This section focuses on auxiliary information of importance for a complete evaluation of the cur-rent stock status. In particular, we highlight information about diseases and other factors that may affect development in stock status, but which are not fully taken into consideration in the current modelling. Likewise, weaknesses in input data and/or difficulties to take into account certain river-specific issues in the modelling might affect the precision of status evaluations, and in the worst case introduce biases. Such shortcomings in the current assessment model are also discussed in this section.
4.4.1 Potential effects of M74 and disease on stock development
Many of the M74-fluctuations seen since the early 1990s have tended to last for some years before changing in direction (Figure 3.4.1.3). After a period with very low M74 abundance in 2011–2015, mortalities increased to higher levels in 2016–2018. In 2019, M74 related mortalities decreased considerably, and mortalities among offspring hatched in 2020 were even lower. The latest thia-mine analyses of eggs spawned in 2020 indicate that M74 mortality among offspring is predicted to decline to close to zero in 2021. Despite the recent positive development, the future occurrence and development of M74 is difficult to predict, which introduces uncertainty in forecasts of the development of salmon stocks. The disease outbreaks reported in several rivers in recent years (Section 3.4.4) is also a concern for the future. The cause(s) of the disease is still unknown, and to accurately quantify the amount of affected or dead salmon in a river appears difficult, if at all possible.
To quantify the effects of health issues among spawners on the recruitment in rivers is difficult.
Existing information indicates that M74 or disease among spawners mainly affect number of eggs deposited or hatched or the number of dispersing fry. That is, losses seem to take place before the offspring reach stages with highest density-dependent mortality. Therefore, a stock with high status is expected to show more resilience against various events that negatively affects early reproduction (i.e. from egg deposition to dispersal of fry), because these effects may partly be compensated by reduced density-dependent mortality among the offspring. In contrast, weaker populations are not expected to have similar ‘buffers’ against such losses.
Average salmon 0+ parr densities in many rivers decreased in 2016–2018 compared to the histor-ically high densities observed around year 2015. In 2019 and 2020, parr densities again increased in many rivers. Part of these fluctuations may be explained by generation effects, i.e. variation in year-class strength among spawners, but mortality due to M74 and/or other disease outbreaks is likely also part of the explanation. Compared to other rivers, the very low parr densities recently observed in Vindelälven and Ljungan are exceptional. In Vindelälven, the average 0+ density dropped drastically, from ca. 40 parr/100 m2 in 2015 to only one parr/100 m2 in 2016, and re-mained at very low levels until 2019 (Table 3.1.2.1). The decline likely reflects a combination of factors. In 2015, only 790 females were counted in the Norrfors fish ladder, which represented 18% among MSW salmon and 11% of the total spawning run (if assuming 6% females among grilse). In 2016, the number of females counted was higher (2741), but a large proportion of the salmon passing the ladder had severe skin problems (fungus infections) and many died soon after having been counted. Female numbers again decreased to 908 in 2017 and 728 in 2018, which represented only 32% and 26%, respectively, among MSW salmon. There are no observa-tions of such skewed sex ratios in the sea or at the river mouth of Umeälven, or in other rivers.
Hence, the recent disease problems in Ume/Vindelälven seem to have prevented particularly females from reaching the spawning areas.
In 2019, the number of MSW Vindelälven salmon counted at Norrfors increased significantly. In total 3389 females, representing 33% of the MSW salmon, passed the counting site. Salmon still expressed symptoms of having health problems, but these were mainly observed during the early part of the migration period. The increase of female MSW salmon in 2019, resulted in a pronounced increase in densities of 0+ salmon in 2020. In 2020, the number of ascending MSW females increased further, indicating that the negative trend in recruitment observed in recent years has likely reversed into recovery phase. Furthermore, as described above, the M74 situa-tion has generally improved with low fry mortalities in 2020 (Table 3.4.1.1) and even lower pre-dicted mortalities in 2021, which will likely improve possibilities for recovery of this river stock.
Also, in Ljungan average 0+ salmon densities in 2017 and 2018 were exceptionally low (<1 parr/100 m2) compared to other rivers in Gulf of Bothnia. There was a slight increase in 2019 and
2020, but the abundance of 0+ parr is still low compared to that of the preceding years (average density of 61 0+ salmon in 2014–2016; Table 3.1.3.1). Notably, the collapsed parr density in 2017 followed after a year with many dead salmon observed in the river, combined with a high ex-pected level of M74-mortality. The very low parr densities in Vindelälven (2016–2019) and Ljun-gan (2017–2020) are expected to result in a successive reduction in smolt production from 2019 and a few years onwards, affecting pre-fishery abundance of salmon from these two rivers from 2021. Because of the exceptional situation for these two rivers, local fishing restrictions, aimed at protecting ascending spawners in the estuarine sea during upstream migration, were enforced in 2019 and were in operation also in 2020. Most likely, these fishing restrictions will continue in 2021.
Although the FLHM cannot in its current form incorporate all details of the specific events af-fecting salmon stocks in Ume/Vindelälven and Ljungan, their consequences for recruitment are incorporated mainly via the time-series of smolt production (including predictions of the near-future production) based on parr densities fed into the river model, as explained and shown in the Section 4.2.2. Also, the recent low success of females to reach spawning grounds in the Ume/Vindelälven is incorporated, but currently there are no methods for predicting the future development of health problems.
4.4.2 Biases in stock status evaluations
The precision in status evaluations of individual river stocks depends to a large extent on the amount of available data. Data from several life stages (parr densities, smolt numbers and num-ber of ascending spawners) and long time-series increase the possibility for an accurate status evaluation, whereas status evaluations of river stocks for which only information on parr densi-ties and/or short time-series is available becomes more uncertain. Also, river-specific factors may introduce uncertainties and/or biases in status evaluations. Migration obstacles, for example fish ways at dams, affect migration possibilities and/or survival of spawners and smolts to a varying extent. If not accounted for (e.g. because of lack of information), such factors may introduce bi-ases in status evaluations. For most stocks included in the FLHM, status evaluations are thought to be reasonably accurate without any severe biases. A few exceptions exist, however, among which particularly Testeboån and Piteälven stand out. A common denominator of these two riv-ers is the occurrence of dams which (to a largely unknown extent) affect migration possibilities and survival of both upstream and downstream migrating salmon. Weaknesses in input data and difficulties to take into account such river-specific issues in the modelling of these two stocks are discussed in more detail below. Although the status evaluations for Testeboån and Piteälven seem to be particularly affected by problems related to migration obstacles, similar issues may at least to some extent exist also in other rivers.
Testeboån was included in the FLHM for the first time in the assessment carried out in 2019. As described in ICES (2019), the PSPC posterior was heavily updated downwards, which resulted in a surprisingly high status of this new wild salmon river, given that salmon parr densities are still comparably low in substantial parts of the river system. The updated PSPC was thought to result from the omission of spawner count data at that time. In 2020, the FLHM was not updated.
This year, spawner count data for years 2016–2020 have been included in the model for the first time, but the PSPC posterior was again heavily updated downwards, resulting in high estimated stock status, similar to the 2019 assessment results. Expert opinions on PSPC in combination with empirical data on the amount of out-migrating smolts, indicate that smolt production in recent years has fluctuated between 20–40% of the PSPC, which clearly deviates from the assessment results suggesting that current smolt production has already approached the river’s production potential.
The reason for the biased results is still not fully understood. A possible explanation is that the time-series on spawner counts is still too short (2016–2020) to provide enough information on the stock–recruit relationship in the river (there are currently only two spawner-recruit data pairs, since only offspring from reproduction in 2016 and 2017 have so far (2020) left the river system as smolts). The possibility for ascending spawners to find their way up and past the power plant in Strömsbro (where the fish counter is situated) to reach the main reproduction areas probably varies considerably between years depending on variation in water flow and op-eration of the power plant. Thus, recruitment of smolts may to a large extent be dependent on migration possibilities rather than the absolute number of spawners that entered the river mouth a few years earlier. The FLHM may interpret this apparent lack of correlation between ascending spawners and subsequent smolt production as if stock status is high (a smolt production close to PSPC). If so, the posterior PSPC may then be updated downwards to match smolt abundances backed up by empirical data. It is also possible that a lack of flexibility in the FLHM when it comes to stock-specific differences in vital rates may affect estimation of stock–recruit parame-ters for river stocks with few stock–recruitment observations. Spawner counts in 2018 (n=22) and 2019 (n=177) represent rather extreme values from a historical perspective, and the resulting smolt production in 2021 and 2022 will hopefully provide useful information on the stock–recruit dynamics, which, in turn, may result in more realistic estimates of the production potential and stock status in the coming years. Until then, status evaluations and projection results for Testeboån must be viewed with caution.
In the 2019 assessment, the modelling of Piteälven was changed so that observations on spawner counts were used directly in the FLHM instead of using them to produce smolt production priors as earlier. The reason for this change was to avoid making assumptions about stock–recruitment parameters outside the model when converting from spawners to smolts. As a consequence of this change, estimates of spawner and smolt abundances as well as stock–recruit parameters were significantly updated (higher stock–recruit steepness, lower PSPC), resulting in changes in status evaluation as compared to the previous year’s assessment. Based on fragmented inde-pendent data currently not used in the model, there is a concern that the status for Piteälven is biased upwards. The reason behind this bias is not known, but similarly as for Testeboån, it may be partially explained by insufficient flexibility in modelling of vital rates and between-river var-iability in the FLHM, which means that smolt-spawner survival is driven by data from other, not necessarily similar rivers. It is also possible that migration problems at the dam at Sikfors (located below the reproduction areas) introduce a similar phenomenon as in Testeboån, i.e. that annual variation in recruitment of parr and smolts to a large extent depends on varying migration pos-sibilities for spawners in the river rather than the absolute number of ascending spawners at the river mouth. As discussed above, the resulting weak correlation between ascending spawners and subsequent recruitment may be interpreted by the model as if the production level is close to PSPC. Although the working group has planned to evaluate the way Piteälven is handled in the FLHM and explore alternative modelling options, this work has not yet been carried out due to time constrains. Therefore, status evaluations and projection results for Piteälven should be viewed with caution.
Independent information, such as river catches and results from genetic mixed-stock analyses from the Main Basin (ICES, 2014), indicate that the smolt production in Ljungan is likely under-estimated by the assessment model, which may also affect status evaluations to an unknown extent. The main reason for the bias is that Ljungan is difficult to electrofish and it is unclear to what extent electrofishing data represent the true abundance of salmon parr. As only electrofish-ing data is currently available from this river, the plan is to start countelectrofish-ing smolts in the near future. Until then, smolt production estimates and status evaluations of this river stock should be viewed with caution.