Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Out of the Black Sea: Phylogeography of the Invasive Killer Shrimp Dikerogammarus villosus across Europe

Out of the Black Sea: Phylogeography of the Invasive Killer Shrimp Dikerogammarus villosus across Europe

  • Tomasz Rewicz, 
  • Remi Wattier, 
  • Michał Grabowski, 
  • Thierry Rigaud, 
  • Karolina Bącela-Spychalska


The amphipod Dikerogammarus villosus has colonized most of the European main inland water bodies in less than 20 years, having deteriorating effect on the local benthic communities. Our aim was to reveal the species phylogeography in the native Black Sea area, to define the source populations for the colonization routes in continental Europe and for the newly established UK populations. We tested for the loss of genetic diversity between source and invasive populations as well as along invasion route. We tested also for isolation by distance. Thirty three native and invasive populations were genotyped for mtDNA (COI, 16S) and seven polymorphic nuclear microsatellites to assess cryptic diversity (presence of deeply divergent lineages), historical demography, level of diversity within lineage (e.g., number of alleles), and population structure. A wide range of methods was used, including minimum spanning network, molecular clock, Bayesian clustering and Mantel test. Our results identified that sea level and salinity changes during Pleistocene impacted the species phylogeography in the Black Sea native region with four differentiated populations inhabiting, respectively, the Dnieper, Dniester, Danube deltas and Durungol liman. The invasion of continental Europe is associated with two sources, i.e., the Danube and Dnieper deltas, which gave origin to two independent invasion routes (Western and Eastern) for which no loss of diversity and no isolation by distance were observed. The UK population has originated in the Western Route and, despite very recent colonization, no drastic loss of diversity was observed. The results show that the invasion of the killer shrimp is not associated with the costs of loosing genetic diversity, which may contribute to the success of this invader in the newly colonized areas. Additionally, while it has not yet occurred, it might be expected that future interbreeding between the genetically diversified populations from two independent invasion routes will potentially even enhance this success.


Biological invasions are the inherent symptom of global changes and a major threat to biodiversity [13]. Alien species may cause irreversible changes to invaded ecosystems, often resulting in reducing distribution or in extinction of native species through direct predation [4], food and shelters competition [5, 6], transmission of parasites [7], or modifications of habitat [8].

Molecular markers proved to be powerful tools in tracking invasion patterns and dynamics [9, 10]. They were useful in identifying: (1) cryptic invasions, when one morphological invasive species is composed of at least two diverging evolutionary units [1113] or when a species thought to be native appears to be an alien [14]; (2) source populations and pathways of introductions [1518]; (3) diversity dynamics including either bottleneck or founder effects [1921], or absence of diversity loss [22], or even diversity enrichment in newly established populations due to multiple introductions from different sources [18, 23]; (4) hybridization, often associated with enhanced invasiveness [2426], involving multiple introduction sources [17, 27], hybridization between native and invasive species [28, 29] or introgression [30].

Such information is still scarce in case of some prominent invaders. One example is the amphipod Dikerogammarus villosus (Sowinsky, 1894), also known as the ‘killer shrimp’, which colonized most of the European main inland waters in less than 20 years [3133]. This species is an efficient, high trophic level predator [3437], characterised by a wide ecophysiological tolerance [3841] as well as by a very high fecundity [4245].

Occurrence of the species in its native area is associated predominantly with brackish lagoons (limans) and lower reaches of large rivers draining to the Black Sea. Its phylogeographic history in the native area is unknown, although it could bring key information for understanding current invasion dynamic, as in the case of other Ponto-Caspian intruders [46,47]. The distribution in the invaded continental Europe was a subject of numerous studies and is well documented (summarised in Rewicz et al. [48]). Based on the distribution pattern, two major distinct routes for the invasion of D. villosus have been proposed [49]. The eastern route would encompass Dnieper, Prypiat, Bug and Vistula rivers (Fig. 1). The western one would be composed of the Danube, Rhine, main French rivers, but also some northern sites of central Europe, such as the Mittelland Canal in Germany, and Oder in Poland (Fig. 1). However, the existence of these two distinct routes has not been firmly tested, and numerous points are still subjects of a debate.

Fig 1. Distribution of Dikerogammarus villosus in its native (N) area in the Black Sea basin (shaded area) and along Eastern (E) and Western (W) routes (R) of invasions in continental Europe as well as in UK.

Dashed lines represent canals. The presumed Western Route is indicated in red, Eastern Route in green. Numbers are dates of the first report of D. villosus at these sites. Black dots are sampling sites used in this study.

First, the origin and genetic diversity of populations found in northern and central Europe (i.e. Oder River and Mittelland Canal) is not clear. Colonization of these sites could be a westward expansion of the eastern route, from the Bug and Vistula rivers, based on the fact that other Ponto-Caspian invasive species followed such direction [49]. However, D. villosus has not been found in the waterways joining the Vistula and the Oder and the dates of first records of D. villosus from both river systems suggest that it was present in the Oder prior to colonization of the Vistula and Bug rivers [49]. Therefore, a secondary eastward extension of the western route was favored to explain D. villosus’s presence in northern Germany and western Poland. If correct, there would be two fronts of invasion in Poland (eastern and western), presently 150 km apart and likely to get in contact in the near future [31].

These two fronts might be genetically distinct. First, they originated from different parts of the native area (Fig. 1). Second, genetic differentiation between source populations of Ponto-Caspian species is already known for other invaders, such as mysids [50] and gobies [47].

A second question is the global impact of the invasion process on the genetic variation of D. villosus. So far, only three studies dealt with its genetic diversity. They focused either on molecular identification of D. villosus versus two congeneric species present in the Danube [51,52] or on its invasion dynamics in south-western Europe [53]. The latter study suggested there was no loss of genetic diversity during the invasion process. However, this study was based on few molecular markers, and it is unknown if this pattern restricts to the western route of colonization, or if it is a general pattern for D. villosus invasion. In particular, genetic variation within the eastern route has not yet been explored. Additionally, the genetic variation of the populations recently established in UK [54], as well as their source, are unknown.

In this study, D. villosus populations from the western part of the Black Sea basin (native area), as well as invasive populations from the presumed western and eastern colonization routes and from UK were genotyped for mtDNA (COI, 16S) and seven polymorphic nuclear microsatellites, in order to answer the following questions: (1) What is the species phylogeography in the native Black Sea area, including the assessment of cryptic diversity (presence of deeply divergent lineages), historical demography, level of diversity and genetic differentiation between populations being potential sources for the two presumed colonization routes? (2) Is it possible to associate a distinctive genetic signature to the two presumed colonization routes in continental Europe? (3) Is there a loss of genetic diversity between source populations and colonized areas, and is there a loss of genetic diversity and an isolation by distance along the colonization routes? (4) What was the source(s) for the newly-established UK populations and was this colonization associated with a genetic bottleneck?

Materials and Methods

Sample collection

Dikerogammarus villosus was collected from 33 sites, both in the native area (hereafter N) and the invaded part of Europe (Fig. 1, Table 1), during expeditions spanning 2002–2012. All the sampling sites were located in public and non-protected areas. No permissions were required for sampling. The study did not involve any endangered or protected species. In the native area, all suitable coastal habitats were surveyed along the western and northern coast of the Black Sea. In the invaded continental Europe, sampling covered both the putative western and eastern routes (hereafter WR and ER, respectively) and the recently invaded UK. One site in UK was sampled twice, in 2010 and 2012.

Molecular analysis

DNA from 876 samples was extracted with a standard phenol-chloroform method after Hillis et al. [54]. Air-dried DNA pellets were eluted in 100 μl of TE buffer, pH 8.00, stored at 4°C until amplification, and subsequently at -20°C long-term storage. A total of 350 specimens were amplified for two mtDNA markers: 16S ribosomal RNA (16S rRNA; ca. 320 bp fragment) with LR-J-GAM/LR-N-GAM primers [51] and reaction conditions following Grabowski et al. [14] and Cytochrome Oxydase subunit 1 gene (CO1; ca. 670 bp fragment) with LCO1490/HCO2198 primers [55] and reaction conditions following Hou et al. [56]. Sequences were obtained using BigDye sequencing protocol (Applied Biosystems 3730xl) by Macrogen Inc., Korea. Sequences were edited and aligned with ClustalW algorithm [57] using BioEdit© 7.2.5, leading to 350 sequences of 16S (303 bp) and COI (654 bp) which were concatenated to perform analyses. Haplotypes were retrieved using DnaSp v5 both for individual markers and concatenated data [58]. Haplotypes for individual markers were deposited in GenBank (accession numbers: KM208862-KM208879).

Seven microsatellite loci (msat) were used as nuclear co-dominant molecular markers: DikS, DikF [52], Dv11, Dv13, Dv17, Dv31, Dv33 [59]. A total of 876 specimens were genotyped, locus DikF amplified only for Danube and Western Route. PCR conditions were described by Wattier et al. [52] and Rewicz et al. [59]. Microsatellite alleles were visualized in 6.5% acrylamide 25 cm long gels on a LICOR 4200 L automated sequencer and scored by eye. Reference individuals were included for inter-gel calibration.

Testing for cryptic diversity

To visualize molecular divergence of mtDNA haplotypes, a Minimum Spanning Network was generated using Arlequin [60]. Pairwise Kimura 2 parameter (K2p) distances were estimated using Mega 6.2 [61]. For analysis based on Bayesian inference, the AICM method of moments’ estimator [62] was used to define best fitting model of evolution. The time calibrated phylogeny was reconstructed in BEAST, version 1.8.1 [63]. The Hasegawa, Kishino and Yano (HKY) model of evolution with proportion of invariable (I) and Yule speciation model were set for priors. The strict clock with rate 0.0142 proposed for the genus Gammarus was applied for the analyses [64]. Two runs of 20 M iterations of Markov chain Monte Carlo (MCMC) sampled each 1000 iterations were performed. Both runs were examined using Tracer v 1.6, all sampled parameters achieve sufficient sample sizes (ESS>200). Tree files were combined using LogCombiner 1.8.1 [63], with removal of the non-stationary 10% burn-in phase. The maximum clade credibility tree was generated using TreeAnnotator 1.8.1 [63]. To add additional support for the tree topology, the same dataset was analyzed with Maximum Likelihood (ML) method based on the General Time Reversible (GTR) model [65] with 10000 bootstrap replicates. Model of evolution was selected using jModelTest2 [66]. ML analyses were performed in the MEGA 6 [61].

Historical demography within the native range based on mtDNA

To reveal historical demography in the Ponto-Caspian region we used 133 individuals from nine localities (Table 1). In order to assess the temporal changes of the effective population size in each of the three phylogeographic lineages (A-C, see results), a set of the Extended Bayesian Skyline Plot (EBSP) analyses [67] was performed in BEAST, version 1.8.1 [63]. The GTR model of evolution was used as the best fitting model. To ensure convergence, four runs of MCMC, 100M iterations long sampled each 1000 iterations, were performed. Both runs were examined using Tracer v 1.6, all sampled parameters achieved sufficient sample sizes (ESS>200).

Allelic/haplotypic diversity and differentiation

Diversity was assessed by calculating: (1) allelic-haplotypic (msat/mtDNA) diversity (k), (2) allelic richness (Ar) and private allelic richness (PAr) corrected for a common sampling size using rarefaction approach [68]. Calculations were performed with Hp-Rare 1.1 [69], differentiation in Ar was tested using the non-parametric Mann-Whitney U-test in Statistica 10 [70], and put in brackets if significant. In addition, observed heterozygosity (HO), expected heterozygosity (HE) and fixation index (FIS) were calculated, when appropriate, for microsatellite markers using Fstat [71]. Pairwise differentiation was determined by two FST estimators: ΘST with Tamura-Nei distance for mtDNA [72] and Θ for microsatellites [73], both implemented in Arlequin, statistical significance being measured using 10000 permutations. Genetic diversity and FST were assessed either pooling sampling sites, or not, according the hypothesis tested, e.g. between fronts in Poland.

Population structure was also analyzed using individual-based Bayesian clustering method implemented in Structure 2.3.4 [74]. Simulations were performed on the full data set including 29 populations and 876 individuals. Runs for each possible value of K (1 to 8) were repeated 20 times. Each run used a burn-in of 500000 iterations, a run length of 750000 iterations. All simulations were performed using the admixture and correlated allele frequencies models with no prior information. Selection of most probable value of K relied on the ΔK method developed by Evanno et al. [75].

Diversity and differentiation along Western Route (WR)

Based on 20 sites along the WR we tested if microsatellite differentiation increased positively with distance between sites (isolation-by-distance, hereafter IBD) but also if diversity (mean allelic richness) was associated with geographical distance from the source area (Danube delta). The distances were estimated using Google Earth v.7.1.2. IBD was tested using Mantel test between FST / (1- FST) and geographic distance as recommended by Rousset [76] for testing IBD in one-dimensional linear systems, with 100000 permutations, using the GenePop on the Web 4.2 [77] and ISOLDE software.


Phylogeography in the native Black Sea area

Out of 133 individuals from 9 sites in the native Black Sea region, a total of 17 haplotypes were identified based on concatenated (303+654 bp) 16S and COI mtDNA sequences (S1 Table). First, the difference observed between the most divergent haplotypes was only five nucleotides (Fig. 2). Second, the mean overall K2p genetic distance between haplotypes was very low being 0.0009 (SD 0.0004). It showed clearly that there is no cryptic diversity involving highly divergent lineages. However, combination of the haplotype network (Fig. 2) and Bayesian phylogenetic reconstruction (Fig. 3A) revealed that the haplotypes may be grouped into three phylogenetic lineages. Their spatial distribution is partly structured geographically. Lineage A includes eight haplotypes (5–12) specific to the Durungol liman in Turkey (2-N). Lineage B includes 7 haplotypes (1, 4, 13, 15–17) and lineage C includes the 3 remaining haplotypes (2, 3 and 14). The Bayesian chronogram showed that C diverged from A+B ca 280 kyr BP, while A and B split ca. 200 kyr BP (Fig. 3A). The results of the EBSP analysis indicated that population of D. villosus in the Durungol liman experienced steady growth for the last 20 ky, while populations of the remaining two lineages in the native area remained stable for most of the last 30k years, with accelerated growth starting less than 10 ky ago (Fig. 3B).

Fig 2. Geographical distribution of D. villosus haplotypes in the native and invaded area.

Numbers near black circles represent sampling localities coded as in Table 1. In upper right corner is the minimum spanning network of mtDNA haplotypes identified in D. villosus.

Fig 3. Phylogeny and demographic history of D. villosus in the native Black Sea area.

(A) Maximum clade credibility chronogram inferred from a strict molecular clock model based on the concatenated COI+16S data set of Dikerogammarus villosus. The two numbers given next to the respective node indicate Bayesian posterior probabilities (> 0.5), and maximum likelihood bootstrap values (> 50%). (B) Multilocus extended Bayesian skyline plots for three linages of Dikerogammarus villosus. Solid lines indicate the median posterior effective population size through time; dashed lines indicate the 95% highest posterior density interval for each estimate.

The highest diversity for mtDNA and msat was observed in Durungol (Ar = 5.3 and 5.73 respectively); the locality harboring also the highest private allelic richness (PAr = 5.3 and 1.08) (Table 2). The potential (i.e. a priori) source areas of invasion, i.e. Dnieper and Danube deltas, did not exhibit the same level of mtDNA diversity (Ar = 2 and 3) but harbored some specific haplotypes (Table 2, Fig. 2). For msat these areas had similar diversities (Ar = 3.94 and 3.57) and low private allelic richness (PAr = 0.05 and 0).

Table 2. Genetic diversity for Dikerogammarus villosus calculated for microsatellites (msat) or mitochondrial DNA (mtDNA) within sampling sites or groups of sites.

Genetic differentiation differed from zero for all area pairwise comparisons, for both msat and mtDNA data (Table 3). However, the level of differentiation was heterogeneous, being the highest between Durungol and Dnieper (FST = 0.180 and ΘST = 0.693) and the lowest between Danube and Dniester (FST = 0.048 and ΘST = 0.103). The results of Bayesian clustering suggest that the four selected areas may represent four genetic clusters, although the division is not strict. The Durungol and Dnieper populations are the most homogeneous ones, while the Dniester and Danube populations show symptoms of migration or very recent common ancestry (Fig. 4A,B).

Fig 4. Bayesian clustering of Dikerogammarus villosus based on six microsatellite loci in 29 populations collected in native and invaded area with K = 3 (A); and K = 5 (B).

Acronyms refer to sites as explained in Table 1. Each individual is represented by a thin vertical line, with proportional membership in different clusters indicated by colors. Black vertical lines separates sampling sites, with site identification indicated below the plot.

Table 3. Genetic pairwise differentiation for Dikerogammarus villosus estimated for microsatellites (FST, below diagonal) or mitochondrial DNA (ΘST, above the diagonal) between sites or group of sites in the native area and invasion fronts in Poland.

Colonization dynamics in Continental Europe

For mtDNA, ER and WR in Poland, i.e. fronts, are not differentiated from their respective putative sources in the native region, i.e. Dnieper (3+4+5) and Danube (6+7+8+9), respectively (Table 3), while differentiation between fronts was significant (Table 3). For msat, although all pairwise comparisons for differentiation were significant (Table 3), the level of differentiation between fronts in Poland and their respective putative areas of origin was low (FST = 0.027 and = 0.019) compared to differentiation among sites belonging to different routes (0.157 < FST < 0.215) (Table 3). Bayesian clustering analysis showed clearly, that individuals from the western front (Poland), the western route and the source in Danube from a homogeneous genetic unit, while the eastern front with its putative source in Dnieper form another homogeneous genetic unit (Fig. 4A,B).

Both fronts in Poland had the same level of diversity for msat compared to their putative source in the native region i.e. in Dnieper and Danube respectively (Table 2). This is also true for mtDNA for beginning the end of the ER, but the WR in Poland seems to present less diversity than the Danube (6+7+8+9) (Table 2). Along the WR, geographic distance from the source area did not explain msat diversity within sites (Fig. 5). Along the 4500 km long route, only one site located on the River Vah in Slovakia (site 16) had lower diversity. No isolation by distance (Fig. 6) was present as no significant correlation was detected between pairwise geographical distances and genetic differentiation (Mantel test, R2 = 0.0047, P = 0.42).

Fig 5. Allelic richness across seven microsatellite loci (mean ± standard deviation) within 20 populations of Dikerogammarus villosus from Western Route (WR) plotted against linear distance from Danube delta to each site along the route (see Fig. 1. for details).

Black diamond indicate Vah river site (site 16), see text for details.

Fig 6. Plot of pairwise FST (seven microsatellite loci) versus pairwise linear distance of 20 populations of Dikerogammarus villosus encompassing source populations for the Western Route (WR), the WR itself and populations in the western front in Poland.

Black diamonds highlight pairwise comparisons with Vah river site (site 16), see text for details.

Source population and diversity for the UK

We observed high genetic differentiation for both mtDNA (0.206 < ΘST < 0.298) and msat (0.138 < FST < 0.172), for pairwise comparisons between ER in Poland and two pooled sites in the Netherlands (23+24) or any UK site. On the opposite, pairwise comparisons between two sites in the Netherlands and each UK site showed no significant differentiation for mtDNA (Table 4). For msat, lower level of differentiation was observed between the UK site 32 (A and B) and the Netherlands (0.026 < FST < 0.035) than between the UK site and ER in Poland (0.138 < FST < 0.172). The UK site 33 showed a less conclusive picture for FST (Table 4). Bayesian clustering analysis (Fig. 4A,B) showed that the UK populations form a homogeneous genetic unit with the western route, and genetically different from the eastern route.

Table 4. Genetic pairwise differentiation estimated for microsatellites (FST, below diagonal) or mitochondrial DNA (ΘST, above the diagonal) between sites or group of sites in the possible source populations for UK and UK sites.

Only one mtDNA haplotype (haplotype 1), the most common in continental Europe, occurred in UK. Haplotype 4, while common in ER, was absent from the UK. In Grafham Water site (32A, 32B) we observed no loss of diversity for msat compared to the Netherlands with Ar values being respectively 3.57 and 3.49, versus 3.8 (Table 2). Diversity in the more recent population (33) was 2.49 but did not differ in the statistical terms from the above Ar values (Table 2).


The Ponto-Caspian region has been recognized as the most prominent donor of non-indigenous hydrobionts to Europe and to the North American Great Lakes system. Their taxonomic spectrum is wide including amphipods, mysids, cladocerans, gastropods and fishes [7982]. Phylogeography and population genetics patterns of these invaders may help in understanding colonization dynamics and in controlling their further spread [16]. Our results confirm that the invasion scheme for one of these species, D. villosus, is complex, with multiple routes, and that the loss of genetic diversity during the course of colonization is weak. We evidenced that colonization of the UK was originating from one (here Western) out of the two genetic units associated with continental invasion.

Phylogeography and contemporary genetic structure in the native region

Level of cryptic diversity for invertebrates in the Ponto-Caspian region is highly variable. Deep level of divergence, but below the species threshold, was observed for e.g. cladocerans [78] and mysids [79]. Cryptic diversity was revealed e.g. in monkey goby [80]. Although the phenomenon is known to occur in several amphipods [81, 82], no cryptic diversity was detected in D. villosus in its native region. The 17 mtDNA haplotypes showed shallow divergence with an overall K2p genetic distance of 0.0009, far below the threshold of 0.03–0.055 identified between crustacean species [83]. Such low divergence may be related to very recent history of the species within the Black Sea region (see below).

Although shallow, the divergence between mtDNA haplotypes is geographically structured with lineage A including a set of 8 haplotypes (out of 17) and being restricted to the Durungol liman. This pool of haplotypes separated from others ca 200 kyr BP. In addition, the other sampled areas are also characterized by high private haplotypic richness. Overall, the level of differentiation (ΘST) between populations is high in the Black Sea area. The turbulent Pleistocene hydro-geological history of the region with recurrent changes of sea level and salinity may be among the most powerful driving forces explaining this pattern [46, 50, 78, 79, 8486]. During the last 670 kyr, there were at least 12 significant saline water intrusions from the Mediterranean Sea, and eight intrusions from the Caspian Lake to the Black Sea [87]. These events caused water level fluctuations and substantial salinity shifts from nearly fresh to full marine conditions that could cause shift ranges and population fragmentations in oligohaline hydrobionts inhabiting this basin [88]. During fully marine salinity stages, slightly brackish estuaries and limans may have become isolated refugia and differentiation centers for local aquatic fauna. The dating of divergence between Durungol liman and others sites coincides with one of the most prominent salinity raises [87]. In other sites presence of shared haplotypes reflects probably both recent and historical migration events among various areas in the native region. However, the overall presence of private haplotypes and high differentiation level indicate possible founder effects at the time of colonization. It could be followed either by restricted gene flow (which is confirmed by msat results) or even by allopatric divergence during stages of raised salinity. The results of EBSP analyses support very recent post-Pleistocene demographic expansion, suggesting that ecological conditions were locally favorable over evolutionary time-scale. Phylogeographic structure was already observed in mtDNA of Ponto-Caspian mysids [50, 79, 84], cladocerans [78] and gammarids [46, 78, 86].

The msat analyses pointed out high level of differentiation (FST) between the four native sampled areas, that were divided in four genetic clusters in Bayesian analyses. If the Durungol liman is clearly isolated from the other sampled areas, the latter show connectivity as pointed out by unclear Bayesian assignment of some individuals. One of the explanations can be a shorter geographic distance between these populations. The north-western Black Sea is also the shallowest and least saline, due to massive sedimentation and inflow of riverine waters from e.g. Dnieper and Danube rivers [89, 90]. Chances for migrations between these populations are high, including anthropogenic transport due to the high ship traffic between local ports.

Invasion routes and dynamics in continental Europe

In our study, combination of mtDNA and msat analyses clearly identified Danube and Dnieper deltas as differentiated sources for the two invasion routes we named “Western” and “Eastern”, respectively. Dikerogammarus villosus has been highly monitored throughout Europe due to its detrimental impact on the ecosystem. Therefore, accurate map of invasion progress can easily be drawn and converted into the most likely scenario for colonization routes [33, 8992]. Agreement of our results with putative routes might seem trivial at first sight. However, few studies upon other species pointed out that molecular data identified routes that were different from the most likely, census-based, scenarios [9, 93]. Based on geographic invasion patterns of several aquatic species, Bij de Vaate et al. [91] defined three invasion corridors from the Ponto-Caspian region into continental Europe i.e. the northern (Volga River, Beloye, Onega and Ladoga lakes, Neva River to the Baltic Sea), the central (Dnieper, Pripyat, Pripyat-Bug channel, Vistula, Oder, Mittelland canal) and the southern (Danube, Rhine) one. Numerous species invasions fitted this pattern [49, 94]. Contrary to other species, D. villosus used only the eastern part of the central corridor and has not passed the Bydgoski channel in Poland which is connecting the Vistula and the Oder rivers (Fig. 1). On the other hand, the western part of the central corridor was colonized eastwards by population which came up the entire southern corridor westward. Possibly, the Bydgoski channel, with its prominently soft bottom, slow current and abundant vegetation is not prone to be colonized by D. villosus [95]. However, the closely related invader, D. haemobaphes, along with some other invasive amphipods, such as Echinogammarus ischnus and Chelicorophium curvispinum, managed to pass the Bydgoski channel and use the entire central corridor westwards. This channel was an important shipping route until mid-20th century, but the traffic now is heavily limited [96]. Yet, we cannot exclude possibility of future contact between these two distinct populations of D. villosus. The two fronts in Poland are differentiated and characterized by level of diversity analogous to their source regions. If the two fronts meet, hybridization will occur as the two populations are not phylogenetically and ecologically divergent which implies the absence of reproduction barrier. This may result in producing a potential “super-hybrid” – an even more effective invader, as it was observed in other cases [25, 26]. Thus, the situation deserves particular surveillance and management to avoid contact between these two fronts.

The WR has a length of about 4500 km from the source population in the Black Sea region to the invasion front in Poland. All mtDNA haplotypes found in the native range were observed in the invaded area, 3 out of 8 non-frequent haplotypes being present in the last 1000 km. In addition, in the Mittelland Canal we found one haplotype (haplotype 18) not even encountered in the native area, probably due to its very low frequency. For microsatellites, we found no loss of diversity along the route and no isolation by distance. Globally, it suggests that no bottlenecks occurred along WR. Similar conclusions were made by Müller et al. [51] and Wattier et al. [53], who conducted research on a smaller scale or with fewer genetic markers. Even if reduction of diversity in the invaded areas was often expected in the literature in 20th century, numerous studies since then have shown it might be far from being the rule [9799]. Lack of diversity loss may result from a very large propagule size i.e. large founding population, and/or propagule frequency i.e. recurrent waves of invaders. We suspect the latter to play an important role for D. villosus. Indeed, no loss of diversity is observed while genetic differentiation between sites is present. This suggests that recurrent waves are both maintaining allelic diversity at a high level, and reshuffling allelic frequencies what generates differentiation.

Only one site (site 16) in the middle section of River Vah, was characterized by both low diversity and high level of differentiation from other sites. Strong founder effect is likely to explain this pattern as, probably, the site was sampled very recently after D. villosus first colonization [48, 100].

Recent UK overseas conquest

The overseas introduction of D. villosus into UK in 2010 was noticed in popular media [54]. It proved clearly that large funds spent on biosecurity programs to prevent the spread of invasive species (i.e. the procedure “check, clean and dry”) have been insufficient to stop the killer shrimp [101, 102]. Both FST and Bayesian clustering of msat allowed us to exclude the populations from ER as donors for the UK sites. In addition, the haplotype 4, both frequent and private to ER, was not detected in UK, although our sampling size was limited in this case.

Diversity in any UK population was not different from the continent and no bottleneck effect was observed. Apparently, the propagule pressure was high enough to alleviate diversity loss. We are not able to conclude whether the introduction to UK was a single event followed by secondary colonization or multiple introductions. Anyway, the killer shrimp is spreading very efficiently throughout the UK. Furthermore, another congeneric invader, the “demon shrimp”, D. haemobaphes, has already been recorded in UK [103]. Based on several possible expansion models of both D. villosus and D. haemobaphes it has been estimated that more than 60% of the UK waterbodies is suitable and vulnerable to colonisation by these two invaders [101, 104107]. Moreover, high popularity of water sports may further accelerate the invasion [108], due to high ability of the killer shrimp to spread via boating and diving equipment [32].


Our results identified impact of the Pleistocene sea level and salinity fluctuations on the phylogeographic structure of Dikerogammarus villosus in the Black Sea native region and presence of two differentiated source populations, i.e. the Danube and Dnieper deltas. These sources are associated with two independent invasion routes (Western and Eastern) in continental Europe for which no loss of diversity is observed. We can expect further spread of the killer shrimp in continental Europe, even in smaller tributaries. MacNeil & Platvoet [95] pointed out that solid objects, like concrete fish passages, could be used by D. villosus as mainstay in smaller tributaries. This may pose a threat to native gammarids occupying such refuges [109].

The UK population has probably originated in the Western Route and despite very recent colonization, no drastic loss of diversity was observed. This recent overseas conquest provides rather non-optimistic message, accounting that the UK authorities implemented preventive biosecurity protocols and risk assessments. The Great Lakes of North America are likely to be the next step, since other Ponto-Caspian invertebrates already managed to reach them [110, 111].

Finally, the Dniester native area is characterized by high msat allelic diversity (including private alleles). Thus, even if at the moment it is not a source population for the colonization of Europe, it may act as a donor for other source areas if the anthropogenic transport increases, enhancing the local genetic diversity.

Supporting Information

S1 Table. GeneBank accession numbers of COI and 16S haplotypes.



Authors thank Alicja Konopacka, Nina Fielding, Paul Stebbing, Jamie Bojko, Krešimir Žganec, Tomasz Mamos and Radomir Jaskuła for their help in collecting the material for this study. Special thanks are due to Tomasz Mamos for his help and assistance in performing the phylogenetic analyses and EBSP.

Author Contributions

Conceived and designed the experiments: KBS RW MG T. Rewicz T. Rigaud. Performed the experiments: T. Rewicz MG KBS RW. Analyzed the data: T. Rewicz RW KBS MG. Wrote the paper: T. Rewicz RW KBS MG. Prepared graphics: T. Rewicz KBS. Conceived the ideas KBS RW MG T. Rewicz T. Rigaud.


  1. 1. Lambertini M, Leape J, Marton-Lefevre J, Mitter-Meier RA, Rose M, et al. (2011) Invasives: A Major Conservation Threat. Science 333: 404–405. pmid:21778381
  2. 2. Richardson DM, Ricciardi A (2013) Misleading criticisms of invasion science: a field guide. Diversity and Distributions 19: 1461–1467.
  3. 3. Simberloff D, Martin JL, Genovesi P, Maris V, Wardle DA, et al. (2013) Impacts of biological invasions: what's what and the way forward. Trends in Ecology & Evolution 28: 58–66. pmid:25619764
  4. 4. Paolucci EM, MacIsaac HJ, Ricciardi A (2013) Origin matters: alien consumers inflict greater damage on prey populations than do native consumers. Diversity and Distributions 19: 988–995.
  5. 5. van Riel MC, Healy EP, van der Velde G, de Vaate AB (2007) Interference competition among native and invader amphipods. Acta Oecologica-International Journal of Ecology 31: 282–289.
  6. 6. van Riel MC, van der Velde G, de Vaate AB (2009) Interference competition between alien invasive gammaridean species. Biological Invasions 11: 2119–2132.
  7. 7. Tompkins DM, Poulin R (2006) Parasites and Biological Invasions. In: Allen RD, Lee WG, editors. Biological Invasions in New Zealand. Berlin, Heidelberg: Springer. pp. 67–86.
  8. 8. Zaiko A, Daunys D, Olenin S (2009) Habitat engineering by the invasive zebra mussel Dreissena polymorpha (Pallas) in a boreal coastal lagoon: impact on biodiversity. Helgoland Marine Research 63: 85–94.
  9. 9. Estoup A, Guillemaud T (2010) Reconstructing routes of invasion using genetic data: why, how and so what? Molecular Ecology 19: 4113–4130. pmid:20723048
  10. 10. Fitzpatrick BM, Fordyce JA, Niemiller ML, Reynolds RG (2012) What can DNA tell us about biological invasions? Biological Invasions 14: 245–253.
  11. 11. Bastrop R, Jurss K, Sturmbauer C (1998) Cryptic species in a marine polychaete and their independent introduction from North America to Europe. Molecular Biology and Evolution 15: 97–103. pmid:9491608
  12. 12. Larson ER, Abbott CL, Usio N, Azuma N, Wood KA, et al. (2012) The signal crayfish is not a single species: cryptic diversity and invasions in the Pacific Northwest range of Pacifastacus leniusculus. Freshwater Biology 57: 1823–1838.
  13. 13. Pilgrim EM, Blum MJ, Reusser DA, Lee H, Darling JA (2013) Geographic range and structure of cryptic genetic diversity among Pacific North American populations of the non-native amphipod Grandidierella japonica. Biological Invasions 15: 2415–2428.
  14. 14. Grabowski M, Rewicz T, Bacela-Spychalska K, Konopacka A, Mamos T, et al. (2012) Cryptic invasion of Baltic lowlands by freshwater amphipod of Pontic origin. Aquatic Invasions 7: 337–346.
  15. 15. Brown JE, Stepien CA (2008) Ancient divisions, recent expansions: phylogeography and population genetics of the round goby Apollonia melanostoma. Molecular Ecology 17: 2598–2615. pmid:18466236
  16. 16. Muirhead JR, Gray DK, Kelly DW, Ellis SM, Heath DD, et al. (2008) Identifying the source of species invasions: sampling intensity vs. genetic diversity. Molecular Ecology 17: 1020–1035. pmid:18261046
  17. 17. Rosenthal DM, Ramakrishnan AP, Cruzan MB (2008) Evidence for multiple sources of invasion and intraspecific hybridization in Brachypodium sylvaticum (Hudson) Beauv. in North America. Molecular Ecology 17: 4657–4669. pmid:18627455
  18. 18. Zhan AB, Perepelizin PV, Ghabooli S, Paolucci E, Sylvester F, et al. (2012) Scale-dependent post-establishment spread and genetic diversity in an invading mollusc in South America. Diversity and Distributions 18: 1042–1055.
  19. 19. Astanei I, Gosling E, Wilson J, Powell E (2005) Genetic variability and phylogeography of the invasive zebra mussel, Dreissena polymorpha (Pallas). Molecular Ecology 14: 1655–1666. pmid:15836640
  20. 20. Dybdahl MF, Drown DM (2011) The absence of genotypic diversity in a successful parthenogenetic invader. Biological Invasions 13: 1663–1672.
  21. 21. Slothouber Galbreath JGM, Smith JE, Becnel JJ, Butlin RK, Dunn AM (2010) Reduction in post-invasion genetic diversity in Crangonyx pseudogracilis (Amphipoda: Crustacea): a genetic bottleneck or the work of hitchhiking vertically transmitted microparasites? Biological Invasions 12: 191–209.
  22. 22. Diez-del-Molino D, Carmona-Catot G, Araguas RM, Vidal O, Sanz N, et al. (2013) Gene Flow and Maintenance of Genetic Diversity in Invasive Mosquitofish (Gambusia holbrooki). Plos One 8. pmid:24482673
  23. 23. Gillis NK, Walters LJ, Fernandes FC, Hoffman EA (2009) Higher genetic diversity in introduced than in native populations of the mussel Mytella charruana: evidence of population admixture at introduction sites. Diversity and Distributions 15: 784–795.
  24. 24. Ayres D, Schierenbeck K, Ellstrand N (2009) Introduction to "Ecological and Evolutionary Consequences after Invaders Hybridize". Biological Invasions 11: 1087–1088.
  25. 25. Benvenuto C, Cheyppe-Buchmann S, Bermond G, Ris N, Fauvergue X (2012) Intraspecific hybridization, life history strategies and potential invasion success in a parasitoid wasp. Evolutionary Ecology 26: 1311–1329.
  26. 26. Facon B, Jarne P, Pointier JP, David P (2005) Hybridization and invasiveness in the freshwater snail Melanoides tuberculata: hybrid vigour is more important than increase in genetic variance. Journal of Evolutionary Biology 18: 524–535. pmid:15842482
  27. 27. Luquet E, Vorburger C, Hervant F, Joly P, Kaufmann B, et al. (2011) Invasiveness of an introduced species: the role of hybridization and ecological constraints. Biological Invasions 13: 1901–1915.
  28. 28. Matsukura K, Okuda M, Cazzaniga NJ, Wada T (2013) Genetic exchange between two freshwater apple snails, Pomacea canaliculata and Pomacea maculata invading East and Southeast Asia. Biological Invasions 15: 2039–2048.
  29. 29. Meraner A, Venturi A, Ficetola GF, Rossi S, Candiotto A, et al. (2013) Massive invasion of exotic Barbus barbus and introgressive hybridization with endemic Barbus plebejus in Northern Italy: where, how and why? Molecular Ecology 22: 5295–5312. pmid:24103005
  30. 30. Muhlfeld CC, Kalinowski ST, McMahon TE, Taper ML, Painter S, et al. (2009) Hybridization rapidly reduces fitness of a native trout in the wild. Biology Letters 5: 328–331. pmid:19324629
  31. 31. Bącela KO, Grabowski M, Konopacka A (2008) Dikerogammarus villosus (Sowinsky, 1894) (Crustacea Amphipoda) enters Vistula—the biggest river in the Baltic basin. Aquatic Invasions 3: 95–98.
  32. 32. Bącela-Spychalska K, Grabowski M, Rewicz T, Konopacka A, Wattier R (2013) The "killer shrimp' Dikerogammarus villosus (Crustacea, Amphipoda) invading Alpine lakes: overland transport by recreational boats and scuba-diving gear as potential entry vectors? Aquatic Conservation-Marine and Freshwater Ecosystems 23: 606–618.
  33. 33. Bollache L, Devin S, Wattier R, Chovet M, Beisel JN, et al. (2004) Rapid range extension of the Ponto-Caspian amphipod Dikerogammarus villosus in France: potential consequences. Archiv Fur Hydrobiologie 160: 57–66.
  34. 34. Casellato S, Visentin A, La Piana G (2007) The predatory impact of Dikerogammarus villosus, a danger for fish. In: Gherardi F, editor. Biological invaders in inland waters: profiles, distribution, and threats. Dordrecht, The Netherlands: Springer. pp. 495–506.
  35. 35. Dick JTA, Platvoet D, Kelly DW (2002) Predatory impact of the freshwater invader Dikerogammarus villosus (Crustacea: Amphipoda). Canadian Journal of Fisheries and Aquatic Sciences 59: 1078–1084.
  36. 36. MacNeil C, Platvoet D (2005) The predatory impact of the freshwater invader Dikerogammarus villosus on native Gammarus pulex (Crustacea: Amphipoda); influences of differential microdistribution and food resources. Journal of Zoology 267: 31–38.
  37. 37. van Riel MC, van der Velde G, Rajagopal S, Marguillier S, Dehairs F, et al. (2006) Trophic relationships in the Rhine food web during invasion and after establishment of the Ponto-Caspian invader Dikerogammarus villosus. Hydrobiologia 565: 39–58.
  38. 38. Brooks SJ, Platvoet D, Mills CL (2008) Cation regulation and alteration of water permeability in the amphipod Dikerogammarus villosus: an indicator of invasion potential. Fundamental and Applied Limnology 172: 183–189.
  39. 39. Bruijs MCM, Kelleher B, van der Velde G, de Vaate AB (2001) Oxygen consumption, temperature and salinity tolerance of the invasive amphipod Dikerogammarus villosus: indicators of further dispersal via ballast water transport. Archiv Fur Hydrobiologie 152: 633–646.
  40. 40. Piscart C, Kefford BJ, Beisel JN (2011) Are salinity tolerances of non-native macroinvertebrates in France an indicator of potential for their translocation in a new area? Limnologica 41: 107–112.
  41. 41. Wijnhoven S, van Riel MC, van der Velde G (2003) Exotic and indigenous freshwater gammarid species: physiological tolerance to water temperature in relation to ionic content of the water. Aquatic Ecology 37: 151–158.
  42. 42. Devin S, Piscart C, Beisel JN, Moreteau JC (2004) Life history traits of the invader Dikerogammarus villosus (Crustacea: Amphipoda) in the Moselle River, France. International Review of Hydrobiology 89: 21–34.
  43. 43. Grabowski M, Bącela K, Konopacka A (2007a) How to be an invasive gammarid (Amphipoda: Gammaroidea)-comparison of life history traits. Hydrobiologia 590: 75–84.
  44. 44. Pöckl M (2007) Strategies of a successful new invader in European fresh waters: fecundity and reproductive potential of the Ponto-Caspian amphipod Dikerogammarus villosus in the Austrian Danube, compared with the indigenous Gammarus fossarum and G. roeseli. Freshwater Biology 52: 50–63.
  45. 45. Pöckl M (2009) Success of the invasive Ponto-Caspian amphipod Dikerogammarus villosus by life history traits and reproductive capacity. Biological Invasions 11: 2021–2041.
  46. 46. Cristescu MEA, Witt JDS, Grigorovich IA, Hebert PDN, MacIsaac HJ (2004) Dispersal of the Ponto-Caspian amphipod Echinogammarus ischnus: Invasion waves from the Pleistocene to the present. Heredity 92: 197–203. pmid:14981532
  47. 47. Ohayon JL, Stepien CA (2007) Genetic and biogeographic relationships of the racer goby Neogobius gymnotrachelus (Gobiidae: Teleostei) from introduced and native Eurasian locations. Journal of Fish Biology 71: 360–370.
  48. 48. Rewicz T, Grabowski M, MacNeil C, Bacela-Spychalska K (2014a) The profile of a ‘perfect’ invader – the case of killer shrimp, Dikerogammarus villosus. Aquatic Invasions 9: 267–288.
  49. 49. Grabowski M, Jażdżewski K, Konopacka A (2007b) Alien Crustacea in Polish waters – Amphipoda. Aquatic Invasions 2: 25–38.
  50. 50. Audzijonyte A, Wittmann KJ, Ovcarenko I, Vainola R (2009) Invasion phylogeography of the Ponto-Caspian crustacean Limnomysis benedeni dispersing across Europe. Diversity and Distributions 15: 346–355.
  51. 51. Müller JC, Schramm S, Seitz A (2002) Genetic and morphological differentiation of Dikerogammarus invaders and their invasion history in Central Europe. Freshwater Biology 47: 2039–2048.
  52. 52. Wattier RA, Beguet J, Gaillard M, Muller JC, Bollache L, et al. (2006) Molecular markers for systematic identification and population genetics of the invasive Ponto-Caspian freshwater gammarid Dikerogammarus villosus (Crustacea, Amphipoda). Molecular Ecology Notes 6: 487–489.
  53. 53. Wattier RA, Haine ER, Beguet J, Martin G, Bollache L, et al. (2007) No genetic bottleneck or associated microparasite loss in invasive populations of a freshwater amphipod. Oikos 116: 1941–1953.
  54. 54. MacNeil C, Platvoet D, Dick JTA, Fielding N, Constable A, et al. (2010) The Ponto-Caspian ‘killer shrimp’, Dikerogammarus villosus (Sowinsky, 1894), invades the British Isles. Aquatic Invasions 5: 441–445.
  55. 55. Folmer O, Black M, Hoeh W, Lutz R, Vrijenhoek R (1994) DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Molecular marine biology and biotechnology 3: 294–299. pmid:7881515
  56. 56. Hou ZG, Fu JH, Li SQ (2007) A molecular phylogeny of the genus Gammarus (Crustacea: Amphipoda) based on mitochondrial and nuclear gene sequences. Molecular Phylogenetics and Evolution 45: 596–611. pmid:17686635
  57. 57. Chenna R, Sugawara H, Koike T, Lopez R, Gibson TJ, et al. (2003) Multiple sequence alignment with the Clustal series of programs. Nucleic Acids Research 31: 3497–3500. pmid:12824352
  58. 58. Librado P, Rozas J (2009) DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25: 1451–1452. pmid:19346325
  59. 59. Rewicz T, Wattier R, Rigaud T, Bacela-Spychalska K, Grabowski M (2014b) Isolation and characterization of 8 microsatellite loci for the “killer shrimp”, an invasive Ponto-Caspian amphipod Dikerogammarus villosus (Crustacea: Amphipoda). Molecular Biology Reports: in press.
  60. 60. Excoffier L, Lischer HEL (2010) Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources 10: 564–567. pmid:21565059
  61. 61. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S (2013) MEGA6: Molecular Evolutionary Genetics Analysis version 6.0. Mol Biol Evol 30: 2725–2729. pmid:24132122
  62. 62. Baele G, Li WLS, Drummond AJ, Suchard MA, Lemey P (2013) Accurate Model Selection of Relaxed Molecular Clocks in Bayesian Phylogenetics. Molecular Biology and Evolution 30: 239–243. pmid:23090976
  63. 63. Drummond AJ, Suchard MA, Xie D, Rambaut A (2012) Bayesian Phylogenetics with BEAUti and the BEAST 1.7. Molecular Biology and Evolution 29: 1969–1973. pmid:22367748
  64. 64. Hou ZG, Li JB, Li SQ (2014) Diversification of low dispersal crustaceans through mountain uplift: a case study of Gammarus (Amphipoda: Gammaridae) with descriptions of four novel species. Zoological Journal of the Linnean Society 170: 591–633.
  65. 65. Nei M, Kumar S (2000) Molecular Evolution and Phylogenetics. New York: Oxford University Press. 333 p.
  66. 66. Darriba D, Taboada GL, Doallo R, Posada D (2012) jModelTest 2: more models, new heuristics and parallel computing. Nature Methods 9: 772–772. pmid:22847109
  67. 67. Heled J, Drummond AJ (2008) Bayesian inference of population size history from multiple loci. Bmc Evolutionary Biology 8. pmid:19113997
  68. 68. Leberg PL (2002) Estimating allelic richness: Effects of sample size and bottlenecks. Molecular Ecology 11: 2445–2449. pmid:12406254
  69. 69. Kalinowski ST (2005) HP-RARE 1.0: a computer program for performing rarefaction on measures of allelic richness. Molecular Ecology Notes 5: 187–189.
  70. 70. StatSoft Inc. S (2011) STATISTICA (data analysis software system), version 10.
  71. 71. Goudet J (2001) FSTAT, a program to estimate and test gene diversities and fixation indices. 2.9.3 ed.
  72. 72. Tamura K, Nei M (1993) Estimation of the Number of Nucleotide Substitutions in the Control Region of Mitochondrial-DNA in Humans and Chimpanzees. Molecular Biology and Evolution 10: 512–526. pmid:8336541
  73. 73. Weir BS, Cockerham CC (1984) Estimating F-Statistics for the Analysis of Population-Structure. Evolution 38: 1358–1370.
  74. 74. Pritchard JK, Stephens M, Donnelly P (2000) Inference of population structure using multilocus genotype data. Genetics 155: 945–959. pmid:10835412
  75. 75. Evanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Molecular Ecology 14: 2611–2620. pmid:15969739
  76. 76. Rousset F (1997) Genetic differentiation and estimation of gene flow from F-statistics under isolation by distance. Genetics 145: 1219–1228. pmid:9093870
  77. 77. Raymond M, Rousset F (1995) Genepop (Version-1.2)—Population-Genetics Software for Exact Tests and Ecumenicism. Journal of Heredity 86: 248–249.
  78. 78. Cristescu MEA, Hebert PDN, Onciu TM (2003) Phylogeography of Ponto-Caspian crustaceans: A benthic-planktonic comparison. Molecular Ecology 12: 985–996. pmid:12753217
  79. 79. Audzijonyte A, Daneliya ME, Vainola R (2006) Comparative phylogeography of Ponto-Caspian mysid crustaceans: isolation and exchange among dynamic inland sea basins. Molecular Ecology 15: 2969–2984. pmid:16911214
  80. 80. Neilson ME, Stepien CA (2011) Historic speciation and recent colonization of Eurasian monkey gobies (Neogobius fluviatilis and N. pallasi) revealed by DNA sequences, microsatellites, and morphology. Diversity and Distributions 17: 688–702.
  81. 81. Mamos T, Wattier R, Majda A, Sket B, Grabowski M (2014) Morphological vs. molecular delineation of taxa across montane regions in Europe: the case study of Gammarus balcanicus Schaferna, 1922 (Crustacea: Amphipoda). Journal of Zoological Systematics and Evolutionary Research 52: 237–248.
  82. 82. Wysocka A, Grabowski M, Sworobowicz L, Mamos T, Burzynski A, et al. (2014) Origin of the Lake Ohrid gammarid species flock: ancient local phylogenetic lineage diversification. Journal of Biogeography 41: 1758–1768.
  83. 83. Costa FO, deWaard JR, Boutillier J, Ratnasingham S, Dooh RT, et al. (2007) Biological identifications through DNA barcodes: the case of the Crustacea. Canadian Journal of Fisheries and Aquatic Sciences 64: 272–295.
  84. 84. Audzijonyte A, Wittmann KJ, Vainola R (2008) Tracing recent invasions of the Ponto-Caspian mysid shrimp Hemimysis anomala across Europe and to North America with mitochondrial DNA. Diversity and Distributions 14: 179–186.
  85. 85. Mordukhai-Boltovskoi FD (1960) Caspian fauna in the Azov and Black Sea Basin. Moscow: Izdatelstvo Akademii Nauk SSSR. pmid:13831310
  86. 86. Nahavandi N, Ketmaier V, Plath M, Tiedemann R (2013) Diversification of Ponto-Caspian aquatic fauna: Morphology and molecules retrieve congruent evolutionary relationships in Pontogammarus maeoticus (Amphipoda: Pontogammaridae). Molecular Phylogenetics and Evolution 69: 1063–1076. pmid:23764337
  87. 87. Badertscher S, Fleitmann D, Cheng H, Edwards RL, Gokturk OM, et al. (2011) Pleistocene water intrusions from the Mediterranean and Caspian seas into the Black Sea. Nature Geoscience 4: 236–239.
  88. 88. Dumont HJ (2000) Endemism in the Ponto-Caspian fauna, with special emphasis on the Onychopoda (Crustacea). Advances in Ecological Research, Vol 31 31: 181–196.
  89. 89. Konikov E, Likhodedova O, Pedan G (2007) Paleogeographic reconstructions of sea-level change and coastline migration on the northwestern Black Sea shelf over the past 18 kyr. Quaternary International 167: 49–60.
  90. 90. Popescu I, Lericolais G, Panin N, Normand A, Dinu C, et al. (2004) The Danube submarine canyon (Black Sea): morphology and sedimentary processes. Marine Geology 206: 249–265.
  91. 91. Bij de Vaate A, Jazdzewski K, Ketelaars HAM, Gollasch S, Van der Velde G (2002) Geographical patterns in range extension of Ponto-Caspian macroinvertebrate species in Europe. Canadian Journal of Fisheries and Aquatic Sciences 59: 1159–1174.
  92. 92. Jażdżewski K (1980) Range extensions of some gammaridean species in European inland waters caused by human activity. Crustaceana Supplement 6: 84–107.
  93. 93. Kolbe JJ, Glor RE, Schettino LR, Lara AC, Larson A, et al. (2007) Multiple sources, admixture, and genetic variation in introduced Anolis lizard populations. Conservation Biology 21: 1612–1625. pmid:18173485
  94. 94. Leuven RSEW, van der Velde G, Baijens I, Snijders J, van der Zwart C, et al. (2009) The river Rhine: a global highway for dispersal of aquatic invasive species. Biological Invasions 11: 1989–2008.
  95. 95. Macneil C, Platvoet D (2013) Could artificial structures such as fish passes facilitate the establishment and spread of the ‘killer shrimp’ Dikerogammarus villosus (Crustacea: Amphipoda) in river systems? Aquatic Conservation-Marine and Freshwater Ecosystems 23: 667–677.
  96. 96. Badtke M (2006) Kanał Bydgoski. Bydgoszcz: EKO-BAD Wydawnictwa Regionalne. 116 p.
  97. 97. Bossdorf O, Auge H, Lafuma L, Rogers WE, Siemann E, et al. (2005) Phenotypic and genetic differentiation between native and introduced plant populations. Oecologia 144: 1–11. pmid:15891837
  98. 98. Hassan M, Bonhomme F (2005) No reduction in neutral variability of mitochondrial and nuclear genes for a Lessepsian migrant, Upeneus moluccensis. Journal of Fish Biology 66: 865–870.
  99. 99. Stepien CA, Brown JE, Neilson ME, Tumeo MA (2005) Genetic diversity of invasive species in the Great Lakes versus their Eurasian source populations: Insights for risk analysis. Risk Analysis 25: 1043–1060. pmid:16268948
  100. 100. Hupało K, Rewicz T, Bącela-Spychalska K, Konopacka A, Grabowski M (2014) First record of the killer shrimp, Dikerogammarus villosus (Sowinsky, 1894), in the Váh River, Slovakia. Lauterbornia 77: 9–13.
  101. 101. Gallardo B, Errea MP, Aldridge DC (2012) Application of bioclimatic models coupled with network analysis for risk assessment of the killer shrimp, Dikerogammarus villosus, in Great Britain. Biological Invasions 14: 1265–1278.
  102. 102. MacNeil C, Boets P, Lock K, Goethals PLM (2013) Potential effects of the invasive ‘killer shrimp’ (Dikerogammarus villosus) on macroinvertebrate assemblages and biomonitoring indices. Freshwater Biology 58: 171–182.
  103. 103. Agency E (2012) Invasive Shrimp, Dikerogammarus villosus Briefing Note 5. 1–7.
  104. 104. Boets P, Pauwels IS, Lock K, Goethals PLM (2013) Using an integrated modelling approach for risk assessment of the ‘killer shrimp’ Dikerogammarus villosus. River Research and Applications: n/a-n/a.
  105. 105. Gallardo B, Aldridge DC (2013a) Evaluating the combined threat of climate change and biological invasions on endangered species. Biological Conservation 160: 225–233.
  106. 106. Gallardo B, Aldridge DC (2013b) The ‘dirty dozen’: socio-economic factors amplify the invasion potential of 12 high-risk aquatic invasive species in Great Britain and Ireland. Journal of Applied Ecology 50: 757–766.
  107. 107. Gallardo B, Aldridge DC (2013c) Priority setting for invasive species management: risk assessment of Ponto-Caspian invasive species into Great Britain. Ecological Applications 23: 352–364. pmid:23634587
  108. 108. Anderson LG, White PCL, Stebbing PD, Stentiford GD, Dunn AM (2014) Biosecurity and Vector Behaviour: Evaluating the Potential Threat Posed by Anglers and Canoeists as Pathways for the Spread of Invasive Non-Native Species and Pathogens. Plos One 9. pmid:25551830
  109. 109. Grabowski M, Bącela K, Konopacka A, Jażdżewski K (2009) Salinity-related distribution of alien amphipods in rivers provides refugia for native species. Biological Invasions 11: 2107–2117.
  110. 110. Ricciardi II, MacIsaac HJ (2000) Recent mass invasion of the North American Great Lakes by Ponto-Caspian species. Trends Ecol Evol 15: 62–65. pmid:10652557
  111. 111. Vanderploeg HA, Nalepa TF, Jude DJ, Mills EL, Holeck KT, et al. (2002) Dispersal and emerging ecological impacts of Ponto-Caspian species in the Laurentian Great Lakes. Canadian Journal of Fisheries and Aquatic Sciences 59: 1209–1228.