire transect according to a randomly determined transect ranking order (to be provided).

Primary Protocols

9 During a discussion of acoustic sampling protocols it was reaffirmed that acoustic data should be collected at all times. The costs of data storage were considered to be relatively cheap when compared to the cost of missing data collection that may prove useful for future analysis. This principle pertains to time periods while calibrating, to noisy conditions during rough seas, to station times, and to transits between sampling transects. The directive, in essence, is to turn on the echo sounder and record data from the time the ship leaves the pier to the time it returns again. It was also noted that additional specifications regarding the characterization of noise and operational guidelines as to its acceptable level should be developed; that guidelines should be developed for simultaneous use of echo sounders and ADCPs; that lists of instrument settings for calibration and underway data collection should be developed and distributed among survey participants; and that during calibration only TS gain and Sv gain be adjusted while keeping axis offset angles (in the case of split beam transducers) set to zero and the beam angles set to the manufacturer’s description, adjusted for sound speed, for the specific transducer. In light of the fact that the acoustic data are critical to the success of CCAMLR-2000, it was also emphasized that data should be recorded redundantly and equipment spares should be aboard each ship. These and other issues are to be addressed in a updated Acoustic Sampling Protocol.

10 During a discussion of net sampling protocols it was noted that some directed net sampling effort would be necessary to reduce the uncertainty associated with the delineation of krill in the acoustic data record. This sampling would be directed at a variety of “acoustic morphs”, some presumed to be krill and some presumed not to be krill, and, as such, would not be appropriate for the primary purpose of the net sampling as stated at the WG-EMM meeting in Cochin; that is, the description of krill population demography. Nonetheless, it was noted by the group that the primary purpose of CCAMLR-2000 is to provide an estimate of B₀ from an acoustic survey and some directed sampling is necessary to achieve this end. Discussion further ensued as to whether the net sampling effort should be increased by reducing the number and/or length of acoustic transects or whether the currently planned net sampling effort (one tow at midnight and another at midday) should be reallocated with some tows used for directed sampling and others as standard oblique tows at pre-determined locations. Again, the primary purpose of CCAMLR-2000 was invoked as a rationale for reallocating net sampling effort rather than reducing acoustic sampling effort. The following fishing strategy was adopted:

At local apparent midnight conduct a standard oblique tow in conjunction with a CTD cast.
From the time of local apparent sunrise to local apparent noon conduct a directed tow if an acoustic morph of interest was detected and a reasonable chance of sampling it existed.
If a directed tow was conducted between local apparent sunrise and three hours before local apparent noon, delay the CTD cast until local apparent noon.
If a directed tow was conducted after three hours before local apparent noon, conduct the CTD cast at the same locale.
If no suitable acoustic morphs were detected by local apparent noon, conduct a standard oblique tow in conjunction with a CTD cast.

Additional issues were raised during discussion of the net sampling protocol including the desire to standardize nets among all participants, the treatment of “other zooplankton”, and the use of additional nets for sampling smaller zooplankton. These issues were addressed in the revised Net Sampling Protocols. It was noted that Japan does not currently have access to an RMT-8 net and that this may be addressed by inviting the participation of an outside expert, together with an RMT-8 net, on the Japanese survey vessel.

11 During a discussion of the CTD protocols, it was noted that both the general flow pattern across the Scotia Sea as well as the position of fronts were important determinates of the dispersion of krill and that describing these should be the objectives of the oceanographic sampling protocol. It was further noted that CTD sampling to the depth of a particular oceanographic feature (eg. a vertical boundary of the circumpolar deep water) may be more rational than sampling to an arbitrary depth of 1000m. This could be considered a plane of no motion for geostrophic calculations as it relates to the structuring of krill habitat. Discussion ensued as to whether this would add to the time required to conduct the CTD casts, but without a detailed analysis of climatic atlases this question could not be readily answered. It was also noted that the U.K. intends to collect ADCP measurements to approximately 400m depth with hull-mounted transducers and that Japan intends to collect LADCP measurements over the full extent of the CTD cast; only the U.S. does not expect to make ADCP measurements. These measurements of absoulute currents may be used to interpret CTD data. It was therefore decided that the current protocol (CTD casts to 1000m depth (or to the bottom if shallower) should stand pending investigation of the climatological depth of UCDW. It was further noted that the position of fronts along the transects could be more accurately described with the use of towed and/or expendable sensors. Issues relating to CTD measurements are addressed in the revised CTD Protocols.

Secondary Sampling

12 S. Hedley presented an overview of the IWC’s objectives, sampling methods, and personnel requirements for their participation on CCAMLR-2000. Discussion ensued as the value of relative versus absolute estimates of cetacean abundance. The IWC’s short-term objective with regard to CCAMLR 2000 is to relate the spatial distribution of baleen whales to krill and other environmental covariates; in this regard relative abundance may be adequate. There is some debate, however, within the IWC scientific community on this point. On the other hand, the IWC’s long range objective is to evaluate the impact of baleen whale consumption on the krill resource (presumably this is of direct interest to CCAMLR as well); in this regard absolute estimates of whale abundance are more appropriate. For CCAMLR-2000 the IWC would like to achieve 100% coverage of all transects using the double platform method which will generate absolute estimates of whale abundance. This would require two 4-observer teams (8 berths) on each survey vessel. Tradeoffs between transect coverage and the proportion of the survey that could be conducted using the double platform method were described if fewer berths were available. It was noted that firm commitments needed to be passed by the CCAMLR-2000 coordinator to Greg Donovan at IWC in time for the annual meeting in May. The final protocols for pelagic krill predator observations may differ between ships and will be determined in consultation with the IWC and published on the web site.

13 The extraordinary opportunity offered by CCAMLR-2000 to sample zooplankton across the Scotia Sea was discussed. Sampling may be accomplished without jeopardizing the primary sampling operations by adding a set of 1 m² 333 micron mesh nets to the RMT-8 sampler which will be used to sample krill and other micro-nekton. Specific protocols were not developed except to note the value to all participants of a common zooplankton database, which could be accessed via the CCAMLR-2000 web site. Phytoplankton sampling was also discussed and it was determined that all three survey vessels will have fluorometers mounted on their flow-through systems as well as CTD instrument packages and will also make chlorophyll extractions from water samples. Additional measurements will vary among vessels and it was agreed that observation protocols would be posted on the web site as they are developed. It was further noted that water samples could be preserved in a Lugol’s solution for subsequent analysis, but that the shelf life of such samples was only two years.

14 Underway observation systems will be maintained by all three survey vessels. Measurements will include wind speed and direction, air pressure, humidity, photosynthetically available radiation, sea surface temperature, salinity, turbidity and fluoresence. In addition, Japan will collect continuous measurements of particle volumes (as a proxy for zooplankton) and dissolved oxygen. The U.K. will also tow an undulating oceanographic recorder which will contain an optical plankton recorder and make additional measurements of photosynthetically available radiance, fluoresence, turbidity, salinity and temperature. It was noted that it would be advisable to standardize averaging intervals among the three vessels. Japan has an ADCP that could be operated continuously, however, current plans call for it to be secured during underway operations and an LADCP used instead in conjunction with CTD casts. The U.K. will operate an underway ADCP but the U.S.A. has no current plans to do so.

15 The potential value of satellite imagery was discussed and was agreed that J. Watkins would investigate the various products that would be appropriate complements to CCAMLR-2000. In this regard it may be necessary to request SeaWiFS ground stations at Palmer and/or Rothera to archive specific imagery.

Data Collection and Archiving

16 With regard to the conduct of acoustic transects it was agreed that after the completion of station observations each survey vessel would relocate to the closest point along the intended transect before proceeding to the next station.

17 The utility of maintaining an underway log was emphasized. Such a log would contain notes regarding the start and stop times of acoustic transects, comments on weather conditions and sea state as they affect the acoustic records, unusual features noted in the acoustic data, and any other details that may be of use when interpreting the acoustic data after the survey is finished. Similar notes obtained in conjunction with net sampling and CTD operations would clearly be useful as well. Discussion ensued regarding routine logs and the various numbering systems for activities and stations employed by different national programs. It was agreed that, as a minimum, the start and end times and positions of all operations should be recorded in an electronic format such that a list could be made up and queried for all activities conducted at a given set of stations or, alternatively, all locations where a given set of activities was conducted.

18 With regard to computer problems associated with the year 2000 (Y2K), all vessel coordinators were encouraged to check the GPS receivers on their ships that will be in use during CCAMLR-2000. These receivers will be the primary source for time stamping the acoustic date set and must be compatible with the millenium change. It is highly desirable that all data collections on a survey vessel are referenced to the same time standard and thus redundant Y2K compliant GPS receivers are critical. Vessel coordinators were also encouraged to check with Simrad and SonarData for assurances that their equipment and software has been tested for Y2K compliance.

Data Analysis

19 The group reaffirmed and strongly endorsed the decisions made at the last planning session for CCAMLR-2000 held during the WG-EMM meeting in 1998 at Cochin, India, that a) the collection of acoustic data, micro-nekton samples from RMT-8 nets, and CTD profiles would form the core data sets, and that b) the analyses and interpretation of these core data sets, and the reporting of results, would be conducted in a collaborative fashion. The core data sets refer to those collected according to the survey design described in Paragraph 7.

20 It was agreed that because an estimate of B0 derived from the acoustic data is expected to be tabled at the meeting of WG-EMM in July 2000, a data workshop should be held sometime during May-June 2000. It was tentatively agreed that the workshop would be held over a one to two week period in La Jolla where computational facilities and other logistic support are readily available. It was emphasized that results from the directed net sampling for identification of acoustic morphs, the oblique net sampling for the determination of krill demographic structure and the oceanographic sampling would be of value in interpreting the acoustic records. As such, it will be highly desirable to include these elements in the workshop. It was also recognized that summary statistics from the regional surveys conducted at South Georgia and the South Shetland Islands would be valuable in the interpretation of the results.

21 With regard to analysis of the acoustic data it was noted the two of the most important tasks will be target strength estimation and the apportionment of backscattered energy to krill and other scatterers of lesser interest. It is anticipated that several methods of accomplishing these tasks will be applied to the data set and results compared as part of the final report to WG-EMM. Accordingly, it was suggested that working papers on various techniques to estimate TS and to delineate taxa in the acoustic data set be invited for the upcoming meeting of WG-EMM in July 1999, that time be requested at that meeting for survey participants to discuss these methods and to select the most promising methods, and that individual scientists be commissioned to develop the computer code required to implement the selected methods on a production basis. This code could then be brought to the workshop and applied to the data sets at hand with the potential of saving a substantial amount of time that would otherwise be spent at the workshop accomplishing these preliminary tasks. The group agreed to this idea in principle.

22 It was also emphasized that the May/June 2000 workshop would be only the first of many workshops and collaborative analyses that may be expected in the aftermath of CCAMLR-2000.

Other Issues

23 J. Watkins reported to the group that D. Miller from South Africa and V. Siegel from Germany had expressed interest in participating in CCAMLR-2000. It was agreed that their participation was very desirable and that recommendations as to how to best deploy additional experts should be made after all expressions of interest are received and a better idea of how they could contribute is in hand. Exchange of personnel between ships was also discussed and it was decided that such agreements would be first explored between national programs and ultimately coordinated by the CCAMLR-2000 coordinator (J. Watkins). M. Naganobu noted that Japan would not be able to send any personnel to other ships but would welcome experts in the field of acoustics and net sampling, particularly if the latter were able to bring an RMT-8 net.

24 The importance of maintaining liaison with other institutions and groups conducting field programs in this sector of the Southern Ocean was recognized. In particular, it was recognized that substantial benefits could accrue by making these groups aware of the developing plans for CCAMLR-2000 and inviting their comments and suggestions. It was recognized that there could be several collaborative opportunities of which we are currently unaware and that individual scientists should actively engage colleagues outside of the CCAMLR community regarding planned operations and the existence of the web site. In particular, the group agreed that the coordinators for SO-GLOBEC (S. Kim and E. Hoffman) should be contacted.

25 M. Naganobu presented a plan to conduct a series of deep CTD casts across the Drake Passage (following the standard WOCE transect) with the intention of describing the flow field. The group recognized the potential value of such a set of observations and enthusiastically endorsed the plan.


Introduction	Itinerary	Station positions	Cruise tracks	Planning Meeting	Sampling Protocols	Participants	Background papers	Contents

Page last updated on 16 March 1999