Ponto-Caspian peracaridan Crustaceans first invaded the Baltic Sea basin from the basin of the Black Sea via artificial canals in the 18th Century. In the 1960s, the number of these species substantially increased due to deliberate introduction of Mysids and Amphipods into Lithuanian waters to enhance fish production. Range expansion of Ponto-Caspian peracaridan Crustaceans is continuing due to secondary dispersal and invasion by new species. Historically the main invasion vectors were inland shipping, deliberate introductions and natural dispersal across inland waterways; more recently the significance of marine invasions related to shipping seems to be increasing. These invaders have had detrimental impacts on local macroinvertebrates and may even have caused local extinctions of native species. Such impacts likely result from predatory and competitive interactions between alien and native species. The magnitude of predatory impacts of omnivorous Ponto-Caspian peracaridan Crustaceans can rise with increasing productivity of the environment. These Crustaceans, especially large-sized species, have proved to be prone to carnivorous feeding and high plasticity when acquiring nutrients for reproduction and growth. It seems that stoichiometric plasticity related to their flexible feeding strategy may favour their invasiveness. Another undesirable outcome of these invasions is distortion of conventional metrics of ecological status for aquatic systems. Meanwhile, at least in lakes, the primary goal of introduction, i.e. enhancement of fish production, has not been achieved. Some invasive Crustaceans seem to be adapting to previously unsuitable habitats in their invaded range, which may exacerbate their negative impacts. Further expansion of accumulated and new Ponto-Caspian peracaridan Crustaceans can be expected in Lithuanian waters as well as in the Baltic Sea basin, and their impacts may increase in the future.
Detrimental impacts of alien species on local biota can substantially alter ecosystems and may even cause damage to human economic interests. Thus, threats from alien invasive species are widely acknowledged. Owing to their wide environmental tolerances and high phenotypic variability arising from the complex geological past of their native region, Ponto-Caspian peracaridan (PCP) crustaceans – mostly Amphipods and Mysids – are among the most successful aquatic invaders (Reid and Orlova, 2002). Recently, they have become widely distributed outside their native range in European waters and even managed to invade the North American Great Lakes (Bij de Vaate et al., 2002).
Ponto-Caspian species have been invading Europe via the three main invasion corridors formed by opening of the canals connecting distinct watersheds in the 18th Century (Bij de Vaate et al., 2002) and by marine shipping (Ketelaars, 2004). However, Lithuania – situated on the south-eastern shore of the Baltic Sea and along the northern branch of the central European invasion corridor – plays an important additional role in the history of invasions of PCPs into Europe (Ketelaars, 2004). During the 1960s, some Ponto-Caspian Amphipods and Mysids invaded the Baltic Sea basin in Lithuania primarily due to deliberate introduction of these peracaridans into the Kaunas Water Reservoir (WR) located on the Nemunas River, the northern branch of the central invasion corridor (Figure 1; Gasiūnas, 1972; Arbačiauskas, 2002). From this Lithuanian WR, these PCPs began their further expansion into European waters. Translocations from their native to the new range had been undertaken in an attempt to improve fish nutrition and consequently increase commercial fish production; such practices were common in the middle of the 20th Century, especially in the former Soviet Union (Arbačiauskas et al., 2010).
PCPs in Lithuania have come under intense scientific scrutiny since the middle of the 20th Century – first as introduced fish fodder animals and later as invasive species. Expansion of alien species is always accompanied by impacts at some level, which are not necessarily the intended ones and are sometimes difficult to predict. Therefore, scientific information on PCPs in Lithuanian waters collected over a 50-year period might have significant value for the understanding of invasion patterns and impacts, and be useful to support predictions and management decisions.
The purpose of this article was to present Lithuanian lessons about PCP invasions and their aftermath. In particular, we review historical and recent invasions, operating vectors and properties of colonised water bodies; the PCP impacts and their consequences are then analysed; and finally, we draw conclusions relevant to environmental management related to biological invasions.
Information compiled in this summary article was acquired through literature surveys and our investigations since the 1990s. PCP distribution data collected between 1998 and 2010 (and summarised previously; see Arbačiauskas et al., 2011b) were updated with more recent information from the surveys conducted during 2012–2015. These investigations included sampling of numerous Lithuanian inland water bodies (rivers, lakes, reservoirs and a lagoon, including previously un-investigated sites) using either a standard dip net, sledge net or dredge. This updated dataset was analysed by means of forward-stepwise logistic regression and classification trees to reveal which water body parameters were the best predictors for establishment success of introduced species. The outcome of classification tree analysis was more straightforward than that for logistic regression as interactions between quantitative predictors can be explored in a dichotomous tree with node rules containing threshold values for the predictors. The pool of explanatory variables (all log-transformed for the logistic regression) included surface area, mean and maximum depth, water circulation, mean multi-annual chlorophyll a and total P concentration. The analyses were performed in the R 3.2.5 environment (R Core Team, 2015) using package fmsb (Nakazawa, 2015) for logistic regressions and package rpart (Therneau et al., 2015) for classification trees, which were simplified by pruning to 3 or 4 terminal branches.
Historical and recent invasions, dispersal vectors and invaded ecosystems
The three main waves of macroinvertebrate invasions into Lithuanian waters were outlined in Arbačiauskas et al. (2011b). These waves are visualised in Figure 1; PCPs recorded in Lithuanian waters within these waves, as well as potential invaders, are given in Table 1.
The first invasion wave, that is, historical invasions, occurred at the end of the 18th Century when the basins of the Nemunas (Baltic Sea drainage) and Dnieper (Black Sea drainage) rivers were connected by the Oginsky Canal. This wave brought three Amphipod species: Chelicorophium curvispinum, Chaetogammarus ischnus and Gammarus varsoviensis. Amphipod C. ischnus recently has not been found in Lithuanian waters, whereas C. curvispinum occurs in the Nemunas River (the largest Lithuanian river), the lower reaches of some of its large tributaries, and the Curonian Lagoon. The Amphipod G. varsoviensis is confined to the middle reaches of the Nemunas River, which is devoid of later Ponto-Caspian Amphipod invaders (Arbačiauskas, 2008). It should be noted that G. varsoviensis only recently has been hypothesised to be a cryptic historical invader of Ponto-Caspian origin due to high reproductive potential and genetic diversity patterns (Grabowski et al., 2012). These historical PCP invaders mostly dispersed naturally and with the aid of inland shipping.
The second PCP invasion wave consisted of deliberately introduced species: three Mysids (Paramysis lacustris, Limnomysis benedeni and Hemimysis anomala) and three Amphipods (Pontogammarus robustoides, Obesogammarus crassus and Chaetogammarus warpachowskyi). These PCP crustaceans were transported by planes from the Ukrainian Dnieper and Simferopol WRs and released into the Lithuanian Kaunas WR located on the Nemunas River, during 1960–1961 (Figure 1). After successful acclimatisation (and rapid downstream self-spread to the Curonian Lagoon) these PCPs were officially transferred to more than 100 Lithuanian lakes, WRs and again to the Curonian Lagoon, as well as some Latvian, Estonian and Russian lakes and WRs (Gasiūnas, 1972; Arbačiauskas, 2002). Thus, deliberate primary and secondary introductions, both official and unofficial, are a distinct feature of this invasion wave. It should be stressed that some introduced PCPs also exhibited good abilities of natural and human-mediated dispersal; recently they have been expanding their range within and outside Lithuania (Ketelaars, 2004; Arbačiauskas and Gumuliauskaitė, 2007; Audzijonyte et al., 2008).
Among the PCP Crustaceans established in Lithuanian waters, Mysid P. lacustris and Amphipods P. robustoides and C. warpachowskyi turned out to be best adapted to fresh water and are rather widely distributed in rivers and lentic waters (Arbačiauskas et al., 2011b). These three PCPs exhibited higher probabilities of establishment after introduction in lentic waters than the Mysid L. benedeni and the Amphipod O. crassus (Figure 2). For decades, the Mysid H. anomala was known in Lithuania only from the Kaunas WR, the place of its primary introduction (Arbačiauskas et al., 2011b). However, recent night-time sampling revealed this nocturnal Mysid to occur in the lower reaches of the Nemunas tributary, close to the Curonian Lagoon (Arbačiauskas and Lesutienė, unpubl. results). This suggests that the species may be quite common in suitable habitats, but its detection warrants appropriate survey methods.
Modern macroinvertebrate invaders in Lithuania are so far only represented by one PCP species, the Amphipod Dikerogammarus villosus, also known as the Killer Shrimp (Šidagytė et al., 2016). This species has amply invaded European waters and was anticipated to breach the Lithuanian border via the central invasion corridor, the Nemunas River in particular (Arbačiauskas et al., 2011b); instead it entered through the Baltic Sea. The only other modern invader, and overall the only peracaridan invader of different geographic origin in Lithuania, is the North American Amphipod Gammarus tigrinus, which also penetrated via the sea (Daunys and Zettler, 2006). Apparently, marine shipping will be the hallmark of modern invaders, and there are more species to come (Table 1). Both of these new Amphipod species are currently established only in the transitional (lagoon and estuarine) waters of Lithuania; however, they are perfectly capable of invading fresh water bodies as well (Grabowski et al., 2007).
The updated information on translocation results of PCPs and identified new invasions in lentic waters allowed for analysis of water body parameters that may support establishment of widespread P. lacustris and P. robustoides. According to the regression analysis (Table 2), the best predictor of successful establishment for both species was larger surface area. The Amphipod also preferred higher total P concentration and deeper water bodies, and the success of the Mysid was further favoured by faster water circulation. The classification trees indicated similar patterns. A preference for deep and large waters was characteristic of the Amphipod, whereas the Mysid was favoured by a larger surface area, increased total P and water circulation (Figure 3).
Data from Lithuanian waters also suggest that, although originating from environments with high mineralisation, the Amphipod P. robustoides was able to establish a viable population under quite low water mineralisation: 160–200 mg l−1 total ion content (Arbačiauskas, 2005). Further, although preferring large warm-water rivers, this Amphipod was also able to colonise a medium-sized cold-water river (Arbačiauskas et al., 2011b). Ponto-Caspian Amphipods are generally characterised as oxyphilic animals. Thus, it has been hypothesised that temporary hypoxia may be a limiting factor in local distribution of these Amphipods in the invaded range (Arbačiauskas, 2002, 2005). Such conditions especially emerge in productive Lithuanian lakes under winter ice cover, with no possibility for aquatic animals to escape them. Interestingly, our recent results have suggested that in some eutrophic water bodies over the 50 years since its introduction, P. robustoides has adapted to decreased oxygen concentration, which may allow the species to colonise previously unsuitable habitats (Šidagytė and Arbačiauskas, 2016). Also, PCPs have demonstrated high stoichiometric (C:N:P) plasticity related to flexibility of feeding strategies, which may favour invasiveness, in particular an ability to invade ecosystems of different trophic status and/or nutrient (N, P) limitations (Arbačiauskas et al., 2013).
Impacts and aftermath of PCP invasions
Native biodiversity and abundance
Deliberate introductions were expected to be beneficial based on the assumption that there was unused niche space in the receiving waters; the PCPs would obtain sufficient nutrition to develop high biomass, supplementing the local benthic macroinvertebrate communities with high-value fish-food items. Consequently, an increase of total benthic biomass was predicted. However, once introduced in Lithuanian lakes, instead of occupying the speculative unused niches, PCPs interacted with resident macroinvertebrates, usually supressing the native community and locally replacing resident peracaridan species. Such assemblage alterations did not result in the expected increase of total benthic biomass in the long term, but diminished macroinvertebrate biodiversity, as demonstrated in the case of P. robustoides (Figure 4). This ecologically aggressive Amphipod was able to decrease the benthic diversity two-fold or more and substantially reduce the abundance of most native macroinvertebrate groups; only worms (Oligochaeta) and beetles (Coleoptera) were not affected by this invader (Gumuliauskaitė and Arbačiauskas, 2008). Predatory impacts on local zooplankton communities are also well documented in the Mysids P. lacustris (Lesutienė et al., 2007) and H. anomala (Ketelaars et al., 1999).
The Amphipod P. robustoides also proved capable of locally eradicating the local Amphipods Gammarus lacustris and G. varsoviensis (Arbačiauskas, 2005, 2008). The new invader D. villosus is also known for its extreme ability to exterminate a number of previous Amphipod invaders in Europe (e.g. Dick and Platvoet, 2000; Kinzler et al., 2009). The extermination of resident Amphipods by newcomers may occur through asymmetry in intra-guild predation (Dick et al., 1999; Kinzler and Maier, 2003).
The invaders P. lacustris and P. robustoides are omnivores with a propensity for carnivory; thus their impacts on local macroinvertebrates primarily depend on the extent of their predation. It was seen that the carnivory of these peracaridans was more pronounced in more productive environments. This phenomenon was explained by high P levels promoting growth rates that subsequently result in higher N demand, which may require a carnivorous diet (Arbačiauskas et al., 2013).
Water quality assessment
Due to their significant effects on macroinvertebrate community structure and replacement of native taxa with different environmental tolerances, invasive species have been recognised as important players in water quality monitoring (Cardoso and Free, 2008). Conventional metrics of water quality assessment are based on affected macroinvertebrate communities or samples identified to a higher taxonomic level (family/genus). Such methods do not distinguish between native and alien species, so invasions can distort metric values even if the water quality remains good. As a result, the metrics may no longer reflect water quality in highly invaded sites (MacNeil et al., 2010).
Further, invasive species may be considered as biological pollution of a community itself. To address this issue, the ‘biocontamination’ concept was proposed, which allows assessment of macroinvertebrate assemblage deviation from a natural community by evaluating its abundance and disparity contamination (Arbačiauskas et al., 2008). Routine macroinvertebrate monitoring data are sufficient for generation of such assessments.
Conventional macroinvertebrate metrics have been demonstrated to be negatively correlated with biocontamination of European waterways (Arbačiauskas et al., 2008). Metrics such as the Biological Monitoring Working Party (BMWP) score, its Average Score Per Taxon derivative and the number of Ephemeroptera, Plecoptera and Trichoptera taxa are negatively correlated with biocontamination in Lithuanian lotic systems, implying they may provide unreliable estimates of water quality when high-impact PCPs are present (Arbačiauskas et al., 2011a). For Lithuanian lentic water bodies, a new multi-metric macroinvertebrate index to address eutrophication pressure has been designed using a dataset of ecosystems invaded by the PCPs, to avoid such an effect (Šidagytė et al., 2013). The Lithuanian monitoring system includes a separate biocontamination evaluation to address biological pressure additionally and to indicate the level of uncertainty in macroinvertebrate-based water quality assessment.
As new invasive species may arrive and establish at any time, risk assessments to identify the most probable invaders need to be carried out (Gallardo and Aldridge, 2013). Further, examination of their potential effects on assessment metrics should be conducted. For example, in a mesocosm study, D. villosus was demonstrated to induce lower BMWP scores than native Amphipods or an absence of Amphipods (MacNeil et al., 2013).
Failed enhancement of fisheries?
The established PCP Crustaceans were immediately assimilated into the diet of some fish species, and a 20% increase in fish production has been proposed in some fishery reports (Arbačiauskas et al., 2010). However, this statement has received no scientific substantiation. Although juvenile European Perch (Perca fluviatilis) readily and substantially assimilate PCPs into their diet, no subsequent influence on their somatic growth rate or littoral catch per unit effort was seen in comparison with perch in lakes devoid of alien Crustaceans (Arbačiauskas et al., 2010; Rakauskas et al., 2010). In addition, neither a comparison of the biomass of a littoral fish community between lakes with and without PCP Crustaceans (Figure 5a), nor commercial catches before and after PCP introduction to a lake with favourable environmental conditions for the PCP species (Figure 5b), demonstrated a beneficial PCP effect (see also Arbačiauskas et al., 2010).
Seemingly, particular fish species might have benefitted from PCP introductions in Lithuanian lakes, but the overall fish production of these lakes has not increased (Arbačiauskas et al., 2010). This suggests that in general, the introductions have not brought the expected economic benefit in Lithuanian lakes. Nonetheless, in the Curonian Lagoon, some fish – especially juvenile European Perch and Pikeperch (Sander lucioperca) – feed predominantly on PCPs, especially Ponto-Caspian Mysids (Ložys, 2003). Consequently, the mass swarming and horizontal autumn migrations of Ponto-Caspian Mysids have been suggested to substantially increase the inshore–offshore habitat coupling when Mysids are consumed by over-wintering fish in deeper areas (Lesutienė et al., 2008).
The three invasion waves to Lithuanian inland waters resulted in nine PCP species persisting to date in inland and transitional waters of Lithuania. This number exceeds the number of native Lithuanian freshwater peracaridan species (six), which include species such as the endangered glacial relict species and species being locally eradicated by the invaders. Most of the PCPs were intentionally introduced, and among these species are the two most widespread alien peracaridans in the country: the Amphipod P. robustoides and the Mysid P. lacustris. The weight of scientific evidence implies that these introductions of PCPs as fish fodder organisms to lakes were ill advised, bringing no genuine economic benefits but resulting in negative environmental impacts. The introduced PCPs did not fill the speculated free niche and did not improve recipient macroinvertebrate communities for fish. Only in some instances – such as in the Curonian Lagoon – did they probably supplement fish diets in a way that has resulted in notably increased production. Instead, they obviously negatively affected native macroinvertebrate diversity and the abundance of some native macroinvertebrate groups. Consequently, conventional macroinvertebrate metrics of ecological status have become unreliable and new assessment systems are warranted. A pressure-specific assessment system is recommended to address water quality using macroinvertebrates and to separately estimate biocontamination as a reliability measure for such assessment. However, as adaptation of plastic invaders may enhance their further spread and impacts, and new invasions are inevitable, the performance of water quality assessment methods needs to be monitored and necessary updates implemented.
This article evolved from a presentation given at the Marine and Freshwater Invasive Species: Ecology, Impact, and Management conference held in Buenos Aires, Argentina, 2–4 May 2016. We thank Francisco Sylvester and an anonymous referee for valuable comments on an early draft of this article.
Parts of this study were funded by the Research Council of Lithuania, Project No. SIT-10/2015.