Haddon, M., Klaer, N., Smith, D.C., Dichmont, C.D. and A.D.M. Smith (2012) Technical reviews for the Commonwealth Harvest Strategy Policy. FRDC 2012/225. CSIRO. Hobart. 69 p.

1 Executive summary

The following sections attempt to identify key points raised in each of the sections of this set of reviews. It should be noted that this material is diverse and relatively complex so, unfortunately, brief summaries of each section are not possible. The following are not conclusions but rather constitute important points that require noting. 1.1 Reference Points Appropriate to Life-History Characteristics The range of suggestions for what would constitute an appropriate target biomass and fishing mortality value is very great but the difficulty in estimating the real risks of running relatively high fishing mortality rates at low stock sizes indicates that the suggestion of B40% rather than something lower is a reasonable compromise. The current default biomass target reference point of B48% would appear to be highly conservative (biologically) for many species, although it may be quite appropriate for slower growing sharks and rays and may not be sufficiently conservative for some key low trophic level species. For example, the Commonwealth small pelagic fishery, in line with a number of regulations world-wide, has adopted a biomass level of at least 80% B0 as the BLIM for each species in this fishery (with higher values in the more data poor situations), and for such ecologically important species such apparent high levels seem appropriate. However, such a level would ignore the fact that such species are naturally highly variable and could quite naturally vary in abundance, sometimes down to very low abundance levels. An alternative could be not to accept a limit with reference to a fixed B0 but rather to only take a standard proportion of available biomass. Such constant escapement strategies are not currently included in the HSP but would be useful for naturally highly variable species such as scallops, small pelagic species, and squid, for which the concept of a stable unfished biomass, B0, may not be meaningful. Full implementation of this would thus mean that management of such stocks would not be in relation to specific biomass limit and target reference points but rather in relation to estimates of current stock size. In addition, such a strategy might need to include some minimum level of predicted harvest before fishing could occur so as to avoid encouraging unprofitable fishing.

For productive species where 0.5BMSY is less than B20% the current HSP suggests that levels of biomass < B20% would be acceptable. Given the uncertainty inherent in estimation of stock productivity, the precautionary approach would firstly require good evidence that 0.5BMSY is indeed below B20%. In the face of these various doubts and uncertainties it would be difficult to argue that there would be no increase in the risk of depletion affecting consequent recruitment levels if the limit biomass reference point was permitted to vary below the current B20%. For small pelagic fisheries, because of ecosystem based fishery management considerations the limit reference point would tend to be either the same as or very close to the target (which has similarities to having a constant escapement strategy.

1.2 Buffered Targets or Meta-Rules

The present arrangements where those harvest strategy control rules in which a break point is clearly defined at the proxy target reference point certainly stabilizes catches and another meta-rule that prevents TACs varying by more than 50% between any two years has also been helpful in preventing serious dislocation and disturbance in the fishery for some relatively unstable species. These particular meta-rules have already been simulation tested using MSE.

If it was decided to pursue the issue of buffers and meta-rules around the targets in an attempt to stabilize catches through time then it would be beneficial to use simulation testing (MSE) to consider the effect of such changes to the expected dynamics of different fisheries.

1.3 Data Poor Fisheries and Tiered Harvest Strategies

We define fisheries or species as data poor if information is insufficient to produce a defensible quantitative stock assessment.

For data poor fisheries, difficulties can arise in almost every component of the harvest strategy – for example, little or no regular monitoring means time series are rare, the assessment method is undertaken with an unknown degree of uncertainty, reference points are poorly defined and the associated control rules do not necessarily address risk clearly. Yet, a recognized component of the present Harvest Policy is the application of a consistent degree of risk across all fisheries, irrespective of fishery type.

Often the efficacy of a data poor harvest strategy can be very fishery specific. The use of a tiered system of assessment methods and associated control rules allows for the development of detailed, integrated stock assessments (Tier 0 and 1) down to the lowest Tiers where data is limited to catch rates, catches, or even just catches (Tiers 6 and 7). Below these tiers is the Ecological Risk Assessment, which aims to determine whether there are particular species that are exceptionally vulnerable to the effects of fishing.

1.4 TAC Setting and Multi-Year TACs

Generally, when TACs are set for individual species, catches of other species are not considered. In multi-species fisheries, there are often technological interactions where fishing effort directed towards one quota species will normally result in a mixed catch of fish that may include other quota species. Fishers can usually ‘target’ to some degree through fishing different areas and depths, seasons, times of day and by modifying gear. But it is the degree to which fishers can target that is the issue. The species mix in catches may not necessarily match the mix in combined TACs or in quota holdings. This difficulty in balancing quotas for multiple species with actual catches may then lead to increased discarding, TAC over-runs, effort restrictions or fishery closures when quota is constrained on some species. It is possible to characterize recent multispecies catch data into primary and companion components. The approach of identifying companion species within a given fishery provides an empirical means to examine the impact of individual species TAC decisions across all of the quota species in a fishery.

In general, multi-year TACs will require a “discount” (reduction) of some level of catch to balance the greater risk associated with less frequent review and adjustment. There are obvious risks of stock depletion if the multi-year TACs are set too high. While there is debate about how best to set multi-year TACs no decisions have yet been made. Currently there has been little testing of the robustness of fisheries to the application of multi-year TACs.

1.5 Rebuilding Strategies and Bycatch-only TACs

A primary objective of the Commonwealth Harvest Strategy Policy (HSP) is to maintain key commercial fish stocks at ecologically sustainable levels and within that context, maximize the economic returns to the Australian community. If a fishery falls below the default limit reference point of B20% the HSP states that: “Typically recovery times are defined as the minimum of 1) the mean generation time plus ten years, or 2) three times the mean generation time.” However, attempting to meet these guidelines has been problematic, for example, in at least three conservation dependent species in the SESSF.

The HSP already states that not all species in a multi-species fishery need be maintained at the target reference point (default of B48% as a proxy for BMEY) as long as all assessed species stay above the limit reference point. So the rebuilding target for each species is not always clear.

The HSP makes the assumption that rebuilding of a depleted species will always occur. However, in a changing marine environment this may not always be true. Potential regime shifts have already been identified in particular species (Jackass Morwong) on Australia’s east coast (a world hot spot for sea water temperature rise) and this provides an example of a species whose long term productivity has declined. There is thus a need to recognize that there are circumstances under which rebuilding to previously experienced levels would not be expected to occur.

It is also possible that some species, particularly when they were fished under a basket species category (e.g. gulper sharks) may have been reduced to such a low level that the probability of them recovering would become influenced by random events. In addition, if the projected timeline for recovery is extremely long it becomes possible that long term changes in the marine environment will become influential on the probability of eventual recovery.

Finally, there are some species which are naturally extremely variable (e.g. squid and scallops). Simulation testing can be used, and has been used, to demonstrate that the harvest strategies in place are potentially capable of achieving the intent of the HSP, even though it is very hard to identify adequate proxies for a particular limit or target biomass reference point. However, some unpredictable events, such as the recent almost complete die-off of scallop beds in south-east Australia, unrelated to any fishing, are not amenable to anything other than reactive management.

1.6 Spatial Management

Spatial management may be applied in various contexts within a harvest strategy. It can form the main harvest strategy framework (such as in a system of rotational closures), it can be used to augment a harvest strategy framework, or spatial management measures can be invoked as a control rule (a variation of rotational closures). For some species a management scheme that controls fishing mortality with large spatial and temporal fishery closures offers a management strategy more robust to uncertainty than direct control of catch, since only a small component of the stock gets exposed to the fishery. However, this relies on good compliance with fixedclosure boundaries (the Commonwealth Vessel Monitoring System ensures this) and is mainly applicable to species that do not move large distances.