Wild Salmon Strategy Science and Evidence Board advice to the Delivery Group: recovery stocking of Atlantic salmon into weak populations

Summary

1. Recovery stocking seeks to rebuild populations of wild salmon that still exist in a river or tributary, are considered to be under threat of extirpation and where the numbers of fish are far below the Conservation Limit (definition from the North Atlantic Salmon Conservation Organisation (NASCO)). The science advice specified here relates to the application of stocking strategies to populations at high risk of extirpation in the face of presently increasing marine mortality. It is acknowledged that there is inevitably uncertainty over whether or not a population would become extinct in the case of no intervention.

2. A key step before any intervention is to identify the pressure(s) driving the population decline. Whilst acknowledging concerns regarding poor marine survivorship, in cases where salmon populations are demonstrably or perceptibly weakened by degraded habitat or obstructions to migration, then actions to improve the habitat or alleviate obstructions are required rather than stocking, because the latter invariably entails risks of damage.

3. In cases where salmon populations are weak due to poor survival at sea, it is important that presumption against stocking is considered as an initial option because this would potentially enable the natural process of population genetic responses, including straying and colonization of salmon incoming from other populations. If the option of not initially intervening is followed then there is inevitably a risk of extirpation (local extinction), in which case stocking may be appropriate for rebuilding a population. “Gene banking” may be considered as a precautionary option for retaining local genetic traits for use in population rebuilding should extirpation occur.

4. Of the stocking methods considered by the SEB, all have risks and require further scientific investigations. No substantial evidence is available to indicate that benefits generally outweigh risks of redistribution of eggs or redistribution of fry; these methods are not recommended for general application. Kelt reconditioning can be beneficial but is resource intensive and logistically highly demanding. Smolt-to-adult supplementation (SAS) is a unique approach in offering the potential to maintain a remnant population in a modified, albeit manipulated, form rather than allowing extirpation during sustained increase in marine mortality. SAS is also resource intensive and logistically highly demanding.

5. A key management decision is required by the Delivery Board as to whether to recommend promotion of the application of SAS or, instead, an approach of allowing river populations to persist unperturbed (but with a risk of extirpation) with potential for undergoing natural recovery and adaptation. If SAS intervention is recommended, then there will be a requirement to specify an appropriate threshold of population strength at which action may proceed. Such might entail annual assessments either of reproductive adults or emigrant smolts.

6. If intervention is applied, it has been recommended that recovery stocking should be used only as a temporary measure, during which the factor(s) limiting a population size are addressed. However, if SAS is used in response to prolonged high marine mortality, it is not clear that there is an appropriately and sufficiently well-defined and recognisable short-term end-point. A recovery target population (reproductive adults and/or emigrant smolts) should be identified a priori, and this used as the determinant for cessation of SAS intervention. Substantial resources and commitment over years to decades will be required to justify the additional risks of SAS over the option of doing nothing, whilst being prepared for rebuilding a population should extirpation occur.

I) Introduction

The Wild Salmon Strategy Science and Evidence Board (SEB) provides scientific advice commissioned by the WSS Delivery Board, which is responsible for developing and deploying management actions in support of Atlantic salmon conservation in Scotland. Outcomes are summarised here of the detailed considerations of the potential application of stocking that has arisen specifically because of the continued decline in populations of wild salmon. The available science is used to identify risks of each stocking approach (set out in Appendix 1), to highlight situations where the SEB concludes that stocking should not be deployed, and to set out options for developing stocking policy.

Stocking is defined by the North Atlantic Salmon Conservation Organisation (NASCO) as the deliberate release of Atlantic salmon into the wild at any stage of its life cycle for any purpose. Introductions of salmon into Scottish rivers are regulated by District Salmon Fisheries Boards and, in cases where no Board exists, by Scottish Government. DSFBs rely on Scottish Government for licencing the removal of brood stock in some circumstances as part of their stocking activities.

There is abundant evidence that some stocked salmon may survive and return to a river as adults. However, stocking generally involves risks of damage to populations, defined here as groups of salmon within a geographically bounded region, such as a tributary or whole river. In particular, stocked fish may displace wild salmon and there can be genetic consequences of artificial rearing, with potential for long-term negative impacts for a salmon population. As set out in the revised NASCO stocking guidelines, stocking of salmon into healthy populations is generally counter-productive and not recommended. However, as stocks of salmon continue on a long trend of decline, an increasing number of populations in small rivers and tributaries are decreasing to lower levels than have been recorded previously. In consultations as part of review of SG stocking policy, the question has arisen as to whether stocking, as a conservation tool, can be applied beneficially under such conditions. This is termed a type of recovery stocking by NASCO and applies to populations which are “considered under threat of extirpation and far below the conservation limit of the home streams”. Hence, there is not a precise recommended threshold at which recovery stocking might be applied, but numbers of smolts produced would be very much lower than would be expected if there was a healthy number of spawning adult salmon. For simplicity, populations at such a level are here termed weak populations. In practice, the setting of a threshold for action would depend on whether a form of recovery stocking is deemed to be appropriate as part of an integrated policy of river management. If stocking is recommended then a clear threshold of a population’s strength below which action could commence would be required.

II) Categorisation of weak populations

Salmon populations may be at risk of extirpation and far below the conservation limit of their home streams for several reasons. A first consideration is to establish why a population is at a weak level. There are four distinct causes, which are not necessarily mutually exclusive:

1. The population persists at a naturally low level due to a habitat bottleneck. This may include, for example, scarcity of suitable spawning habitat, or natural processes including availability of space and food and losses to predators.

In this case, recovery stocking of any form is not recommended.

2. The population is weak due to man-made obstructions to access by salmon as smolts and/or returning adults.

If so, then in many cases passage of salmon should be facilitated, for example by alleviation of obstructions, removal of weirs or installation of fish passes. Recovery stocking is not recommended.

3. The population is weak because of low carrying capacity due to habitat degradation within the spawning/rearing areas.

If so, habitat within the stream/river should be improved, for example by the reduction of diffuse pollution, increases in fast-flowing riffle and glide habitats, or management of invasive non-native species. Recovery stocking is not recommended.

4. The population is weak due to a progressive decline in numbers of spawning salmon associated with increasing marine mortality.

The SEB has considered whether, in this case, recovery stocking could be recommended for the purpose of sustaining the population. This might allow (i) improvement in conditions at sea, and/or (ii) time for the salmon population behaviours, physiology and the associated gene constitution to respond to changes in the environment.

III) Options for recovery stocking into populations at a low level due to a progressive decline in numbers of spawning salmon associated with increasing marine mortality.

Option 1: Do not stock salmon unless extirpation occurs

The first, and most important, option for consideration is to do nothing. It is not possible to determine in advance whether a declining population of salmon will become extinct or will persist due to natural adaptations to a changed environment. Although groups of salmon may become locally adapted to the area of river to which a large proportion of individuals home to spawn, there also is straying among distinct populations that are part of a larger metapopulation. This natural process is important because frequencies of certain traits (including sea age at maturity and return migration) within some populations fluctuate substantially over decadal timescales. Hence, it may be that some local populations naturally decrease as the environment changes, and subsequently strengthen again as gene frequencies alter due to colonising strayers. In this case, attempting to conserve a declining population by stocking with locally obtained salmon may obstruct the evolution of a population that would have adapted to changed conditions at sea and/or in the river.

If the option of doing nothing is followed during a sustained period of severely limiting environmental pressure, then in time there is a possibility of extirpation (the local population becomes extinct). If extirpation occurs, then stocking may be considered to attempt to re-establish a population that is more viable under the prevailing environmental conditions at sea and in the river. The question of how a regenerated population would be developed is beyond the scope of the present advice and would require further substantial consideration. However, in principle, it may involve using salmon from populations that have adapted to, or remained resilient to, changes in marine conditions combined with genes that are adapted to local freshwater conditions. Hence, “gene banking” as a precautionary measure may be important to preserve original genetic material including specific local freshwater adaptations.

Option 2: Intervene and stock salmon

Several types of stocking have been considered by the SEB:

a) Kelt reconditioning.

This approach involves the in-river capture of brood stock prior to spawning, their prolonged maintenance in captivity (often beyond the natural lifespan of mature salmon), artificial mating and introduction of their offspring back into the river population of origin over multiple years. The approach needs to be applied before a population is so weak that removing broodstock from the wild population would comprise a significant proportion of the natural spawning cohort.

In this case, the potential successful application can outweigh the risks (Appendix 1), but only where suitable facilities are available, alongside skilled and experienced personnel. It also requires access to appropriate levels of funding and should be considered only where numbers of broodstock to be removed are small in relation to population sizes.

b) Interception of returning adult salmon, artificial insemination and stocking to change the local distribution of eggs, and bypass supposed high mortality in natural redds or limited fry rearing habitat.

This approach involves intercepting adult salmon and rearing their artificially inseminated offspring in a hatchery before introduction into the wild. Where the method has been used to redistribute eggs - due to patchy spawning locations, and tested with scientifically robust monitoring - it did not deliver predicted benefits. The approach has also been advocated to avoid in-redd mortality due to low oxygen and supposed loss of salmon redds in high water flows; however, available evidence suggests that salmon are well adapted to survive extreme floods in naturally mobile riverbeds. The approach has substantial risks and, even if successful, would have only short-term benefits for a declining population. It therefore is not recommended unless clear robust scientific evidence is produced to indicate that it would be successful over the long term if applied in any specified situation.

c) Redistribution of fry from high to low density areas of streams.

This approach involves electrofishing salmon fry adjacent to redds and moving them to areas further from redds. There are substantial risks (Appendix 1) but little or no scientific evidence of general benefits. The approach has substantial risks and, even if successful, would have only short-term benefits for a declining population. It therefore is not recommended unless clear robust scientific evidence is produced to indicate that it would be successful if applied in any specified situation.

d) Introduce salmon from another population.

This approach is not recommended if there is a possibility of the native population adapting to changing environmental conditions and strengthening. However, it may be applied should the population become extirpated as part of a recovery exercise, potentially including native banked genes (see Option 1).

e) Smolt-to-adult supplementation (SAS).

This approach involves capture of parr or smolts to rear in (fresh or salt water) captivity before release of mature adults (or their eggs or offspring) back into the river. The method is unique in potentially allowing a sustained population to persist in the river even with no return of adults from sea. The approach might allow retention of a distorted population beyond the point that the original population would have become extirpated. It would potentially allow adaptation to change in the environment in the river but not at sea. There are many uncertainties and risks of SAS and there is a potential for rapid and serious damage to weak populations (Appendix 1). Substantial scientific research is required to evaluate the method and alleviate concerns of risk. It also requires access to appropriate levels of funding and should be considered only where numbers of broodstock removed are small in relation to population sizes.

Of these approaches, and in the face of extirpation as a consequence of increasing marine mortality an insufficient time for adaptation, only SAS has potential to maintain a remnant population, albeit in a manipulated form. In this extreme context, the SEB recommends that SAS could be considered by the Delivery Board as a management option for conserving populations. Kelt reconditioning can provide a period of elevated smolt production but requires replenishment of broodstock from returning adult salmon. The SEB recommends that kelt reconditioning can have a role to play in recovery stocking in some situations. However, further research would be recommended to assess performance in the wild of salmon produced by kelt reconditioning.

IV Choosing between Options 1 and 2.

There is no established guideline or policy position to facilitate choice between Option 1 (do nothing) and Option 2 (deploy recovery stocking in the form of SAS and/or kelt reconditioning). The Scottish Code for Conservation Translocations (SCCT) has been advocated by some parties as a set of underpinning guidelines, but it seems to promote a precautionary approach that might be inappropriate for SAS, which remains at an untested and experimental stage. Similarly, the International Union for Conservation of Nature Guidelines for Reintroductions and Other Conservation Translocations states that “where risk is high and/or uncertainty remains about risks and their impacts, a translocation should not proceed”. The SEB notes that salmon stocking was specifically excluded from consideration when the SCCT was drafted but recognises that some of the principles from the SCCT could form part of a plan should SAS be applied.

Because of the risks involved (Appendix 1), the SEB recommends that if there is a policy for application of SAS then the method should first be developed and tested experimentally before general application in open natural systems. In this context, the SEB recognises the potential for research using high quality captive salmon rearing facilities (e.g. at University of Stirling) coupled with isolated river sections currently devoid of salmon, for example due to hydro obstructions, to generate test populations. Such an approach has been employed in other contexts, such as testing the effects of nutrient additions on wild salmon. In the case of SAS, such an approach might test the abilities of reared salmon to survive and spawn successfully after release, for example, and to monitor the performance of any progeny that result. Other aspects of SAS, such as interactions between wild and reared salmon may need to be tested in rivers with robust existing wild populations. Kelt reconditioning is better established than SAS as a stocking tool and might be applied where suitable infrastructure, resource and staff are available and perhaps while SAS is further tested and developed.

If there is a policy for application of SAS to weak salmon populations in nature, then SEB recommends that meaningful monitoring is undertaken for adaptive management to address the many risk areas (Appendix 1). A detailed, transparent and freely-available deployment and monitoring plan should be formulated in advance and adjusted to the satisfaction of the statutory authorising/licensing body. Detailed monitoring and quantification of annual smolt emigration might provide the most effective and objective tool in assessing the success or failure of SAS intervention. The exercise should be subject to annual review and termination if failing to deliver objectives.

As a general principle, the SEB recommends that stocking should be applied only as a short-term measure with a clearly defined endpoint. However, given that the aim of SAS is to preserve a population until conditions at sea return to a previous beneficial state, it is difficult to envisage how to define a population target and meaningfully define a temporal intervention endpoint. If the perception of managers is that the focus of the environmental “problem” lies at sea, then it is highly unlikely that the precise causal factor(s) (ocean temperatures? prey availability? fishery by-catch? disease?) will be known. Therefore, it is important that there is commitment for the substantial ongoing funding and support that would be required over many years/decades. From a simple logistic viewpoint, the SAS approach is highly demanding and inadequate infrastructure, facilities, support and technical expertise would risk damaging the original population (which may have recovered if left undisturbed (as in Option 1)) and for no potential benefit.

V. Conclusions and options for consideration by the Delivery Board.

The SEB process has resulted in some clear recommendations of forms of stocking that should be avoided, and types of weak population where habitat restoration should be prioritised, as outlined above. The only forms of stocking recommended for consideration in weak declining salmon populations are kelt reconditioning and SAS.

The SEB recommends that kelt reconditioning may be deployed provided that the substantial logistical and financial support required is in place and monitoring and adaptive management is deployed to address risks (Appendix 1) where possible. There would be the need to specify a threshold level of population size below which it may be deployed, and consideration given to the levels at which suitable broodstock are available.

With regard to the application of SAS, the SEB recommends that a management judgement is required to select among the following options:

Option 1. Do not stock but invest research effort in considering how new populations could be built if local extirpations occur. (This option risks losing the native population constitution unless gene-banking is employed).

Option 2a. Deploy SAS, but only after successful scientific development. (This option risks the opportunity to preserve populations in a manipulated form while science progresses, but gains benefit of allowing potential natural changes to new marine environments during that time).

Option 2b. Deploy SAS now more generally, or in specific designated locations, with strict management guidelines, including availability of guaranteed adequate financial and logistical support, detailed monitoring and adaptive management. (This option risks losing the potential for self-sustaining populations adapted to changed marine conditions. This option delays a potential process of reconstituting a population that is fit for changed marine conditions whilst retaining only a remnant of a population that would have failed).

If Option 2b is adopted, then there also will be a need to specify a threshold level of population size (mature adult and/or emigrant smolt abundances) below which it may be deployed.