From North West Angling Trust Fisheries Consultative Council

North West Angling Trust Fisheries Consultative Council

part of the Angling Trust and AT North West Freshwater Forum
14h September 2019 NWATFCC & CPWF recommendations to drafting of NASCO 2019 – 24 Implementation Plans

re: Improvements to Conservation targets, Stock reporting, Methodology and Decision process

Purpose; A short discussion paper to identify and highlight the scope of improvements to proposing a Quality Assured, open to view system and opportunity to harmonise mainland UK conservation principles and strategy

Validation – Independent validation of existing or new system procedures is required with periodic review to provide quality assurance.
Reason – the present system of river classification and decision structure process first introduced in 2004 (and updates) is not validated or subject to periodic review. The present systems frame recently introduced Regulatory measures and are both open to and currently being challenged.
In implementing National England & Wales Salmon Byelaws, the EA & NRW share common regulatory undertakings framed by NASCO approved salmon conservation objectives, reporting methodology and decision process, yet have introduced quite separate national conservation strategies in relation to fisheries management of stocks. Specifically C & R measures and method restrictions.

Reaching a single Biological Reference Point interpretation of a Rivers spawning target (Conservation Limit, Management Target etc)
Reason – a single BRP will provide clearer focus to meeting and achieving conservation objectives. Conservation Limit and higher river specific Management Target is frequently misquoted and misunderstood in relation to river Management Objectives and incorrectly associated in generic terms with reporting “management targets”.
Scotland and Ireland have undergone this process, defining MSY (Minimum Sustainable Yield) as the single spawning CL biological reference point.

Review and revision of individual River CL`s
Reason – in both the stock recruitment curve (SRC – egg to smolt) and replacement line (RL – marine, post smolt to returning adult) phase, assumptions in setting and defining CL are being questioned as regards the need for updating Salmon Action Plan estimates and targets. The accuracy of RL & SRC model principles and predictions in times of significant trend change has to reassessed, as with decisions made in 2004 to reduce national River CL to lowering marine survival estimates that have significantly worsened since that time. The mechanism of estimating actual individual CL from total wetted area and density (eggs /100m2) has to somehow balance and take a account of waterbodies that are not and unlikely to meeting WQ standards, suffer from degraded spawning/nursery habitat and significant un-quantified predation losses.

Redefining and simplifying Management Objectives and River Status methodology
Reason – set out below are the fundamental weaknesses and flaws with the current River classification methodology – use of the Linear Regression trend and five year forward prediction of probability of meeting Management objectives.

1. What is a rivers Management Objective and is it statistical precise ? Originally defined as a River that meets or exceeds its CL in 4 years out of 5 on average or 80% of the time. Found to be statistically imprecise and has recently been amended in the published 2018 Stocks Assessment Report to exceeding CL in 4 years out of 5 or 80% of the time. The question that is asked is whether the regressed Linear trend actually expresses MO ?
The alternate model being proposed is that a rivers MO exceeds CL on average in the last 5 historic years. This is statistically precise and provides the best mean actual 5 year trend data to an exact % CL attainment. Appropriate river classification management decisions and C & R measures can then be framed around precise real time trend performance.

2. Applying Variable corrections to spawning stock and River egg deposition estimates. Seven variables are applied to the majority of rivers reliant on rod catch returns to quantify final female 1SW/MSW component numbers and their weight/egg bearing capacity to estimate a Rivers total spawning egg deposition to CL. The original 2004 methodology and subsequent annual Assessment reports clearly state the accuracy of egg deposition estimates and predictive trend forecasting required incorporation of revising seasonal variations of rod exploitation on individual rivers, as one of seven variables. In relation to annual reviewing of rod exploitation rate variables, these are specifically rod effort, river flow conditions and run timing/out of season proportions of the run. Rod exploitation rate estimates have not been reviewed and revised nationally in a time series 1994 – 2018 when Salmon Action Plans and CL were originally framed around 2000. Rod effort regionally and nationally has dropped alarmingly to 27 – 35% of 1994 estimate. The Appendix Tables 6, 8 & 21 indicate that reduced rod effort is the major and principle cause of declining rod catch and CL attainment performance. River flows are threatened by abstraction pressures and extreme rainfall events and extended periods of rainfall shortage. These in turn shorten the window of angler participation and perceived favourable salmon running and take conditions. A quality assured system has to place periodic reviews at the heart of its system procedures and with it the accuracy of species rod effort, catch reporting, weight scale determination of components and fecundity. We recommend a single reporting statement of key River variables along similar lines to the Table.22 Information request to NRW of their annual review and revising policy for assessing stocks.

3. What is the 20 percentile of the linear regression trend and is it expressing a Rivers MO ? Hugely complex area of Bayesian analysis of time series trend distribution which few fisheries managers can interpret or explain. In my simple terms the Linear regression trend line is depicting a rivers 10 year historic performance where 80% of a Rivers annual attainment in its moving 10 year timeframe is above its 20th percentile trend line. Concern being that the trend line is not statistically expressing MO (its median or midpoint value in its best 80% years attainment), but as we know a much reduced precautionary trend based on higher management target. The proposed alternate model – 5 year MAT of % CL (Table 24 Option 2 with current model comparisons) simplifies and precisely maps actual river attainment.
4. Should we continue to use current methodology where “predicted” trend forecasts map a 15 year Linear (straight-line) projections ? No, River trends are not linear and if we make that assumption there is a high risk that predictions will fail. Salmon River performance map cycles of abundance, with good years followed by low or lower abundance. Fig. 6, 7 & 8 illustrate how 15 year linear predictions fail with rivers that exhibit longer and stronger salmon cycles. The NW river graphs for the Eden, Ribble & Lune illustrate these trough predictions in comparison with the River Wye in that same timeframe with over optimistic predictions of improving stocks. Rivers with good CL attainment performance over 10 historic years do not become extinct in a short 5 year forward projection window as predicted in several NW Rivers over the years 2014, 2015 & 2016. The use of the proposed 5 year historic average or mean CL River attainment makes no forward prediction of stock performance, relying on most recent real time data.
The model produces the necessary dampening effect in strong cycles and assurance required in decision making to account for seasonal variations that in some years can be significant.

5. Comparisons of predictive to actual recorded status in 5 years time forecasts have been poor. Table. 23 is an analysis of the accuracy of EA & NRW five year forward predictive stock status forecast for the current 64 English & Welsh monitored Rivers in the 2008 – 2017 timeframe, illustrating low accuracy on which regulatory stock management decisions are made

6. Use of log transformed and non log assessment charts – in recent years Log transformed charts have superseded actual scale log assessment charts. The visual scale of actual trend prediction and a Rivers annual CL attainment has been lost in this process. Log scale assessments are clearly masking Linear regression trends that for some Rivers cause unrealistic and extreme predictive stock forecasts. See Fig. 6, 7 & 8 again.
7. Do the four risk status groups and specifically PaR & PNaR status that form the 90% range of probability that a river will meet its MO in 5 years time, provide a good interpretation of a Rivers performance? Rivers that can report an exact % estimate of actual historic 5 year MAT attainment as proposed, will provide clearer and more accurate interpretation of actual trend performance with greater clarity in national/regional trend reporting. Precise Rivers stock assessment classifications bands (e.g. 10% bands) can then more clearly define decision making step procedures. An example of this model classification option with applied Decision process – C & R measures is set out in Table. 24 The Option 2 model was produced in February 2018. In essence, similar to the “Scottish style” grading approach with classification assessments that can be reviewed annually, based on 5 year MAT % CL river performance and visible and easily interpreted by fisheries managers and anglers.
The Option 2 model example guidelines for the 64 English & Wales monitored Rivers are set out below and framed on current CL target values and applied variables and SAP estimates.
Rivers exceeding 5 year MAT of CL target (> 100%) – required to meet 80% C & R target

“ 80 to 100% 5 year MAT of CL target (80 >100%) “ “ “ 90% C & R target

“ below 80% 5 year MAT of CL target (< 80%) “ “ “ 100% Mandatory C & R

Resetting the Decision Structure Process framework – using clear precise river classification status & C & R measures strategies
For England and Wales the Decision Structure Process can be reset in the decision flow chart to a Rivers MAT % CL attainment, whilst maintaining and reviewing the CL principles & targets and applied variables. Annual review of river performance to CL attainment will flow visibly from simplified annual assessments and the focus and effort can shift to ensuring River data (rod catch, rod effort/river conditions/rod exploitation rates, weight, fecundity etc) are clean and accurate.
In tandem that Index River counter/trap datasets are providing outputs that are reviewed and revised to linked reference Rivers and SAP`s.
The Scottish grading system underwent a consultation process that generated a 5 year MAT CL equivalent model, on clear lines with yearly review of river Status and applied measures. The Scottish model accounts for seasonal factors – river flow, out of season runs etc and the Consultation enabled a review and major revision of wetted area River estimates, with MSY defining River CL targets.

Conclusion
These proposals and model suggestion allows UK jurisdictions to maintain existing CL setting and annual attainment to CL and then build in appropriate break points for setting probability or risk of meeting a precise MO as the Rivers actual 5 year MAT performance. Annual revision of status will flow from this and necessary revision of seasonal factors and variable corrections.
The new methodology procedures would remove the Linear Regression trend and 5 year forward predictive forecasting that is widely criticised.
Focus can then concentrate on ensuring rod catch, counter/trap, variable and seasonal data is clean and accurate.
An opportunity might then be realised to harmonise mainland conservation strategies and systems with obvious potential operational and reporting benefits. Mike Ashwin. redfa/nwatfcc 14.09.19