(from "Report of the Scientific Committee", 1997 meeting)

12.1 Advice on the effect on stock(s) of scientific permit catches
Last year, the Committee had had extensive discussions of this Item (IWC, 1997e, p.93). It had agreed that the matter should be considered further this year and had encouraged members to submit documentation presenting further proposals. In the absence of such proposals the Committee did not consider this Item but agreed to keep it on the Agenda for next year's meeting, noting that it will only be considered if documentation is available.

12.2 Review of results from existing permits
The Working Group met at the Mariners' Court Hotel, Tokyo, on 12-16 May 1997. The review meeting was convened by Schweder. Reilly was elected chairman.

Bannister participated in the opening, where he outlined some of the considerations that had led to the structure of the draft agenda, in particular the need to reflect in the discussions two separate components arising from the review meeting's terms of reference:

the specific research carried out, and its results;
the contribution made by those results to the stated objectives of the research programme, and to the aims of the IWC as expressed in its Resolutions.

The review meeting completed all but the final task, i.e. judging the merits of the results in terms of the Commission's resolutions. That more general task was forwarded to the full Committee.

Five components of JARPA were reviewed: sighting surveys and abundance estimation; stock structure; biological parameter studies; marine ecosystem studies; and those addressing environmental change. For each component the Group considered the following: its background, including original and additional research objectives; methodology of data collection; data analysis; results; and potential of results in the context of the objectives of JARPA and of stock management.

Ohsumi reported that Japan's original objectives for the research had been:

estimation of biological parameters to improve the stock management of the Southern Hemisphere minke whale;
elucidation of the role of whales in the Antarctic marine ecosystem.
Subsequently, as part of the natural evolution of the programme and in response to developing requirements, two further objectives had been added:

elucidation of the effect of environmental changes on cetaceans;
elucidation of the stock structure of the Southern Hemisphere minke whales to improve stock management.

Fujise presented a brief summary. Two feasibility studies had taken place in 1987/88 and 1988/89, with the full scale 16-year research commencing the following season and alternating each season in Areas IV and V. In 1995/96 and 1996/97, coverage was extended to Areas III E and VI W respectively, for a limited period feasibility study of stock structure. He stated that the full programme has two components: a sighting survey whose primary purpose is the estimation of trends in abundance; and a sampling component to allow biological parameter values to be estimated given also the abundance information provided by the sighting survey.

In the programme as originally proposed, it was planned to take 825 animals in any one season from either Area IV or Area V. For two years of feasibility studies, 300 animals (with an allowance of ± 10%) were planned to be sampled in parts of Areas IV and V. Following the results of the feasibility studies, considerations of the balance between the expected precision of estimates of the mortality rate and the available research capability led to the decision to set the number of animals to be sampled each season at 300 with an allowance of ± 10%. In 1995/96, additional samples of 100 animals with an allowance of ± 10% were planned for Area III E, and subsequently 100 ± 10% in Area VI W in 1996/97, for studying stock structure (see Item 3, SC/49/Rep 1).

Initially, three vessels (plus the mother ship) had been employed, but a fourth had been added in 1995/96 to allow for an increase in searching effort in the sighting survey. This and some other changes during the progress of the programme had been made in response to comments from the Committee. Initially, both dwarf and ordinary forms of the minke whales had been sampled, but sampling of the former had ceased in 1993/94. A total of 1,546 (Area IV), 1,546 (Area V), 110 (Area III E) and 110 (Area VI W) ordinary and 16 dwarf form minke whales had been sampled by the end of the 1996/97 season.

The review meeting noted that while both the quantity and quality of the scientific work had been commended by the Committee, differing views had been expressed about its relevance to management considerations. A list of specific issues in contention was compiled from past Committee reports and is given in annex F of SC/49/Rep 1.

Noting that the objectives for the research did not include the provision of abundance estimates (for input, perhaps, to the RMP) per se, clarification was sought on the intent of this component of the programme. Ohsumi stated that though the primary reason for the sighting surveys was their contribution to Objective 1 (the estimation of biological parameters), their pertinence to the RMP and the associated implementation process for Southern Hemisphere minke whales should be seen as a derivative objective.

The closing procedures (e.g. to determine school size) were identical to those used in closing mode for the IDCR surveys, excepting for the additional time spent on sampling, and closing being limited to minke whales. The sighting and sampling vessels engaged in such sampling ('SSVs') surveyed along parallel tracklines, but from the 1991/92 season one vessel at any one time was devoted to sighting only, in order to investigate the effect of sampling activity on abundance estimates.

Certain changes in JARPA survey procedures over time were noted but the review meeting considered that adequate comparability over time in data collection had been achieved.

One important difference in survey methodology from that used on the IDCR cruises was noted. On IDCR cruises, survey on one day commences from the position reached at the end of the previous day. However, for the JARPA programme, a target distance per day was established; if this distance has not been achieved by the end of the day, the remainder of the planned trackline for that day is not covered, and survey the next day starts from the 'targeted' (not the actual) position for the end of the previous day. This procedure can lead to under-surveying of higher density areas, because more time is required for closing on and sampling whales, so that less survey distance is covered on the day in question. This has implications for abundance estimates that are discussed immediately below, and for sampling representativeness.

Two papers provided analyses of the sighting data to estimate abundance. The key differences between the methods of analysis in SC/M97/1 and SC/M97/23 are that the former stratifies by school size in estimating effective search half-width (w), whereas the latter estimates w for all schools combined and then multiplies an estimate of the total number of schools by another of mean school size. Furthermore, SC/M97/23 stratifies data to a lesser extent, based on the AIC statistic, to avoid the high variances associated with models that over-fit data. Both analyses assumed g(0)=1.

In addition, SC/M97/23 attempted an approach to adjust for the under-surveying of higher density areas in the JARPA surveys arising from the difference in survey methodology described in the final paragraph of the preceding section. This involved scaling the number of sightings made each day upward by the ratio of the target distance for that day to the searching distance actually achieved. Concern had been expressed previously in the Committee that estimates of minke whale abundance from JARPA reported at that time had been considerably lower than IDCR estimates. One of the motivations for this part of the analysis given in SC/M97/23 was to investigate whether that factor was a primary cause of the difference.

The authors of SC/M97/23 suggested that:

their unadjusted estimates are in broad agreement with those of SC/M97/1; and
their adjusted estimates (with one exception) are not systematically lower than comparable IDCR closing mode abundance estimates, although confidence in this conclusion would be enhanced if analysis of the JARPA surveys in seasons other than the two analysed had yielded similar results.

In respect of (1), the review meeting felt that no strong inferences could be drawn given the limited number of comparisons.

In discussions arising from (2), a number of concerns were raised as to whether the 'adjusted' approach applied to the JARPA analysis is an appropriate means of correcting for the higher-density-under-surveying effect to yield comparable estimates of absolute (or even relative) abundance. These concerns arose primarily from considerations of the mechanisms that lead to clustering of whales, and hence areas of higher and lower density. The review meeting considered that further research to develop an approach to correct for the effects of clustering was required. A specific suggestion was that this be based on modelling the extent of the clustering each year by a Neyman-Scott process.

In relation to annex D of SC/49/Rep 1, the review meeting agreed that this topic was pertinent to items B 2.1, 2.4, 3.1 and 3.2. Relevance to item 2.4 related to the agreed research requirement of consolidating the Southern Hemisphere minke whale abundance estimates provided by different programmes for different areas of coverage and using different methodologies (e.g. IDCR, JARPA, JSV).

The fact that JARPA provides more frequent repetitions of surveys of the same localities than the IDCR programme would facilitate estimation of the extent of inter-year variability in local abundance, which would in turn lead to improved results from the consolidation exercise.

The review meeting agreed that more research was required to develop a reliable method for adjusting for the higher-density-under-surveying feature of the JARPA survey design. Once this had been achieved, the resultant abundance estimates should be useful both as absolute and relative indices.

One of the original objectives of JARPA was to elucidate the stock structure of minke whales to improve stock management. The stock identification objectives were being addressed by both genetic and non-genetic techniques.

Information on stock structure presented at the 1990 Comprehensive Assessment was based on biological samples collected during commercial whaling operations. Because of the nature of this fishery, these samples tended to be from animals distributed along the pack-ice edge and were not evenly distributed within Areas IV and V. In contrast, during JARPA 2,887 minke whales were systematically collected from between 1987/88 and 1995/96, where survey effort was evenly distributed within the study area. A variety of tissue samples and body measurements was taken from each whale.

A simulation exercise had been conducted which indicated that a sample of at least 150-200 individuals using mtDNA was needed to detect significant differences between putative stocks (SC/M97/3).

Genetic studies, based on mitochondrial DNA (mtDNA) were conducted to investigate the phylogenetic relationships of ordinary and dwarf forms of minke whale and stock structure in the ordinary form. Two methods (using mtDNA) indicated a clear genetic differentiation between dwarf and ordinary forms. The sequencing analysis also indicated that the southern dwarf, southern ordinary, North Pacific and North Atlantic forms represent independent genetic populations. Further, based on genetic data, the dwarf form was found to be more closely related to minke whales in the North Atlantic and North Pacific than to the ordinary form.

Four studies of stock identity of the ordinary form were reported. In the first study, stock structure was evaluated using chi-square statistics for heterogeneity on the observed haplotype frequencies. The resulting chi-square was significant and supported the alternate hypothesis that sub-structure was evident in Areas IV and V (i.e., the western and eastern strata probably contained minke whales that were genetically distinct, but mixing occurred in the central stratum).

A second and third study considered both geographical and temporal factors. The results of both were consistent with the hypothesis that different stocks occurred in Areas IV and V, where the composition of animals from the two putative stocks changed both longitudinally and temporally during the feeding season.

A fourth study reported the results of an analysis of molecular variance (AMOVA). A priori strata were established using four longitudinal sectors and two time periods. Of 137 haplotypes identified, none were unique to any single geographic stratum. The AMOVA test results were significant between Area IVW-early and all other strata. Excluding stratum IVW-early, none of the other contrasts between strata were statistically significant.

It was reported that a comparative study of both dwarf and ordinary forms, involving body colouration, and morphometric and skeletal measurements, is underway.

With regard to the study on stock identity in the ordinary form, a single morphometric analysis grouped 326 animals into three strata (Area IVW-early, Area IVW-late, and Area IVE-early). A multivariate analysis revealed that the three strata were not separated exactly; nonetheless, whales from Area IVW-early stratum were found to be significantly different from whales found in the other strata.

The authors noted that they were not proposing new stock boundaries at this time, but that their results were consistent with the hypothesis that minke whales collected in the western part of Area IV, early in the feeding season form a stock distinct from whales taken in strata further east or later in the feeding season.

There was discussion among review meeting participants as to what level of genetic distinctness was significant at the stock and species level. It was noted that general standards for species-level differences have been established based on a comparative approach (i.e., the genetic distance between 'good' species in a taxon is typically used as the standard for putative species). Regarding stock structure, the answer is less clear.

Taylor (annex G of SC/49/Rep 1) commented that the number of individuals exchanged between populations exists on a continuum in nature. In statistical terms, the amount of difference required before a particular grouping of individuals can be designated as one stock or two is referred to as effect size. It is important that effect size is explicitly stated prior to undertaking research on stock structure, as it is not possible to determine the sample size necessary to reliably detect a specific effect size without specifying it. Taylor recommended that the Committee should develop criteria regarding the effect size required to designate two putative stocks as separate management units.

Several members noted that for the Implementation Simulation Trials for the RMP the key issue regarding stock identity was the number of breeding groups and not the number of, or distribution on, feeding areas. However, the distribution and number of breeding groups for the ordinary form of minke whale is poorly understood.

It was noted that only information on mtDNA had been used so far to investigate stock structure of the ordinary form, but that efforts were underway to use existing tissue samples to look for stock structure using nuclear markers. It was also recognised that the statistical analysis of the genetic data should consider the inclusion of school size as a covariate because: (1) schools of different sizes are not detected with equal probability; and (2) of the schools encountered, animals from smaller schools are oversampled relative to animals from larger schools.

Participants agreed that the following points listed in annex D of SC/49/Rep 1 were pertinent to items A (Objectives 1 and 4) and B 1, 2.1, 2.4, 3.2 and 3.3. It was noted that research on stock structure is clearly related to Objective 4 of annex D, but is also important regarding the manner in which specific biological parameters are both estimated and interpreted.

There was general agreement in the review meeting that the data presented on stock structure, particularly the new genetic data, were important contributions to the objectives of JARPA and stock management. It was further noted that based on the new genetic information, at least some of the historic management Areas were inappropriate for stock definitions for Southern Hemisphere minke whales.

Two additional lines of research were also recommended. First, development of more theory on the use of genetic information for estimating mixing rates between putative stocks. Second, participants supported an earlier recommendation of the Committee that efforts should be undertaken to collect tissue samples from minke whales on Southern Hemisphere breeding grounds to allow contrasts of distribution and frequency of specific haplotypes from the breeding and feeding grounds.

Regarding the RMP, the new genetic information indicated that there was a temporal component to the stock structure of the ordinary form of the Antarctic minke whale in Area IV, which had not been recognised at the time of the Comprehensive Assessment. In the long term, genetic information could be used in implementing an improved version of the RMP.

It was also agreed that a protocol should be developed that specifies how such data (i.e., on genetic relatedness of putative stocks from the breeding or feeding areas) would be used in either developing specific Implementation Simulation Trials or the general management of commercial harvests.

In addition, it was noted that the information discussed during this section of the review meeting did not exclude the possibility that there are more than two genetically distinct stocks of the ordinary form of minke whale in Areas IV and V. Additional analyses, including for example the use of nuclear DNA, could reveal additional stock structure.

To avoid a repetition of past debates within the Committee regarding methods alternative to lethal removals, the proponents of the two different viewpoints summarised their views. These summaries are presented in SC/49/Rep 1, annex H .

Estimation of biological parameters, especially natural mortality rates, was originally the main research objective of the JARPA, since knowledge of some of these parameters was at the time considered to be necessary for a rational management of whale stocks by many scientists.

From 1987/88 to 1991/92, sampling involved taking up to two whales from the targeted school, but from 1992/93 the protocol was to take one whale from each school. Sampling success from the targeted school was 0.52 - 0.75 for the scheme of taking up to two whales from a school, but increased to 0.8 or more for one whale per school. An analysis concluded that no substantial differences in the samples taken from schools of size 2 and above had resulted from the change in sampling scheme.

It was pointed out that the sampling method resulted in over-sampling from small schools, and also that another source of sampling bias results from the aspects of the survey and sampling methodology. Although the protocol induces under-surveying in areas of high whale density, it will result in over-sampling where density is high.

Three papers discussed catch-at-age data.

SC/M97/6 presented an extension of methods previously presented to the Committee (Butterworth and Punt, 1996; Butterworth et al., 1996) for joint analysis of catch-at-age and abundance data. The extension involved taking account of assumed separability of the fishing mortality matrix for the ages of 16 and above in the commercial catch. The method was applied to catch-at-age (both commercial and JARPA) and sightings survey (both IDCR and JARPA) data for both Areas IV and V. The Area IV analysis provided an estimate for natural mortality (M) of 0.057 yr-1, and of the trend in recruitment over the 1947-68 period showing an increase of 5.5% yr-1. A number of possible reasons were advanced for the marked drop in recruitment from 1970 to the mid-1980s evident in the results.

SC/M97/11 presented an interim result for the estimation of the natural mortality rate by the method originally proposed by Tanaka (1990) using the JARPA age data. The paper used the estimated age composition of these data taking account of whale abundance, selectivity and ageing error. Resultant estimates of average natural mortality rate ranged from 0.0165 (SE=0.13) to 0.167 (SE=0.116).

SC/M97/21 analysed the age data obtained from JARPA expeditions 1987/88 through 1995/96 in two stages. First, the analyses found that the proportion of young (under 10 yr) animals was related to latitude and school size, and further that this dependence varied from year to year. An analysis was then conducted of the matrix of samples by age and year for age groups 10-30. The authors concluded that it is not possible to estimate recruitment and mortality separately, because each term in the model involves pairs of aliassed parameters: mortality is aliassed with the rate of change (first derivative) of recruitment, the first derivative of mortality is aliassed with the second derivative of recruitment, etc. Nevertheless linear combinations can be estimated and some example results were shown.

There was considerable discussion of the implications of the results of papers SC/M97/6 and SC/M97/21. Two opposing views crystallised. The first held that the VPA-type analyses of SC/M97/6 did provide valid results for an extensive list of reasons given in SC/49/Rep 1, item 4.3.1. The second held that results of the catch-at-age analyses of both methods were compromised to some extent by the aliassing of estimates of mortality and recruitment parameters. In spite of this difference of opinion, there was consensus on some aspects of the results, as noted below.

The review meeting felt there was merit in pursuing the approaches of SC/M97/6 and SC/M97/21 further, but that estimates from the application of such methods could not be considered reliable until difficulties associated with the estimates of abundance from JARPA (see SC/49/Rep 1, item 2.4) had been resolved.

Two papers considered transition phase data, SC/M97/7 and SC/M97/22. They had used different, but overlapping, datasets. Both analyses were restricted to Area IV, but that in SC/M97/7 was based on data from commercial whaling and from the first two JARPA expeditions to Area IV. The analysis in SC/M97/22 used only JARPA data, and from all five expeditions to Area IV. Both analyses showed decline in age at transition phase, but, for the same period of cohorts, 1950-70, the analysis of SC/M97/7 indicated a decline in age at transition phase roughly double that estimated in SC/M97/22.

The review meeting agreed that it was important to resolve the differences between the two transition phase studies, but that the transition phase observations could not be explained as an age related effect alone. A number of participants considered that the overall evidence was sufficient to conclude that there had been a real decline in age at maturity over the cohorts studied.

Results were also presented regarding sexual maturity, apparent pregnancy rate, length and age at sexual maturity, annual ovulation rate and growth.

With respect to the JARPA objectives, the review meeting agreed that the papers presented gave valuable information on recruitment, natural mortality, decline in age at sexual maturity and reproductive parameters of minke whales in Areas IV and V. However, there are some unresolved problems in the analyses, and further work is necessary.

One of the specific objectives of the programme was to collect random samples for the estimation of biological parameters. The results had demonstrated that this was a more difficult task than had been envisaged. Despite the considerable attention given to the sampling scheme, it has not been completely successful at obtaining random samples. The review meeting further noted that the geographical delimitation of the sampling areas has not resulted in either distinct biological populations being sampled or the entire ranges of the population being sampled. The implications of this for the representativeness of the sample should be given further consideration. However, it was noted that the VPA analyses (as, for example, SC/M97/6) required representative sampling over only part of the range of ages in the population. There were no indications that this had not been achieved for animals of age 10 and above. The review meeting also considered that the results of the genetic studies should be used to redefine the geographical boundaries for any future analysis. It noted that there was still uncertainty as to whether information that fully represents a biological stock could be obtained, but considered that much progress had been made towards that end. Before JARPA was initiated, whales occurring in Area IV and Area V were managed as different stocks, but a clearer picture about the biological stocks in these Areas was now emerging. Although the present state of knowledge still leaves much to be desired, considerable data reflecting the status of the whale stocks occurring in Areas IV and V have been collected, and have produced many valuable results.

The review meeting noted that there were non-lethal methods available that could provide information about population age structure (e.g. natural marking) but that logistics and the abundance of minke populations in Areas IV and V probably precluded their successful application.

With respect to the relevance of the work for stock management, in 1993, the Committee proposed mechanisms for amendment of the RMP (IWC, 1994a). It distinguished between mechanisms for the amendment of case-specific implementations and mechanisms for amendment of the RMP itself. The review meeting discussed the relevance of a better knowledge of biological parameters to management objectives in this context, i.e. it distinguished between short-term improvements, which would be amendments of the case specific implementations, and long-term improvements which could imply more fundamental changes to the RMP itself.

In the short-term perspective the three key considerations identified were: (1) changes in the definition of Small Areas; (2) changes to the selection between RMP options such as catch-capping and catch-cascading, and (3) changes in the range of plausible MSYRs to use in Implementation Simulation Trials.

In the longer time perspective better knowledge of biological parameters could lead to modifications of the CLA in the RMP.

The review meeting noted that the recruitment data from analyses such as those in SC/M97/6 could be fitted by stock recruitment models to provide estimates of MSYR once reliable input data are available. Trends in recruitment from SC/M97/6 and SC/M97/21 could be used directly when conditioning future Implementation Simulation Trials for Southern Hemisphere minke whales, and if the caveats expressed concerning abundance estimates from JARPA could be resolved, together with some further methodological development in estimating the essential biological parameters, the results from the JARPA could be directly relevant for management, both in the short term and the long term.

The second of the two original JARPA objectives was: 'Elucidation of the role of whales in the Antarctic marine ecosystem'.

In the 1996/97 research plan this objective was restated as: 'Elucidation of the role of whales in the Antarctic marine ecosystem through whale feeding ecology'. The research plan concentrates on the feeding ecology of minke whales by the analysis of stomach contents and blubber volume.

Data on the feeding ecology of minke whales is obtained from the weight and species composition of stomach contents of the sampled whales. The whole body mass of each whale is measured using a weighing platform.

An index of body fat condition has been calculated as the ratio of mean girth to body length. Data analyses of feeding rates (SC/M97/17) used three different methods.

Of the six years included in the study of body fat condition, stomach contents and distribution, two were categorised as years of poor body fat condition and three as years of good condition. Estimated body weight gain during the feeding season in poor years was estimated to be 25% lower than in good years. In Area IV and the northern part of Area V, krill (Euphausia superba) was the dominant prey species, but in the southern part of the Ross Sea (in Area V), Euphausia crystallorophias was the dominant prey species. Distribution of minke whales showed greater interannual variability in Area V than in Area IV, reflecting a greater degree of variability in sea ice extent in Area V. In Area V, in years of high sea ice extent, the krill-rich slope region of the western part of the area is covered by ice. This leads to poor food availability and results in a very low density of minke whales along the ice edge. The Ross Sea zone was an area of low food availability throughout the study period. Paradoxically, this zone always contained numerous whales, especially pregnant females.

Daily food consumption estimates ranged from 3 to 4% of body mass. The annual consumption estimates of prey for Area IV ranged from 1.42 to 1.78 million tonnes. For Area V, the range was 5.98 to 7.49 million tonnes. The value for Area IV is roughly 25% of the total estimate of krill biomass in the area. Consumption of krill by minke whales in Area V was an order of magnitude greater than that estimated for Adelie penguins and crabeater seals.

No difference was found in the fatness index between the sexes. From seasonal changes in the index and from its distribution by foetus size it was suggested that some whales over-winter in Antarctica and that others arrive on the feeding grounds late in the season. Analyses from earlier commercial catches and from the JARPA samples show a gradual decline in blubber thickness since 1978/79.

The review meeting agreed that the following points listed in SC/49/Rep 1, annex D were pertinent to items A2, B1, B2.3, B2.5, B3.4.

The review meeting noted the striking similarity in the results obtained from the three methods for estimating daily food consumption reported in SC/M97/17. It was agreed that these estimates could be used with confidence for the estimation of total food consumption by Antarctic minke whales.

The review meeting considered that the body condition index presented in SC/M97/18 required further refinement. It was not convinced that the analyses could be used to infer that some whales over-winter in high latitudes and that others arrive late in the feeding season.

The review meeting agreed that the studies being undertaken were contributing to the objective of the 'elucidation of the role of whales in the marine ecosystem through whale feeding ecology'. However, it was suggested that elucidating the role of whales in the marine ecosystem also requires concurrent studies on the distribution and abundance of prey species, and process-oriented studies integrating information from physical and biological oceanography with zooplankton and predator studies would be useful. Such studies should be conducted on a smaller scale, possibly using radio tagging, than that covered by JARPA, perhaps of the order of ten to one hundred kilometres. Such studies should be set up to examine specific hypotheses about ecological interactions. The review meeting agreed that the JARPA studies provided useful information for both the formulation of such hypotheses and for the selection of study areas. The marginal sea ice zone is an obvious candidate for process oriented studies. The review meeting noted that such studies would be of interest to CCAMLR and Southern Ocean GLOBEC.

The review meeting noted the reported decrease in blubber thickness since the late 1970s. Such information could contribute to the specification of a range of krill-surplus hypotheses for use in further implementation trials for the RMP.

The 1995/96 research plan added the following objective: Elucidation of the effect of environmental changes on cetaceans.

In addition to data collected for studies on environmental change, data have been collected on marine debris and body burdens of pollutants, including organochlorines and heavy metals. Tissue samples, including liver, muscle, kidney and blubber have been collected for these analyses. Marine debris observations have been based on visual observations, but recently nets have been introduced to estimate prevalence of smaller items. Air and sea water samples have been collected to monitor environmental pollutant levels.

The atmospheric and sea-water concentrations of organochlorines such as PCBs in the Southern Hemisphere were lower than in the Northern Hemisphere, except for HCB. Levels of DDT showed no yearly variation, but an increasing trend in PCB levels was detected in the period 1984 - 1993. This implies the continuing discharge of PCBs into the Southern Hemisphere.

Analyses of hepatic mercury concentrations were grouped by sex, geographical position and by time within season. Hepatic mercury concentration in the younger animals seems to have decreased in the last decade. This suggests that the increased mercury intake had begun to decrease in that decade.

It was noted that organochlorine concentrations in blubber are strongly influenced by seasonal variation. Analyses should therefore include a correction for the effects of increasing blubber thickness during the season.

The review meeting referred to the recommendations of the Bergen Workshop on Chemical Pollution and Cetaceans (IWC, 1998). It considered that the pollutant studies under JARPA were pertinent to these recommendations.

One of currently contentious issue in Antarctic research is the relative weight to give to the competing hypotheses that changes in abundance of Antarctic predators are due to either 'krill surplus' or the effects of environmental change. The review meeting recognised that distinguishing between these hypotheses will be difficult. In the meantime, the observations on changes in blubber thickness and variations in recruitment should be used to formulate specific hypotheses on the possible effects on cetaceans of environmental change, for use in constructing scenarios for further RMP implementation trials for Southern Hemisphere baleen whales.

Several main points were agreed to by workshop participants regarding the contribution of JARPA to minke whale management in the Antarctic.

First, under the objective of 'Estimating biological parameters', the information produced by JARPA has set the stage for answering many questions about long-term population changes regarding minke whales in Antarctic Areas IV and V. Not surprisingly, at this halfway point in the JARPA program there are few definitive answers because of the timescale required to obtain sufficient age distribution and abundance data, and because of unanticipated problems in designing representative sampling regimes and in understanding the stock structure of minke whales in the Southern Hemisphere. For example, JARPA has already made a major contribution to the understanding of certain biological parameters (e.g., direct measures of the age at sexual maturity) pertaining to minke whales in Areas IV and V, yet such analyses have not fully addressed potential problems related to stock structure.

Second, under the objective of 'Elucidating the role of minke whales in the Antarctic ecosystem', JARPA has collected data on body condition that, in conjunction with the data on biological parameters as noted above, should result in an improved understanding of the status of minke whales in these Areas. These data are likely to be useful in testing various hypotheses related to aspects of the 'krill surplus' model.

Third, under the objective of 'Elucidation of the effect of environmental change on cetaceans', there is considerable uncertainty in how biological parameters of minke whales may vary in relation to environmental change. This is exacerbated by lack of knowledge regarding processes related to environmental change (e.g., interdecadal signals, global warming, etc.). For example, long-term trends related to the annual positioning of the extent of the pack-ice during the feeding season have implications regarding the interpretation of trends in various biological parameters. Therefore, more effort is needed to develop mesoscale studies to integrate physical and biological oceanography and prey distribution with minke whale studies.

Fourth, under the objective of 'Elucidation of the stock structure of minke whales to improve stock management', deciding on the amount of genetic data required to meet this objective is difficult because the Committee has provided only a vague definition as to what constitutes a stock. Proper delineation of stocks has implications for interpretations of data gathered for all other JARPA objectives.

Finally, the results of the JARPA programme, while not required for management under the RMP, have the potential to improve the management of minke whales in the Southern Hemisphere in the following ways: (1) reductions in the current set of plausible scenarios considered in Implementation Simulation Trials; and (2) identification of new scenarios to which future Implementation Simulation Trials will have to be developed. COMMITTEE DISCUSSION
The Chairman thanked Reilly for his hard work in chairing the review meeting and complimented the participants for producing a thorough report. He believed the review meeting had significantly improved the Committee's understanding of the issues involved in the JARPA programme. He reiterated that time constraints had prevented it from completing item 8 on its Agenda. This would be taken up by the Committee under its Item

Before discussing the report the Committee considered two new papers relevant to the catch-at-age issue. SC/M97/6(Rev) and SC/49/SH22 reported analyses arising out of the discussions at the Review Meeting. The former provided results on the qualitative implications of possible biases in the JARPA estimates of abundance, and of trends in natural mortality with time, for estimates of the historic trend in recruitment of minke whales in Area IV. Although these factors can change the values estimated, the authors argued that they were hardly able to alter the qualitative conclusion of the paper of an increasing trend in recruitment until the late 1960s. SC/49/SH22 fitted these recruitment estimates by a slight variant of the BALEEN II model which allowed for an increase in carrying capacity from 1930 to 1960. It showed that patterns of decreasing recruitment after 1970 indicated by the analyses of SC/M97/6 were as might be expected from the joint consequences of catches and supercompensation. It also demonstrated the possibility of estimating MSYR from the results of the SC/M97/6 analyses, and hence provided a potential link between the results of that analysis and information pertinent to RMP Implementation Simulation Trials.

Discussion of SC/49/Rep 1 in the Committee concentrated on two topics. The first concerned issues of stock structure arising out of, inter alia, annex G to the report and the discussions on stock structure summarised under above. The JARPA review report had commended the work carried out but noted the need for future work. Questions such as what comprises a management stock, how this relates to a biological stock, what inferences can be drawn from the inability to detect significant genetic differences and the relationships between temporal and geographical boundaries, have taxed the Committee for several years in many of its sub-committees. The other issue concerned the problems associated with obtaining representative samples and their implications for the programme.

In view of the importance of this, two Working Groups were established under Polacheck: (1) to address general issues of stock identity and the representativeness of samples in the JARPA report; and (2) to specifically consider a paper submitted to the present meeting on the issue of stock identity and the use of historical samples (SC/49/SH28). The latter report, modified slightly to incorporate some factual information (e.g. with respect to gels) is given as Annex U1. It was agreed that the report of the former group should be incorporated into the main report of the Committee.

In discussing Annex U1, the Committee agreed that the computer simulations suggested in SC/49/SH28 to try to address the value of commercial samples to answer stock identity problems were feasible. Provided the relevant information was available, they should be carried out.

However, several issues remain with respect to the commercial samples, not the least being their availability/existence. It was recognised that creating a sample inventory for the commercial catches was a major task. The Japanese scientists agreed that they would undertake this exercise for at least some of the Areas. The Committee welcomed this.

The final issue discussed in the Annex U1 concerned questions of the need for further catches in the context of whether genetic studies of commercial samples are found to be suitable and adequate. This is discussed under Items and 12.3.1.

Turning to the more general issues the Committee recalled that the review meeting had provided the following overview of its view on the JARPA objective of estimating biological parameters:

The information produced by JARPA has set the stage for answering many questions about long-term population changes regarding minke whales in Antarctic Areas IV and V. Not surprisingly, at this halfway point in the JARPA programme there are few definitive answers because of the time-scale required to obtain sufficient age distribution and abundance data, and because of unanticipated problems in designing representative sampling regimes and in understanding the stock structure of minke whales in the Southern Hemisphere. For example, JARPA has already made a major contribution to understanding of certain biological parameters (e.g., direct measures of the age at sexual maturity) pertaining to minke whales in Areas IV and V, yet such analyses have not fully addressed potential problems related to stock structure.

The question was raised as to whether these unanticipated problems had been resolved and if not how this would effect the ability of the JARPA programme to meet its objectives.

With respect to stock structure - the representativeness of the sampling and bias in the JARPA estimates of abundance - unresolved questions still remain. There is an interaction between these questions as abundance estimates and the representativeness of sampling need to be evaluated relative to the stock being sampled.

The Committee noted that the problem of stock identity is common to almost all cetacean assessments. The data collected and the research carried out in the JARPA programme along with historic commercial catch samples are uniquely valuable in attacking this problem.

The Committee identified ten main areas to address these unresolved problems (Table 2). Fujise indicated that work on all of these is either in progress, has recently been initiated or is at the planning stage.

Table 2

Future work to address outstanding issues.

1. Abundance estimates
- Development of method to correct bias of abundance estimate
2. Stock structure
- Stock definition*
- Statistical analysis of mtDNA data considering the inclusion of school size as a covariate
- Pilot study on nuclear DNA analysis on JARPA minke samples
- Effort to obtain biological materials for genetic analysis from low latitude areas of the Southern Hemisphere**
- External morphology/morphometry analysis
- Examination of possible stock boundaries (geographical and temporal) in Areas IV and V
3. Biological parameters
- Segregation study
- Recalculation of biological parameters by biological stocks
4. Marine ecosystem and environmental change
- Meso-scale survey plan
* The lack of a working definition of stocks and sub-stocks is a general problem, not for JARPA alone, and therefore, needs to be addressed by the Committee (SC/49/Rep 1).
** A preliminary paper was presented last year with the results of a survey of biological material among researchers and institutions of the Southern Hemisphere (IWC, 1997h, pp. 132-3)

The current JARPA survey design resulted in the under-sampling of high density areas. The method described in SC/M97/23 represented an initial attempt to develop an unbiased abundance estimation method with this design. It was noted that work was planned to test the bias of this method and to further develop unbiased abundance estimators as discussed in Annex E.

Achieving adequate precision in the abundance estimates is critical in order to be able to estimate biological parameters (e.g. natural mortality rates) with adequate precision to meet the JARPA objective. Correcting for the biases may decrease the currently estimated precision. Investigations need to be carried out to determine how this might affect the ability to meet the JARPA objectives. Tanaka stated that if this bias can be accounted for, his method (Tanaka et al., 1992 pp. 531-6) can be applied. However, he noted that since there was no finalised method for correcting the bias, it was not certain if the abundance estimates would have sufficient precision over the course of the experiment, although he was optimistic that this would be the case.

SC/49/Rep 1 had identified that 'the geographical delimitation of the sampling area has not resulted in either distinct biological populations being sampled or the entire ranges of the population being sampled'. The primary factor was a lack of sampling from lower latitudes that are known to contain a higher proportion of juveniles. Non-representative sampling of whales by school size and areas were also identified. The major identified consequence of the lack of non-representative sampling is that juveniles are under-represented in the sampling to a variable extent, as a result of the lack of sampling in lower latitudes and other factors. Analyses conducted to date, indicate that sampling among the post-juveniles (i.e. ages 10 and above) is consistent with representative sampling with respect to age. If this is the case, it should be emphasised that the JARPA age compositional data are adequate in this regard for estimating some important biological parameters (e.g. natural mortality rates) for these older ages.

Japanese scientists reported that they plan the following process to resolve the problem of sampling biases:

the quantity of the biases will be evaluated using a resampling simulation model;
post-modelling methods will be applied to determine whether it is possible to resolve or reduce biases;
if this is not effective, modification of the sampling scheme will be considered, and the effectiveness and practicability of such a modification will be evaluated; while simultaneously,
comparability between the current and any new scheme will be carefully evaluated.

With respect to item (iii), Annex U2 contains one approach for a possible modification of the current sampling scheme to be considered if post-modelling is found not to be effective.

With respect to the catch-at-age analyses, the JARPA review meeting concluded

that there was merit in pursuing the approaches of SC/M97/6 and SC/M97/21 further, but that estimates from such methods could not be considered reliable until difficulties associated with the estimates of abundance from JARPA (see Item 2.4) have been resolved.

The difficulties with the abundance estimates are being addressed (see above). Another concern raised in the JARPA review with respect to the VPA analyses was the problem of 'aliassing' or the confounding of recruitment trends with trends in time in natural mortality rates. This question was addressed further in SC/M97/6 (Rev) presented to this year's Committee meeting.

One factor not completely considered with respect to the catch-at-age analyses was the interaction between these analyses and the uncertainty about stock structure. The analyses were made by analysing the data by conventional management units. However, the JARPA review indicates that the genetics data are not consistent with these units. There is a need to consider the implication of alternative stock definitions for the catch-at-age analyses and the ability of these analyses to estimate biological parameters.

With respect to the VPA analyses, the uncertainty about stock structure applies not only to the data collected under JARPA but also to the historical catch-at-age data. In this regard, it would be extremely valuable to obtain genetic analyses of the stored biological samples from the commercial catch.

The Committee noted that JARPA is at the half-way point and has provided substantial improvement in the understanding of stock structure. It is anticipated that as stock structure becomes better clarified, the information will be incorporated to provide analyses of biological parameters by stocks. For some of these analyses, this may not be straightforward.

In conclusion, the JARPA review had identified areas of additional future work that could contribute to resolving some of the unresolved and unanticipated problems in sampling and stock structure that could limit the ability of JARPA to estimate biological parameters. Effort in response to all of these identified areas is being undertaken and this additional work may improve the value of the JARPA data and results. Initial results from the simulation studies being undertaken to examine post-stratification and alternative sampling designs will be presented to the Committee as soon as possible. Based on these results, consideration should be given as to whether the sighting and biological sampling designs should be modified to achieve more representative sampling.

Finally, the Committee agreed that none of the sampling and stock identity problems that had been identified either in the JARPA review or subsequently, would in principle prevent JARPA from achieving its objectives in terms of estimation of biological parameters. All of the identified problems appear to be addressable. Most members were optimistic that the JARPA data, in conjunction with additional work planned, would allow estimation of the biological parameters with reasonable levels of precision. However, others thought that the problems associated with bias and the level of variance in the JARPA abundance estimates, and with interactions between catch-at-age analyses and uncertainty in stock structure, mean that it is not yet possible to determine whether reasonable levels of precision will ultimately be achieved.

In addition, aspects of JARPA relevant to abundance estimation were considered in Annex E. A brief summary of those discussions is given below.

SC/49/SH10 described sighting data collected during the 1996/97 JARPA programme. Work was conducted in Area V and the western part of Area VI over 103 days. Three vessels were engaged in both sighting and scientific whaling, and only one vessel conducted a closing mode sighting survey. The ships maintained roughly simultaneous parallel tracklines despite their different tasks by working in target-distance mode, whereby the survey begins each day at targeted positions rather than where each ship had completed work the previous day (SC/49/Rep 1, item 2.2). The ships surveyed 17,755 n.miles of tracklines, with 850 primary sightings of schools totalling 2,508 animals.

A proposal to develop unbiased abundance estimators from the JARPA sightings data was presented in SC/49/SH30. The JARPA survey design results in an undersampling of high density areas. SC/49/SH30 aims to complete the work in two phases. The first phase would be completed and presented to the Scientific Committee for comment at its 1998 Annual Meeting. The Committee agreed that the proposed framework has the potential, in principle, to obtain unbiased abundance estimates using the JARPA data and would be useful for obtaining an improved understanding of those data. It noted that the likelihood of being able to develop unbiased estimates with reasonable levels of precision depends on the degree to which the spatial process can be modelled with the available covariates and to some extent on the stationarity of the process. DISCUSSION OF ITEM 8 FROM SC/49/REP 1
The Committee agreed to address this issue by providing a summary of the Commission's resolutions with respect to JARPA and relevant comments from the JARPA report and its discussions under Item

The Commission had recommended that

scientific research intended to discuss the Comprehensive Assessment and the implementation of the RMP shall be undertaken by non-lethal means

that scientific research involving the killing of cetaceans should only be permitted in exceptional circumstances where the questions address critically important issues which cannot be answered by the analysis of existing data and/or use of non-lethal research techniques

and then requested the Committee to

undertake a comprehensive review of existing programmes and report its view as to whether they remain justifiable in the light of the above recommendations, especially whether lethal research substantially contributes to answering critically important questions that cannot be addressed by other means.

This is discussed under Items and

The Committee was also requested to structure its reviews in the manner given in (a)-(c) below.

Identify the relationship between programme objectives and research needs previously identified by the Scientific Committee (see also SC/49/Rep 1, annex D; annex F; item 2.5; item 3.5; item 4.5; item 4.5 B2; item 5.5; item 6.5; item 7).

The relationships are identified in Table 3.

Table 3

Relationship between JARPA research objectives and Scientific Committee research needs.

JARPA Research Objectives Scientific Committee's priorities
Estimation of biological parameters to improve stock management.
RMP: Relevant to MSYR discussions, Implementation Simulation Trial scenarios.
Elucidation of the role of whales in the Antarctic marine ecosystem.
Relevant to work discussed at the Climate Change Workshop, particularly with respect to filling in identified gaps in knowledge (Rep. int. Whal. Commn 47:309-10); multi-species issues.
Elucidation of the effect of environmental changes on cetaceans.
Related to above and the Pollution Workshop.
Elucidation of the stock structure of minke whales to improve stock management.
RMP: Relevant to Implementation Simulation Trial discussions and specification of Small Areas.

Evaluate the likelihood of the programme meeting its objectives by providing reliable answers to the questions asked (see also SC/49/Rep 1, annex D; annex F; item 1.6; item 2.5; item 3.5; item 4.5; item 5.5; item 6.5; item 7).

SC/49/Rep 1, annex F summarises the Committee's earlier discussions on this matter. The Committee had no agreed consensus on this in the past.

In SC/49/Rep 1 (item 2.5), it was noted that the frequent sighting surveys in the same localities would facilitate estimation of interannual variability in local abundance which would in turn lead to improved overall results from combining them with IDCR/SOWER and/or JSV data, for example. However, improvements in methodology were suggested.

Under SC/49/Rep 1, item 3.5, there was general agreement that the stock structure data were of value to management. However additional research was recommended. It was also agreed that the information was relevant to improved Implementation Simulation Trials and an improved RMP in the longer term.

In SC/49/Rep 1, under item 4.5, it was agreed that the programme provided valuable information on a number of biological parameters (recruitment, natural mortality, decline in age at sexual maturity and reproduction). However, it identified the need for further work particularly in view of the difficulties in obtaining fully representative samples. Although there is much still to be done, it was agreed many valuable results have been obtained. It was noted that the results in the short term could be valuable with respect to several aspects of the RMP, provided certain identified problems were resolved.

Under SC/49/Rep 1, item 5.5, it was agreed that estimates of daily food consumption could be used with confidence for estimating total food consumption. However, the Committee had concerns over the use of a body condition index for inferring information on the migration of minke whales with respect to timing of arrival on the feeding grounds and over-wintering in high latitudes. The review meeting agreed that the studies were contributing to Objective 2 (p.96). However, additional studies were recommended. It was also noted that the information obtained would be of interest to CCAMLR and Southern Ocean GLOBEC.

Under SC/49/Rep 1, item 6.5, the review meeting had agreed that the work was pertinent to Recommendations 1, 4 and 5 of the Pollution Workshop. However in the Plenary some concerns were expressed on the extent to which the work on minke whales directly addressed Recommendation 1 (see Item 6).

Under SC/49/Rep 1, item 7, the review meeting had recognised that this was a long-term programme that had only reached its half-way point. In several cases therefore, it could be said to have set the stage to answer many questions about long-term population changes. It also noted that while JARPA results were not required for management under the RMP, they had the potential to improve it in the following ways: (1) reductions in the current set of plausible scenarios considered in Implementation Simulation Trials; and (2) identification of new scenarios to which future Implementation Simulation Trials will have to be developed (e.g. the temporal component of stock structure). The results of analyses of JARPA data could be used in this way perhaps to increase the allowed catch of minke whales in the Southern Hemisphere, without increasing the depletion risk above the level indicated by the existing Implementation Simulation Trials of the RMP for these minke whales.

In the Committee's discussions (Item, it had considered the implications of identified problems in stock structure and sampling in terms of the ability to achieve stated objectives. Most members were optimistic that JARPA data, in conjunction with the additional work planned, would allow estimation of biological parameters with reasonable levels of precision. Others however believed that the identified uncertainty meant that it is not yet possible to determine whether reasonable levels of precision will ultimately be achieved.

Identify non-lethal methods and alternative sources of data that might be used in meeting the research objectives (see also SC/49/Rep 1, annex D; annex F; annex H ; item 3.5; item 4.5 B1).

SC/49/Rep 1, annex H , provides summary statements supporting/refuting the use of lethal removal pertaining to the collection of stock structure information.

Under SC/49/Rep 1, item 4.5 B.1, the review meeting noted that there were non-lethal methods available that could provide information about age-structure (e.g. natural marking) but that logistics and the abundance of minke whales in the relevant Areas probably precluded their successful application.

In the Committee's discussions ( and Annex U2), the question was raised as to whether there were adequate suitable samples already available from the commercial catches for examining the question of stock identity. It is not yet clear which of the commercial data can be used for this process but a procedure to try to determine this is in place.

The aspect most relevant to JARPA appears to be:

that the Scientific Committee should develop practical means to address questions raised by these exchanges where 'these exchanges' refers to contact with other organisations and exchange of information on the effects of global environmental change in the Antarctic that may be relevant to whale stocks (also see SC/49/Rep 1, item 5.5; item 6.5; IWC, 1997d).

Under SC/49/Rep 1, item 5.5, the meeting noted that the information obtained would be of interest to CCAMLR and Southern Ocean GLOBEC.

The aspect of Resolution 1995-10 most relevant to JARPA is:

providing information on the potential effects, both direct and indirect, of pollutants on cetaceans as they become known

The aspects of Resolution 1996-8 most relevant to JARPA are to:

Increase collaboration and co-operation with other organisations; and
Consider and act on, as appropriate, the recommendations of the Climate Change and Pollution Workshops and other items identified as requiring additional information, so as to develop non-lethal means of assessing the impact of environmental change on cetaceans (see also SC/49/Rep 1, item 6.5; Pollution Workshop recommendations).
Under SC/49/Rep 1, item 6.5, the review meeting had agreed that the work was pertinent to Recommendations 1, 4 and 5 of the Pollution Workshop (IWC, 1998). However, the Committee expressed some concerns over the extent to which the work on minke whales directly addressed Recommendation 1 (see also Item 6).

Under SC/49/Rep 1, item 5.5, the review meeting agreed that the studies were contributing to Objective 2. However, additional studies were recommended. It was also noted that the information obtained would be of interest to CCAMLR and Southern Ocean GLOBEC.

12.2.2 Japan - North Pacific
A number of papers were presented arising out of the Japanese Research Programme in the North Pacific (JARPN). These include SC/49/NP2, NP8, NP9, NP10, NP11, NP12, NP13 and NP14.

These were considered by the relevant sub-committees and are also considered under Item 8.1.

12.3 Review of new or revised proposals
12.3.1 Japan - Southern Hemisphere
The Government of Japan (1997) presented the 1997/98 JARPA Research Plan (SC/49/SH3). This is a continuation of the programme discussed previously by the Committee, although for the coming season sampling will not only be conducted in Area IV but also in the eastern half of Area III. Samples of 300 minke whales will be taken in Area IV and 100 in Area III E: all will be ordinary form minke whales and both samples will have an allowance of ± 10%. Sampling strategies and methodologies will be the same as in the previous research plan.

The objective of sampling again in Area III E is to investigate inter-year variability in the occurrence of stocks W and C in Areas III and IV. When Area III E was sampled previously (in 1995/96) it was announced that this would be for 'one year only at this stage', but the results of that survey had shown a completely different pattern in stock distribution from that expected from previous commercial samples. The new samples were required to test for further variability in stock distribution in these two Areas.

The Committee refers the Commission to its views expressed in previous years on the applicability of non-lethal methods, possible effect on the stock and opportunity for participation by foreign scientists, which were considered unchanged from last year (e.g. IWC, 1997e, p.96).

In discussion, reference was made to Annex U1 with respect to the use of commercial catches, and in particular whether it was possible for samples from these to replace those animals to be taken in Area III. Annex U1 had noted that there was general agreement to use existing samples where they would provide equivalent information relative to a particular experimental objective and design before collecting new samples through lethal removals. It was noted that the inventory and computer simulations could not be completed before the 1997/98 sampling year. Therefore Yagi stated that it was Japan's intention to proceed as proposed with the 1997/98 research plan (SC/49/SH3) including the Area III E component.

Donoghue suggested that if no lethal takes were taken this year this would allow time for the work suggested in Annex U1 to be undertaken. In response, Fujise noted that JARPA was being undertaken for a number of purposes in addition to stock identity. In particular, the extension into Area III had been in response to problems about the representativeness of the catch. Even with respect to stock identity he noted that a range of techniques, not exclusively genetic, was being used. Another factor being investigated concerned the identification of 'core' areas and a further elucidation of stock structure in light of the annual variation identified from previous years. Hatanaka pointed out that existing commercial samples from Area III had already been analysed and the results presented to last year's Committee meeting (IWC, 1997g).

12.3.2 Japan - North Pacific
SC/49/NP1 described the continuation of a programme, begun in 1995 after a feasibility study in 1994, to examine: (i) whether sub-stocks of minke whales exist in the Okhotsk Sea - Western Pacific stock (O stock); and (ii) whether an additional minke whale stock (W stock) exists in the central North Pacific, and if it does, the rate of mixing with O stock. One hundred animals will be sampled in two to three areas (of a total of 13 Sub-Areas to be sampled in due course). The Committee noted that it had addressed the proposal in detail previously. It refers the Commission to its previous comments (see IWC, 1997e, item 14.4.2).