Reproductive characteristics of invasive gammarids in the Rhine-Main-Danube catchment, South Germany

Transcription

1 ARTICLE IN PRESS Limnologica 36 (26) Reproductive characteristics of invasive gammarids in the Rhine-Main-Danube catchment, South Germany Axel Kley, Gerhard Maier Department of Ecology of Animals, University of Ulm, Albert-Einstein-Allee 11, 8969 Ulm, Germany Received 13 September 25; received in revised form 25 January 26; accepted 25 January 26 Abstract The gammarid composition at 25 sites in the rivers Danube, Main and the southern reaches of the Rhine were studied during the years Dikerogammarus villosus and Echinogammarus ischnus were the most frequent species prevailing at 17 sites. Sympatric occurrence of D. villosus and E. ischnus was observed at 12 sites. Dikerogammarus haemobaphes was recorded at 7 sites; this species prevailed in the Danube, west of the Weltenburger Enge and in the Isar mouth where it co-existed with native species (Gammarus pulex and/or G. roeseli) at 6 sites. Dikerogammarus bispinosus and E. berilloni were found at only 1 site, where they co-existed with D. villosus and E. ischnus, and with D. villosus and native species, respectively. Investigation of reproductive characteristics at 3 sites showed that females of D. villosus and D. haemobaphes produced the biggest clutches with more than 1 eggs. Females of E. ischnus produced much smaller clutches (1 35 eggs on an average), but very big eggs. Clutch sizes and egg volumes of D. bispinosus and E. berilloni resembled those of native species. Our results suggest that the most successful invaders (D. villosus, D. haemobaphes and E. ischnus) display reproductive traits that facilitate their success. Both Dikerogammarus sp. allocate energy into production of many but small eggs, thus maximizing offspring number, while E. ischnus allocates its energy into production of fewer but large eggs which could be beneficial at sites where food is scarce. r 26 Elsevier GmbH. All rights reserved. Keywords: Invasive amphipods; Dikerogammarus spp.; Echinogammarus spp.; Distribution; Reproduction; Running waters; South Germany Introduction Central European large rivers were invaded by numerous non-indigenous species in the last 2 decades as a result of shipping and building of canals which connect river systems (e.g. Bij de Vaate, Jazdzewski, Ketelaars, Gollasch, & van der Velde 22; Kinzelbach 1995; Tittizer 1996). In South Germany, the opening of Corresponding author. Tel.: address: gmaier.limnos@t-online.de (G. Maier). the Main Danube canal in 1992 facilitated the invasion of species from the Ponto Caspian region. Meanwhile, more than 15% of species, and often more than 9% of macroinvertebrate individuals in the Rhine river are invasives (Kinzelbach 1995). Among crustaceans, amphipod species of the genus Dikerogammarus and Echinogammarus mainly from the Ponto Caspian region have invaded many reaches of large rivers in Germany. Dikerogammarus haemobaphes Eichwald, 1841, was the first of the genus Dikerogammarus which invaded South German rivers in the 197s (Schleuter, /$ - see front matter r 26 Elsevier GmbH. All rights reserved. doi:1.116/j.limno

2 8 ARTICLE IN PRESS A. Kley, G. Maier / Limnologica 36 (26) 79 9 Schleuter, Potel, & Banning 1994; Tittizer 1996; Tittizer, Leuchs, & Banning 1994). Meanwhile, its relative, Dikerogammarus villosus Sovinskij, 1894, which appeared later, in the early 199s, has replaced D. haemobaphes in many reaches (e.g. Kley & Maier 23). Dikerogammarus bispinosus Martynov, 1925, a third Dikerogammarus invader, was recorded from the German Danube near the Isar mouth in 1998 (Eggers & Martens 21). Echinogammarus ischnus Stebbing, 1899, was recorded from the Rhine, the Main and the Danube since late 198s (Scho ll 199; Tittizer 1996 and citations therein), and Echinogammarus berilloni Catta, 1878, a Mediterranean species, has been recorded from the Rhine, Saar and Mosel (Mauch 1963; Tittizer 1996). Although distribution and life histories of Echinogammarus and Dikerogammarus spp. from their places of origin have already been studied (e.g. Bikuna & Asensio 1991; Chovet & Lecureuil 1994; Konopacka & Jesionowska 1995; Musko 1993), and although distributions of invaders in large European rivers are known (e.g. Bollache et al. 24; Devin, Beisel, Bachmann, & Moreteau 21; Dick & Platvoet 2; Tittizer 1996 and citations therein, Van den Bossche 22) there is lack of knowledge about reproductive characteristics of Central European populations of invaders. In an earlier paper, we studied life history characteristics of D. villosus and E. ischnus in the Main river (Kley & Maier 23). We showed that D. villosus produced much bigger clutches than species native to Central Europe, and that E. ischnus produced large eggs in comparison with D. villosus and native species (cf. Po ckl 1993a, b). These differences in reproductive mode between gammarid species point to differences in reproductive strategies (Kley & Maier 23). This paper has two aims: firstly, to study current gammarid composition in reaches of the large rivers in South Germany; and secondly, to study reproductive characteristics of 5 invasive gammarids, 3 Dikerogammarus sp. (D. haemobaphes, D. bispinosus, D. villosus) and 2 Echinogammarus sp. (E. ischnus and E. berilloni), from sites in the Rhine and the Danube. We expected that species which are particularly successful invaders, would exhibit life history traits that facilitate their rapid and successful spread. Methods Distribution The composition of gammarid communities at 25 near-shore sites along South German rivers (13 sites in the Danube, 8 in the Main/Main Danube canal, and 4 sites in the southern parts of the Rhine system) were investigated by sampling with kick-net and pond-net, turning over rocks, dislodging gammarids and rinsing them into a net (mesh-sizes of nets 25 4 mm), and if appropriate, with the use of a surber sampler. At each site two to six areas of.5.5 m ( 1.5 m depth) were carefully searched for gammarids. Sites were sampled 1 14 times during the years After their Table 1. Background characteristics for sites where gammarids were collected for investigation of reproductive parameters. The parameters were measured/analyzed between April 22 and March 24 Characteristics Sites Donau Deggendorf Isar Plattling Rhine-tributary Iffezheim Gammarids D. villosus, D. bispinosus, E. ischnus D. haemobaphes E. berilloni Altitude (m a.s.l.) Position N, O N, O N, O Aspect SE NE SW CV (m s 1 ) p.5 p.4 p.5 T(1C) Cond. (mscm 1 ) O 2 (%) ph Tot. hard (meq l 1 ) Ca 2+ (meq l 1 ) Alcal. (meq l 1 ) PO 4 -P (mgl 1 ) NO 3 -N (mg l 1 ) Chloride (mg l 1 ) Makrophytes No No Yes Shipping Yes No No Abbreviations: CV, current velocity; T, temperature; Cond., conductivity; Tot. hard., total hardness; Alcal., alcalinity.

3 ARTICLE IN PRESS A. Kley, G. Maier / Limnologica 36 (26) collection, gammarids were immediately preserved in 7% alcohol. In the laboratory, gammarids were identified to species and counted. The proportion of each gammarid species in relation to other gammarids present (all gammarids at one site ¼ 1%) was calculated for each site. Life history characteristics To get life history characteristics, gammarids were sampled at 3 sites during the years Some relevant site characteristics are summarized in Table 1. The sites vary in morphology, but are similar in physical and chemical characteristics. The Rhine tributary is a smaller stream, while the Isar and the Danube are large rivers. The water at all sites is hard and nutrient-rich. Oxygen saturation in near-shore areas is 47%. The substrate at the shore-line consists of boulder while gravel and larger stones (grain size 1 15 cm) prevail in greater depths. Macrophytes are only present in the Rhine tributary. D. haemobaphes was collected from a site where the Isar flows into the Danube, near Plattling. D. villosus, D. bispinosus and E. ischnus were obtained from the Danube, near Deggendorf, and E. berilloni from the Rhine tributary, near Iffezheim. Samples were usually taken at monthly intervals. Sampling was not possible during some time in autumn/ winter of 22 and autumn of 23, respectively, because of high water. However, care was taken that one complete year-cycle was obtained for each species. To get quantitative estimates of number of males, females, precopulatory pairs, and ovigerous females, one to four near-shore areas of.5.5 m (to a depth of approximately 1.5 m) were carefully searched for gammarids by pond-net and/or kick-net sampling (meshsizes of nets 25 mm), as well as by turning over rocks and larger stones, dislodging gammarids and rinsing them into 25 mm nets. Use of surber samplers was not possible due to the rocky substrate at the shore-line. Sampling procedures were repeated until no more gammarids were obtained from the areas. In addition, 1 2 adult gammarids, dependent on their density, were taken randomly. This additional sample served to N Hamburg Frankfurt Berlin Weser Main Elbe Rhein Bamberg Main Würzburg Mannheim Stuttgart Donau Rhein (Rhine) Neckar MDK Nürnberg Karlsruhe Regensburg Strasbourg Tributary Stuttgart Weltenburger Enge Ingolstadt Kehlheim Deggendorf Ulm Donau (Danube) Isar Donau Passau Dikerogammarus villosus Dikerogammarus haemobaphes Dikerogammarus bispinosus Echinogammarus ischnus Echinogammarus berilloni Gammarus roeseli Gammarus pulex 1 km Fig. 1. Map showing distribution and relative abundances of gammarids in some reaches of rivers in South Germany. MDK means Main Danube canal. Framed area in the map of Germany marks region from where samples were taken.

4 82 ARTICLE IN PRESS A. Kley, G. Maier / Limnologica 36 (26) 79 9 get sizes of adults, clutch and egg sizes. In the laboratory, gammarids were separated into juveniles (individualso.5 cm), adult males, adult females, ovigerous females (females with dark coloured pouch of eggs between the limbs), and precopulatory pairs. Clutch sizes were obtained by removing eggs/neonates from the brood chambers of 2 ovigerous females per sampling date with pins, and by counting the number of eggs present. Egg/neonate stages were identified according to Weygold (1924) and Skadsheim (1982), and egg sizes were obtained by measuring the longest and the shortest axis of stage-2 eggs with a Leitz stereomicroscope equipped with an ocular micrometer (magnification 1x). Stage-2 eggs are commonly used to determine egg size of gammarids (Kley & Maier 23; Po ckl 1993b; Sutcliffe 1992). Eggs from different brood chambers, i.e. from different ovigerous females, were measured to get variation of egg size within the same species. Weight of adults was obtained by weighing 2 males and 2 females per sampling date with a Sartorius MP 24 balance (accuracy 1 mg). Water attached to gammarids was removed with filter paper before weighing them. To get egg volume (EV), fecundity index (FI) and reproductive effort (RE), the following formulas were Table 2. Results of the two-way ANOVA concerning differences in wet weights and reproductive characteristics between months and species Parameter Factor time Factor species F (df) P F (df) P WW (males) 71.6 (13/746) (2/746).1 WW (females) 67.5 (13/738) (2/738).1 CS 38.5 (7/416) (2/416).1 EV 17.7 (7/24) (2/24).1 FI 6. (7/416) (2/416).1 RE 4.9 (7/23) (2/23).6 WW, wet weight; CS, clutch size; EV, egg volume; FI, fecundity index; RE, reproductive effort. 1 5 Females [%] PC-Pairs (no.m -2 ) Species Dv Db Dh Ei Eb 1 A M J J A S O N D J F M A M J Months / Years J A S O N D J F M Ov.Females [%] A M J J A S O N D J F M A M J J A S O N D J F M Months / Years Fig. 2. Sex ratio (percentage of females in relation to adults), percentage of egg-bearing (ov.) females (in relation to adult females), and number of precopula (PC) pairs of 5 invasive gammarids in large rivers in South Germany during the years Dv, Dikerogammarus villosus; Db, Dikerogammarus bispinosus; Dh, Dikerogammarus haemobaphes; Ei, Echinogammarus ischnus; Eb, Echinogammarus berilloni.

5 ARTICLE IN PRESS A. Kley, G. Maier / Limnologica 36 (26) employed (cf. Po ckl 1993b): EV ¼ pab 2 =6, (1) FI ¼ CS=WW, (2) RE ¼ EV FI, (3) where a is the longest and b the shortest axis of the egg, CS the clutch size and WW the wet weight of a female. To get possible differences in weights and reproductive characteristics between different seasons/months and species a two-way ANOVA and one-way ANOVAs followed by Tukey s post-hoc test were used. Pearson s correlation coefficient was employed to test for correlations between wet weight and clutch size, wet weight and egg volume, and clutch size and egg volume. Results Distribution D. villosus and E. ischnus were the dominant gammarids at the sampling sites in the rivers Rhine and Main (Fig. 1). In the Danube, both species were confined to reaches where shipping takes place; they were not present in the western part beyond the socalled Weltenburger Enge. Co-existence of D. villosus and E. ischnus was observed at 12 (out of 25) sites. D. haemobaphes was recorded from the Isar mouth and from sites west of the Weltenburger Enge, where it coexisted with the native species Gammarus pulex and/or Gammarus roeseli. D. bispinosus was only found at one Species Dv Db Dh Ei Eb 4 2 A M J J A S O N D J F M A M J J A S O N D J F M Months / Years Fig. 3. Mean wet weight (WW) 7SD of five species of invasive gammarids in large rivers in South Germany during the years Abbreviations for species as in Fig. 2.

6 84 ARTICLE IN PRESS A. Kley, G. Maier / Limnologica 36 (26) 79 9 site in the Danube, near Deggendorf, where it co-existed with D. villosus and E. ischnus. E. berilloni was recorded in the Rhine tributary near Iffezheim, but not in the Rhine itself; it co-existed with D. villosus and the native species, G. pulex and G. roeseli. Reproductive characteristics and weights Sex ratio of all gammarids was between 4% and 8% (Fig. 2). In most species (except D. haemobaphes), and most part of the year, sex ratio was slightly skewed in favour of females. The number of precopulatory pairs and percentage of ovigerous females varied seasonally in all species (Fig. 2). Number of precopulatory pairs and percentage of ovigerous females were highest in spring. In late autumn and early winter months, no precopulatory pairs and ovigerous females were found (Fig. 2). Mean wet weights and mean reproductive characteristics varied significantly with time of the year and with species (Table 2). Seasonal patterns of mean wet weights and mean reproductive characteristics are shown in Figs. 3 and 4. Although there was great variation in CS (eggs ov.female -1 ) EV [mm 3 ] FI Species Dv Db Dh Ei Eb RE A M J J A S O N D J F M A M J J A S O N D J F M Months / Years. A M J J A S O N D J F M A M J J A S O N D J F M Months / Years Fig. 4. Mean clutch size (CS), egg volume (EV), fecundity index (FI) and reproductive effort (RE) 7SD of five species of invasive gammarids in large rivers in South Germany during the years Abbreviations for species as in Fig. 2. Table 3. Results of Tukey s post-hoc test (Po.5) showing month(s) where wet weights and reproductive characteristics were maximum Parameter Dv Dh Db Ei Eb WW (males) May March, May, June, November, December May April, May, June May WW (females) May March, May, June, November, December May April, May, June, July April, May, June CS April, May March, April, May April, May May April, May, June EV March December March, April, July n.s. February, March FI April April June, August May All months4february RE March, April March, April, August March, July May March Abbreviations for weight and reproductive parameters as in Table 2. Dv, Dh, Db, Ei and Eb are abbreviations for different Dikerogammarus (D. villosus, D. haemobaphes, D. bispinosus) and Echinogammarus species (E. ischnus, E. berilloni). n.s. means no difference between months.

7 ARTICLE IN PRESS A. Kley, G. Maier / Limnologica 36 (26) weights and reproductive characteristics, throughout the year some trends could be detected. Mean wet weights of D. villosus, D. bispinosus, E. ischnus and E. berilloni were highest in spring, and those of D. haemobaphes in spring and November/December, respectively (Fig. 3; Table 3). Clutch sizes of all species also peaked in spring (Fig. 4; Table 3). Mean egg volume, FI and RE displayed a less-clear seasonal pattern than weights and clutch sizes. Although in D. villosus and E. ischnus the values for these variables were highest in spring, there are deviations from this pattern. Comparatively large eggs, high FI and RE in summer were recorded for D. bispinosus, and high FI throughout the year for E. berilloni (Fig. 4; Table 3). That FI and RE decreased to in late autumn (Fig. 4) originates from the lack of reproductive females during that time (cf. Fig. 2). Species ranks in weights and reproductive characteristics are shown in Fig. 5 and Table 4. Adults of WW-male WW-female 1 1 [mg] 8 6 [mg] CS.25 EV 1.2 (eggs ov. female -1 ) [mm 3 ] FI.2 RE Dv Dh Db Ei Eb. Dv Dh Db Ei Eb Species Species Fig. 5. Box-whisker graphs showing mean wet weights (WW) and reproductive characteristics (CS, clutch size; EV, egg volume; FI, fecundity index and RE, reproductive effort) of five species of invasive gammarids in large rivers in South Germany during the years Abbreviations for species as in Fig. 2.

8 86 ARTICLE IN PRESS A. Kley, G. Maier / Limnologica 36 (26) 79 9 Table 4. Results of Tukey s post-hoc test concerning differences in weights and reproductive parameters between species Parameter WW (males) WW (females) CS EV FI RE D. villosus and D. haemobaphes were much heavier than adults of the other species, while weights of E. berilloni were lowest. Clutches produced by D. haemobaphes and D. villosus were greater than those of the other species; comparatively smaller clutches were produced by E. ischnus and D. bispinosus.. Volumes of eggs produced by E. ischnus were much greater than volumes of eggs produced by the other species. The fecundity index was highest in E. berilloni and lowest in E. ischnus; reproductive effort was higher in D. villosus and D. haemobaphes than in D. bispinosus and E. berilloni. Clutch sizes were closely correlated with weights of gammarids (Fig. 6). The steepest increase of clutch size with increasing weight was observed in E. berilloni, and the shallowest in E. ischnus. Egg volumes of all species (except E. berilloni) also increased with increasing wet weight (Table 5). However, no significant correlation (except in D. villosus) was found between clutch size and egg volume. Discussion Species ranking (Tukey Po.5) Dh4Dv4Db, Ei4Eb Dv4Dh4Ei4Eb, Db Dv4Dh4Eb4Ei, Db Ei4Dh, Db, Dv4Eb Eb4Dh4Dv4Db4Ei Dv, Dh, Ei vs. Ei, Db, Eb Abbreviations for weight, reproductive parameters and for species as in Tables 2 and 3, respectively. D. villosus is a very successful invader of rivers in Europe and expected to spread to North America (e.g. Bij de Vaate et al. 22; Bollache et al. 24; Devin et al. 21, Devin, Piscart, Beisel, & Moreteau 23; Dick & Platvoet 2, 21; Mu ller, Schramm, & Seitz 22). Due to its predatory habit (Dick & Platvoet 2, 21; Kinzler & Maier 23; Krisp & Maier 25), D. villosus can pose a threat to the native fauna. Our results show that D. villosus has become the most successful gammarid in the rivers of South Germany, having displaced native and invasive gammarids in many reaches. There are only some reaches in the Danube (beyond the Weltenburger Enge and near Deggendorf) and in the Isar mouth where D. haemobaphes, native, species and D. bispinosus could subsist, respectively. It is worthy to note that D. haemobaphes could surpass the Weltenburger Enge, which, due to its high current velocity and the lack of shipping was expected to be a barrier for invasive species. That E. ischnus could coexist with D. villosus at many sites was expected, because both species co-exist in their place of origin (Thienemann 195). Co-existence of D. villosus and E. ischnus may be possible by niche segregation. Kley & Maier (25) showed that in the Danube, E. ischnus lived at the uppermost shore-line in boulder substrate, whereas D. villosus prevailed in somewhat greater depth in gravelly substrate. Thus, both species are spatially segregated, which may facilitate their co-existence. According to Tittizer (1996), E. berilloni is present in the Rhine river. That we did not find E. berilloni in the Rhine itself, but only in a Rhine tributary, may be a result of the comparatively low number of sites we studied in the Rhine system. Another explanation could be that the small species, E. berilloni was displaced in the Rhine by predatious Ponto Caspian species, in particular by D. villosus (cf. Dick & Platvoet 2, 21; Kinzler & Maier 23). D. bispinosus, which was first recorded in the Danube in 1998 (Egger & Martens 21), obviously could not spread since that time. Kley & Maier (25) showed that D. bispinosus lived in the same microhabitat as D. villosus. They further showed opposite trends in abundance of both species at the site near Deggendorf. Increase in D. villosus abundance in spring of the year 23 was followed by a decrease in D. bispinosus abundance, which suggests some interactions between both species. That weights, number of precopulatory pairs, percentage of ovigerous females and clutch size are highest in spring, and that reproduction ceases in late autumn, as well as a balanced or slightly female-biased sex ratio throughout the year has been shown for many populations of European gammarids and seems to be characteristic in this group (e.g. Devin, Piscart, Beisel, & Moreteau 24; Hynes 1955; Iversen & Jensen 1977; Kley & Maier 23; Konopacka & Jesionowska 1995; Musko 1993; Teichmann 1982; Po ckl 1993a). The increase of egg number with increasing size of the mother is also common in many crustaceans including gammarids (e.g. Kolding & Fenchel 1981; Po ckl 1993b). That egg volume was not (negatively) related to clutch size in our study contrasts the results on some brackish marine gammarid species (Kolding & Fenchel 1981; Skadsheim 1984), but supports the results in some freshwater gammarids (Sheader 1983). Our results show that D. villosus can produce clutches of up to 16 eggs, which is high in comparison to the literature (Cioplan 1987; Devin et al. 24; Musko 1989). This high reproductive capacity together with rapid growth, early maturation, tolerance of a wide range of temperature and salinity as well as predatory strength (Bruijs, Kelleher, van der Velde, & Bij de Vaate 21; Devin et al. 24; Dick & Platvoet 2; Kinzler &

9 ARTICLE IN PRESS A. Kley, G. Maier / Limnologica 36 (26) Y = *X r =.779 P < 8.56e-39 Dv 2 Y = *X r =.757 P < 3.17e - 24 Dh CS (eggs ov.female -1 ) 15 1 CS (eggs ov.female -1 ) CS (eggs ov.female -1 ) Y = *X r =.57 P < 1.72e -17 Db CS (eggs ov.female -1 ) Y = *X r =.815 P < 2.24e - 44 Ei Y = *X r =.824 P < 2.8e -45 Eb CS (eggs ov.female -1 ) Fig. 6. Correlation (Pearson) between clutch size (CS) and wet weight (WW) with correlation coefficient, P-level and regression equation. Abbreviations for species as in Fig. 2. Maier 23) may have contributed to the rapid spread and finally to the success of this species in German and many other European waters (Bij de Vaate & Klink 1995; Dick & Platvoet 2; Devin et al. 21; Musko 1993). That D. haemobaphes can produce almost the same big clutches as D. villosus may have facilitated its rapid spread in the early 9 s. Why D. haemobaphes was displaced by D. villosus is unknown. Whether inferiority in predation, which is regarded as a key factor leading to elimination of species (cf. Dick 1992; 1996; MacNeil,

10 88 Table 5. ARTICLE IN PRESS A. Kley, G. Maier / Limnologica 36 (26) 79 9 Correlation (Pearson) between clutch size (CS) and egg volume (EV), and between egg volume and wet weight (WW) Species CS EV EV WW r P-level Y ¼ A þ B X r P-level Y ¼ A þ B X D. villosus Y ¼ :46 þ 525:65 X.4.1 Y ¼ :9 þ :4 X D. haemobaphes (ns) Y ¼ :9 þ :4 X D. bispinosus (ns).18.5 Y ¼ :1 þ :4 X E. ischnus (ns) Y ¼ :18 þ :6 X E. berilloni (ns) (ns) Dick, & Elwood 1997), has led to the decline of D. haemobaphes has to be cleared up in laboratory experiments. That E. ischnus produces large eggs in comparison to other gammarids has already been shown in an earlier study (Kley & Maier 23). Kley & Maier (23) suggested that the production of large eggs, which are often equipped with large amounts of reserves can be advantageous when food resources are sparse, since juveniles which hatch from eggs may benefit from reserves. D. villosus and D. haemobaphes seem to allocate their energy into producing many but small eggs thus maximising offspring number, while E. ischnus seems to allocate its energy into few but large eggs, possibly to increase juvenile survival in reaches with harsh food conditions (cf. Kolding & Fenchel 1981; Skadsheim 1984). The different reproductive strategies of D. villosus and D. haemobaphes vs. E. ischnus are reflected in the FI (weight-specific number of eggs), which is high in the former and low in the latter. Different reproductive strategies in D. villosus and E. ischnus may facilitate their frequent sympatric occurrence. Possibly, E. ischnus is able to survive in low-food microhabitats in the presence of D. villosus. This is a suggestion, however, which has to be examined in further experiments. Reproductive characteristics of E. berilloni and D. bispinosus resemble those of natives. Mean clutch sizes ranged between 1 and 4 and mean egg volumes approximately between.7 and.8 mm 3 and between.8 and.15 mm 3, which is roughly within the range of clutch sizes and egg volumes reported for native species (Po ckl 1993a, b; Teichmann 1982; Ward 1986). In summary our results show that the most successful invasive gammarids in Southern German streams (D. villosus, E. ischnus and D. haemobaphes; cf. Fig. 1) show particular reproductive characteristics, with large clutches or large egg volumes and high reproductive effort, which may support their success. E. berilloni and D. bispinosus, whose reproductive characteristics resemble those of native species, seem to be less successful invaders. Further, we think that simple habitat structures (e.g. the lack of aquatic vegetation; cf. Table 1) have promoted the success of invasives. Acknowledgements We want to thank the Landesgraduiertenfo rderung for financial support. We also thank anonymous referees for their valuable critical comments on an earlier draft of this manuscript and Frank Pa tzold who helped to find a site where the gammarid E. berilloni was present. References Bij de Vaate, A., Jazdzewski, K., Ketelaars, H. A. M., Gollasch, S., & van der Velde, G. (22). Geographical patterns in the range extention of Ponto Caspian macroinvertebrate species in Europe. Canadian Journal of Fisheries and Aquatic Sciences, 59, Bij de Vaate, A., & Klink, A. G. (1995). Dikerogammarus villosus Sowinsky Crustacea: Gammaridae a new immigrant in the Dutch part of the Lower Rhine. Lauterbornia, 2, Bikuna, B. G., & Asensio, R. (1991). Variaciones estacionales en la distribucion y abundancia de dos especies de anfipodos (Echinogammarus tarragonensis Pinkster, 1973) y Echinogammarus berilloni Catta (1878) (CL. Crustacea) en la cuenca del rio mercadillo (Pais Vasco). Zonacion y competencia, Graellsia, 47, Bollache, L., Devin, S., Wattier, R., Chovet, M., Beisel, J.-N., Moreteau, J.-C., & Rigaud, T. (24). Rapid range extension of the Ponto Caspian amphipod Dikerogammarus villosus in France: Potential consequences. Archiv für Hydrobiologie, 16, Bruijs, M. C. M., Kelleher, B., van der Velde, G., & Bij de Vaate, A. (21). Oxygen consumption, temperature and salinity tolerance of the invasive amphipod Dikerogammarus villosus: Indicators of further dispersal via ballast water transport. Archiv für Hydrobiologie, 152, Chovet, M., & Lecureuil, J. Y. (1994). Repartition des Gammaridae epiges (Crustaces, Amphipodes) dans la Loire et les rivieres de la Region Centre (France). Annales de Limnologie, 3, Cioplan, O. (1987). Zur Fortpflanzung von Dikerogammarus villosus Sow. Mart. Crustacea. Amphipoda im Eisernes Tor Stausee Bahna-Golf, Revue Roumaine de Biologie, 32, Devin, S., Beisel, J. N., Bachmann, V., & Moreteau, J. C. (21). Dikerogammarus villosus (Amphipoda: Gammaridae):

11 ARTICLE IN PRESS A. Kley, G. Maier / Limnologica 36 (26) Another invasive species newly established in the Moselle River and French hydrosystems. Annales de Limnologie, 37, International Journal of Limnology. Devin, S., Piscart, C., Beisel, J. N., & Moreteau, J. C. (23). Ecological traits of the amphipod invader Dikerogammarus villosus on a mesohabitat scale. Archiv für Hydrobiologie, 158, Devin, S., Piscart, C., Beisel, J. N., & Moreteau, J.-C. (24). Life history traits of the invader Dikerogammarus villosus (Crustacea : Amphipoda) in the Moselle River, France. International Review of Hydrobiology, 89, Dick, J. T. A. (1992). The nature and implication of differencial predation between Gammarus pulex and G. duebeni celticus (Crustacea: Amphipoda). Journal of Zoology, London, 227, Dick, J. T. A. (1996). Post-invasion amphipod communities of Lough Neagh, Northern Ireland: Influences of habitat selection and mutual predation. Journal of Animal Ecology, 65, Dick, J. T. A., & Platvoet, D. (2). Invading predatory crustacean Dikerogammarus villosus eliminates both native and exotic species. Proceedings of the Royal Society of London, Series B, 267, Dick, J. T. A., & Platvoet, D. (21). Predicting future aquatic invaders; the case of Dikerogammarus villosus. Aquatic Nuisance Species Digest, 4, Eggers, T. O., & Martens, A. (21). Bestimmungsschlu ssel der Su ßwasser-Amphipoda (Crustacea) Deutschlands. Lauterbornia, 42, Hynes, H. B. N. (1955). The reproductive cycle of some British freshwater Gammaridae. Journal of Animal Ecology, 24, Iversen, T., & Jensen, J. (1977). Life-cycle, drift and production of Gammarus pulex L. (Amphipoda) in a Danish spring. Freshwater Biology, 7, Kinzelbach, R. (1995). Neozoans in European waters Exemplifying the worldwide process of invasion and species mixing. Experientia, 51, Kinzler, W., & Maier, G. (23). Asymmetry in mutual predation: Possible reason for the replacement of native gammarids by invasives. Archiv für Hydrobiologie, 157, Kley, A., & Maier, G. (23). Life history characteristics of the invasive freshwater gammarids Dikerogammarus villosus and Echinogammarus ischnus in the river Main and the Main-Donau canal. Archiv für Hydrobiologie, 156, Kley, A., & Maier, G. (25). An example of niche partitioning between Dikerogammarus villosus and invasive and native gammarids: A field study. Journal of Limnology, 64, Kolding, S., & Fenchel, T. M. (1981). Patterns of reproduction in different populations of five species of the amphipod genus Gammarus. Oikos, 37, Konopacka, A., & Jesionowska, K. (1995). Life history of Echinogammarus ischnus (Stebbing, 1898) (Amphipoda) from artificially heated Lichenskie Lake (Poland). Crustaceana, 68, Krisp, H., & Maier, G. (25). Consumption of macroinvertebrates by invasive and native gammarids: a comparison. Journal of Limnology, 64, MacNeil, C., Dick, J. T. A., & Elwood, R. W. (1997). The trophic ecology of freshwater Gammarus (Crustacea: Amphipoda); problems and perspectives concerning the functional feeding group concept. Biological Reviews of the Cambridge Philosophical Society, 72, Mauch, E. (1963). Untersuchungen u ber das Benthos der deutschen Mosel unter besonderer Beru cksichtigung der Wassergu te. Mitteilungen des Zoologischen Museums Berlin, 39, Musko, I. B. (1989). Amphipoda Crustacea in the littoral zone of Lake Balaton Hungary. Qualitative and quantitative studies. Internationale Revue der gesamten Hydrobiologie, 74, Musko, I. B. (1993). The life history of Dikerogammarus haemobaphes (Eichwald) (Crustacea: Amphipoda) living on macrophytes in Lake Balaton (Hungary). Archiv für Hydrobiologie, 127, Mu ller, J. C., Schramm, S., & Seitz, A. (22). Genetic and morphological differentiation of Dikerogammarus invaders and their invasion history in Central Europe. Freshwater Biology, 47, Po ckl, M. (1993a). Beitra ge zur O kologie des Bachflohkrebses (Gammarus fossarum) und Flussflohkrebses (Gammarus roeseli). Natur und Museum, 123, Po ckl, M. (1993b). Reproductive potential and lifetime potential fecundity of the freshwater amphipods Gammarus fossarum and G. roeseli in Austrian streams and rivers. Freshwater Biology, 3, Schleuter, M., Schleuter, A., Potel, S., & Banning, M. (1994). Dikerogammarus haemobaphes Eichwald, 1841 (Gammaridae) aus der Donau erreicht u ber den Main-Donau-Kanal den Main. Lauterbornia, 19, Scho ll, F. (199). Erstnachweis von Chaetogammarus ischnus Stebbing im Rhein. Lauterbornia, 5, Sheader, M. (1983). The reproductive biology and ecology of Gammarus duebeni (Crustacea: Amphipoda) in southern England. Journal of the Marine Biological Association of the United Kingdom, 63, Skadsheim, A. (1982). The ecology of intertidal amphipods in the Oslofjord. The life cycles of Chaetogammarus marinus and C. stoerensis. Publicazioni Stationi Zoologici di Napoli (Marine Ecology), 2, Skadsheim, A. (1984). Coexistence and reproductive adaptions of amphipods: the role of environmental heterogeneity. Oikos, 43, Sutcliffe, D. W. (1992). Reproduction in Gammarus (Crustacea, Amphipoda): basic processes. Freshwater Forum, 2, Teichmann, W. (1982). Life history of Gammaridae under environmental conditions. Archiv für Hydrobiologie, 64(Suppl.), Thienemann, A. (195). Verbreitungsgeschichte der Su ßwasserwelt Europas. In A. Thienemann (Ed.), Die Binnengewässer, 18 (p. 89). Stuttgart: E. Schweizerbart Verlag. Tittizer, T. (1996). Vorkommen und Ausbreitung aquatischer Neozoen (Makrozoobenthos) in den Bundeswasserstrassen. In H. Gebhardt, R. Kinzelbach, & S. Schmid-Fischer (Eds.), Gebietsfremde Tierarten (pp ). Landsberg: Ecomed Verlag.

12 9 ARTICLE IN PRESS A. Kley, G. Maier / Limnologica 36 (26) 79 9 Tittizer, T., Leuchs, H., & Banning, M. (1994). Das Makrozoobenthos der Donau im Abschnitt Kehlheim Jochenstein (Donau-km ). Limnologie aktuell, 2 (pp ). Stuttgart: G. Fischer. Van den Bossche, J.-P. (22). First records and fast spread of new (1995 2) alien species in the river Meuse in Belgium: Hypania invalida, Corbicula fluminea, Hemimysis anomala, Dikerogammarus villosus and Crangonyx pseudogracilis. In Proceedings of the Symposium Status and trends of the Belgian fauna with particular emphasis on alien species (pp ). Ward, P. I. (1986). A comparative field study of the breeding behaviour of a stream and a pond population of Gammarus pulex (Amphipoda). Oikos, 46, Weygold, P. (1924). Die Embryonalentwicklung des Amphipoden Gammarus pulex. Zoologische Jahrbücher, 77, (Abteilung Allgemeine Zoologie und Physiologie der Tiere).