https://doi.org/10.5061/dryad.m0cfxppbg

In this study we take advantage of an extensive database of more than 200,000 angler catch reports from Swedish recreational inland fisheries and relate it to monitoring (fisheries independent) data on the associated fish communities. To this end, we first compared relative species abundance and the trophic level of angling catches to monitoring data across 62 distinct fisheries (lakes) where data overlapped. Next, we assessed size selectivity of these fisheries on European perch (hereafter perch, Perca fluviatilis, Percidae) and pikeperch (Sander lucioperca, Percidae), two predatory focal species in the fisheries. Finally, we explored the effects of C&R practices on selectivity by disentangling angler harvest from angler catch (including released fish) in all analyses. The results provide quantitative and robust estimates of species and size selective harvest in recreational fishing with implications for management of stocks targeted by rod-and-reel fisheries.

Description of the data and file structure

The two datasets consist of monitoring data (monitoring_data.csv) from the national database for standardised survey fishing with Nordic multi-mesh gillnets (National Register of Survey test‐fishing ‐ NORS, 2021) as well as angling data (angling_data.csv) from catch reports submitted to the online fishing license sales platform iFiske AB (https://www.ifiske.se/. Both datasets are fully anonymised and lake names have been substituted with numerical labels. Each row in the datasets represens a reported fish

Monitoring data

• Lake_ID (anonymised numerical label instead of lake name)
• Year
• SpeciesName
• cmklass (standard length in cm)

Angling data

• Lake_ID
• Year
• SpeciesName
• Weight (individual weight in grams derived from either individual or joint reports)
• Report (Individual = the row represents a fish reported with an individual weight; Joint = the row represents a fish reported with a joint weight)
• Fate (the fate of the individual; either harvested, released, or no answer)

Prior filtering

Monitoring data has been filtered to only include survey fishing performed according to the standard Swedish method for freshwater fish. Angling data have been filtered so that reported catches with passive gears (e.g. gill nets, longlines, and fyke nets) were excluded. Data were restricted to the years 2010-2020. Small Swedish lakes are typically managed jointly with additional waters by a single local organisation, and consequently, catch reports from such waters cannot be linked to specific lakes. Therefore, lakes with an area less than 15,000 m2 was excluded from the initial lake selection (number of lakes in the monitoring dataset was reduced from 842 to 128). Remaining lakes were matched to the angling dataset resulting in 113 lakes where both monitoring and angler catch data were available. Among these, 78 were managed by distinct local management organisations, each covering the entire lake (constituting at least 90 % of the water area). We considered each of these lakes as distinct fisheries. Two adjacent lakes, both having monitoring data from the same years, were managed by a single local organisation and consequently, we merged monitoring data and treated the two lakes as a single Fishery in the analyses. At last, fisheries with fewer than 100 reported angling catches were excluded from the selection to improve the overall quality of catch data, resulting in a total of 62 distinct fisheries included in the analyses. Anglers often caught several individuals of the same species and reported the total mass together with the number of caught individuals of the species-specific catch (~50 % of reports were of combined body mass after excluding zero catch reports). Individual body mass from these reports were calculated as averages.

Sharing/Access information

Data was derived from the following sources:

• National Register of Survey test-fishing - NORS. (2021). Swedish University of Agricultural Sciences, Department of Aquatic Resources. http://www.slu.se/sjoprovfiskedatabasen
• iFiske AB (https://www.ifiske.se/) - Data not publicly available

Code/Software

The R script file (Code for Recreational fisheries selectively capture and harvest large predators.R) consist of all scripts used in our analyses. The scripts were created using version 4.3.1 (2023-06-16). Annotations are provided throughout the script.

Versions of packages used:

• DHARMa 0.4.6 @DHARMa

• ggeffects 1.3.1 @ggeffects

• ggpubr 0.6.0 @ggpubr

• ggsignif 0.6.4 @ggsignif

• lme4 1.1.35.1 @lme4

• Matrix 1.6.4 @Matrix

• mgcv 1.9.0 @mgcv2003; @mgcv2004; @mgcv2011; @mgcv2016; @mgcv2017

• nlme 3.1.163 @nlme2000; @nlme2023

• rstatix 0.7.2 @rstatix

• tidyverse 2.0.0 @tidyverse