Katamysis Warpachowskyi Sars, 1877 (crustacea, Mysida) Invaded Lake Constance

The mysid Katamysis warpachowskyi Sars, 1877 originated in the Ponto-Caspian region and the associated river systems. The first evidence of its transgression of the limits of the watersheds of its natural Ponto-Caspian origin was found when three individuals were recorded in October 2009 in eastern Lake Constance (Austria). In March 2010, K. warpachowskyi comprised 10% of the mysid assemblage and was mainly found in rocky habitats. On both sampling dates, breeding females were present. Before this invasion, Limnomysis benedeni Czerniavsky, 1882, was the only mysid in Lake Constance and is also distributed in the Rhine river system. Since the two mysids coexist in Lake Constance, K. warpachowskyi will likely become established in the lake and further expand into the main part of the Rhine. K. warpachowskyi is known as a benthic organism and feeds on detritus and small algae; therefore, the impact on the lake ecosystem should be weaker than that of pelagic mysids.


Introduction
The number of invasions by alien species in freshwater ecosystems has increased over the last decades with increasing connection of water systems by humans to transport goods (Kinzelbach 1995;Bij de Vaate et al. 2002;Chandra and Gerhardt 2008).Several new species also invaded the littoral zone of Lake Constance.Recently, three species were recorded for the first time in the eastern part of the lake (Vorarlberg/Austria): the bivalve Corbicula fluminea Muller 1774 in 2003 (Werner and Mörtl 2004), the mysid Limnomysis benedeni Czerniavsky, 1882 in 2006 (Fritz et al. 2006), and the amphipod Crangonyx pseudogracilis Bousfield, 1958Bousfield, in 2008 (Hanselmann and Gergs 2008).All these species are common invaders and are distributed in the Rhine below Lake Constance (Bernauer and Jansen 2006, Leuven et al. 2009).The invasions are most probably due to human impact, e.g., via boat transports or ballast waters, because Lake Constance is separated from the main part of the Rhine by a natural barrier, a waterfall (Kinzelbach 1995;Rheinhold and Tittizer 1998;Martens and Grabow 2008).
In Central and Western European freshwaters systems, the invasive mysids are Hemimysis anomala Sars, 1907 and L. benedeni (Kelleher et al. 1999;Ketelaars et al. 1999;Wittmann and Ariani 2008).These mysids have been distributed in parts of the Rhine-Main-Danube system for more than 25 years (Schleuter et al. 1998;Bij de Vaate et al. 2002;Wittmann 2007).The mysid Katamysis warpachowskyi Sars, 1877 originated in the Ponto-Caspian region, i.e., mainly in the Caspian Sea and the euryhalin regions of the Danube and Dnjestr (Wittmann 2002, and older Russian literature cited therein).The species was first recorded in the Don in 2000 (Daneliya 2001) and in the upper parts of the Danube (km 1769) in 2001 (Wittmann 2002).In 2008, it reached the winter harbor of Passau/Germany (km 2228) (Wittmann 2008).The mysid, however, had not transgressed the limits of the watersheds belonging to its exclusively Ponto-Caspian range (European Watershed) and therefore was not considered as a neozoan invader for a long time (Wittmann 2007).Here I report the first recordings of this mysid outside of its Ponto-Caspian range, i.e. its jump over the European Watershed.

Materials and methods
The study site is located on the eastern part of Lake Constance, near the confluences of the rivers Alpen-Rhine and Dornbirner Ach, on the shoreline of a dam of anthropogenic origin (Hard,near Bregenz,47°29'53.8"N,9°40'50"E).The substrate on the sampling site consisted of areas with stones of different sizes (gravel, stones from 2 to 25 cm, rocks).During continuous studies on Limnomysis benedeni (e.g.Gergs et al. 2008), quantitative samples (three replicates each) were taken on July 29, 2009, October 12, 2009, and March 12, 2010.These samples were taken in the stony habitat with an underwater surber sampler (area of 625 cm², height of 40 cm, mesh wide 200 µm, Baumgärtner 2004;Mörtl 2004) at a water depth of 0.5 m.In the laboratory, the number of all mysids was counted.Additional samples were taken by kick-sampling between the rocks and directly fixed in 96% ethanol to measure the population structure.Photographs were taken with a fire-wire camera (Imaging Source) a stereomicroscope and a computer.Body length was measured from the top of the rostrum to the end of the telson (e.g., Schleuter et al. 1998).As mysids carry their brood in a brood pouch, the marsupium (Mauchline 1980), larval stages could be identified according to Wittmann (1981).The species L. benedeni and K. warpachowskyi were identified by their characteristic telson (Figure 1).Males and females were differentiated by the marsupium of the females or the enlarged 4th pleopod of the males (Bacescu 1954).Temperature, pH, conductivity, and oxygen concentration (mg l -1 and % saturation) were measured in the field (WTW MultiLine F/Set-3).

Results
In mysid samples from Lake Constance in July 2009, only Limnomysis benedeni was present.On October 12, 2009, I found the first specimens of Katamysis warpachowskyi in Lake Constance (Figure 1).Most of the mysids were found in shallow areas (about 0.5 m depth) with larger stones, especially between the rocks.The composition of the mysid assemblage changed with the habitat.254 individuals of L. benedeni and 3 individuals of K. warpachowskyi (~ 1 % of the mysid assemblage) were recorded in October between the rocks.The 3 K. warpachowskyi mysids consisted of 1 juvenile and 2 females; the marsupium of the larger female was filled with nauplioids (Table 1).The abundance of L. benedeni in October 2009 on the stones was 75 ± 56 individuals m -2 , and of K. warpachowskyi, one single animal was found in one of the three replicates (5 ± 9 individuals m -2 ).On March 12, 2010, the proportion of K. warpachowskyi within the mysid sample from the rocks increased to more than 10 %.The mean body length of K. warpachowskyi females was higher than that of males (Table 1).No juveniles were found on that date, but every female (n = 6) had a marsupium, which was not always filled.All recorded broods were characterized as embryos.The abundance of L. benedeni on the stones on March 3, 2010 was 171 ± 74 individuals m -2 , and that of K. warpachowskyi again 5 ± 9 individuals m -2 .None of the L. benedeni females carried a brood in March.
The abiotic factors in October 2009 and March 2010 were similar (Table 2), except that the temperature was higher in October 2009.On both dates, the oxygen content exceeded 100 %.

Discussion
The recording of Katamysis warpachowskyi on October 12, 2009 in Lake Constance is the first outside the Ponto-Caspian drainage system.After Limnomysis benedeni, K. warpachowskyi is the second mysid found in Lake Constance.The immigration of K. warpachowskyi must have been enabled by human impact, probably by moving boats from one water body to another (Martens and Grabow 2008), because the nearest incidence of the mysid is in the Danube near Passau (Wittmann 2008), around 300 km in linear distance away.This supports the hypothesis of Wittmann (2008) that human impact is responsible for the range expansion of K. warpachowskyi in the Danube.
The low proportion of K. warpachowskyi in Lake Constance in October 2009 and the low abundances suggest that the mysid invaded Lake Constance recently.The higher proportion on the mysid assemblage in the rocks in March 2010 indicates that the mysids reproduced in the lake.It is surprising that K. warpachowskyi carried broods in March since L. benedeni does not reproduce at this temperature (unpublished data), but there are no detailed reports on the reproduction period of K. warpachowskyi.Daneliya (2001) only one reported from the Don a smaller brood size in summer than in spring.
The dispersion of the mysids differed between the habitats.K. warpachowskyi was mainly found between the rocks, whereas L. benedeni was found everywhere.Both mysids are nectobenthic, but K. warpachowskyi seems to be adapted to higher water velocity.Their body is much more depressed, and they were not found in drift (Wittmann 2002).The abiotic factors at the sampling site (Table 2) were similar to those measured at other locations where K. warpachowskyi has been found (Wittmann 2002(Wittmann , 2007)).K. warpachowskyi often coexists with L. benedeni (Wittmann 2007).In Lake Constance, the two mysids were found together, but L. benedeni dominated the mysid community.The environmental demands of the two mysids are comparable, except for oxygen.K. warpachowskyi requires at least 6 mg l -1 , whereas L. benedeni requires only 4 mg l -1 (Bacescu 1954;Wittmann et al. 1999).
The mean body length of adult K. warpachowskyi in March 2010 in Lake Constance was longer than that recorded in May 1995 in the Danube by Wittmann (2002), but the range of body length is within these previous published data.In Lake Constance, the females were longer in body length than the males, which is typical for the species (Bacescu 1954;Wittmann 2002).The maximal brood size in Lake Constance was higher than expected by Wittmann (2002) and recorded by Bacescu (1954) and comparable to the data from Daneliya (2001).
The feeding behavior of K. warpachowskyi has not been studied well, and only a few observations and stomach analyses are available.The mysid feeds on detritus and small algae and is not a predator (Wittmann 2002); it likely feed the same way as L. benedeni (Gergs et al. 2008).Therefore, the impact of K. warpachowskyi should be comparable to that of L. benedeni, in keeping with the hypothesis of Wittmann (2002).K. warpachowskyi will probably not have such a high impact as, e.g., the pelagic predator Hemimysis anomala, introduced in the Netherlands (Ketelaars et al. 1999).Nevertheless, it is probable that K. warpachowskyi will establish in Lake Constance, colonize the lower parts of the Rhine easily, and then spread up through the associated water systems, if enough oxygen is available.Therefore, the monitoring of the macrozoobenthos community of this water system should be continued.

Table 1 .
Body length and brood size of Katamysis warpachowskyi in Lake Constance given as mean ± standard deviation if n > 3 or single measurements if n  3.

Table 2 .
Abiotic factors at the sampling site in Lake Constance.