New England Fishery Management Council DRAFT MEMORANDUM

Transcription

1 New England Fishery Management Council 50 WATER STREET NEWBURYPORT, MASSACHUSETTS PHONE FAX John F. Quinn, J.D., Ph.D., Chairman Thomas A. Nies, Executive Director DRAFT MEMORANDUM DATE: April 13, 2017 TO: Herring Committee FROM: Herring Plan Development Team (PDT) SUBJECT: Updated analysis for development of acceptable biological catch (ABC) control rule alternatives for Amendment 8 to the Herring Fishery Management Plan (FMP) The Herring Committee met on April 5, 2017 to develop a range of ABC control rule alternatives for Amendment 8. Several specific motions passed recommending the Council include three individual control rule shapes, and another motion passed tasking the PDT with identifying control rule alternatives that meet specific criteria. The Management Strategy Evaluation (MSE) models developed for this action can be used to identify control rule alternatives based on specific objectives. Example objectives identified through this process are: preventing overfishing, maximizing yield, maintaining sufficient herring for forage and ecosystem needs, etc. For each objective there are corresponding metrics, or means to quantitatively measure each objective. The Herring Committee has identified a handful of key metrics they recommend the Council use to identify ABC control rules that could be included in the range of overall alternatives considered. The specific metrics recommended by the Committee are: 1. Constrain %MSY to be 100%, with an acceptable level as low as 85%; 2. Variation in annual yield set at a preferred level <10%, acceptable level as high as 25%; 3. Probability of overfished set at 0%, with an acceptable level as high as 25%; 4. Probability of herring closure (ABC=0) set between 0-10%. The Committee has tasked the PDT to identify the range of ABC control rule alternatives that meet these specific metrics. During the Herring Committee meeting, Dr. Jonathon Deroba on the Herring PDT displayed a figure similar to the one below. Figure 1 shows the 71 control rule shapes that fit the metrics requested. However, including all 71 as individual alternatives in Amendment 8 is not practical. Therefore, the PDT has reviewed the shapes in more detail and has recommended a method the Council could consider to identify a handful of potential alternatives within the criteria requested. In addition to that recommendation, the PDT has several others relative to the range of ABC control rule alternatives. The PDT recommendations are underlined on pages 9-10.

2 Figure 1 Control rule shapes that fit the criteria identified by the Herring Committee The PDT had a conference call on Tuesday, April 11, The request from the Committee was reviewed and the PDT discussed possible ways to refine the range of alternatives that meet these criteria. One specific approach was presented that is based on three systematic steps: 1. Step One: Remove any shapes that have an upper biomass parameter < 0.5. The MSA requires entering a rebuilding program if a fishery is overfished, or when biomass is less than 50%SSBmsy. Holding the target fishing mortality constant at any biomass above this value implies there is no increased risk of being overfished at any biomass. Thus, control rule shapes that have an upper biomass parameter less than 0.5 may not respond to declining stock sizes before separate rebuilding requirements would be required. Control rules that have an upper biomass parameter > 50%SSBmsy use a linear decrease in fishing mortality as biomass declines, which 2

3 should reduce the likelihood of a stock becoming overfished and needing a rebuilding plan. These control rules are therefore more likely to achieve rebuilding requirements and are more consistent with the goal to avoid an overfished status. Applying this step reduced the number of potential control rule shapes to Step Two: Set probability of overfished equal to zero. A common objective from most stakeholders in this process has been that the control rule should have zero, or very little probability of causing the stock to be overfished. When that metric is set at 0, rather than allowing some probability up to 25%, the number of potential control rule shapes is further reduced to Step Three: Set proportion of MSY to be 88% or greater The Committee motion recommended the proportion of MSY be set at 100%, with an acceptable level as low as 85%. When the lower tolerance is adjusted to 88%, rather than 85%, the number of control rule alternatives reduces from 30 to 6. The final six shapes are in Figure 2. 3

4 Figure 2 Herring Committee Recommendations for the potential range of ABC control rule alternatives, incorporating updated PDT analysis of performance based control rules with the other Committee recommendations (Strawman A, Strawman B, and alternative that identifies parameters upfront) Strawman A (Yellow) Parameters upfront (Black dotted) Range of performance based alternatives (6 Total) Strawman B (Grey dashed) 4

5 For comparison, the control rule parameters of all the potential control rule alternatives have been summarized in Table 1. Some include a fishery cutoff, a lower biomass parameter greater than 0, and some do not (set the lower biomass parameter equal to 0). The range of values for the lower biomass parameters for this group of potential CR alternatives is The range of values under consideration for the upper biomass parameter is This value specifies the ratio of SSB/SSBmsy where a control rule begins to reduce fishing mortality. For example, for the alternative that sets parameter values upfront, the upper biomass parameter is 0.7, so fishing mortality would be set lower than Fmax (in this case set at 0.9) when biomass is lower than 70% of Bmsy. Finally, the range considered for Fmax for this set of potential CR alternatives is between 0.5 and 0.9. Fishing mortality would not be able to exceed those levels, regardless of whether estimated biomass was very high (e.g. >2.0 SSB/SSBmsy). Table 1 Summary of control rule parameters for the potential range of alternatives Upper Biomass Parameters Lower Biomass Parameters Max F Strawman A (SQ) Strawman B (~75/40) Parameters upfront MeetCriteria MeetCriteria MeetCriteria MeetCriteria MeetCriteria MeetCriteria Preliminary analyses The preliminary results for the six potential performance based CR alternatives are summarized below, along with the other three CR alternatives the Committee recommends including for consideration (Strawman A, Strawman B and the Parameters Upfront alternative). The PDT has only included a handful of example tradeoff plots to date; additional analyses will be prepared later for other metrics of interest once a final range of alternatives is identified. When reviewing these results it is important to keep in mind that these are not predictions of what the system would actually look like under different control rules. Instead, the results represent a long-term probability of the system performing a certain way, given that the true system dynamics are reasonable approximated by the operating models. For example, even if the MSE projects that the probability of herring being overfished is zero, that does not mean that will never happen under a certain control rule. The probability of that happening may be very low in some cases, but there are many factors involved that are not fully captured in these models. Likewise, the projected yields may not reflect actual catches in the short term. These models are constructed to simulate performance of control rules for the long-term, 150 years in this case. 5

6 Therefore, it is important to keep in mind that the utility of these results is to help illustrate the relative performance of various control rules with absolute performance (e.g. future herring biomass and catches) potentially deviating from that projected by the MSE, especially in the short-term. Overall, the results for Strawman A and the alternative that sets parameters upfront have similar results to each other and generally fall on one end of the spectrum. Strawman B has different results, on the other spectrum from the previous two. Finally, the performance based alternatives generally have results somewhere in the middle, and the differences among the six are very small, essentially overlapping with each other. Figure 3 summarizes the tradeoff between yield and biomass as well as the tradeoff between yield and frequency that SSB is less than 30% unfished biomass (a threshold similar to that used in CR literature and of interest to some herring stakeholders). Strawman A and the CR alternative that sets parameters upfront have higher yields on average (about 85-95% of MSY) with biomass projected to be about 20-40% of unfished, on average. Strawman B is on the other spectrum at about 40-90% of MSY and 45-75% unfished biomass; the results for this alternative are more uncertain between operating models than the other alternatives considered (results produce a larger polygon compared to other alternatives). This figure also shows the tradeoff between yield and frequency that biomass falls below 30% unfished. Again, Strawman A and the alternative that sets parameters upfront is on one spectrum, ranging from about 10%-60%, and Strawman B is tighter for this metric, with 0-10% frequency that biomass may fall below 30% unfished. The performance based CR alternatives range between about 0-40%. It is important to keep in mind that for herring the current estimate of Bmsy is at about 30-40% of unfished biomass. Therefore, if CRs are set to attain at least 85% of MSY, the probability of biomass falling below Bsmy some proportion of years is expected. Figure 4 summarizes the tradeoff of yield and variation in annual yield, as well as yield and frequency of fishery closure, or setting ABC=0. For variation in yield, most of the alternatives have similar results of 20-30% variation, the alternative that sets parameters upfront is slightly higher (25-35%), and Strawman B falls off the plot above 50% variation in annual yield. For the plot of frequency ABC=0, essentially all of the alternatives have very low probability that there would be no herring fishery, ABC=0. Figure 5 shows that all of the alternatives have a low probability of overfished, most are at or near zero, with Strawman A and the alternative that sets parameters upfront having up to a 15% probability of overfished under some operating models. Finally, this figure also has the results for the frequency that a modeled tern population would be able to replace itself based on herring abundance, and in this case, all of the alternatives have a high probability of meeting that criterion (80-100%). The variation is in proportion of unfished biomass, which ranges from about 25% to 75%. 6

7 Figure 3 Tradeoff plots for Herring Yield/MSY and SSB/Unfished biomass (top) and Yield/MSY and Frequency that biomass is <30% of unfished biomass (bottom) Yield vs. Biomass Top Left: more yield / less biomass Bottom right: less yield / more biomass Size of polygon matters: larger more uncertain Yield vs. Frequency B<30% unfished Most have relatively high % MSY As you move from left to right: the frequency that B falls below 30% unfished threshold increases. Differences driven by operating model more than CR alternative (high production operating model shifts frequency B<30% unfished to the right). 7

8 Figure 4 Tradeoff plots for Herring Yield/MSY and Variation in Yield (top) and Yield/MSY and Frequency that fishery will close of ABC = 0 (bottom) Yield vs. Variation in yield Most have relatively high % MSY. As you move from left to right: the annual variation in yield increases. Note: Strawman B has fallen off figure, >50% variation in yield. Yield vs. Frequency ABC =0 (no fishery) Most have relatively low frequency that ABC=0, no fishery (far left of plot near 0). 8

9 Figure 5 Tradeoff plots for Herring Yield/MSY and Frequency Herring is Overfished (top) and Frequency that tern production >=1 and SSB/unfished biomass (bottom) Yield vs. Frequency Overfished Most have relatively low frequency of overfished (far left of plot near 0). Frequency Terns can replace population vs. Biomass Most have relatively high probability that tern production will stay at replacement level. As you move from left to right: biomass increases. 9

10 PDT input 1. Refine the range of performance based alternatives The PDT recommends the Committee and Council refine the range of alternatives using these three relatively straightforward steps. The Council may want to narrow the range down further, but the PDT would not recommend expanding it to include more options. In addition to these six that meet the criteria identified by the Herring Committee, there is the No Action alternative, as well as three other specific alternatives recommended by the Committee. In addition, the Council has already recommended that there be annual biomass based alternatives, as well as 3- year biomass based alternatives, so that will double the number of alternatives, potentially from 9 shapes to 18 alternatives. The PDT notes that while some of the more predator driven metrics were not included in the handful of key metrics used to identify potential control rule shapes, ecosystem and predator metrics have been explicitly considered throughout this process. The models and metrics for three predators (tuna, terns, and dogfish) suggest that they are generally insensitive to a broad range of herring abundance and so generally insensitive to the range of control rule alternatives here. While some control rules considered during the MSE did induce negative consequences for the predators, the reduction in herring abundance had to be relatively extreme, and such control rules have generally been eliminated from consideration (e.g. variants of biomass based with 15% restriction and constant catch). Furthermore, most predators in the region are generalists and the ecosystem has experienced extreme lows in herring abundance in the past, and growth and reproduction of some predators was better then than in recent years with higher herring abundance. While the data and modeling we currently have suggest an ecosystem robust to a broad range of herring abundances, there may be ecosystem services that can be gained by maintaining relatively high herring abundance. Therefore, including a relatively broad range of alternatives would enable the Council to use the MSE analysis to evaluate the potential tradeoffs of control rule alternatives on different aspects of the ecosystem. 2. Input on 1 vs. 3 year alternatives The PDT clarified the main differences between an annual biomass based control rule and a three year control rule. If the Council agrees with the PDT recommendations, specifications would be set similar to how they are now, every three years. An annual biomass based control rule would set ABC one year at a time, based on an updated assessment of the herring stock. This approach requires more resources than are currently dedicated to the herring assessment and management process. Each year herring biomass would be estimated based on updated fishery and available data. Based on the updated estimates, the Council would develop annual specifications. The PDT does not believe this alternative is feasible given current resources; therefore, a modified alternative is recommended below (a three year biomass based approach that uses an annual application based on the most recent herring assessment and short-term projections). A three year biomass based approach would set ABC at the same level for three years (consistent value in MT for three years straight). Specifications would be set for three years at a time using the most recent herring stock assessment information available. In terms of timing, this is similar to how herring specifications are set now, every three years. 10

11 A three year biomass based approach with an annual application, specifications would be set every three years, but ABC would not be the same value. Each year the ABC value could change. ABC would be set each year based on the most recent herring assessment and short-term projections. The short-term projections would apply the selected ABC control rule in each projected year. This process is essentially identical to what is done for specification setting in many other fisheries in the region. For example, some groundfish stocks not in rebuilding plans use 75%Fmsy in short-term projections to specify annually varying ABCs. The PDT recommends that the Committee/Council replace annual biomass based control rules with this approach instead. The performance of several examples from the MSE has shown that most control rules perform similarly when using an annual biomass based control rule or a three year control rule. Generally, the performance for most metrics slightly degrades when switching from an annual to a three application, with the slight costs of using a three year application coming to the benefit of short-term fishery stability. Given the robust nature of the control rules to annual or three year applications, the Council could choose the general control rule shape (i.e., a set of biomass based control rule parameters) in Amendment 8, but then choose separately whether to apply the control rule annually or in three year blocks during each specifications cycle. The long-term performance of switching between annual or three year application would likely fall within the combined range of uncertainty for the annual and three year block performance for the given control rule in the MSE. For example, the minor differences are displayed in Figure 6 for Strawman A and Strawman B. 3. Input on adding text to Amendment 8 about rebuilding requirements Regardless of the chosen ABC control rule, deviations from the control rule may be required to meet the legal requirement of the MSA. Should application of the control rule result in the probability of overfishing exceeding 50%, then the ABC will have to be reduced until that legal restraint is met. Should the stock go below 50%SSBmsy and enter a rebuilding plan, two outcomes are possible: 1) The ABC control rule results in an ABC that does not meet required rebuilding time frames and so the ABC would be reduced to meet the legal restraint; or 2) The ABC control rule results in an ABC that is lower than required for rebuilding, and this would require choosing between an ABC the meets the minimum rebuilding time frame or the lower ABC produced by the control rule. The PDT recommends that Amendment 8 should include details about which approach is preferred if biomass falls below 50% SSBmsy and specify how and when the Council may deviate from the control rule in the final description of CR alternatives. In addition, if the performance of a control rule deviates drastically from what was anticipated (likely due to large changes in the true system dynamics), it may be useful to have a backup plan that can be temporarily applied until the MSE can be updated and re-evaluated. The PDT can develop draft language for the Council to consider for these issues in the coming months when the document is being prepared for public hearings. 11

12 Figure 6 Comparison of using annual (black) or three year (red) biomass based control rules for Strawman A (top) and Strawman B (bottom) 1 vs. 3 year (Strawman A) Minor differences in tradeoff of yield vs. biomass Some reduction in yield for 3 year compared to some addition of biomass for 1 year option. 1 vs. 3 year (Strawman B) Minor differences in tradeoff of yield vs. biomass Some reduction in yield for 3 year compared to some addition of biomass for 1 year option. 12