conservation case for introducing THE WALES ROD AND LINE - - PowerPoint PPT Presentation

A statistical summary review of the NRW technical conservation case for introducing THE WALES ROD AND LINE (SALMON AND SEA TROUT) BYELAWS 2017 THE WALES NET FISHING (SALMON AND SEA TROUT) BYELAWS 2017 Presentation by Mike Ashwin, as part of WPWF objections by Chris White, John Eardley & Rueben Woodford

In the short window period (20-30 years) since formal stock assessments were introduced, fisheries scientists, teams and experts have yet to encounter or monitor the longer span climatic and ecological impacts on our migratory stocks We are only now only beginning to see the major 1SW & MSW component transitions emerging and predicted over 50 – 60 year cycles, associated with the North Atlantic oscillation of cooling and warming currents and availability of food sources Our monitoring of migratory life cycle populations is at best incomplete, with little if any conclusions or actions taken for the important marine phase of their life cycle. There are major resourcing issues to overcome and parties at all levels need to work together. Monitoring and stock assessment reporting systems must be reviewed and tuned in to adapt to these trend changes if individual river stock performance data is to be used with confidence to guide and set Regulatory measures

Significant trend changes reported in the 2017 national Cefas, NRW & EA Salmon Stocks & Fisheries and Fisheries Statistics reports “True” exploitation by Net & Rod fisheries ( salmon caught and killed) has reduced by 90% from the early 1970`s (page 23 - ACC/28) In Wales the number of rod caught salmon released annually has improved dramatically from 7% in 1993 to 86% in 2017 (page 27 - ACC/28). CPWF maintain that Anglers are adopting conservation codes and being guided by peer pressure and advice from Regulatory & conservation bodies In the same period Angler days fished for Salmon & Sea Trout in Wales has dropped by 60% from 118,862 (1994) to 47,092 in 2017 (page 17 ACC/28) Yet over the long term combined annual and short term salmon & migratory trout rod licence sales in England & Wales have remained reasonably constant 37,728 in 1994 compared with 35,162 in 2017 (page 15 ACC/28) Marine smolt survival rates are reported have crashed by 90% - WHY ? 1970`s 1SW - 25% & MSW - 15% 2011 – 15 !SW - 2.5% & MSW - 2% (Dee, Tamar, Frome estimates 5 yr mean, page 56 – ACC/28)

CPWF Objections and concerns are presented in four sections. In simple terms the four stages of determining a rivers stock status & implementation of measures where appropriate The report will conclude with presentational issues and a short analysis of alternate modelled systems - summarising CFWF & NWATFCC recommendations made to Defra and Welsh Ministers to improve and harmonize Welsh & English systems and introduce a single conservation

• strategy. Precautionary principles are considered also.
• 1. the principles and methods adopted for setting and monitoring

conservation targets - Conservation limit setting (CL) and influence of salmon stock dynamics and recent River smolt deficits

• 2. identifying and highlighting concerns with the capture and use of

core data for reporting Annual River egg deposition estimates to Conservation limit (CL)

• 3. explaining weaknesses and flaws in the methodology and systems

procedures where the national River Classification model is used to designate River Stock status

• 4. the final decision structure process and measures applied

1. Conservation Limits, salmon stock dynamics and smolt deficits Conservation targets and Biological Reference Points (BRP`s) for managing and maintaining sustainable salmon stocks were adopted in England & Wales for monitored salmon rivers, following transportation of the River Bush statistical model in the 1990`s. Revised model principles being adopted from the 1997 EA commissioned Wyatt & Barnard Technical reports BRP`s and minimum sustainable CL targets (River egg depositions and juvenile recruitment from spawning stock) were set above which, stocks should be maintained in the long term. Management Targets (MT) being a higher management target to aim for. In terms of accurately assessing stocks we cannot physically count  a River spawning egg depositions or how many actually successfully fertilize and develop to juvenile stage. What are used are estimates, either :-

• based on declared Welsh rod angler catches (21 of 23 individual Rivers)

and corrections with total spawning fish estimated using a variable factor expressed as the rod exploitation rate or multiplier

• or for two Welsh rivers, that the “efficiency” of counters or trap is

accurately monitoring a proportion of the “validated” adult run estimate

The conservation model in England & Wales was introduced in the 1990`s and a River Classification 10 year trend model incorporated to this in 2003. The 2003 trend model did not undergone Validation and its use for designating five year forward River “trends” and a Rivers stock status in the proposed regulatory measures is hotly contested and under scrutiny Today the Conservation model uses Sea Survival (Replacement Line) estimates that are out of alignment with current forecasts and rod catch and exploitation estimates that can require significant correcting or in some cases cannot correct for important seasonal factors (e.g. rod effort, river flows, closed season runs) The threats and & challenges that our salmon stocks face are set out in the NRW conservation case BUT the Technical case document does not assess or quantify “Life Cycle” stage impacts or prioritise these in terms of their total impact on

• stocks. Doing so would identify strategies that should be prioritised above others

Rod exploitation (retained fish) would appear well down that list (estimates of 1- 2% of spawning stocks) and CPWF maintain Rod fisheries actually do and can have a positive impact on fish stocks – Anglers guarding stocks against poaching, avian predators, unlawful river polluters, unregulated abstraction and contributing to habitat improvements before even the socio/economic benefits

Salmon population dynamics and use of graphs and trends as indicators of stock health Migratory (Anadromous - river and sea) salmon & sea trout have populations that naturally fluctuate, as do most species Salmon have a propensity to fluctuate in cycles of high and lows that mirror the 4/5 year and life cycle stage of eggs to returning adult spawning maturation On many Rivers Salmon peaks and abundance often repeat in 5 or longer 10 year cycles peaks/troughs, with even longer 50 – 60 year population abundance associated with the transition of different component groups or stocks of that annual spawning run – this is the Rivers proportion or dominance of either 1SW (returning

• ne sea winter adults) and MSW (multi sea winter adults)

Higher stock proportions of longer sea phase MSW make that pattern more complex Yearly graphs – depicting 10 years or more of a rivers estimated egg deposition depicts the “health of stocks” and a trend over the time series is used to indicate salmon health or that “estimated trend” The wide scatter and random nature of some river estimates of egg deposition causes difficulties of interpretation and confidence in these trends and future stock

• predictions. Conversely, more regular River populations with small annual variability

make trend predictions easy & accurate to forecast going forward

 Refer to Flip charts  Depicting time series trends & Actual River egg deposition trends as

• pposed to

Modelled Regressed Linear trends and use of the 20th percentile as expressed as Rivers “regressed 10 year historic & projected linear 5 year forward predictions” What is a 20th percentile ? It is a trend used in a Rivers formal stock assessment compliance that is widely misunderstood and difficult to grasp or explain IT IS NOT as many think - a rivers historic actual performance to CL expressed as its 4 out of 5 years or 80% trend IT IMIGHT BEST be described as the Rivers actual 10 year historic trend lowered or regressed to reflect a higher management target for the River. In effect the Precautionary Approach that is already being applied in the Assessment

And what is a 20th percentile river trend ??? Or a rivers Confidence Limit, higher Management Target , Management Objective , River assessment confidence limits and risk status categories bands defining “the probability of river meeting its Management Objective in 5 years time” ?? Its hugely complex, not understood by most fisheries staff - even at a national level and difficult to be understood or explained because by its very nature the Management Objective is statistically ambiguous and imprecise . How do you statistically define a river that is exceeding its CL in 4 out of 5 years, or 4 out of 5 years on average ? Which is it ? They are fundamentally different. Which 4 out of 5 years are you assessing the River meeting its Management Objective over the long term. The best 4, the last 4 of 5, Any 4 ? And can you say this trend assessment actually depicts the probability of a River meeting its CL - 4 out of 5 years or 80% of the time in 5 years time ????? How were the confidence risk status bands set - do they actually mean anything to fisheries staff or rod fisheries/anglers ? A probability or risk status assessment that uses actual averaged historic attainment to CL that reports an actual statistical MAT performance desirable ? One that everyone understands, is not shrouded in complexity and provides visible risk status targets and measures (mandatory & voluntary) that can be observed and regulated ?

50 100 150 200 5,000 10,000 15,000 20,000 25,000

Adults Smolts (thousands)

Conservation Limit = 5,000 adults Management Target = 6,200 adults Max sustainable catch = 7,700 adults Replacement point = 15,000 adults Max smolt prod = 11,500 adults

An example Conservation model illustrating the Stock Recruitment SR curve (blue line - river phase) and Replacement Line RL (red line marine phase) to derive

Conservation Limit and other targets Y axis - river annual smolt output and dotted lines smolt output CL and maximum smolt output. This graph is highly unusual in that it depicts more tangible spawning adult numbers on the X axis – normally depicted as the rivers annual egg deposition values The RL line is sea survival estimated at 10% for outgoing smolts i.e. 50,000 smolts would actually produce 5,000 returning adults. But a river actually has a capacity from 5,000 adults to produce far more smolts - in the SR curve pristine and production approx 130,00 smolts In the SR curve the River Maximum smolt output of 160,000 smolts (its carrying capacity) would be produced from 11,500 spawning adults

The conservation model illustrating how changing sea survival trends impact

• n demand for increased smolt output beyond its optimum carrying capacity

The strong red line as before 10% sea survival ( 50,000 smolts provide 5,000 adults) As sea survival diminishes and the RL steepens we reach a point at 2.5% sea survival (red dotted line – far left) where the smolt output to meet the same 5,000 adults at CL is now 200,000 smolts. And far exceeds the Rivers optimum or maximum smolt

• utput - 160,000 smolts.

At the higher MT of 6,200 adults this would require 248,000 smolts !

At the example CL target of 5,000 spawning adults the rivers smolt output requirement to meet estimated sea survival losses has increased from 1990`s - 25,000 smolt output (average 20% sea survival) 2000 - 50,000 “ “ “ 10% “ “ 2016 - 200,000 “ “ “ 2.5% “ “ Even with 12,500 spawning adults and a River at pristine juvenile condition the River only has a maximum carrying capacity to produce 160,000 smolts. Sea survival losses are

• utstripping the Rivers ability to compensate even at its
• ptimum carrying capacity

It is widely accepted that sea survival has now reached levels where migratory populations are at risk and may not be sustained even at CL targets.

• 2. Capture and use of core data in reporting Annual River egg

deposition estimates to Conservation limit (CL)

A complex matrix of River datasets and variables are used to produce a Rivers total annual estimate of spawning adults and eggs deposited. These include :-

• Rod catch estimates from angler declared catches (or counter/trap data for Dee

& Taff)

• Corrections applied for angler under reporting
• Rod exploitation rate multipliers for rod catch derived Rivers
• Mortality adjustments for natural River losses and C & R handling
• Age/weight scale calculations to estimate the proportion of 1SW & MSW

component stocks

• proportions of female spawners in each 1SW & MSW class
• Fecundity or female egg depositions calculated from the estimated age/weight
• f rod caught fish

We will look at the first three of these as they form the core data or building blocks for stock estimates and directly relate to angler & fisheries activities

!!! How accurate are current Angler rod catch estimates ? !!! Historically and until 2015 NRW & the EA processed angler catch returns and applied a standard 1.1 multiplying factor to take account and correct what was estimated to be under declared catches (estimated to be 91% of their real value) The anglers who were reporting assumed to have caught little or no catch. Predominantly short term licence holders & juniors and for this reason reported as not materially affecting the provisional total declared rod catches. In 2017 in England & Wales 41% - 14390 of the total 35162 season and short term licences issued did not have an angler catch return received or processed At the close of 2015 a number of River fisheries representatives who compiled reliable independent total river fisheries returns started to report unusually high discrepancies between the provisional reported river angler declared catches and their own. The national estimates appeared to be 20 – 40% below their figures. No definite explanation was provided for this, but the under reporting error continued into 2016 & 2017 and is now believed to be partly attributable to a new national “On line” angler catch reporting system that was introduced. Steps were taken for rod catch uplifts to be applied to most Rivers in E & W to compensate for this. On Average : - 2015 - 1.36 uplift (an additional 36% applied) 2016 - 1.32 2017 - 1.28

CPWF & NWATFCC believe the current annual assessments have been compromised by reporting or system failures that are difficult to quantify. Importantly our own investigations on two major fisheries with reliable long term data show that we believe the low reporting at significant levels may pre date the 2015 adjustments and materially affect CL attainment in those earlier years and the 2016 River byelaw stock assessments and stock status

Table 4 Independent Rod Fisheries catch comparisons to NRW Angler declared catch illustrating variance to NRW Underreported angler catch corrections in River CL estimates

How can rod estimates be improved ??

CPWF & NWATFCC have made recommendations to NRW & the EA that effective penalties should be introduced for anglers failing to report catches. Reporting of paper and On line system need to be integrated so that the national reporting offices at Bristol work to real time information where only non reporting anglers are sent reminders and penalties are issued as required. We have already expressed serious concern that a new licence system, introducing 12 month migratory licenses that can be purchased at any time in the calendar year will cause considerable confusion as anglers are now required to submit two returns for that licence over split years. NWATFCC worked with the EA to self audit their processing system, providing 40 NW angler contacts who resubmitted their 2015 & 2016 returns to the EA to audit their historic data inputs.

At this time we are not confident the cause or causes have been identified or measures are now in place to provide reliable and accurate capture of rod catch data

Do rod exploitation rate multipliers accurately reflect individual river estimates ?

Once rod catch estimates are corrected and with fish assigned in age/weight proportions of 1SW & MSW component stocks, Rod catch is then flexed by individual rod exploitation factors to arrive at the estimated total numbers salmon spawners in the river (this is combined f/m estimates and before mortality losses are deducted) It is widely recognised and reported in the NRW, EA & Cefas Salmon Stocks Assessment of 2017 and Annual reports going back to 2004 that rod exploitation rates do not account for a number of key factors that influence rod catch as the baseline metric for spawning estimates. The EA & Cefas undertook to improve the application of these factors in 2004 and restated their intention to introduce procedures post 2008. However no systems have been introduced to quantify and correct rod exploitation rates to account for significant trends in fishing effort and weather conditions on River catches and spawning stock estimates

The 2008 & 2017 National Background Assessment report states as follows: “Annual compliance with the CL is estimated using egg deposition figures. These are derived from returning stock estimates (Table 27), where such data are available. However, for rivers without traps or counters, the usual procedure for estimating egg deposition involves derivation of run size from rod catch using estimates of exploitation. Currently, these procedures do not take into account annual changes in fishing effort. In years when effort was low – such as the ‘low-flow’ year of 2003 and the foot-and- mouth disease year of 2001 – this approach has probably resulted in rod exploitation being over-estimated on a number of rivers and hence escapement and egg deposition being under-estimated. In wetter years, when conditions are more favourable for angling and fishing effort increases, the opposite is likely to be true. An improved procedure is being developed by the Environment Agency to take account

• f annual changes in fishing effort, as well as partitioning effort between salmon and

sea trout (no distinction is currently made between these species when reporting fishing effort, but see analysis in Section 1.2.4). This procedure will be used to improve assessments in future reports. Many rivers, and particularly some of the smaller catchments on the west coast of Wales, support relatively small salmon stocks and are principally regarded as sea trout

• rivers. Current procedures may also fail to take adequate account of this”.

5000 10000 15000 20000 25000 30000 20000 40000 60000 80000 100000 120000 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

Salmon & S/T Days fished Salmon declared catch Sea trout dec catch

Angler Days fished combined salmon & sea trout Angler Declared Catch salmon red line sea trout yellow line

Welsh River Days Fished and Angler Declared Catch 1994 – 2017 rod effort Does NOT report individual salmon or sea trout days fished

Table 5. Reducing Angler Rod effort - figures supplied by Environment Agency Wales for the Rivers Mawddach & Wnion between 1994 - 2009 to illustrate its trend and impact on reported rod catch

As acknowledged above, these omissions are principally Rod Effort (time & species split), Weather/River conditions and actual numbers of fish that may be “missed” in river spawning estimates when they run the river after the season closes. It is very clear that rod effort has declined significantly and potentially exploitation rates also. These trends need to be reviewed and where necessary flexed annually so that rod catch and total spawning estimates take account of these seasonal changes. Appendix A Table 2 CPWF`s own analysis of species rod effort on individual River show that there is considerable variation in angler effort between salmon and sea trout and that this adds another layer of uncertainty to reported catches and the accuracy of River rod exploitation estimates. This variation reflects the wide diversity

• f river habitats and species populations that characterise Welsh game rivers.

“No two rivers are the same” Catch per Unit Effort (CPUE – number of days per salmon or sea trout caught) is used as a national indicator of rod effort/stock presence in the river but uses an aggregated salmon & sea trout number of days fished. Angling effort by species is not recorded and the CPUE output is questioned. CPWF and NWATFCC have recommended that in future angler returns include separate sections for reporting day’s effort by species.

CPWF note in Ian Davidson Rebuttal evidence - January 2019 (NRW/2R) on failure by the EA & NRW to review River exploitation rates (for Rivers without counters & traps) that a review is now being reported (within the last two weeks) as being underway. It states :- 3.8 A review of the models used hitherto to derive angling exploitation rates on rivers without traps or counters is currently underway. This is being conducted jointly by NRW, EA and Cefas, and is supported by a statistical modeller working with the Game and Wildlife Conservation Trust (“GWCT”). The review aims to develop a single model to apply to all rivers - building on and refining existing procedures with the aim of utilising a version of this model in the 2018 assessment. 3.9 This refined model will draw on recent developments by NRW in this area; namely the exploitation rate model used to derive sea trout returns/spawner numbers from rod catch

• data. The sea trout model responds to annual changes in species-specific fishing effort and

includes a variable to account for in-season flow conditions. 28 The model can be readily adapted for salmon but (pending the above review) hasn’t yet been formally applied to this species. However this will Not Correct historic Assessments contained in the Welsh Byelaw proposals which are acknowledged as being calculated on exploitation rates that are inaccurate.

This is the reason that CPWF wrote to the Minister in April & Sept 2018 requesting that interim Byelaw measures are introduced until new assessment procedures are agreed and in place.

Welsh Dee Chester Weir trap operations & validated river estimates

• how the Dee Angler log Book system & trap operations defines the efficiency of the

counter and the Rivers whole validated run size estimates

• determines and sets rod exploitation rates for other Rivers

The Index River Dee are critical to National stock monitoring, providing the longest England & Wales River datasets and trends. Welsh Dee exploitation rates are used for estimating CL for the Wye & Usk ( using 5 year Dee rolling averages), Dyfi, Mawddach, Galsyn, Dwyfawr, Seoint, Ogwen, Clwyd (10% early year estimates) and other welsh Rivers based regression modelled effect of fishing effort to its exploitation rate. The Dee also provides and sets rate estimates used for NW Rivers. And nationally age/weight and smolt survival trend data Information requests made by CPWF to NRW reveal that rod exploitation variables for the Welsh Dee are flexed annually on total rod catch estimates of 1SW & MSW rod caught salmon based on their proportions to the whole river estimates (determined by the log book/trap operations) NRW in their 2017 Stock Assessment Programme Angler Report indicate that rod exploitation estimates for the Dee (to produce the total run size estimates) only use the core rod catches from anglers using the Dee Log book scheme. It is not reported how many Dee log book anglers use the scheme, what the total scheme annual catch is or whether this is a suitable sample size in relation to what are understood to be substantial numbers of other Dee anglers not in the scheme. In 2017 the national Statistics report 656 Dee anglers declared rod effort and 5835 rod days were fished.

What is the “efficiency” of the Dee trap and estimates of total river run size ?

The Dee trap (count) efficiency IS the reported 20 – 30% proportion of tagged to untagged salmon caught by Dee log book anglers. To then estimate the annual total spawning run size, the total number of Dee tagged fish released from the Chester Weir trap is flexed by this proportion. The assumption is that the reported tagged and untagged salmon proportions caught by Dee log book anglers accurately mirrors the same trap to whole river proportions ! In theory the trap efficiency accuracy should self compensate for rod effort changes and catch success providing the Dee log book anglers :-  are a good sample size i.e. not to small, so that catches from numbers of reliable anglers are not biased to individual angler reported outputs  rod proportions between tagged and untagged fish are not influenced by the methods fished or effort expended by those anglers or that tagged and untagged fish behave identically to anglers “lures” ! The real concern is over the 40% downtime of the Dee trap. What total numbers of salmon run in these periods when the trap is out of action ? - Is this significant and are these missing 40% downtime estimates factored into the final spawning estimates. How is this factor determined, particularly as component trends and weather factor triggers are altering normal migration timing. Without further clarification there is uncertainty of the casual influence of these underlying factors on this important facility for setting Welsh exploitation rates.

• 3. Validation of the national River Classification stock status Model

The current National River Classification model first introduced in 2003 is not validated. There is uncertainty concerning the process of independent scrutiny and peer review of the model and modified national system

• procedures. CPWF and Welsh fisheries are concerned that the proposed

regulatory measures are open to challenge. A national model of this importance to recreational and commercial fisheries, must receive routine validation procedures to provide assurance of the river Classification model data inputs and “prediction” outputs. Had validation taken place, improvements to the model would have been identified and appropriate actions taken at an early stage. We believe these concerns are in tune with leading experts who speak of the need for future change and improvement to salmon stock monitoring and River assessments procedures CPWF has collaborated closely with NWATFCC (North West Angling Trust Fisheries Consultative Council) since February 2018 and with support of the Angling Trust commissioned independent statisticians to undertake a statistical evaluation and analysis of the River Classification model.

Timeline modifications to the England & Wales (EA) for formal Compliance using River Classification models

Before 2003 - Compliance based on a simple historic 3 year average attainment to Conservation Limit – expressed as a River “ PASS, FAIL, or UNCERTAINTY” 2004- First year of using a Rivers 10 year historic trend and extrapolation beyond the current year. Bayesian regression analysis is applied to egg deposition estimates on the assumption that there is an underlying trend.10 year. The method fits a 20th percentile regression line to the data and calculates the probability thst the regression line is above the CL, and that CL will be exceeded in 4 out of 5 years. FAIL is <5% probability of the 20th percentile above CL. Uncertain is 5% > 95% probability and PASS is >95% probability 2007 - Introduction of the formal annual assessment and Decision Structure Process using “Status of Stock” as the probability of a river meeting its management Objective -exceeding CL in 4 years out of 5 2008- First year of introduction of the four banded Risk designations - AR, PAR, PNaR, NAR 2009 - First reference to Management Objective as exceeding 4 out of 5 years on average

The Statisticians Qualified Statement & full Report can be found in Appendix B - River Classification Model. In summary the report and its “Evaluation of the model and its use” confirm there are fundamental flaws and weaknesses, which would have become self evident with independent validation. CPWF, NWATFCC believe a review of the current methodology for Stock Assessments is long overdue. Significantly the draft NASCO 2019 to 2024 Implementation Plan for England & Wales circulated for comment in November 2018 incorporates

• bjectives for such a review within a three year timetable, excerpts below :-

CNL (18)50 NASCO Implementation Plan for period 2019 - 2024 What are the objectives for the management of the fisheries for wild salmon? There are plans to review the current methodology for assessing salmon stocks, as well as the associated compliance scheme and decision structure, and to consider the need for possible improvements (see Section 2.8). The aim is to undertake this within the next three years; this Plan will be updated at that time, as necessary, to reflect changes. CNL(18)50 NASCO Implementation Plan for the period 2019 – 2024 See also: Fisheries Management Focus Area Report for EU-UK (England and Wales) (IP(08)05(rev) [ http://www.nasco.int/pdf/far_fisheries/FisheriesFAR_EnglandWales.pdf

NWATFCC & CPWF have submitted a joint response to NASCO with comments and

• ur recommendations concerning improvements sought to the current methodology

for assessing stocks and expressed our wish to be consulted as important stakeholders with a direct interest. The independent statisticians are able and willing to engage in this process We have received assurance from the England & Welsh - NASCO representative

• ur submission will be considered, and that from the EA , NWATFCC has received

confirmation that we will be invited to participate in the rod exploitation review process The Statistician Statement and conclusions have been forwarded to Defra and NRW, EA & Cefas. Only one response from Cefas and Defra was received by NW fisheries

• n the 25.05.18 to a NW Chairman of the Ribble FA containing a Cefas critique of

statisticians Report. The Cefas critique was forwarded to the Statisticians, who replied and rejected many of the comments it contained. NWATFCC forwarded the Statisticians response to Cefas, EA & NRW and we have received no further communications. NWATFCC who initiated the statisticians report and communicated their report to the EA , have not received the J. Barry report of the 23rd August - NRW/3E. Had we done so the Statisticians and NWATFCC would have responded. We do not accept that linear 5 year forward stock status provides sufficient accuracy and assurance that precautionary action principles are always the correct action.

The national Conservation Limit & Stock Assessment workshop - July 2016

Prior to these events the National Institute of Fisheries Management, Atlantic Salmon Trust and Environment Agency (IFM/AST/EA) held an important national workshop in July 2016 to consider the future of Conservation Limit (CL) and Stock Assessments Leading experts in this field were present and key note speakers gave presentations focusing on the concerns associated with Assessing Stocks and the need for future changes and improvement to the national Stock Assessment process, including use

• f the River Classification Model. NRW experts and specialists were present and

contributed to presentations by Ted Potter of Cefas. Ivor Llewelyn (Director AST) in his circulated summary workshop paper titled: Possible changes to Conservation limits and Stock assessment in England made a number of key guiding observations and recommendations, excerpts include :- Compliance Assessment The present system is complicated and hard to understand, even for fisheries managers and has important flaws.

Its essence is genuinely simple however: the current objective of managers in England and Wales (the Management Objective) is that stocks should not fall below their CL’s in more than 1 year out of 5, or 20% of the time over the long term. To achieve this, the stock size that managers aim to achieve (the Management Target) is set well above the CL; it is typically around 35% higher than the CL. However additional complexity has been introduced into the system, linking trends in river stock performance to predictions of future abundance, in order to forecast stock levels against CL’s in five years’ time. “This makes assumptions about continuing linear stock trends which are untenable”. The figures below show the percentages of rivers in the different assessment categories in the two most recent Annual Assessments, with the percentages forecast five years’ previously (in 2009 and 2010 respectively).

% of Rivers in 2014 & 2015 Actual classifications to earlier 5 year forward status “predictions of status (from 2009 & 2010)

2014 (2009) 2015 (2010)

Not at Risk 0 (9) 0 (19) Probably not at Risk 10 (36) 5 (31) Probably at Risk 67 (39) 45 (34) At Risk 24 (16) 50 (16)

Given the differences between the ‘forecasts’ and the actual figures, the value of this particular exercise is questionable due to the difference between the predicted and actual forecast after five years. Furthermore, the probabilistic reporting metrics used i.e. probably at risk, probably not at risk, etc. are also confusing. The benefits of the simple compliance test (above) appear to have been lost in the attempt to make the process more comprehensive. When evaluating the status of a stock it is nevertheless informative to consider the recent trend in stock abundance as well as the current level of abundance, so alternative ways of doing this need to be considered. One option would be to use a 5 year period of observations, and/or non-linear analysis, which would better reflect recent trends in the stock size; this, though, would greatly increase the uncertainty of the assessment. In terms of the decision structure currently, decisions are linked to the forecast probability of falling within a particular category in five years’ time. Given the weakness of the current forecasts (discussed above), consideration could be given to basing decisions on current assessments, with various decision trees linked to the actual status of the stock i.e. At Risk, Probably at Risk etc. CPWF analysis of the accuracy of NRW 5 year forward stock status “predictions” of Welsh rivers in the four years from 2013 to 2016 shows that there has been a high level of inaccuracy, with :

2013 - 41% correct, 2014 - 41%, 2015 - 27%, 2016 - 18% correct

These stock status “predictions” are the central performance indicator of stock health

• n which the Byelaw measures are proposed and a major cause of concern.

Recent Welsh River stock classifications are incorrect in four out of five cases. Basing legislation on a on five year forward ‘prediction’ of stock index trends that have such poor accuracy and correlation to actual stock events is extremely worrying The letter from Mark Lloyd of the Angling Trust to the Welsh AM Cabinet Secretary

• n 17.09.18 (CPWF/12) outlines specific assessment flaws and refers :-

We have reviewed the CEFAS response to the independent statisticians’ critique of the current methodology, but we have outstanding concerns about:

• validation of the system;
• the ambiguous and imprecise compliance to management objective;
• use of linear regression trend and five year forward extrapolation;
• the impact of changes to rod effort, river flow, out of season runs and imprecise rod

angler catch reporting;

• poor correlation between past predicted and actual forward stock performance.

We understand that it takes time to introduce new methodologies, but we need to see a clear acceptance from government that such a change is necessary and a commitment of funding to pay for staff to develop a new system in which there can be greater confidence. We would be happy to work with EA and NRW staff to develop this in partnership.

!!! How reliable are Five year forward extrapolated stock predictions ? !!! The Wye has a history of meeting CL only once in the last 10 years, with stock status predicted to be PNaR in 2022. The 2016 - 2021 stock assessment chart illustrated below, indicates (red arrow) where 2017 & 2018 compliance is anticipated based on 96% compliance in 2017 and rod catch estimates of 50 – 60% Compliance for 2018 As the statisticians conclude - using 10 year Linear Regression trends with an assumed 5 years additional bolted on trend is not a reliable predictors of forward trend where rivers do not exhibit steady trend patterns

River comparisons of CL attainment and trend performance

2016 2021 showing Percentage of Conservation Limit Attainment %

Compl Predicted 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Compl Mawdach 169 148 99 236 199 199 73 75 106 130 PaR PaR 5 year MAT 176 161 156 130 117 Glasyn 96 146 106 78 122 107 193 104 147 68 PaR PaR 5 year MAT 112 121 121 135 124 Ogwen 302 365 216 347 244 231 112 39 135 101 PaR PaR 5 year MAT 281 230 195 152 124 Conwy 207 212 153 331 200 164 107 76 100 135 PaR PaR 5 year MAT 212 191 176 129 116 Wye 40 59 35 25 50 79 79 43 95 131 PaR PNaR 5 year MAT 50 54 55 69 85

This table illustrates other important Welsh Rivers with Good stock performance

• annual CL compliance and 5 year MAT trend yet designated as PAR with

mandatory measures applied compared to Wye CL performance and PNaR status

We believe there is every reason to question the use of the model in the proposed byelaws and make again our request for an independent stock reporting review

• 4. Decision Structure Process and NRW presentation of River stock health

performance to support its Option 2 measures Salmon - Egg deposition ranking tests. NRW present a case for mandatory 10 year byelaw measures using a combination

• f stock health indicators ranked in order of weighted significance.

a) River stock status classification b) 10 year long term linear trend egg depositions c) 5 year short term annual egg deposition shortfall comparisons to MT Each of these performance tests actually use the same base River egg deposition

• metric. For b) the 10 year linear trend is already assessed in the River stock status

classification test a). Likewise for c) in b) & a) Using a single metric individual river value (annual egg depositions) and presenting this as three separate ranking tests is unorthodox and is questioned by CPWF. Has this been presented in this way to confer added emphasis that all rivers stocks are vulnerable and require additional regulatory protection ? Why was a the more usual performance test omitted measuring a Rivers Management Objective achievement ? i.e. number of years exceeding CL or an individual years actual 5 or 10 year MAT attainment to CL . Is it that a very different picture would have emerged i.e. some rivers would clearly be meeting their MO of exceeding CL in 4 out of 5 years ?

WHY is the critical Decision Structure process NOT included in any NRW, Cefas or S&TCUK Inquiry documents ? CPWF have provided what appears to be the only detail of binding measures Policy for NASCO, Wales & England in its SD/3 – draft NASCO 2019 – 2024 IP CPWF notes - the NASCO guidance that Wales & England adopts is that CNL(18)50 NASCO Implementation Plan for the period 2019 – 2024 2.1 What are the objectives for the management of the fisheries for wild salmon? The current management objective’ (MO) for each salmon river stock when reviewing management actions and regulations is that the stock should be meeting

• r exceeding its CL in at least four years out of five (i.e. >80% of the time)

Further - NRW Annexe 4 - Salmon Stock Management system Explicitly states in its Decision Structure Process for PAR (12 in Wales) and AR Rivers ( 8 in Wales) that after

• 2. Second stage – initial screening for potential options (blue boxes)

This stage screens options appropriate to those rivers that have a <50% probability

• f failing the management objective (NAR & PNaR rivers) taking into consideration

socio-economic concerns and stakeholder support. Management options that would not be supported by stakeholders can be ruled out. One of the possible options is to ‘do nothing’. For rivers where there is >50% probability of failing the management objective (i.e. PAR & AR Rivers), all options must be carried through to the next (evaluation) stage.

The next options stage

• 3. Third stage - option evaluation (purple boxes)

The purpose of this stage is to set out and evaluate options to realise the required changes in exploitation. For rivers where 50%≤p<95% (PAR Rivers i.e where p= probability of failing the management objective and the trend is down and with an annual catch of >20 salmon and C&R rate < 90%, then voluntary C&R will promoted for 1 year. If this fails to significantly improve C&R rates, mandatory C&R or closure of the fishery will be considered. Protected rivers such as SACs (Special Areas of Conservation) are given particular emphasis For rivers where the above criteria apply, except that the annual mean salmon catch is <20 salmon, voluntary measures will be promoted For rivers where p>95% (AR Rivers i.e the management objective is clearly being failed) and with an annual catch of >20 salmon and a C&R rate < 90%, then voluntary C&R will promoted for 1 year. If this fails to significantly improve C&R, mandatory C&R or closure of the fishery will be considered. Clearly NRW are not following the Decision Structure Process for the twelve PAR Rivers or those Rivers with an annual salmon catch below 20 fish - Aafan, Aeron, Ystwyth, Rheidol, Dysynni, Arto, Dwyafr, Llynfni, Gwyrfai, Seniot and others

Why are NRW proposing Option 2 mandatory measures and method restrictions and rejecting Voluntary Options and agreement with stakeholders & Anglers ? NRW Technical Case - Page 13 - Executive summary Option 2 states:- Reduce exploitation by nets and rods through a combination of byelaw controls on rods and nets and / or fishing method control by voluntary catch-and-release fishing

NRW dismiss the voluntary C & R option with the statement Despite good uptake of this by most anglers in many rivers, it is clear that the urgently required reduction in exploitation is not evident. CPWF refute NRW`s statement.

There is vastly improving voluntary C & R observance, (possibly not all rivers) but reported as 86% in 2017 and this can be improved. 90% is a recognised target and is achievable. At 90% C & R and rod catch being 10% of the total spawning run the contribution of mandatory measures is 1% to stocks (2% where rod catch is 20% of the annual run) The contribution of 1-2% is minimal compared to the consequential losses from introducing these measures

Stating further page 96 - in their Salmon preferred option

The preferred option for salmon at present is Option 2 to reduce exploitation of fish in both the net and rod fisheries, through a considered package of measures so that all salmon are released, and have a good chance of survival. These measures will allow many of the social and economic benefits of the two fisheries to be maintained.

Further - whilst Wales and England both use identical River classifications models, NRW are proposing more stringent measures with all 20 Rivers (Wye & Usk excluded) being subjected to mandatory C & R measures for salmon and further method restrictions. In Wales that is 8 At Risk rivers and additionally 12 Probably at Risk classified Rivers plus all other smaller non monitored river with 10 year mandatory measures In England the EA reconsidered their proposals for their 40 monitored Rivers and with updated 2017 assessments, four At Risk Rivers have mandatory measures applied and the remaining 36 (including PAR Rivers) have Voluntary measures CPWF is asking WHY have NRW not observed their own Annexe 4 Decision Structure Process or NASCO guidance contained in the Cefas, NRW & EA Annexe 7 - 2017 Background Report for Assessment of Salmon Stocks and Fisheries in England & Wales

• I. Russel – NRW/4a states - 3.5 NRW’s proposed Byelaws are entirely consistent

with the NASCO guidelines and with a precautionary approach. They respond to the widespread poor status of stocks in Wales, have been developed in accordance with the current national decision structure and are aimed at conserving the resource, which is the overriding management requirement. CPWF do not believe the Byelaws are consistent

Other jurisdictions

Although listing English, Scottish & Irish conservation policies as countries with alternate conservation strategies, NRW have not explored the benefits of these or whether adoption would alter the current stock status classification of Welsh Rivers CPWF has analysed and considered alternative modelling scenarios and presented three alternate River Classification options in comparison to current River status Classification modelling by using historic CL datasets. CPWF Table 8. One being the Scottish style model recently introduced after public consultation based on 5 year historic MAT (with current CL compliance) as one option by the

• Statisticians. The Scottish model, using the same historic Welsh CL datasets

would reclassify eight Welsh rivers into different Voluntary C & R measures. This system would allow an annual review and measures mechanism and has adaption's for river flow in Scotland. We are advised, in I.Davidson`s rebuttal evidence NRW/2R - 3.8 that a review of rod exploitation rate is now underway for E & W Rivers for 2018 assessments . We welcome this development, but ask that predated assessments are reviewed

There are very obvious operational, resourcing and reporting benefits from E & W adopting a common Conservation strategy, with important cross border rivers a part of this single Byelaw approach

Are Precautionary Principles being observed in consideration of the proposed Byelaws ? Ref STC/6 NASCO Precautionary Approach (STC/4) CPWF enclose a detailed review of points raised that confer responsibilities on NRW and conservation organisations that require attention These are contained in the pages following Key Points and concluding

• statements. Example to the

CNL(98)46 Agreement on Adoption of a Precautionary Approach

• NASCO and its Contracting Parties agree to adopt and apply a Precautionary Approach to

the conservation, management and exploitation of salmon in order to protect the resource and preserve the environments in which it lives. Accordingly, NASCO and its Contracting Parties should be more cautious when information is uncertain, unreliable or inadequate. Yes caution is required in using uncertain rod based estimates for applying mandatory measures - particularly as rod catch corrections, rod exploitation rate and methodology reviews are under review.

• The absence of adequate scientific information should not be used as a reason for

postponing

• r

failing to take conservation and management measures. Not where the observed impact is significant and of long duration and effective actions and measures have not been initiated or brought about improvement. Scottish salmon farming and diffuse pollution are two such threats

Key points & concluding statements

An opportunity for anglers to be involved and engaged in a consultation process “with Options” has been missed - Anglers are not the cause of stock failures and for the future of our rivers, catchment parties must agree actions and measures

The real value that rod fisheries bring to managing and improving our fisheries has been

• verlooked and must be recognised to regain the lost support of anglers

Voluntary C & R measures, where good gains have been made in recent years have been dismissed . They present the obvious and most viable Option The byelaws lean heavily on stock performance assessments for Rivers that are uncertain, too complex and are timetabled for review and improvement. The national River Classification model is not validated and open to challenge. Interim measures warrant consideration, where historic data and assessments are questioned CPWF have made improvement recommendations to NASCO, NRW and the Welsh Secretary. These offer scope for further discussions and one path towards agreement. Harmonisation of Welsh & English byelaws must be desirable and with common stock systems and three major cross border Rivers must be a desirable objective The mandatory C & R button has been pressed for salmon, reinforced with a range of method restrictions that will impact heavily on angler participation, enjoyment and engagement CPWF believe Voluntary Option 2 measures were available, and can still be developed with rod

• fisheries. The current Byelaws cross clear national and NASCO decision criteria and pathways.

Precautionary principles place fisheries management obligations that have not been met. The Inquiry can kick start a new direction where parties can work together to conserve stocks

CNL(98)46 Agreement on Adoption of a Precautionary Approach 1. NASCO and its Contracting Parties agree to adopt and apply a Precautionary Approach to the conservation, management and exploitation of salmon in order to protect the resource and preserve the environments in which it lives. Accordingly, NASCO and its Contracting Parties should be more cautious when information is uncertain, unreliable or inadequate. Yes caution is required in using uncertain rod based estimates for applying mandatory measures - particularly as rod catch corrections, rod exploitation rate and methodology reviews are under review.

• 2. The absence of adequate scientific information should not be used as a reason for

postponing or failing to take conservation and management measures. Not if the observed impact is significant and of long duration and effective actions and measures have not been initiated or brought about improvement. Scottish salmon farming and diffuse pollution are two such threats

• 3. The Precautionary Approach requires, inter alia:

a) consideration of the needs of future generations and avoidance of changes that are not potentially reversible; Does this refer to future generations of anglers or fish ? 10 year measures will impact on older anglers and method restriction on the infirm and for all anglers of rivers that cannot be effectively fished with fly or lure. b) prior identification of undesirable outcomes and of measures that will avoid them or correct them; NRW have failed to identify strategies that achieve the same or better goals i.e. voluntary measures – thus reducing the undesirable outcomes of reduced angler participation and revenue losses. The undesirable prospect of reduction in angler support to enforcement and habitat improvement programmes and predation control has not been identified or remedied

c) initiation of corrective measures without delay, and these should achieve their purpose promptly; 1. Failure to address sea survival – E & W have restrictions on inshore salmon farming - no attention to threat to Welsh SAC rivers from emergence of recognized and observed impact of Scottish farms on stocks in migration pathways.. In E & W there are restrictions applied by NASCO 2. Hydro, water pollution and River flow legislation. 3. EA & NRW failure to implement Cefas, NRW & EA commitments to improve catch reporting systems and rod exploitation reviews - i.e. these have compounded under reporting and CL failures. 4. Voluntary C & R measures (mandatory if targets not met) with effective controls should be directed at those rivers with lower conservation compliance – not on all rivers d) priority to be given to conserving the productive capacity of the resource where the likely impact of resource use is uncertain; presumably this is fish as The Resource – but equally it could be the fisheries resource NRW & EA capacity to monitor Rivers and provide accurate whole river spawning data or electrofishing result is diminishing and not being conserved. e) appropriate placement of the burden of proof by adhering to the above requirements. There is clear evidence and NRW admission that anglers are not the cause of “stock depletion” – the burden of proof has not been met. NRW have not identified the scale of fisheries impact on spawning stocks (1 – 2 % being 90% C & R with a rod exploitation rate of 10 to 20% of stocks) or the benefit that anglers contribute to fish stocks from presence on water(poaching)/avian predation/river watch for polluters & unlawful abstraction and contributions to river habitat works

• 8. The application of the Precautionary Approach to salmon fishery management is an

integrated process which requires at least the following: a) that stocks be maintained above the conservation limits by the use of management targets; Some Rivers are achieving CL in the long term, yet are packaged with measures designed for Rivers that are failing. Is this equitable or desirable ? b) that conservation limits and management targets be set for each river and combined as appropriate for the management of different stock groupings defined by managers; They are set for 1SW & MSW component stock groups (1990 SAP`s etc) where CL & MT`s are out of alignment with changing component stock proportions and sea survival estimates c) the prior identification of undesirable outcomes including the failure to achieve conservation limits (biological factors) and instability in the catches (socio-economic factors); Undesirable outcomes mentioned 3.b. As regards the instability of catches then the proposed measures will impact heavily on angler participation and future catch statistics as an accurate angler based rod catch for CL assessment purposes d) that account be taken at each stage of the risks of not achieving the fisheries management objectives by considering uncertainty in the current state of the stocks, in biological reference points and fishery management capabilities; No life stage mortality or strategy to prioritise actions or benefit of Byelaw measures is

• given. The uncertainty of not achieving MO is clearly apparent and the contribution from

the proposed Byelaw measures small, if not negative in terms of wider protection of stocks.

e) the formulation of pre-agreed management actions in the form of procedures to be applied over a range of stock conditions; rod fisheries and anglers are not in agreement with pre-agreed management actions where NO consultation Options have been presented f) assessment of the effectiveness of management actions in all salmon fisheries; f) how has or is this to be conducted - it appears there is NO assessment process of effectiveness of management actions - Voluntary C & R observance has been effective but other prior actions appear to have failed over the long term ? Stock assessment process and accuracy of stock status predictions is identified as needing urgent improvement h) stock rebuilding programmes (including, as appropriate, habitat improvement, stock enhancement and fishery management actions) be developed for stocks that are below their conservation limits. Stock enhancement programmes utilizing wild broodstock has been written out as a viable option. Egg placement in unpopulated streams via Scoty box

• r other juvenile introductions that have been successful in other English rivers has been

dismissed.

• 9. The management procedures for all salmon fisheries could include the following elements:

a) definition of target spawning stock levels in the relevant rivers; b) definition of pre-fishery abundance of individual salmon stocks or groups of stocks

• ccurring in the relevant fishery;

c) utilisation only of the surplus according to a) and b) above; d) socio-economic factors. EA commissioned a national England report on the Socio-economic impact of the byelaw measures – April 2018. It appears no specific Welsh report has been commissioned. The EA report refers to undesirable and significant impact in certain Regions from introducing Mandatory C & R and identifies the NW and Solway Rivers in particular.

• 10. New fisheries targeted on salmon or which could result in a by-catch of salmon should be

subject to cautious conservation and management measures. In accordance with Article 2, paragraph 3 of the Convention, the Parties shall invite the attention of non-Contracting Parties to any significant by-catch of salmon by its vessels.

• 11. Efforts to minimise unreported catches, and to improve estimates of them, are consistent

with the Precautionary Approach. NASCO and its Contracting Parties agree to evaluate and report on progress in this area. E & W – increasing number of unreported licence holders – 59% in 2017 - this is not minimizing unreported catches which have been estimated at 20 – 35% between 2015-8. Current licensing and reporting failures have not been resolved

• 12. ICES or other scientific advisors should be requested, inter alia, to:

a) provide stock conservation limits and management targets for all river stocks; b) advise on the risks of not achieving the objectives of NASCO or its Contracting Parties by considering uncertainty in the current state of the stocks, in biological reference points related to specific management objectives and in fishery management capabilities; c) provide catch options or alternative management advice with associated risk assessments for the fisheries regulated by NASCO and homewater fisheries for all salmon stocks; Alternate modeling systems and rod catch reporting improvements which afford improved accuracy have not been considered. Rod exploitation rate improvements recommended and committed to by Cefas, NRW & EA over many years back to 2004 and identified in NASCO – IP`s have not been actioned and are the cause of significant uncertainty with historic individual river assessments d) advise, in the light of current conditions in the freshwater and marine environment, on stock rebuilding programmes including, where appropriate, habitat improvement, stock enhancement, disease prevention and fishery management actions; f) identify the monitoring and data collection required to better achieve the objectives of NASCO and its Contracting Parties; Simpler and more effective communication of stock reporting data to rod fisheries is required g) advise on the impacts on salmon stocks of existing and new fisheries for other species, and of salmon fisheries on non-target species. No impact assessment or analysis of Scottish salmon farming