|
This assessment arose in response to the growing concerns of all stakeholders in 2002 for the sustainability of the Australian east coast spotted mackerel (Scomberomorus munroi) fishery, and the necessary requirements of the Commonwealth EPBC Act. The fishery is comprised of a single unit stock that undertakes seasonal spawning and feeding movements along the Queensland and New South Wales coasts. The highly aggregated, near surface schooling behaviour of the stock coupled with its predictable seasonal movements along the east coast allows ease of targeting by both the commercial and recreational sectors; thereby making the stock susceptible to over-fishing. In 1999-2000, commercial catches of spotted mackerel increased significantly in response to the development of valuable overseas export markets, where subject to management intervention in 2002, an increase in effort was likely to continue while attractive prices were being offered and overseas markets continued to expand. Concurrently, anecdote suggested that recreational catches decreased significantly, leading to major concerns about the ecological sustainability of the fishery. In this assessment, therefore, we evaluate the Australian east coast spotted mackerel fishery from northern Queensland to northern New South Wales waters, incorporating all available biological data, and commercial and recreational catches for the fishing (i.e., financial) years 1960-2002.
Spotted mackerel spawn between August and October in northern Queensland waters. Peak spawning occurs in September. Following spawning, the majority of the stock appears to undertake a summer feeding movement to southern Queensland and northern New South Wales waters, returning in autumn and early winter. The timing and extent of these movements are most likely related to water temperature and clarity, and baitfish distributions.
Spotted mackerel grow quickly for the first three years of life, and demonstrate sex-specific growth rates, with females tending to grow faster and to larger sizes. Considerable variation in length is found for any given age of spotted mackerel, where they have been aged up to 7 years and observed to 105 cm total length (TL) and 7.4 kg. Female and male spotted mackerel reach maturity at about 60 cm and 52 cm TL, respectively, within 1-2 years of age. Because of the significant differential growth between the sexes, sex-specific age length distribution keys (ALK) were derived for spotted mackerel and used to estimate age structures from length distributions collected in 1991-1996, 1999 and 2002. Sex-specific age structures derived from the respective ALK were combined to form a single age structure for each year’s catch in which there was fish length data and used as input into the assessment. Overall, the spotted mackerel population was highly dependent on young fish, with 1-3 year olds being the dominant age groups.
Although the Australian east coast spotted mackerel fishery extends into northern New South Wales waters, the majority of the fishery occurs in Queensland waters. Spotted mackerel are caught by both commercial and recreational fishers mainly around Bowen and Innisfail during winter and early spring (July-September), and from Hervey Bay and Moreton Bay during late spring and summer (November-February). Few by-product or by-catch species are caught when targeting spotted mackerel, although this will vary depending on method of fishing, and may include Spanish, school, grey and shark mackerel, long-tail and mackerel tuna, bonito, shark and trevally. In addition, relatively few sub-legal or undersize fish are discarded when targeting spotted mackerel, although potential numbers will most likely increase for line caught spotted mackerel with the introduction in 2002 of a 60 cm TL minimum legal size. Furthermore, little is known about the post release mortality of line caught and subsequent release of spotted mackerel, although anecdote suggests this to be significant.
Mackerel have been commercially caught from Queensland waters since at least 1945, with reported landings up to 855 t in 1974. Anecdote suggests that most of these historic landings were Spanish mackerel, with targeted commercial fishing for spotted mackerel commencing in about 1960. Commercial catches of spotted mackerel in Queensland waters have increased over the years, reaching a peak of 410 t in 2000. Prior to management intervention and the banning of targeted netting of spotted mackerel in 2002, the use of ring nets was the main method of capture by the commercial sector. Significant quantities of unspecified mackerel have also been reported in the compulsory commercial logbooks each year, ranging from 10-131 t. In contrast, spotted mackerel comprise a relatively small component of the total mackerel catch in northern New South Wales waters, with a peak catch of about 55 t in 1999.
Recreational fishing for spotted mackerel is similar to the commercial sector in being highly localised and seasonal, reflecting the spatial and temporal availability of the species. Recreational fishing for spotted mackerel is particularly important south of Townsville, and also for visiting interstate recreational anglers who target the winter run of spotted mackerel in northern Queensland waters. Recreational catches of spotted mackerel were estimated to be about 31,000, 210,000, 86,000, 129,000 and 53,000 fish in 1995, 1997, 1999, 2000 and 2002. As with the commercial sector, significant quantities of unspecified mackerel catches were also reported in the State and National recreational fishing surveys, of which a proportion of these were estimated to be spotted mackerel. The average length and weight of a recreationally caught spotted mackerel was about 65 cm TL and 1.64 kg, in contrast to the commercial sector which typically caught a 70 cm TL and 1.92 kg spotted mackerel.
The reported commercial and recreational spotted mackerel catches were used as the basis for this assessment because it was assumed that these were a function of fishing effort and abundance of the population, where the level of total catch over time may reflect changes in the proportion of the population caught, changes in the abundance of spotted mackerel, or both. However, significant quantities of unspecified mackerel were reported in the commercial logbooks and recreational fishing diaries each year , where it was assumed a proportion of these were spotted mackerel, which in turn needed to be estimated to determine total catches. Binary regression models, therefore, were used to allocate reported unspecified mackerel catches to spotted mackerel catches.
The binary regression models identified significant changes in the probability of commercial and recreational catches of mackerel being spotted mackerel between fishing regions, gears, months, years, number and weight of mackerel caught, time spent fishing, etc. The probability of a Queensland commercial catch of mackerel being reported as spotted mackerel was significantly greater when caught using ring nets, and from the traditional spotted mackerel fishing regions of Moreton Bay and Hervey Bay during summer, and Bowen during winter. As well, the probability of a recreational catch of mackerel being spotted mackerel was lower for longer times spent fishing, and higher when large numbers of mackerel were caught. The proportion of unspecified mackerel allocated as spotted mackerel varied from 1-42 t in any given year for the Queensland commercial sector, less than 1 t for the New South Wales commercial and Queensland charter sectors, and up to 94 t for the recreational sector.
Standardised annual catch rates (i.e., catch per unit effort; CPUE) of spotted mackerel were used in the assessment as a relative index of population abundance. The standardisation analysis considered a number of different climate variables thought to affect the catchability (and subsequent catch rate) of spotted mackerel including the Southern Oscillation Index (SOI), wind speed and direction, sea surface temperature and lunar phase. The analysis also used Queensland and New South Wales spotted mackerel commercial and recreational catches from 1988 to 2002. The analysis examined only those spotted mackerel catches where fish were caught and retained (i.e., catches > 0). No data were available on searching effort to find spotted mackerel or on fishing effort where no spotted mackerel were caught. The final model considered the reported spotted mackerel catch associated with the number of units and type of fishing effort (eg. days fished by commercial line operations or the number of fishers in a recreational fishing group), fishing year, month, region, lunar phase and climate/weather conditions. A range of weighting values were also examined to reduce the hyperstability effect of certain fishing effort types or target/non-target fishing for spotted mackerel.
Spotted mackerel catch rates varied according to fishing years, regions, months, strength of north-south winds, lunar cycle, SOI and the amount and type of fishing effort. Catch rates appeared to decline from 1990 to 1995, increasing slightly thereafter, before remaining relatively stable. As expected and supporting the anecdote, catch rates of spotted mackerel increased with favourable weather conditions such as light northerly winds (i.e., <20 km/hr). Favourable weather and sea conditions would both increase the targeting efficiency of fishers and maintain the integrity of near surface feeding schooling mackerel; thereby leading to greater catch rates. Estimated average relative fishing power was also considerably higher for ring net fishing than the other line fishing effort types, suggesting potential hyperstability issues with net based catch rates. The pseudo target/non-target weightings had little influence on the analysis. Commercial line fishing catch rates, therefore, were considered best to reflect the underlying population abundance of spotted mackerel due to the elevated hyperstability of net fishing, and was used in the assessment as an annual relative index of abundance.
The annual total catches for the fishery that were used in the assessment included data from the Queensland and New South Wales commercial and recreational sectors. Total catches were estimated separately for two periods of the history of the fishery (1960-1987; 1988-2002) because of data availability issues. Total catches of spotted mackerel for each year of the historic period (1960-1987) included: 1) Queensland commercial catch from Queensland Fish Board data (1960-1980); 2) New South Wales commercial catch of 1 t based on the average catch of spotted mackerel for the initial years of the logbook data (1984-1988); and 3) average total catch for years in which no data were available. In those years where there were no data available (1981-1987) or only commercial data (1960-1980) a generalised linear model was used to estimate the average total catch; being fitted to the nominal total catches for 1988-2002 and projected back to 1960.
Total catches of spotted mackerel for each year of the more recent period (1988-2002) included: 1) actual reported catches for spotted mackerel from the Queensland and New South Wales commercial logbooks (plus binary model allocated unspecified mackerel catch); 2) Queensland recreational survey catch estimates for 1995 (52 t), 1999 (201 t), 2000 (265 t) and 2002 (180 t); 3) New South Wales recreational survey catch estimates for 1993 (5 t), 1994 (1 t) and 2000 (27 t); 4) estimated Queensland recreational catch for 1988-1994, 1996-1998 and 2001 when there was no actual survey conducted; and 5) estimated New South Wales recreational catch for 1988-1992, 1995-1999 and 2001-2002 when there was no actual survey conducted. In those years when there were no surveys conducted, recreational catch estimates were based on the average relative recreational effort for the years in which data were available. The estimated total catches for spotted mackerel increased significantly since the fishery was assumed to have commenced in 1960. Total catches reached a peak of about 755 t in 2000, just prior to the investment warning in 2002. Major uncertainties, however, exist in the total catches and relate mostly to the magnitude of the historical catches and those of the recreational sector.
The assessment used all available and relevant biological and fisheries data to provide an indication of the current level of exploitation and sustainability of the spotted mackerel fishery. A sex-specific age-structured population dynamics model was used to evaluate the status of the fishery. The main data sources for the model were the total catches, catch rates and age structures. Results from the model suggested that the stock is most likely being harvested near or exceeding maximum sustainable levels, and is at risk of being over-fished; albeit given the data and model assumptions and uncertainties. Biomass trends demonstrated significant declines in the stock over the past 10 years, particularly during the mid-1990s to early-2000s when catches were at their peak, with current levels estimated to be at 33-63% of unfished or virgin biomass (i.e., B 0) levels. Sensitivity analyses demonstrated that the assessment model was quite robust with similar biomass ratios and management quantities estimated for a variety of model runs. This included testing a range of input parameters such as natural mortality and stock-recruitment steepness. The estimated catches (or yields) when fished at a level to attain maximum sustainable yield (i.e., Y(F MSY)) varied from 296-570 t for the different model runs, although the preferred base models fitted to the age structure and CPUE data, and age structure data alone, estimated Y(F MSY) to be about 366 t and 296 t, respectively. Only for the latter base model (i.e., Y(F MSY) = 296 t), was the spotted mackerel exploitable biomass in 2002 predicted to be below that which would sustain MSY (i.e., B 2002 < B MSY). Furthermore, projections based on this model suggest that future catches of less than 350 t will likely increase biomass back to B MSY and above. Management advice derived from the age structure only model was the preferred option because of uncertainties associated with the potentially hyperstable catch rates; a directive in accordance with the precautionary approach.
Several alternate assessment models (e.g., surplus production (ASPIC), virtual population analysis ( VPA/ADAPT), statistical catch-at-age (ASAP)) were also examined to evaluate the relative performance, robustness and uncertainty associated with the population trends derived from the age-structured population dynamics model. Results from the surplus production model (ASPIC) were in contrast to those expected for the life history of spotted mackerel. Biomass estimates from ASPIC were at extremely high and unrealistic levels as the model attempted to fit to the relatively flat catch rate time series. In contrast, results from the VPA/ADAPT and ASAP models were relatively similar to those from the population dynamics model tuned only to age structures. Overall, results from the alternate models showed the sensitivities associated with some of the key input data and assumptions. In particular, results from the models suggested that the relatively flat standardised catch rate data may still not be a good indicator of population abundance; undoubtedly being affected by hyperstability.
A hierarchical approach to reference point estimation, which dictates that reference points be determined by the method that most reliably captures the salient population and fishery dynamics given the data available, was used to evaluate a range of reference points (including those derived from the age-structured population dynamics model) for determining sustainable catch strategies of spotted mackerel. Reference point estimation was assessed in terms of data type, quantity and quality to enable the associated uncertainty to be evaluated in a transparent, hierarchical framework. Uncertainty in the data types was a common problem and varied according to representativeness of data coverage, degree of extrapolation for years of missing data, use of proxies or species analogies when no data were available, etc. Most of the data types were considered to have moderate to high uncertainty. A suite of target (i.e., desirable) and limit (i.e., avoidable) candidate reference points were chosen that may be considered appropriate for the management of spotted mackerel. At the lower tier of the hierarchy, based on simple historical proxies for sustainable catch levels, a target total catch for the fishery was estimated to be about 200 t with a limit of 333 t. In contrast, the more complex age-based production model estimated a target total catch for the fishery of 277-282 t and a limit of 296 t. The nominal 2003 total catch of 350 t (if the TACC of 140 t was fully realised) for the fishery was above all the estimated candidate reference points, irrespective of the data and models used. Management of the spotted mackerel fishery, therefore, may need to consider more prudent actions in the future, in accordance with the precautionary approach to ensure the long-term sustainability of the fishery, even given the recent management intervention. Also, s election of candidate reference points should be informed by the use of alternate methods and their assumptions, which in turn should be used to identify critical assumptions in the estimation process and to direct future research and monitoring programs.
The uncertainty associated with this assessment and subsequent reference points and management advice derived from the related analysis and models was a function of the quality and extent of the input data. Future research and monitoring programs, therefore, should be directed towards providing the necessary data required for improved model parameters and reference point estimation. Various forms of monitoring and opportunistic or project oriented research provided the necessary data used in this assessment of spotted mackerel. The assessment data came from a range of sources, but relied heavily on Queensland’s commercial and recreational fishing databases, as well as fishery-dependent spotted mackerel length and age data . The continuation of a monitoring program for spotted mackerel, therefore, is essential if further data- or model-based assessments are to be conducted for this fishery. Consequently, w e used random effects modeling and power analysis to determine an effective and optimal monitoring strategy for spotted mackerel in the future.
Results from the analysis showed the importance of monitoring spotted mackerel lengths (and ages) every year, with samples ideally collected across fishing regions from all line fishing sectors. We recommend the minimum sampling of at least 600 fish from both the recreational and commercial line sectors across two broad regions: 1) northern Queensland (Townsville – Bowen); and 2) south east Queensland ( Hervey Bay – Moreton Bay). This equates to a minimum sample each year of 600 fish distributed across the 2 fishing sectors and 2 regions (i.e., sampling target of 150 fish per sector per region); all fish should be sexed and aged. Furthermore, based on the median daily catch by sector, a minimum of 20 catches should be sampled from the commercial sector and 150 from the recreational sector. Overall, the need for more age-structured data collected from both the commercial and recreational sectors is of greatest importance to improve this assessment in the future.
All indications from this assessment, besides the relatively flat CPUE time series, suggest that the spotted mackerel stock is most likely being harvested near or exceeding maximum sustainable levels, and is at risk of being over-fished. Although the best available data were used to determine the status of the stock and inform an appropriate level of risk, there was an inherent level of uncertainty associated with the data and model assumptions that need to be considered for management advice and future assessments. Major levels of uncertainty exist in the key biological parameters of natural mortality, stock-recruitment and reproductive output, as well as in the fisheries data of the historical and recreational catches. The transparent and comprehensive nature of this assessment should enable all stakeholders and managers involved in the spotted mackerel fishery to make more informed decisions concerning the management of the resource, with an understanding of the associated uncertainties and risks. The choice of management actions to implement in the future should be examined in a management strategy evaluation framework, similar to the approach used in this assessment, to determine the trade-offs between particular management actions and the management objectives to be met; coupled with the associated levels of risk.
In addition, to improve and develop this assessment in the future, we recommend the need for: 1) a more comprehensive and structured monitoring approach to the collection of appropriate age-structured data from both the commercial and recreational sectors; 2) the recording of a better measure of effort and species identification in the commercial logbooks and recreational diaries to provide a more reliable indicator of CPUE; 3) a review of the historical catch data to confirm the assumed commencement of the fishery and magnitude of the catches; 4) a robust evaluation of the selectivity functions for the different fishing gears; 5) an appraisal of the protocols used to age spotted mackerel; 6) a fishery-independent measure of changes in stock size; 7) investigations into the fecundity, spawning, recruitment processes and environmental-catch distributions of spotted mackerel; and 10) a periodic review and update of the data and models used in the assessment via a systematic and transparent stock assessment review process.
This stock assessment is the most comprehensive attempt to evaluate the status of the Australian east coast spotted mackerel fishery. The assessment used all available biological and fisheries data to provide an indication of the current level of exploitation and sustainability of the Australian east coast spotted mackerel fishery. The results, however, need to be tempered with the uncertainty associated with the various data and model assumptions; although this should not be used as a basis for management inaction. Indeed, the precautionary approach dictates that management should be more prudent given greater uncertainty. The transparent and comprehensive nature of the assessment should enable all stakeholders involved in the fishery to make more informed decisions concerning the management of the resource, with a thorough understanding of the associated uncertainties and risks. Overall, the analyses and modeling facilitated a critical assessment of the spotted mackerel fishery; thereby, making more effective use of the catch data and past biological research on the species. The assessment has provided a basis for Queensland and New South Wales fisheries managers, and their relevant advisory committees to consider sustainable levels of fishing and management objectives for the fishery. Operational objectives and trigger points for the fishery, however, need to be defined to guide future management strategies. Recent management measures also need to be assessed in the future, and more prudent actions may be needed, if fishing pressure increases in the recreational sector or the commercial catch quota is exceeded.
For a pdf file (7MB) of the report contact CRC Reef on info@crcreef.com.
|
