Within the genus Carassius Jarocki, 1822 , the crucian carp (C. carassius L., 1758) occurs naturally in the northern part of Middle Danube Basin (Austria, Morava, Slovakia). This species has the least concern status in this region, but observations in the last decades suggest that it is very close to extinction here. The distribution of crucian carp is limited to a small number of vanishing lentic habitats (oxbow lakes, marshlands). These biotopes are in the last stage of succession due to the drying up of the landscape and a reduction in the creation of new natural alluvial habitats. The non-native cyprinid, C. gibelio (Bloch, 1782), known as gibel carp and Prussian carp, has gradually become eudominant in a wide spectrum of habitats/biotopes since the 1960s Several biological adaptations of non-native species are generally considered the strong basis for the mass spreading in the invaded area. The other side of the expansion of non-native C. gibelio is affected by anthropic activities associated with fish farming, translocation and stocking the fish in open water ecosystems. In this study, we analysed historical scientific data on the distribution of Carassius spp. published from the 19^th century to the present from the mentioned areas. The results suggest that the number of records of invasive C. gibelio has gradually increase in rivers, regulated channels and creeks, which could be considered as natural pathways of spreading. However, the presence of invasive C. gibelio in artificial biotopes (fishponds, reservoirs) is continuous from the 1960s. In the area mentioned, the artificial biotopes are managed by national fisheries associations and relate to the historical way of farming in Central and Eastern European countries. To show the current state of the fishing grounds of the Slovak Angling Association, we a created the distribution map based on the Carassius spp. catches recorded in last two decades.

Danube basin, crucian carp, angling, fish farms, habitat lost, climate

Fish species of the genus Carassius Jarocki, 1822, have acquired a wide range of statuses across the world, including ornamental (Chen et al. 2020), non-native (Morgan and Beatty 2007; Hamilton 2021), and species with high invasive potential (van der Veer and Nentwig 2015). The crucian carp (C. carassius L, 1758) was a common species in the Middle Danube area until the 1950s (Holčík 1966; Sedlár and Amena 1989). The species occurred mainly in numerous village ponds and alluvial habitats (Baruš and Oliva 1995). Since the nineteenth century, the number of suitable habitats has significantly decreased (Holčík 1996), due to the gradual liquidation of village ponds, the construction of flood defences and the subsequent loss of alluvial habitats caused by large-scale river channel regulation measures linked to planned agriculture (Jurík et al. 2018). Currently, the crucian carp occurs only sporadic in secondary habitats (e.g., regulated channels, small water reservoirs) and in the isolated river branches, usually with a very low density (authors´ obs.). Currently, the conservation status of this species within the Europe is that of least concern (Freyhof 2010). However, in several European countries, this species has started to be critically endangered. Nonetheless, the survival of C. carassius populations now depends on human management (Copp et al. 2008; Sayer et al. 2020), or facilitation of a natural flood regime (Guti 2020). A noticeable change occurred after the 1950s, which is probably linked to the importation of Asian cyprinids through the former Soviet Union (Slynko et al. 2011). Based on published data, non-native Carassius species entered the Slovak region of the Danube River in 1962 (Balon 1962). However, the mass spreading of this non-native species in throughout the Middle Danube appears to have begun in the 1950s, primarily underpinned by fish farms operating in Bulgaria and Hungary at the time (Holčík and Žitňan 1978; Holčík 1980). It is also possible that in the Middle Danube, as well as in the other parts of Europe, the spread of common carp Cyprinus carpio L., 1758, was supported by Benedictine monks. It is possible that Carassius species were at times also accidentally transported as contaminants of purposeful transfers of other species (Rylková et al. 2013). However, the mass occurrence of non-native Carassius spp. is not known from the literature to have occurred before the early 21^st century. However, records from Europe are convoluted as in some European regions C. gibelio (C. auratus complex) was considered native, based on the description of this species by Bloch (1872). However, Bloch’s museum material has been found to consist of C. carassius and hybrid specimen (Paepke 1999), and the taxonomic status was redetermined by Kalous et al. (2012) based on the C. gibelio specimen from the Silesian region of Czechia.

In the present study, to identify proportional temporal changes in species distributions for native and non-native Carassius species in relation to habitat (biotope), we analysed historical spatio-temporal distribution data from parts of the Middle Danube basin (i.e., presence/absence data for Slovakia and the surrounding river catchments). Further, we considered fisheries Carassius catch data recorded by the Slovak Angling Association (Slovak AA) during the last two decades. In doing so, we aimed to comparatively assess historical and recent presence/absence data pertaining to native and non-native Carassius spp.

For the period 1875–2021, a total of 836 sources for presence/absence records for the distribution of Carassius species in Slovakia and surrounding watersheds (Middle Danube basin) were assessed (Suppl. material 1). The analysed data contained information concerning faunistic observations, grey literature, and our own unpublished observations from the last 35 years. Literature data included the authors’ information about the biotope/habitat recorded or were obtained based on coordinates from historical maps and authors´ database. We excluded from the analyses the less numerous biotope observations, such as those recorded in the mountain and peak lakes (Suppl. material 1, obs. number 26, 47, 49, 89, 190, 242, 262, 308, 367, 384, 412, 797). In our analysis, we considered C. carassius as native, and based on the latest designation (Papoušek 2008; Rylková et al. 2013; Pakosta et al. 2016; Knytl et al. 2022). We designated all species/lineages within the C. auratus complex (C. auratus, C. gibelio, C. langsdorfii, lineage “M”) that have been recorded in the Danube basins as non-native Carassius. Three distributional maps were created based on presence data recorded in A: 1875–1961 (observation of C. carassius before first record of non-native Carassius spp.); B: 1962–1990 (a period of large-scale river regulation work in Slovakia and intensive imports of Asian cyprinids); C: 1991–2021 (a period of global temperature increase; www.eea.europa.eu/ims/global-and-european-temperatures). We created a distribution map of Carassius spp. based on 12,060 Carassius catch records from 1,064 Slovak AA fishing grounds (river sections in the administration of the Slovak AA) recorded between 2003 and 2021. Data from the fishing association includes the spatial distribution of fish of the genus Carassius (without species identification). Distributional maps were created with the use of QGIS.

Statistical analysis

For analysis of the data we used the non-linear Generalized Additive Models (GAMs) with a binomial family (R environment, packages – mgcv(), visreg()). Changes in the presence/absence of Carassius spp. were modelled depending on the time factor in interaction with the habitat type (Table 1 – explanatory variables). The number of dimensions (k-model rank) in each GAM model was identified automatically based on the gam.check() function. The quality of the models was determined based on a decrease in the Akaike information criterion AIC(). The results and trend lines of the GAM plots (Figure 1, 2) were constructed based on statistical outputs (Table 1, GAM-b in Results) in the R environment with the use of the visreg() package.

Figure 1. 

The proportion of presence (1) and absence (0) data (y-axis, one meaning 100% of the information) from the literature that mentions native (C. carassius) and non-native Carassius species between the years 1875 and 2021 (x-axis) in the river basins of Slovakia and border regions (Middle Danube basin).

Figure 2. 

Non-linear GAM-b presence/absence (y-axis) plot for the historical information (x-axis; 1955–2021) of the Carassius species: A) native C. carassius; B) non-native using the two-way interaction between historical data and type of habitat (biotope) recorded (type of habitat in multi-plots). The red curves describes statistically significant changes in the proportion of information from the specific habitats.

Table 1.

Download as

CSV

XLSX

Approximate significance of smooth terms of generalized additive models (GAMs, binomial family) and a comparison of the statistical models (related to Figure 2A and B) (Deviance explained = 42.7%).

Model Fit	GLMM family	Response	Explanatory	edf	χ2	p	AIC	df
3	binomial (logit)	presence/absence of Carassius spp.	s(year) - spline parameter year, k = 16 - Model rank, by = biotope				649.1	41.5
			+ biotope
			s(year):biotope channels	8.6	40.5	< 0.001
			s(year):biotope creeks	1	13	< 0.001
			s(year):biotope reservoir	2.8	0.01	> 0.99
			s(year):biotope oxbows	8.3	11.8	> 0.32
			s(year):biotope rivers	9.9	23.6	< 0.05
			s(year):biotope ponds	2.8	0.06	> 0.99
			s(year):biotope pits	1	8.2	< 0.005
2			s(year) - spline parameter year, k = 16 - Model rank				652.3	19.1
			+ biotope
1			s(year) - spline parameter year				666.2	14
			+ biotope

Analysis of 836 scientific observation-based data recorded between 1875 and 2021 revealed changes in the proportion of native and the non-native Carassius species recorded in literature sources (Figure 1). The analysed data show a significant non-linear trend (GAM-b, χ2 4.91 = 93.8, p < 0.001). A significant decrease in the literature information (Table 1 – model fit 3; Table 2) for native (C. carassius) and an increase for non-native Carassius spp. is visible in the interaction with biotope recorded since the 1950s (Figure 2A, B). The amount of information shows a different trend for each biotope, with a significant trend in the rivers, gravel/mining pits, regulated channels and creeks. River oxbows and river side-arms represent the last refuge as well as the biotope of co-occurrence of native and non-native Carassius species. Non-significant information trends for fishponds and reservoirs show the permanent presence of non-native Carassius spp. and the absence of C. carassius in biotopes with high human influence. Spatio-temporal visualisation of the scientific observations of Carassius spp. recorded 1875–2021 is shown in the three-distributional maps (Figures 3A, B, C). The extensive predominance of non-native Carassius species since the 1990s is visible (Figure 3C). This situation has occurred just 30 years after first official record of non-native Carassius species in 1962. A spatial trend map based on the catch data form Slovak AA is shown in Figure 3D. Overall, 1,893 fishing grounds managed by the Slovak AA represents 92% of the total available fishing grounds in Slovakia (data for 2022). The last mentioned figure shows 12,060 catches of Carassius spp. recorded by the anglers of the Slovak AA between 2003 and 2021. These specimens were caught in the 1,064 fishing grounds of the Slovak AA, which represents 56.2% of all fishing grounds managed by the Slovak AA. Based on the analysed data, we can confirm the presence of non-native Carassius spp. in the Danube basin as well as in a small part of the River Vistula basin in the north of Slovakia (i.e., Dunajec and Poprad River basin).

Figure 3. 

Spatio–temporal patterns of the distribution of the native C. carassius (green squares) and non-native Carassius spp. (red points) in the river basins of Slovakia (constructed based on scientific literature data – Suppl. material 1) A) 1875–1961; B) 1962–1990; C) 1991–2021; D) distributional map of Carassius species according to Slovak Angling Association data recorded 2003–2021: blue colour – main catchments of Slovakia (fishing grounds of the Slovak AA), red colour – showing 56.2% of fishing grounds with the catch (presence) of the Carassius specimens (constructed based on 12,060 fisheries catches); shades of green color indicate elevation (dark green – mountains, light green – lowlands).

The distribution area of limnophilic fish species (e.g., C. carassius, Misgurnus fossilis (Linnaeus, 1758), Umbra krameri (Walbaum, 1792), etc.) in the Slovak Danube basins is one of the northernmost occurrences within the Middle Danube catchment area (Freyhof and Kottelat 2008). The presence of this species is connected to river side-arm formations associated with the dynamics of natural lowland rivers (Lenhardt et al. 2020). During the last 40 years, a gradual decrease of C. carassius has been recorded in the Danube River basin (Holčík 1996; Telcean and Cupşa 2009; Ferincz et al. 2012; Bănăduc et al. 2016, 2020) as well as in the other basins across Europe (Sayer et al. 2010; Mezhzherin et al. 2019). The most probable causes are competition of the native C. carassius with other – mainly Asian – cyprinids (Carassius, Cyprinus, Ctenopharyngodon, Hypophtalmichtys). The competitive disadvantages of C. carassius include limitations in food resources (Busst and Britton 2015), altered hydrology (swirling of sediments, eutrophication) and the behaviour of stocked non-native cyprinids (Richardson et al. 1995). However, factors influencing the process of extinction of the crucian carp are diverse. The most serious factor in the decline of crucian carp in the Middle Danube region has been historical regulation of the river since nineteenth century (Hanušin 1949), and the increase of this regulation associated drainage of the land as part of the agricultural plan of the Communist era, beginning in the 1950s (Jurík et al. 2018). A higher number of literature records of the crucian carp from the 1950s are linked to their presence in regulated channels (Figure 2A). These secondary biotopes represented a refuge for this species after major dewatering of marshlands. The succession and extinction of the last natural habitats have likely influenced the presence of the crucian carp (Figure 2A – river oxbows, side arms). Currently, this trend is being accelerated by climate change (Gómez-Baggethun et al. 2019), which has caused the complete drying of the few remaining lowland habitats in recent years. Historically, the proportion of available information concerning the presence of native C. carassius varied across the nineteenth century (Figure 1), with records largely being associated with river regulation. Consequently, the predominance of non-native Carassius spp. is continuous from the 1960s, with a short decreasing trend in the 1990s. Here we can assume two waves (peaks) of dispersal that may coincide with the spread of asexual (gynogenetic) females in the 1960s and a second wave of expansions (1990s) connected to the formation of the mixed sexual-asexual Carassius populations and the presence of sexual males (Lusková et al. 2004, 2010; Barbuti et al. 2012). In general, localities with the co-occurrence of native and non-native Carassius spp. are not numerous (Artaev and Ruchin 2016). Here we hypothesize that the co-occurrence is beneficial for the spreading of clonal non-native Carassius females. Polyploid females are reproducing by gyno/hybrid-genesis and during the expansion, they can incorporate part of the genome of other Carassius spp, or use the sperm of native males of C. carassius for the production of their clonal offspring (Gui and Zhou 2010; Tapkir et al. 2022). In contrast, F1 crossbreed offspring of native C. carassius are probably sterile or they do not reach sexual maturity and their occurrence is rare (Papoušek et al. 2008).

Trends in the literature for creeks and gravel (mining) pits are continuous and increasing, which indicates the spatial and temporal spread from main sources (rivers, ponds, reservoirs). In the case of reservoirs and ponds, temporal changes as recorded within the literature are not significant (Table 1, Figure 2B), indicating that invasive individuals have been continuously present in this habitat type since the 1950s´. Based on this information, we assume different pathways of introduction have occurred for non-native Carassius spp.. Firstly, we suspect these include natural dispersal across the river network from the lower Danube basins (Balon 1962; Holčík and Žitňan 1978), with gradual expansion by the non-native Carassius spp. from the main river network into smaller streams, as well as into inundation pits (mining, gravel) as a result of flooding events. Secondly, as records for the presence of non-native Carassius spp. in reservoirs and fishponds is continuous, from the 1960s onwards (Figure 2B), we infer that their spread in these spread across these environments is underpinned by fish farms and managed fishery biotopes (Mišík and Holčík 1962; Dorko and Terek 1976). The third pathway is the spreading of Carassius baitfish by anglers, which was first recorded in the 1960s by Mišík and Holčík (1962) for the Orava Reservoir. Nowadays, according to the applicable law – the Act on Fisheries of the Slovak Republic (No. 216/2018) – the use of live baitfish is still possible, as is the translocation of baitfish between fishing grounds. Consequently, the non-native Carassius populations have continued to radiate from their source populations (e.g., fish farms, fishing grounds, pet shops) into the surrounding river basins (Sedlár et al. 1976; Holčík and Žitňan 1978). A temporal increase for records of non-native Carassius also reflects the fisheries activities, which are connected to post-communist planning regimes in the Eastern Bloc countries (Britton and Gozlan 2013). Based on our analyses of fisheries data (Figure 3D), the spreading of Carassius spp. is likely to be nationwide, with a higher intensity in lowland habitats (lowlands occur in the south part of Slovakia). Moreover, the slow running and stagnant waters, such as reservoirs, small ponds and mining pits in river inundation, are favoured by carp anglers. Fishing grounds with the presence of Carassius species represent more than 56% of all fishing grounds managed by the Slovak AA (Figure 3D). Whereas, current scientific data (Figure 3C) indicates that most of the Carassius spp. in Slovak River basins are probably non-native specimens of the C. auratus complex. Uncontrolled stocking of non-native Carassius spp. and their spread from production ponds and subsequent penetration to natural biotopes has also been confirmed in other parts of the Middle Danube basin, such as in Hungary (Takács et al. 2017), Serbia (Lenhardt et al. 2011) and in the Transcarpathian Ukraine (Koščo et al. 2004). Higher haplotype diversity of non-native Carassius spp. has been observed, especially within large waterbodies under human pressure, such as tourism, fishing, fish farming (Keszte et al. 2021). The high number of haplotypes is probably a result of introduction from several sources; with subsequent hybridization and occupancy of a wide spectre of ecological niches (Faud et al. 2021). Aside from the human influence, the spread of non-native Carassius spp. is accelerated by wide spectrum of the biological adaptations (Tcherfas 1971; Lusková et al. 2010; Fagernes et al. 2017) as well as gynogenetic reproduction of polyploid females (Xiao et al. 2011).

Based on historical data analysis, we can consider the Slovak population of C. carassius as critically endangered, with a negative outlook. In conclusion, based on historical data, we call for an urgent re-assessment of the IUCN status for C. carassius throughout the Danube River basin. In the case of non-native Carassius species, we suggest: i) the prevention of their spread from fish farms into open waters through improved biosecurity protocols; ii) a reduction in population density across biotopes connected to common carp farms (e.g., drainage channels), with management by angling associations for fishing grounds; and iii) acceleration of the adoption of appropriate legislative to prohibit unethical live bait fishing, with strict implementation for compliance.

This work was supported by grants of the Scientific Grant Agency of the Slovak Republic under the grants VEGA 1-0364-20 and the Slovak Research and Development Agency under the project APVV SK-AT-20-0009.

JF conceived and designed the study, gathered literature data, conducted statistical analysis and wrote the paper. PK gathered data of the Slovak Angling Association and created the distributional maps. JK provided his own databases and literature and supervised the study concept.

Own data provided in this study were authorised through a permit from the Ministry of Environment of the Slovak republic, special yearly permits for electrofishing and in the accordance with all ethical principles.

We would also like to thank the National Scholarship Programme of the Slovak Republic (App. no. 39062) and all the colleagues who helped us or provided data. We are also especially grateful to the reviewers and the editor for their careful reading of our manuscript and their helpful comments.