The commercial value of brown trout Salmo trutta, along with its significant importance to humans, is reflected in various aspects, particularly its widespread artificial propagation in fish farms globally. Farming is carried out for various purposes, of which stocking notably impacts the survival of native, genetically pure, and geographically unique brown trout populations.

Numerous studies on this species, still considered the S. trutta species complex, have shown that the life history, phenotypic plasticity, and high adaptability to diverse environmental conditions have not yet been clarified, more than likely as a consequence of their exceptionally high morphological and genetic variability (Behnke, 1972; Weiss et al., 2001; Sanz, 2018; Škraba Jurlina et al., 2018). The Western Balkans exhibit a high degree of endemism in salmonid genera and species (Behnke, 1986; Oikonomou et al., 2014). Together with the Iberian Peninsula and the Apennines, the Western Balkans is considered the area with the greatest phenotypic and genetic diversity among trout populations (Bernatchez, 2001) with complex evolutionary mechanisms, including the occurrence of secondary contacts between ancestral lineages, as well as local adaptation (Sanz et al., 2002; Snoj et al., 2008; Vera et al., 2010). The Western Balkans is the home range for populations of three phylogenetic lineages of Salmo cf. trutta which are defined based on the control region (CR) of mitochondrial DNA (mtDNA): Danubian (DA), Adriatic (AD) and marmoratus (MA). The DA lineage inhabits the northern, eastern, and central parts of the Balkan Peninsula, i.e., the Black Sea basin, and the AD lineage inhabits the northwestern and southwestern regions, including numerous lotic and some lentic habitats in the continental part of the Adriatic and Ionian Seas’ basins (Georgiev, 2003). The MA lineage is found close to the dispersal range of brown trout of the Danubian and Adriatic lineages, in the narrowest area in the Western Balkans, inhabiting the Adriatic drainages in Bosnia and Herzegovina and Montenegro (Snoj et al., 2010; Mrdak eta al., 2012; Škraba Jurlina et al., 2020; Righi et al., 2023). The status of the MA lineage, which is still often identified as marble trout, has not yet been resolved in Croatia (Ćaleta et al., 2019).

The rivers of the Western Balkans within the Black Sea basin drain a considerable territory and are very important in the context of studying the diversity and life history of different phylogenetic lineages (Škraba Jurlina et al., 2018). The DA lineage, native to the Danube basin, consists of numerous haplotypes, with the Da1 haplotype being the most widespread. Still, not all of the DA haplotypes are considered native to the Western Bakans area/region. The presence of non-native DA haplotypes, as well as other phylogenetic lineages, in most of the studied rivers presents a significant threat to the native population’s genetic structure (Snoj, 2004; Marić et al., 2006; Jadan et al., 2007; Tošić et al., 2016; Simonović et al., 2017a; Buj et al., 2020; Ivić et al., 2021).

Breeding brown trout in the Balkans is a profitable enterprise which makes it highly popular across all countries in the region. With the aim of intensive artificial propagation, and thus, a greater commercial profit (Simonović et al., 2017b), the initially domesticated native DA lineage (Gridelli, 1936; Kohout et al., 2012) proved to be insufficient. Consequently, breeders introduced breeding technology from Western Europe and the Atlantic lineage of brown trout suitable for this purpose (Taler, 1949; Pofuk et al., 2017; Piria et al., 2020).

Human-induced translocations of organisms and anthropogenic alterations of natural habitats are the most common causes of hybridization and introgression of non-native genotypes into native populations (Allendorf et al., 2001). Salmo trutta is characterized by genetic diversity at the intrapopulation level, which is the reason for the existence of genetically differentiated local populations. The loss of these unique populations could lead to a significant reduction in genetic variability within the species (Laikre et al., 1999). Stocking became a common practice to increase and support wild stocks of many fish species, including brown trout (Righi et al., 2023). In the beginning of the 20th century, it was pointed out that supportive breeding programs could also pose a threat to gene-level biodiversity, by releasing cultured individuals into the wild, with the aim of increasing the size of natural populations (Laikre et al., 2008). These activities could lead to manipulation of reproductive rates, causing reduced effective population size, increased rate of genetic drift and loss of genetic variability (Ryman et al., 1995; Wang and Ryman, 2001; Laikre et al., 2008), which is observed in many populations of brown trout in the Balkans (Jadan et al., 2007; Buj et al., 2020; Piria et al., 2020; Škraba Jurlina et al., 2020).

Although the research on genetic variability of brown trout has been extensively conducted to assess the introgression of farmed lineages, gene-level monitoring programs are still notably absent, despite repeated calls for their implementation (Laikre et al., 2008; Tošić et al., 2016; Simonović et al., 2017a; Škraba-Jurlina et al., 2020). To ensure the survival of a population or species, it is necessary to preserve its genetic diversity, and thus its evolutionary potential (Ryman et al., 1995). Unfortunately, the genotyping of brown trout individuals in brood stocks has been limited to a small number of fish farms in the Western Balkans.

In Croatian fish farms, brown trout of the AT lineage are mainly bred, so consequently the rivers are most likely stocked with these individuals due to their availability (Marić et al., 2022). In this way, non-native genetic material enters natural watercourses (Piria et al., 2020). Brown trout of the AT lineage are already recognized as the main cause of the loss of the original genetic diversity of brown trout (Weiss et al., 2001; Simonović et al., 2017b) and show an invasive character (Simonović et al., 2013; 2015; Piria et al., 2020), but their stocking is not regulated by any law in Croatia.

The goal of this research was to determine the genetic structure of brown trout populations in the Croatian part of the Danube basin, and to assess the degree of introgression of non-native genetic material within studied populations, as well as its possible origin.

The results of this study indicated a decrease in genetic variability in the Orljava River and Jankovac Stream. Analysis of CR mtDNA revealed that only one haplotype is present in both populations, non-native Da22 in Orljava River and Da1 in Jankovac Stream (Table 1). Neither Jankovac Stream nor Orljava River population proved to be genetically “pure”, considering that DA-AT hybrids were identified there (Kanjuh et al., 2020). In this regard, a bottleneck effect was observed for these two populations. Therefore, to continuously follow the status of these two populations, ongoing monitoring and further investigation are necessary. Evaluating the bottleneck effect in other populations revealed a heterozygosity deficit in the Brzaja, Curak and Kupčina rivers, indicating population expansions, most likely due to the introduction of the non-native farmed At1 haplotype. The same was observed in the “Vrabac” fish farm, but that was to be expected knowing that within the fish farm non-native At1 individuals are farmed.

In comparison with all the analysed populations, it was shown that the Slapnica River population is genetically different. The presence of AT lineage individuals was not recorded there, so this population consists only of individuals of the DA lineage. Nevertheless, the presence of non-native Da2 and DA-AT hybrids in the Slapnica River indicates that its native character was disturbed (Kanjuh et al., 2020). Population from the Slapnica River most likely separates due to the presence of private alleles in hybrid individuals, at microsatellite loci Str73INRA, Ssa410Uos, SsaD190 and OMM1064 (Supplementary Table S1). In all other populations, a large overlap was observed (Figure 2; Figure 3), both within geographically close and among geographically distant ones, as clearly evidenced in populations from the Čabranka River (Gorski Kotar) and Jankovac Lake (Mt. Papuk). This level of overlap indicates a high genetic similarity of the populations, which is a consequence of stocking from hatcheries that import and breed, most likely, the same (AT) phylogenetic lineage of brown trout (Piria et al., 2020). In fish farms throughout Croatia, brown trout carrying the At1 mtDNA haplotype is probably the most farmed and used for stocking of streams attractive for fishing. Minor genetic differences between populations observed in this study, most likely caused by stocking with the same genetic material, were additionally confirmed by the results of the population structure analysis. Separation of only two genetic clusters indicates strong introgression of the non-native genotypes, which is consistent with molecular variance results for CR mtDNA haplotypes (Kanjuh et al., 2020), which show little variation among populations. Structuring of only DA individuals also showed genetic similarity between populations. This result indicated the stocking with not only Atlantic (AT) lineage, but also non-native Danubian (DA) haplotypes of brown trout. Two non-native DA haplotypes were recorded in the analysed rivers of the Danube basin in Croatia (Kanjuh et al., 2020). The Da22 haplotype was limited to three localities (Veličanka River, Orljava River and Brzaja River) at the northern slopes of the Mt. Papuk (Kanjuh et al., 2020). This haplotype is native to the Lohnbach and Daglesbach rivers in Austria (Duftner et al., 2003) and in the Una River drainage in Bosnia and Herzegovina (Škraba et al., 2017). The non-native Da2 haplotype was much more frequent than Da22 haplotype (recorded in seven out of fifteen analysed populations) (Kanjuh et al., 2020). The Da2 haplotype is native to the streams and rivers in southern Germany (Bernatchez, 2001) and streams belonging to the Austrian part of the Danube drainage (Weiss et al., 2001). The stocking of various freshwaters of Europe during the 19th century was carried out with fish material from southern Germany (Kohout et al., 2012). It is most likely that these stockings included the territory of today’s Croatia. Precise data on fish farm breeding of non-native Danubian haplotypes of brown trout do not exist, but first data on the presence of the Da2 haplotype in freshwater of Croatia was recorded in the Gacka River, together with the non-native At1 haplotype (Jadan et al., 2007). The separation of DA individuals from the Curak River into a single genetic cluster (Figure 6B) is interesting information from the conservation point-of-view. All DA individuals from the Curak River belong to the native Da haplotype and are homozygous to the native LDH-C*100 allele (Kanjuh et al., 2020). It is possible that the influence of the non-native At1 haplotype was suppressed in this population, considering that out of a total of twelve individuals, only one was genetically pure AT and only one was AT-DA hybrid (Kanjuh et al., 2020).

Genetic research of brown trout and their susceptibility to the presence of the introduced AT lineage in Croatia was until now mainly analysed using cytb and CR mtDNA as molecular markers. Buj et al. (2020) and Ivić et al. (2021) emphasized the negative impact of non-native AT lineage on the native populations of the DA lineage (nominal Salmo labrax) and the need for conservation of its genetic character in the Plitvice Lakes National Park and Žumberak-Samobor Hills Nature Park, respectively. The presence of the At1 haplotype has been reported for many streams attractive for fishing throughout the Western Balkans (Snoj, 2004; Marić et al., 2006; 2012; Jadan et al., 2007; Simonović et al., 2015; Tošić et al., 2016; Škraba Jurlina et al., 2018; 2020). In this regard, it is significant that apart from the At1 haplotype, the presence of the rare At17 haplotype was recorded in the upper reaches of the Mrtvica River (Adriatic Basin), undoubtedly indicating a unique event of stocking of this river (Škraba Jurlina et al., 2018). As its presence was not detected in the lower part of the river, the authors assume that the trout spawned only in the upper part of the river, producing hybrids with brown trout of the resident Adriatic lineage. They also pointed out that gene flow between the populations in the upper and lower parts of the river was low (despite the absence of physical and/or reproductive barriers), indicating the impossibility of individuals from the upper reaches of the Mrtvica River to migrate downstream and hybridize with the brown trout of the native Adriatic lineage consisting of both resident (predominantly males) and lake-dwelling, migratory (predominantly females) fish (Škraba Jurlina et al., 2018). The negative influence of the At1 haplotype was also recorded in non-fishing rivers, which are characterized by the presence of specific and narrowly distributed haplotypes. One such example is the population in the Vratna River, a tributary of the Danube River at the downstream border of the Iron Gate Gorge, where the rare haplotype Da23c occurs, which is currently threatened by the appearance of the stocked brown trout of the At1 haplotype (Tošić et al., 2016; Škraba Jurlina et al., 2020).

In addition to the genetic analyses, the invasive character of the AT lineage has been highlighted in recent research, which showed the rapid adaptation of this lineage to the different conditions in natural watercourses, the consumption of a wide range of available food and competition for food and space, as well as diet overlap with native DA lineage individuals (Piria et al., 2020). Also, in assessment of invasiveness risk in the Danube and Adriatic basins across four Balkan countries, out of 13 existing and four horizon non-native salmonid species, S. trutta of the AT lineage showed a very high risk of invasiveness in the current climate conditions (Marić et al., 2022). In addition to At1, the non-native Da2 haplotype has also a significant impact on disrupting the genetic structure of the native Croatian populations of brown trout. Therefore, it would be important for future research to focus on obtaining data on the potential breeding origin and ways of introduction of the Da2 haplotype, but also on assessing the risk of invasiveness to the native genetic pool of the brown trout complex. Apart from the presence of the Da2 haplotype in the rivers of Croatia (Kanjuh et al., 2020), the presence of Da2 haplotype was recorded in the rivers of Serbia (Marić et al., 2006; Simonović et al., 2015; Tošić et al., 2016), Bosnia and Herzegovina (Mrdak et al., 2012; Škraba et al., 2017), and Montenegro (Mrdak et al., 2012).

Everything said above points out to a long history and the current management practice of stocking with the hatchery-reared brown trout. There are clear warnings of the negative impact of the AT hyplotypes on native and wild brown trout populations. Although the detrimental effects of strong introgression in wild brown trout populations are visible, management procedures in many countries are apparently not strict enough and properly applied. Despite proposals for potential management plans, the native gene pool of brown trout remains under serious threat. Studying the wild, native, and locally specific populations of brown trout on the territory of Croatia, determining of their status and genetic structure should be a priority to maintain and preserve their original genetic stocks. Genotyping the remaining brown trout stock from inland waters is essential. It is to be implemented into the national legislation, in order to set a mandatory genotyping of hatchery brood stocks and to identify and register native brood fish. This practice is currently weakly enforced in Croatia, and in many countries of the Balkan Peninsula. It is important to examine the genetic structure of brown trout grown in fish farms, as well as the stocks used for stocking, which, so far, has not been legally regulated or implemented as a necessary practice in majority of countries in the Western Balkans. Although some authors point out that stocking as a measure of fisheries management should be re-examined (Škraba Jurlina et al., 2018), for the recovery of the degraded status of the brown trout populations, stocking would be desirable. Stocking should imply a controlled, strictly defined, and planned approach, and the stocked material should be bred as a “foundation brood stock”. In addition, all of the populations that would be studied and marked as genetically “pure” and/or unique to a certain locality should be protected from the negative influence of fishing pressure. Implementing a mandatory unconditional “catch-and-release” angling regime (Simonović et al., 2018) as a compulsory legal protection measure is proposed as a possible solution that would ensure the self-sustainability of the trout stock without the need for restocking, but also the sustainability of the fishery. Many authors additionally pointed out that long-term stabilization of the density of trout populations is often not possible without environmental improvement strategies, such as habitat restoration, removal of migration barriers, improvement of hydrological regimes, which are generally more effective than stocking (Cowx, 1994; Fjellheim et al., 2003; Oosterhout et al., 2005; Ferguson, 2007; Škraba Jurlina et al., 2018), and the brown trout habitats are certainly special in their characteristics, being very sensitive to all kinds of anthropogenic pressures.