The negative association of sea lice from fish farms on recreational fishing catches of Atlantic salmon
Abstract
The question of whether and to what extent sea-louse (L. salmonis) infestations from salmon farms influence wild Atlantic salmon survival has been subject to sustained scientific debate and political controversy. Documenting the population-level effects of sea lice on wild salmon remains inherently challenging. We employ comprehensive sea-lice data and recreational catch data from Norway to assess the impact of farm sea-louse infestations on wild salmon catches in different production areas (PAs). Our analysis finds a significant correlation between declines in wild Atlantic salmon catches and increasing amounts of adult female sea lice per km2. The effect is most pronounced in PA 4 on the west coast of Norway, an area within the government’s ”traffic light” management system where out-migrating salmon smolts are deemed to face high exposure to louse-induced mortality risk. Our model predicts below-average catches when the total sea louse load exceeds the government’s limit of 0.1 average adult female sea louse per farmed fish within some production areas. Furthermore, our results indicate that the risk of below-average catches increases by approximately 47% when salmon farms exceed this limit (estimated risk ratio of 1.47, 95 % CI [1.10, 1.96]).
Synthesis and Applications: Our study expands the existing body of evidence demonstrating a negative association between fish farming and the ecosystem services provided by wild salmon stocks. It has important implications for aquaculture management. First, it shows that farm sea lice directly affect peoples, rights holders, and interest groups, such as landowners and anglers, who rely on viable populations of wild salmon. Second, it suggests that a >0.1 sea lice limit may be insufficient to prevent subpar catches in some areas. Furthermore, in the context of allowing further growth in farmed salmon biomass, setting absolute sea-louse limits for entire production areas may be a more effective regulatory instrument than setting average lice limits per farmed fish. Our findings contribute to advancing a scientific basis for setting appropriate louse limits on farm and area scales.
https://doi.org/10.5061/dryad.905qftttr
Description of the data and file structure
The dataset consists of 6 columns of data:
- Catch: The dataset comprised official government statistics that documented the nominal river catches of adult Atlantic salmon, caught by rod, spanning the period 2006 to 2022 (Norway, 2021). Catch statistics offer insights into total catches in terms of both numbers and kilograms, encompassing catch and release data from 2009 onwards. As a substantial proportion of released fish are not recaught (Thorstad et al., 2020a), we used total catch as a metric to describe annual riverine catch. The age of the fish was inferred from the size categories that were registered in the catch data, where a fish < 3 kg was assumed to have spent one winter at sea (1 SW), and one between 3-7 kg was assumed to have spent two winters at sea (2 SW) and one heavier than 7 kg was assumed to have spent three winters at sea (3 SW). The absolute number of caught salmon per year varied among the SW classes and differed across PAs (see Table SI 1 no. 3 and Fig. 2 a). To facilitate comparative evaluations of catch trends across PAs, we aggregated the catches for each SW class from all rivers within each PA. The aggregated values were then standardized (subtracting the mean and dividing by SD) across the period examined. This standardized variable resulted in a normally distributed continuous variable that expresses both positive and negative values, signifying years when catch levels for each specific SW-class within a PA deviate from the average.
- Lice: The estimated total louse level at individual sites was aggregated to calculate the cumulative amounts of lice within each corresponding PA for each year. To facilitate a comparison of relative lice levels across PAs, we factored in the size of PAs and computed the total number of female sea lice per km2, denoted as the farm sea-louse load.
- sw: This classification is based on the number of winters a salmon spends at sea before returning to freshwater to spawn (1SW, 2 SW, or 3SW). The sea-winter class (SW) of the fish was inferred from the size categories that were registered in the catch data, where a fish < 3 kg was assumed to have spent one winter at sea (1 SW), and one between 3-7 kg was assumed to have spent two winters at sea (2 SW) and one heavier than 7 kg was assumed to have spent three winters at sea (3 SW). The absolute number of caught salmon per year varied among the SW classes and differed across PAs
- pa: PAs serve as delineated management zones within the traffic light system (FOR-2017-61, 2017), and are strategically demarked to minimize cross-contamination and infection between neighbouring production areas.
- lice_year: The year the smolt migrates from the river and risks being infested with sea lice
- catch_year: The year the salmon are caught in the river
Sharing/Access Information
Data was derived from the following sources:
- Barentswatch: https://www.barentswatch.no/nedlasting/fishhealth/lice?lang=no
- Statistics Norway, Table 08991: https://www.ssb.no/statbank/table/08991/
- Lusedata: https://lusedata.no/statistikk/excel/
The data pertains to the total number of sea lice (L. slamonis) in salmon farm production areas and wild Atlantic salmon catch statistics for Atlantic salmon in Norway.