The rivers under study are right-bank tributaries of the Oder estuary. The largest among them, the Ina River, flows into the Oder between the outflow from Lake Dąbie and the southern part of the Szczecin Lagoon. Slightly further north is the mouth of the Gowienica River, which flows into the bay of the Szczecin Lagoon near the Stepnica village. In turn, the Wołczenica River flows into the Dziwna Strait, connecting the waters of the Szczecin Lagoon with the Baltic Sea (Robakiewicz 1993) (Fig. 1).

Map showing the locations of sampled sites in the Ina, Wołczenica, and Gowienica basins (pink dots), and areas stocked with resident brown trout progeny (green ellipses) and sea trout progeny (orange ellipses)

The Ina River is 128 km long, and its drainage basin covers an area of 2151 km2. Its average gradients in the upper and lower courses are 4‰ and 0.2‰, respectively. The average flow at the estuary is 10 m3/s. Currently, thanks to a fishway system, the river is potentially open for migration along almost its entire length up to the mill of Rybaki (Fig. 1).

The Gowienica River is 51 km long with a drainage basin of 368 km2. It is fully accessible for fish migration up to the old mill in Babigoszcz (19.5 km) (Jankowski et al. 2017). The Wołczenica River is 52 km long with a drainage basin of 462 km2. It is accessible for fish migration up to the small hydroelectric facility in Derkacz (25.5 km) (Tański et al. 2018). In a recent fish fauna study on the Ina River, 33 species of fish and lampreys were identified, with migratory trout, perch, and bullhead dominating (Keszka et al. 2013).

All three rivers have populations of brown trout. The resident form is most numerous in the upper reaches of the studied rivers, often isolated by hydrotechnical structures (Fig. 1). All studied rivers are under the management of the Polish Angling Association District in Szczecin and are stocked with migratory trout (from the Rega River population) and resident brown trout (often from the Folusz Breeding Center) of non-local genetic lines. Stocking of the Ina River with Rega sea trout has been practiced by the Polish Angling Association Region Szczecin since the 1960s. Between 250,000 and 500,000 migratory trout fry and around 40,000 resident brown trout alevins are introduced into the Ina River annually. Gowienica and Wołczenica are stocked annually with ~ 50,000 migratory trout fry and 8000 resident brown trout alevins (HELCOM 2011; ICES 2023).

First, samples were collected from stocking material from anadromous and resident forms, before they were released into the Ina, Wołczenica, and Gowienica in the spring of 2021 as fry (Table 1). Stocking material for sea trout was from spawners collected in the neighboring Rega River. Resident brown trout were from the Folusz hatchery located in southern Poland (Bernaś and Wąs-Barcz 2020).

Field sampling began in summer 2021 when 178 brown trout were collected from six locations. These were fish of various ages between 0 and 2 + years. Additionally, 20 anadromous adults were collected from the lower part of the Ina River between the river mouth and Goleniów during electrofishing for telemetry survey in autumn 2022 (Fig. 1, Table 1). The mean size of anadromous spawners was 63.8 cm (52 − 80 cm). To facilitate interpretation and verify actual relationships, the last group added to the analysis was a sample of 40 adult sea trout from the Rega River, caught during spawning migration in Trzebiatów during autumn 2021.

Fin clip samples (approximately 2 − 5 mm2) were collected from 425 individuals. Genomic DNA was extracted from fin tissue preserved in 96% ethanol using a Genomic Mini Kit (A&A Biotechnology, Gdynia, Poland) and diluted to a concentration of 30 − 100 ng/μl. Sampling details are presented in Table 1. Regarding sample abbreviations, the first two letters indicate the name of the river, the third the locality or origin, and the fourth additional details (E, electrofishing; T, telemetry; S, stocked).

For microsatellite analysis, a set of 13 fluorescently labeled polymorphic microsatellite loci (OneU9, Strutta58P, Ssosl438, Ssosl311, Str15INRA, Str543INRA, Str60INRA, Str73INRA, Ssosl417, Str85INRA, Ssa85, Bs131, and Ssa407) were amplified by single multiplex PCR using a Qiagen Multiplex PCR Kit (Qiagen, Hilden, Germany). Multiplex PCR (7 μl) contained ~ 100 ng of template DNA, multiplex PCR master mix, and 0.2 − 0.6 μM of each primer. Amplifications were carried out in a TProfessional Basic Gradient thermal cycler (Biometra) with an initial denaturation step at 95 °C for 5 min followed by 38 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 90 s, and extension at 72 °C for 60 s. Reactions were terminated after 30 min, and a final extension was performed at 60 °C. PCR products were genotyped via single capillary electrophoresis on an ABI Prism 3130xl genetic analyzer (Applied Biosystems) along with GeneScan 600LIZ size standards (Applied Biosystems). DNA fragments were estimated using a Peak Scanner v1.2 (Applied Biosystems).

Observed and expected heterozygosity and the mean number of alleles (total number of alleles at all loci divided by the number of loci) were calculated using Arlequin 3.5.2.2 (Excoffier and Lischer 2010). Population-specific FIS, pairwise-weighted FST values over all loci based on the number of different alleles, and Nei’s genetic distances were also determined with this software. Departures from the Hardy–Weinberg equilibrium (HWE) were detected with Chi-square tests in GenAlex 6.5 (Peakall and Smouse 2012). HPRARE was used to calculate allelic richness (which allows comparison of allele numbers without the bias associated with different sample sizes) and the richness of private alleles (alleles limited in a single population) (Kalinowski 2005). Overall, the F-statistic (FST, FIT, FIS) was estimated by analyzing molecular variance (AMOVA) implemented in Arlequin 3.5.2.2.

STRUCTURE 2.3.4 was applied to detect genetic structure and gene flow (Pritchard et al. 2000). The Evanno method (ΔK) (Evanno et al. 2005) was chosen to infer the highest number of clusters (K) based on the rate of change in log probability among consecutive K values. Five iterations of each K were performed with 200,000 burn-ins and 200,000 Markov chain Monte Carlo (MCMC) repetitions. Clumpak was then used to identify the optimal alignment of inferred clusters across different values of K (Kopelman et al. 2015).

In order to perform family structure parentage analysis, Colony 2.0.6.6. (Jones and Wang 2010) was employed for each electrofished sample together with corresponding samples from anadromous and resident forms released in the area of the given location. This analysis was also performed for adult sea trout from the Ina and Rega Rivers, which were caught during their spawning migration in 2021 and 2022. This was to determine the actual population parameters in a random sample, the most representative among the population in the Ina. We applied non-default COLONY job settings including typing error rate 0.001, mating system I with male and female polygamy, mating system II with inbreeding, run length medium, and analysis method FL, with defaults for other settings. The main goal was full-sib and half-sib dyad detection, determination of the number of families, and estimation of the effective population size Ne.

Additionally, to better understand and present genetic diversity, the obtained genotypes were compared with brown trout stocks used for stocking in Poland (Bernaś and Wąs-Barcz 2020). A neighbor-joining (NJ) phylogenetic tree was constructed based on pairwise Nei’s standard genetic distance without sample size correction (DST) with 10,000 bootstrap replications using PopTree2 software (Takezaki et al. 2010).