In species with complex life cycles, life history theory predicts that fitness is affected by conditions encountered in previous life history stages. Here, we use a four-year pedigree to investigate if time spent in two distinct life history stages has sex-specific reproductive fitness consequences in anadromous Atlantic salmon (Salmo salar). We determined the amount of years spent in fresh water as juveniles (freshwater age, FW, measured in years), and years spent in the marine environment as adults (sea age, SW, measured in sea winters) on 264 sexually mature adults collected on a river spawning ground. We then estimated reproductive fitness as the number of offspring (reproductive success) and the number of mates (mating success) using genetic parentage analysis (>5000 offspring). Sea age is significantly and positively correlated with reproductive and mating success of both sexes whereby older and larger individuals gained the highest reproductive fitness benefits (females: 62.2% increase in offspring/SW and 34.8% increase in mate number/SW; males: 201.9% offspring/SW and 60.3% mates/SW). Younger freshwater age was significantly related to older sea age and thus increased reproductive fitness, but only among females (females: -33.9% offspring/FW and -32.4% mates/FW). This result implies that females can obtain higher reproductive fitness by transitioning to the marine environment earlier. In contrast, male mating and reproductive success was unaffected by freshwater age and more males returned at a younger age than females despite the reproductive fitness advantage of later sea age maturation. Our results show that the timing of transitions between juvenile and adult phases has a sex-specific consequence on female reproductive fitness, demonstrating a life-history trade-off between maturation and reproduction in wild Atlantic salmon.

Anadromous adults were sampled in September-October 2011-2014 at the lower Utsjoki spawning grounds at the mouth of the Utsjoki tributary of the Teno River in northern Finland (69°54'28.37''N, 27°2'47.52''E, see Mobley et al. (2019) for further details on sampling location). Adults were weighed, and total length was recorded. Condition was calculated as the residual from a linear model of weight predicted by length for each sex and spawning cohort (Mobley et al., 2019; Patterson, 1992). Scales were collected for age analysis and a small piece of anal fin was collected for genetic analysis prior to release near the site of capture. Juveniles were sampled by electrofishing shallow areas in the region of the spawning grounds 10 to 11 months later, which is two to three months after they are expected to have emerged from the nests in the stream bed gravel (Mobley et al., 2019). Genetic samples were collected from all juveniles by collecting a small piece of adipose and/or anal fins, after which they were immediately returned to the river (Mobley et al., 2019). Four parent-offspring cohorts were sampled in this manner between 2011 and 2015.

Age determination

Freshwater age, defined as the number of years spent in freshwater prior to migrating to sea, and sea age, defined as the number of years an individual overwintered at sea before returning to spawn, was determined for adults captured on the spawning ground using scale growth readings as outlined in Aykanat et al. (2015). Freshwater age could not be determined on 25 individuals (3 females, 22 males) using scale data. Sea age could not be determined for 16 adults > 1SW (1 female, 15 males) using scale data and was therefore extrapolated based on calculated distributions of weight of known sea age individuals (see Mobley et al., 2019, Supplementary Materials, Table S4). However, freshwater age was not extrapolated based on weight due to the poor relationship between weight and freshwater age (see Results). Therefore, these individuals were excluded from statistical analyses. Repeat spawners that were spawning for a second time were also determined using scale data. Thirteen individuals (6 females, 7 males) were identified as repeat spawners by scale aging analysis. The mean sea age of repeat spawning females was 3.2 ± 0.4 SE (range 2-4 SW) and all repeat spawning males had spent one year at sea before the first spawning migration and another year at sea before returning to spawning for the second time (i.e., all male repeat spawners were 2 SW).

Parentage analysis

Molecular parentage analysis was conducted according to Mobley et al. (2019). Briefly, all adults and juveniles were genotyped using 13 microsatellite loci previously used for parentage analyses in this species (Aykanat et al., 2014). Pedigrees were constructed for each parent-offspring cohort separately using the package MasterBayes V2.55 (Hadfield, Richardson & Burke, 2006) in the R programming environment (R Core Team, 2018). Genotyping error rate was calculated as per Mobley et al. (2019). The distribution of unsampled population sizes (mothers and fathers separate) were given a prior mean of twice the sampled population size, following Aykanat et al. (2014), with a variance calculated as 1.5 – 0.25 * sampled population size, which encompassed likely parameter space. The pedigree was run for 30,000 iterations after a burn-in of 5,000. We then extracted the mode of the posterior distribution of pedigrees, and removed assignments with a likelihood of less than 90%. Offspring assigned to a known parent were either confidently (>90% likelihood) assigned to two sampled adults, or one parent confidently assigned to a sampled adult and the other confidently assigned to an unsampled adult. In this manner, an offspring that contributed to our reproductive fitness measures were either assigned to both a sampled sire and a sampled dam, or to either a sampled sire or a dam and an unsampled adult (Mobley et al., 2019).

Reproductive fitness estimates

Reproductive success was quantified as the number of offspring assigned to an adult, following parentage assignment of all offspring. Mating success was estimated as the number of unique mates per individual identified within our sample by parentage analysis (Mobley et al., 2019).

Citations

Aykanat T., Johnston S. E., Cotter D., Cross T. F., Poole R., Prodőhl P. A., ...Primmer C. R. (2014) Molecular pedigree reconstruction and estimation of evolutionary parameters in a wild Atlantic salmon river system with incomplete sampling: a power analysis. BMC Evolutionary Biology 14, 68. doi: 10.1186/1471-2148-14-68

Aykanat T., Johnston S. E., Orell P., Niemelä E., Erkinaro J., Primmer C. R. (2015) Low but significant genetic differentiation underlies biologically meaningful phenotypic divergence in a large Atlantic salmon population. Molecular Ecology 24, 5158-5174. doi: 10.1111/mec.13383

Hadfield J. D., Richardson D. S., Burke T. (2006) Towards unbiased parentage assignment: combining genetic, behavioural and spatial data in a Bayesian framework. Molecular Ecology 15, 3715-3730. doi: 10.1111/j.1365-294X.2006.03050.x

Mobley K. B., Granroth-Wilding H., Ellmen M., Vaha J.-P., Aykanat T., Johnston S. E., ...Primmer C. R. (2019) Home ground advantage: local Atlantic salmon have higher reproductive fitness than dispersers in the wild. Science Advances 5, eaav1112. doi: 10.1126/sciadv.aav1112

Patterson K. R. (1992) An improved method for studying the condition of fish, with an example using Pacific sardine Savdinops sagax (Jenyns). Journal of Fish Biology 40, 821-831. doi: 0022-1 112/92/060821

R Core Team (2018) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.

UtsAdults.csv

This file contains all phenotypic and reproductive fitness data for each individual adult salmon sampled at the lower Utsjoki study site over four cohort years (2011-2015). ID corresponds to ID in DRYAD submission: 10.5061/dryad.3ss2t53.

UtsParentageAssignments.csv

This file contains the output of the pedigree fit, i.e. parentage assignments, for all sampled offspring sampled at the lower Utsjoki study site over four cohort years (2011-2015). ID corresponds to ID in DRYAD submission: 10.5061/dryad.3ss2t53.

Utsreadme.txt

Describes the data and abbeviations in the above datasets.