(from "Report of the Scientific Committee", the version distributed at 2001 meeting)

Analysis had been conducted using DESS with options that have become standard⁴ for analyses of these surveys (notwithstanding the fact that the analysis methods are currently under review). Alternate options might also be reasonable. While it was agreed that the analyses of these data might be particularly sensitive to changes in the options used, the results did not suggest that the standard estimate was inadequate. The effect of adequate alternate analysis options on the abundance estimate is to be investigated intersessionally.

³ The 1984/85 survey covered only the eastern half of Area IV
⁴ See Annex G, Item 6.2.2 and Appendix 5

10.2.2 Updated estimates by Area
10.2.2.1 OVERVIEW OF DATA COLLECTION AND TRAINING METHODS
10.2.2.1.1 IDCR/SOWER CRUISES
Changes in the IWC/IDCR and SOWER Antarctic survey designs and data collection protocols were reviewed. Notwithstanding the importance of maintaining consistency over time, there have been two major modifications and various minor modifications of survey design during the course of the surveys.

The programme was modified from a combined Discovery marking and sightings program to a rigorous and systematic sightings programme from the second circumpolar set of surveys (starting 1985/86 and referred to as CPII for brevity). At that point, strict identification guidelines were established for sightings of Antarctic minke and Southern bottlenose whales. In the third set of circumpolar surveys (starting in 1991/92; CPIII for brevity), the survey design was modified to cover the whole region south of 60°S, at the expense of full longitudinal coverage of Management Areas in a single year.

10.2.2.1.2 JAPANESE SCOUTING VESSEL DATA
Japanese scouting vessel (JSV) data comprise daily summaries of sightings from four kinds of vessels (full-time scouting vessels, operating catcher boats, national dedicated survey vessels and IDCR vessels), but mainly from full-time scouting vessels. For a number of reasons (there is no systematic track design, no record of sighting distance information, no identification of primary/secondary sightings), the JSV data are not suitable for direct density estimation. However, unlike IDCR/SOWER data, they cover latitudes north of 60°S. Although they do not extend later than the 1987/88 season, they are potentially useful for extrapolating IDCR/SOWER abundance estimates northwards and for examining seasonal migration. Some members expressed reservations about using JSV data for extrapolation because weather conditions change with latitude. JSV data with minke whale sighting records are available from the 1970/71 season; data with less systematic and reliable recording of minke whales are available from the mid-1960s. Dwarf minke whales are not distinguished from Antarctic minke whales in the data.

10.2.2.1.3 JARPA
Methodology from the sighting and sampling surveys in the JARPA programme was reviewed and the similarities and differences in procedures between the JARPA and IDCR/SOWER programmes were reviewed. Both involved searching at about the time of the peak abundance of Antarctic minke whales on the feeding grounds. IDCR/SOWER cruises had shorter research period (early January to late February) than the JARPA cruises, which typically started in late December and ended in early March (see Appendix 4 of Annex G). JARPA surveys involved simultaneous searching by different vessels in closing mode ('SV mode' in JARPA terminology - where 'SV' stands for 'sighting vessel') and in 'search and sampling' mode (SSV mode). In SSV mode animals are caught so that this mode involves more time off-effort in order to close on and catch whales. JARPA surveys covered only Areas IV and V, whereas the IDCR/SOWER surveys covered all Areas. The areas covered by the JARPA surveys in Area V were almost always smaller than those covered by the IDCR surveys.

10.2.2.2 REVIEW OF CURRENT IWC LINE TRANSECT ANALYTICAL METHODOLOGY
The development of the 'standard' analysis methodology was reviewed; its evolution within the IWC Scientific Committee is summarised in Annex G (Appendix 5). Use of the word 'standard' does not imply that this methodology could not be improved further; indeed a substantial component of the current review of Antarctic minke whale abundance estimates involves a review of the methods and development of improved methodology. The methods are 'standard' insofar as they have evolved by agreement within the Committee (and have been applied consistently over a number of IDCR/SOWER surveys in the past).

Branch and Butterworth (2001b) report the results of applying this standard methodology to minke whale sightings data from all IDCR/SOWER surveys up to and including the 1997/98 survey, using the IWC software DESS (Version 3.0). It is possible that the resulting Antarctic minke whale estimates may include a very small proportion of dwarf minke whales. Estimates of abundance obtained for three circumpolar sets of surveys: 1978/79-1983/84, 1985/86-1990/91 and 1991/92-1997/98 (*still incomplete) and are 608,000 (CV = 0.130), 766,000 (CV = 0.091) and 268,000* (CV = 0.093) respectively. These surveys have covered 65%, 81% and 68%, respectively of the areas south of 60°S and the ice edge. These estimates are negatively biased estimates of Antarctic minke whale abundance because areas inside the pack ice cannot be surveyed, not all minke whales are in the survey region, the assumption is made that all whales on the trackline are sighted and 'like minke' sightings are omitted.

The Committee welcomed the estimates, which had been updated to accommodate concerns and suggestions made by the Committee on the estimates presented to it in 2000. It expressed its appreciation of the work involved in revising the estimates intersessionally.

It was suggested that alternative methods of analysis, stratification, detection functions, strip width models, pooling methods and school size estimation were unlikely to make a large difference in either the magnitude of the estimates or the relative estimates in the three surveys. By contrast, the potentially large differences in the proportion of the minke whale populations covered on each survey were thought likely to affect the estimates to a much greater extent.

It was noted that violation of the assumption that all animals on the trackline are detected (g(0)=1), on which the estimates of Branch and Butterworth (2001b) are based, would affect the estimates in ways that are not entirely predictable. It was also noted that it would be preferable to estimate abundance without the assumption that g(0)=1. Several attempts to estimate abundance without the assumption in the past have been unsatisfactory, but recent methodological developments are promising in this regard.

10.2.2.3 FACTORS THAT MAY INFLUENCE ABUNDANCE ESTIMATES AND THEIR TRENDS
10.2.2.3.1 SCHOOL SIZE ESTIMATION
The performance of the DESS school size estimation rule was examined against two alternatives. It was noted that the standard methods of estimating school size depend on g(0)=1 and that this was certainly not true for all schools in all conditions. It was also noted that the regression method used to estimate mean school size might be sensitive to the form of the detection function used. Investigation of alternate estimation methods that do not rely on the g(0)=1 assumption was considered important.

The Committee recommends that the modified approach to estimating school size suggested in Branch and Butterworth (2001b) should be implemented in DESS and should be used in future analyses from IDCR/SOWER surveys pending the development of models which either provide satisfactory estimates of g(0) or do not build certain detection on the trackline into their assumptions.

10.2.2.3.2 STRATIFICATION/POOLING
Sample size considerations limit the degree to which stratification can be used in estimation. Following recommendations made last year. Branch and Butterworth (2001b) investigated pooling options for estimation of effective strip width (esw) and mean school size and suggested that the AIC criteria usually used to decide on pooling should be replaced by an alternative rule outlined in Annex G. In addition, other possibilities were suggested in discussion. The Committee agreed that the use of this new method needed to be considered further. It also noted that methods that model detection probability as a function of covariates should require less pooling of data and should be investigated and applied to these data.

10.2.2.3.3 OBSERVER EFFICIENCY
Average observer experience on the third set of circumpolar surveys is lower than that on previous surveys. Results of analyses investigating the relationship between observer experience and sighting efficiency were presented to the Committee. It was noted that this analysis suggests a substantial and significant difference in the sighting rates with experience when observers work alone in the IOP. However, it is not clear what the overall effect of this is in practice because other observers may sight schools that the less experienced observer might have missed. The data are inadequate to determine if this is a cause of the lower abundance estimates obtained from the third set of circumpolar surveys because the increased proportion of less experienced observers since 1990 has resulted in there being too few observations involving experienced observers alone in the third circumpolar survey.

The effect of two methods of estimating observer efficiency on circumpolar estimates of abundance was examined. Two analyses based on different assumptions indicated either: (a) that changes in observer efficiency changed overall abundance estimates by only a few percent; or (b) they resulted in an increase in the ratio of third to second circumpolar abundance estimates from some 35-40% to 45-50%. It was noted that the effect of observer efficiency on g(0) may be more complicated than is assumed in the above analysis.

10.2.2.3.4 CHANGES IN 'LIKE-MINKE' CLASSIFICATION
When observers believe a sighting is of a minke school, but have some degree of uncertainty about this, the sighting is classified as 'like-minke'. Results were presented showing that the proportion of 'like minke' sightings had increased over time, from negligible levels in the first circumpolar set of surveys to over 30% in IO mode in the third. If 'like minkes' are included in abundance estimates, those for closing mode increase only slightly because esw estimates also increase. The IO mode estimates for the second and third circumpolar sets of surveys increase by some 10% and 20% respectively.

Analyses examining the relationship between observer experience and the proportion of 'like-minke' sightings concluded that the increase of 'like minke' whale sightings in recent years seems not to be caused by the introduction of inexperienced observers. Other reasons, such as the increasing number of species codes and the extent of survey coverage to the northern stratum (in which weather is worse and school sizes smaller on average) may have caused difficulties in determining the species.

10.2.2.3.5 ANIMALS MISSED ON THE TRACKLINE AND DUPLICATE IDENTIFICATION
'Standard' analyses assume that all animals on the trackline are detected - that g(0)=1. Probability of detecting animals on the trackline is likely to depend on school size and the assumption that g(0)=1 for small schools at least, is almost certainly violated; abundance estimates are biased as a result, although the bias may be small. There was considerable discussion about whether multi-platform analyses could yield estimates of g(0), given the problems encountered in attempting this in the past. One of the major problems discussed was that of non-independence of sightings from the IO and barrel platforms due to the availability of visible cues. The approach outlined in SC/53/IA31 could in principle accommodate this difficulty (although it requires accurate species identification) and the Committee recommends that this method be applied to the IDCR/SOWER data.

Net bias in abundance estimates when g(0) is less than 1 is necessarily negative, even though g(0) depends on school size and so estimation of mean school size by extrapolating to perpendicular distance zero (as the regression method does) provides a positively biased estimate of mean school size (see Appendix 15 of Annex G).

Crude estimates from SC/53/IA27 suggest an effect of about 15% due to assuming g(0)=1. Crude analyses of duplicate proportions suggest that g(0) might have decreased by as much as 40% in Area I from CPII to CPIII (see Appendix 6 of Annex G).

10.2.2.3.6 EFFECT OF ENVIRONMENT ON SIGHTING CONDITIONS
Results were presented indicating that esw depends on sighting conditions, as expected. The Committee recommends that methods be developed and applied that model detection probability as a function of covariates (including sightings conditions) and that further effort be invested to develop methods that would allow reliable estimation of abundance without assuming g(0)=1.

10.2.2.3.7 TIMING OF SURVEYS AND ENVIRONMENTAL FACTORS
Analyses on the timing of migration to the Antarctic were presented, using: (a) JSV data from 1966/67 to 1987/88 in the waters south of 40°S in the Areas III, IV and V; and (b) using CPUE and sightings survey data information. Both indicated a peak abundance of whales in the waters south of 60°S in late January. There is also indication of a rapid decrease in the proportion of whales in this region in February. This suggests that a longer research period in February in the recent SOWER surveys may have resulted in increased negative biased abundance estimates from recent surveys. Although taking account of the migration trends indicated by the CPUE and sightings data increases estimates of abundance by roughly 25%, it affects the ratio of the estimates for the third and second circumpolar sets of surveys negligibly. This may be a consequence of Area-specific differences in migration timing and the fact that the Ross Sea, which contains high abundance, was surveyed relatively later.

It was noted that the spatial modelling methods outlined in SC/53/IA29 could be used to estimate temporal as well as spatial trend in abundance in the survey regions and should be less susceptible to any bias resulting from possible differences in migration timing in different regions.

Results were presented detailing changes in the sea ice extent from 1979 to 2000. During the period of ice retreat, the monthly extent of the ice between mid-December and mid-February was highly variable, as expected. There was also considerable variation in its extent between years, with most variability in Areas II, V and VI. It was noted that the nature of the variability in sea ice concentration within years differed between geographic areas, suggesting the need to consider the effects of the ice configuration in its totality, not only its latitude, on abundance estimates. The Committee noted the importance of estimating minke whale abundance south of the ice edge, but in the absence of more information, no quantitative conclusions about this abundance could be drawn. It recommends that this issue be addressed intersessionally.

10.2.2.3.8 CLOSING-PASSING MODE CALIBRATION
Since 1991, IDCR/SOWER analyses have applied a correction factor, R, to calibrate Closing mode estimates (which are acknowledged as potentially biased) to those estimates that would have been obtained had the Closing mode survey been in IO mode. Currently, a single estimate of R is calculated, this estimate being updated annually (although the estimate of R used in analyses of IDCR/SOWER data, such as those presented in SC/53/IA3, uses data from surveys up to and including the 1988/89 survey.) However, the Committee noted that the extent of bias in closing mode is believed to be density dependent and thus that it was unreasonable to assume that R is constant across years. The change in the proportion of 'like-minke' sightings adds further evidence to support this conclusion and it was agreed that alternative methods which allow a calibration to be estimated separately for each survey should be investigated.

10.2.3 Inter-year comparisons and trend
10.2.3.1 DEALING WITH VARIABLE PARTIAL COVERAGE
10.2.3.1.1 EXTRAPOLATION TO UNSURVEYED REGIONS
Problems with comparability between estimates from the three circumpolar sets of surveys arise because of two factors relating to survey coverage. Firstly, most surveys in the first two circumpolar sets did not completely cover the full latitudinal range to 60°S. Secondly, the third circumpolar set of cruises has not yet completed a full circuit of the Antarctic - the longitudinal ranges of 140°W-160°W and 80°E-130°E have yet to be surveyed. The Committee considered three possible approaches for estimating density in these unsurveyed regions. The first assumes that the unsurveyed northern areas have the same density of whales as the northern surveyed strata in each survey and uses the density in the covered part of these northern strata to extrapolate abundance estimates to a common area south of 60°S. Extrapolating in this way and correcting for different longitudinal coverage results in comparable circumpolar estimates of minke whale abundance on CPI, II and III of 729,000 (CV=0.150), 824,000 (CV=0.117) and 359,000 (CV=0.108). These estimates apply to 280° of a possible 360° longitude for complete circumpolar coverage.

Exploratory analysis of the utility of two spatial modelling methods for extrapolating and interpolating into unsurveyed regions were presented. Both performed well on the data they were applied to and the Committee recommends that these approaches be developed further intersessionally.

10.2.3.1.2 JSV DATA
SC/53/IA11 used JSV data in the waters south of 40°S in Areas III, IV and V from 1966/67 to 1987/88 to extrapolate abundance estimates from IDCR/SOWER surveys north of 60°S to 40°S. A GLM with research period, season, latitude and longitude as explanatory variables and presence/absence as the binary response variable was fitted to the JSV data. Using JSV daily encounter rates pooled over longitude, together with the fitted GLM, relative abundance is predicted over the range of latitudes spanned by the JSV data. This is used to extrapolate the IDCR/SOWER estimates north of 60°S. Of the minke whale population in these Areas, 30% are estimated to be north of 60°S. There are a number of sources of possible bias associated with this extrapolation method, some positive and some negative.

Butterworth briefly outlined the results from an analysis of JSV data from 1978/79 (Borchers et al., 1990). He reported that including minke whales (and dwarf minke whales) that are distributed north of 60°S would result in an increase of about 10% to the estimates using data from south of 60°S only.

10.2.3.1.3 ANIMALS WITHIN THE PACK-ICE
Results relating to the estimation of minke whale abundance south of the ice edge were presented. The Committee welcomed this first attempt at estimating the number of minke whales that may be present within the pack ice, but was unable to reach any conclusions other than noting that some minke whales are present within the ice, but that their numbers are unknown. It encouraged efforts to ensure that more information on surveys within the ice are presented at next year's meeting. The Committee supported the conclusion in SC/53/IA14 that surveys within the ice would provide useful data on Antarctic minke whale abundance. The Committee recommends that the minke whale data available from the APIS program be analysed intersessionally to provide estimates of abundance south of the ice edge.

10.2.3.1.4 JARPA DATA
The Committee noted the likely utility of the spatial modelling methods discussed under Item 10.2.3.1.1 above, for extrapolation and interpolation of JARPA survey data into unsurveyed regions on IDCR/SOWER surveys within and between years.

10.2.3.2 TRENDS IN ABUNDANCE
At its last meeting, the Committee noted the fact that the overall abundance estimate from crude extrapolations from the third circumpolar set of IDCR/SOWER surveys⁵ was appreciably lower than the total of the previously agreed point estimates by Area from the 1990 Comprehensive Assessment (IWC, 2001i, p.189). It also noted that there are a number of factors that make interpretation of this difficult. The Committee reconsidered the issue this year, in the light of the substantial new analyses that had been conducted intersessionally.

The possible utility of population dynamics models for the review of Southern Hemisphere minke whales had been noted last year (IWC, 2001j, p.199). At that time, initiation of work on population dynamics modelling had been considered premature. Having made substantial progress on other aspects of the review since then, the Committee revisited the issue.

It agreed that the interpretation of trend in a meaningful way required some form of population dynamics modelling and that this should take account of sightings and other data in addition to those provided by analyses of the IDCR-SOWER sightings data. While opinions differed on the likelihood of being able to interpret the results of such modelling in a way that would conclusively explain the trend in estimates (since the distinction between process error and sampling error was based on the modelling assumptions), the Committee recommends that population dynamics models should be examined at the earliest opportunity.

SC/53/IA13 examined year to year trends in the biological parameters of Antarctic minke whales from 1971/1972 to 1999/2000, using both commercial and JARPA data. The authors believed that these data provide no evidence to support a decline in Antarctic minke whale abundance after the early 1970s and in particular not since 1987.

Branch and Butterworth (2001b) contains updated estimates of abundance from the IDCR/SOWER surveys from 1978/79 to 1997/98 using currently standard⁶ estimation methodology. It also contains estimates for each survey that have been corrected for different survey coverage of the population in a crude but reasonable way. The total estimated abundance from CPIII is significantly lower at the 5% level than the comparable total for CPII. This is true both when the estimates are corrected for different survey coverage and the inclusion of 'like-minke' sightings (but no other factors) and when they are not.

SC/53/IA27 considered the effect of the following factors on the abundance estimates: an increase in the proportion of 'like-minke' sightings over time; different survey coverage; changes in the method used to estimate mean school size; the decreasing trend in observer experience; changes in the timing of the surveys. Results indicated that the effects with largest impact on abundance estimates were the change in the area covered by the survey and decreased observer efficiency⁷, although no firm conclusions could be drawn in this regard, given the simple and exploratory nature of the analyses. Choices for the various factors in combination, which likely err towards positive bias, suggested an increase in the ratio of abundance estimates for the third compared to the second circumpolar set of surveys from 35-40% to 65-75%.

SC/53/IA12 presented estimates of the abundance of Antarctic minke whales in Areas IV and V from 1989/90 to 2000/01 using sighting data obtained from JARPA surveys. The standard methods were used (for comparability with analyses of IDCR/SOWER data) and trends in abundance estimates were examined. It was assumed that abundance estimates from SV data can be treated as a relative abundance index, as proposed previously (IWC, 2001i, p.188). Since the time series of sightings data from Sighting and Sampling Vessels (SSVs) is longer than that for Sampling Vessels (SVs), the SSV data were converted to pseudo-SV data, using the method described in Haw (1991) to extend the time series available for examining trends in abundance. The trend in relative abundance using the SV and pseudo-SV abundance estimates was estimated to be - 0.04% with 95% CI (-4.32%, 2.90%) in Area IV and 0.83% with 95% CI (-2.44%, 4.19%) in Area V; in neither Area is the estimated trend significantly different from zero. The authors conclude that neither a significant increase nor a significant decrease in the abundance of minke whales has occurred in Areas IV or V since the JARPA surveys commenced.

In discussion of these results, it was noted that the JARPA estimates had not been considered in the detail in which the IDCR/SOWER survey estimates had been considered at this meeting. It was also noted that some considerations that apply to the IDCR/SOWER survey estimates do not apply to the JARPA estimates (the major changes in survey design between circumpolar sets of surveys is a case in point) and vice-versa. It was also considered important that direct comparisons between JARPA estimates and IDCR/SOWER estimates for Areas IV and V be presented to the Committee's next meeting. The Committee recommends that an attempt should be made before next year to determine whether there is a statistically significant difference in trend between density estimates from JARPA survey data and those from IDCR/SOWER surveys in Areas IV and V and that IDCR/SOWER estimates be presented for Areas excluding IV and V. It further noted that application of spatial modelling methods to IDCR/SOWER datasets, as detailed in Annex G (Appendix 10) could provide estimates at higher spatial resolution.

The lack of a trend in these estimates contrasts with the fall in abundance estimates reported in Branch and Butterworth (2001b), although it should be noted that the JARPA surveys occur only in Areas IV and V, whereas the IDCR/SOWER surveys cover all Areas. The timing of JARPA surveys has remained approximately constant throughout, whereas since 1994/95, the IDCR/SOWER surveys have started later (see Appendix 4 of Annex G). Concerns were expressed about this change of timing as evidence of peak abundance in the survey region (see section 10.2.2.3 above) suggests that the change to a later date may have led to some of the survey areas being covered after the peak in more recent surveys rather than at about the time of the peak on earlier surveys. Furthermore, the JARPA surveys spanned a longer period of time in the survey areas so one would expect that they might be less sensitive to the location of the peak. The Committee agreed that if it was logistically feasible, future IDCR/SOWER surveys should revert to the time schedule of earlier surveys. Although taking account of the migration trends indicated by the CPUE and sightings data increases estimates of abundance by roughly 25%, it affects the ratio of the estimates for the third and second circumpolar sets of surveys negligibly.

It was noted that even when the IDCR/SOWER abundance estimates of Branch and Butterworth (2001b) are adjusted using what in the context of results from SC/53/IA27 are likely positively biased correction factors, the total corrected abundance estimate from CPIII (corrected for the factors indicated in Branch and Butterworth, 2001b) remains lower than that from CPII. Interpretation of this result is difficult. The Committee did not have the necessary results to determine whether the difference between the corrected estimates of abundance for CPII and CPIII were statistically significant.

After considering many of the factors affecting abundance estimates that were identified last year, there is still evidence of a decline in the abundance estimates from CPII to CPIII, although it is not clear how this reflects any actual change in minke abundance. Three hypotheses that might explain these results were identified:

(1)

a real change in minke abundance;

(2)

changes in the proportion of the population that is present in the survey region at the time of the survey;

(3)

changes in the survey process over the course of the surveys that compromise the comparability of estimates across years.

Fig. 1 (a) and (b): Estimates of relative abundance from the three sets of circumpolar IDCR/SOWER surveys of Antarctic minke whales from the three circumpolar sets of IDCR/SOWER surveys. Because of incomplete longitudinal coverage, the estimates account for only 280° of a possible 360° longitude. Estimates are shown separately for (a) Closing mode and (b) IO mode. Within each mode, estimates are shown which (i) include corrections for incomplete survey coverage south of 60°S and the inclusion of "like-minke" (these are called '2 Factors' in the plots), and also for the case in which (ii) correction is made for incomplete survey coverage south of 60°S and inclusion of "like-minke", school size estimation method, timing of the survey relative to the migration peak of minke whales south of 60°S, and observer efficiency (These are called '5 Factors' in the plots). Vertical bars represent approximate 95% confidence intervals. These are likely too narrow because they do not include uncertainty due to estimation of the correction factors. It should be noted that the estimates have not been corrected for any change in g(0) (effect on abundance estimates could be large: 15 and 40%), or the proportion of animals south of the ice edge that may have occurred. Estimates might change somewhat if different analysis options are used in obtaining the point estimates.

Fig. 1 (c) and (d): Estimates of relative abundance in (c) Area IV and (d) V only, from JARPA data, together with estimated 95% confidence limits. These estimates do not include corrections for any of the factors noted.

Fig. 1 shows the comparable estimates of the relative abundance of Antarctic minke whales from the three circumpolar sets of IDCR/SOWER surveys and estimates of relative abundance of Antarctic minke whales in Areas IV and V only from JARPA survey data. Note that the IDCR/SOWER survey estimates cover 280° of the possible 360° of longitude spanning the Antarctic, whereas the JARPA estimates cover only Areas IV and V.

⁵ The third circumpolar set of IDCR/SOWER surveys is abbreviated to 'CPIII' in what follows; the second set to CPII and the first set to CPI.
⁶ See Appendix 5 of Annex G
⁷ The size of the observer effect can be estimated in two ways, depending on assumptions about the way it operates. The correction factor used here was obtained using those assumptions that result in it having the larger effect.

10.2.4 Plans for completion of minke review
10.2.4.1 DESIGN ISSUES
The importance of maintaining a consistent survey design throughout the period under study was emphasised and it was noted that there are uncontrollable spatial and environmental factors which affect the estimation of trend and interpretation of results. The Committee noted the substantial process error associated with estimates of trend and abundance from the IDCR/SOWER surveys. Some design considerations are given in Annex G (Appendix 9), including outlines of two possible alternative transect designs.

10.2.4.2 WORKPLAN
Annex G (Appendix 10) details the tasks identified by the Committee to further the review of Antarctic abundance estimates, together with an indication of priorities for the next year. Noting the need to resolve the issue of whether or not minke whale abundance has declined in recent years, the Committee strongly recommends that substantial progress be made on high priority tasks by the 2002 meeting of the Committee. It also recommends that JARPA estimates of abundance and trend in Areas IV and V should be evaluated in a similar way to the way the IDCR/SOWER survey results were evaluated this year.

The Committee recalled that at its last meeting it had strongly recommended holding an intersessional workshop to consider estimation methods and estimates for Antarctic minke whale abundance and trend, but that this had not been funded. It reconsidered the issue this year and drew the Commission's attention once again to the importance of supporting the Committee's work under this Item if the review of Antarctic minke whale abundance and trend is to be successfully completed in the near future. It repeated that the review would necessarily span a number of years. It also noted that the process of methodological development was open-ended and that there was a need to specify a realistic deadline for completion of the review. Given that the third set of circumpolar surveys will be completed in 2002/3 and that the review cannot be completed by its 2002 meeting, the Committee proposes that estimates of Antarctic minke whale abundance and trend using the improved methodology developed in the course of the review be presented for the three full circumpolar sets of IDCR/SOWER surveys at its 2003 meeting.

Recognising also that some further analyses might be required in the light of these results, the Committee proposed further that the review be finalised at its 2004 meeting. The Committee will develop more detailed plans for intersessional work related to the review at its 2002 meeting and, if necessary, at its 2003 meeting.

In order to provide sufficient time at its next meeting to consider the results of intersessional work and to plan its work for the remainder of the review, the Committee strongly recommends that a full two days be set aside immediately before the 2002 meeting specifically to address methods and results relating to the estimation of Antarctic minke whale abundance. It further noted that the arrangement this year, in which less than two days were set aside for this purpose and the second of these overlapped the first day of the Scientific Committee meeting, had compromised the effectiveness of the meeting.

The budget requests are considered under Item 21. They include successful completion of the 2001/02 SOWER survey, and recommends a two-day meeting and other high priority items from Annex G (Appendix 10).

10.2.5 Other
The results of genetic analyses in SC/53/IA17 re-emphasised the importance of collecting genetic samples from whales in putative breeding grounds in lower latitudes. It was also noted that the results suggest that estimation of population dynamics model parameters may be more complicated than previously envisaged.

_