Comparative biology of the predatory cladoceran Cercopagis pengoi from Lake Ontario,

Baltic Sea and Caspian Sea

Igor A. Grigorovich, Hugh J. MacIsaac¹, Irina K. Rivier, Nikolai V. Aladin,

and Vadim E. Panov³

Great Lakes Institute for Environmental Research, University of Windsor, Windsor, Ontario

With 8 figures and 3 tables in the text

Cercopagis pengoi, an onychopod cladoceran native to the Ponto-Caspian-Aral region, was first found in the Baltic Sea in 1992 and in Lake Ontario in 1998. Morphological studies of C. pengoi from Lake Ontario, Baltic Sea and Caspian Sea revealed a high degree of variability in body length, structure and length of the caudal process, and shape and size of the brood pouch. The female brood pouch changed dramatically in size and shape between instar stages, resulting in distinctive appearances. Males had significantly shorter body length, caudal process length and distances between adjacent articular spines than did females. Instar III parthenogenetic females dominated populations of C. pengoi from all geographical localities. Body length of all instar stages, genders, reproductive modes from Lake Ontario was smaller than that of comparable individuals from European localities. The ratio of caudal process length to body length also varied geographically, with highest values observed for individuals from Lake Ontario. Clutch size for instar III parthenogenetic females from Lake Ontario was lower than values reported from the Baltic Sea and Caspian Sea. Observed differences in morphological and reproductive characteristics of geographically remote populations may reflect C. pengoi’s vulnerability to zooplanktivory or responses to environmental conditions.

Running Head: Comparative biology of Cercopagis pengoi

The Great Lakes have been invaded by a large number of species, including those from the Ponto-Caspian (Black, Azov and Caspian Seas) region (Mills et al. 1993). Ponto-Caspian species that have recently become established in the Great Lakes include zebra mussels (Dreissena polymorpha), quagga mussels (D. bugensis), round gobies (Neogobius melanostomus), tube-nose gobies (Proterorhinus marmoratus), amphipods (Echinogammarus ischnus), and onychopod cladocerans (Cercopagis pengoi) (MacIsaac & Grigorovich 1999). Zebra and quagga mussels, round gobies, and Echinogammarus amphipods are now ecologically dominant species in the lower Great Lakes, particularly in Lake Erie.

Cercopagis pengoi (Ostroumov, 1892) is native to the Black, Azov, Caspian and Aral sea basins (Sars 1897, 1902, Meisner 1908, Mordukhai-Boltovskoi 1960, 1968a). In 1959, the species was first recorded in the Kakhovka reservoir and in other downstream localities in the Dnieper River, Ukraine (Tseeb 1963, 1964). Damming and construction of reservoirs and canals on the Dnieper, Don and Manych rivers has facilitated C. pengoi dispersal in eastern Europe (Tseeb 1964, Zhuravel 1965, Mordukhai-Boltovskoi & Galinskiy 1974, Mordukhai-Boltovskoi 1979). This expansion was likely aided by shipping traffic and by intentional stocking of invertebrates (Tseeb 1964, Zhuravel 1967, Mordukhai-Boltovskoi & Galinskiy 1974). The species’ current distribution in the Black and Azov sea basins includes all reservoirs on the Dnieper River, the Tsimlyansk and Veselovsk reservoirs in the Don River drainage, and numerous canals in southern Ukraine (Zhuravel 1965, 1967, Glamazda 1971, Kaftannikova & Basilevich 1972, Volvich 1978, Gusynskaya & Zhdanova 1978, Gusynskaya 1989, Zimbalevskaya 1989, Shevtsova & Primak 1990).

C. pengoi was first reported outside of the Ponto-Caspian basin in coastal waters of the eastern Baltic Sea during 1992 (Ojaveer & Lumberg 1995). The species has since been reported from other coastal and open-water regions of the sea (reviewed by Krylov et al. 1999). The first report of Cercopagis in North America was made during summer 1998 in Lake Ontario, where the species was reported in relatively high numbers throughout the lake (MacIsaac et al. 1999). Lake Ontario was previously invaded in 1985 by a related cercopagid, Bythotrephes, though this species occurs rarely in the plankton (Johannsson et al. 1998). Both cercopagid species are believed to have been transported to the Great Lakes from Europe via ocean-going vessels (Sprules et al. 1990; MacIsaac et al. 1999).

C. pengoi occurred at an average density of 170 ind. m-3 in western Lake Ontario during August 1998 (MacIsaac et al. 1999). Ecologically, the species is a voracious predator on crustacean plankton (Mordukhai-Boltovskoi 1968a, Rivier 1998). Given its diet and abundance in this lake, C. pengoi could potentially impact the native plankton assemblages including carnivorous onychopods. For example, Bythotrephes and Cercopagis have inverse patterns of abundance in the Neva River estuary (Baltic Sea) (V.E. Panov, personal observation).

Knowledge of Cercopagis’ life history is essential to the development of predictive models of population growth and to evaluate the potential impact of this species on invaded ecosystems. Some aspects of C. pengoi biology and ecology have been reported for the Caspian Sea (e.g. Rivier 1967, 1998, Mordukhai-Boltovskoi & Rivier 1971, 1987) and Baltic Sea (Krylov & Panov 1998). For example, C. pengoi may reproduce both parthenogenetically and sexually, respectively producing embryos or resting eggs which develop in a dorsal brood pouch (see Rivier 1969, 1974). However, Cercopagis morphology and phenotypic variability have not been studied as extensively. Species boundaries in the genus Cercopagis are not clearly understood, as some diagnostic morphological features such as the structure and length of the caudal process and brood pouch shape exhibit variability (Rivier 1998). As a result, there is no clear consensus as to the geographical distribution of representatives of the genus. The valid records beyond the Ponto-Caspian-Aral region have been provided only for two taxa of Cercopagis - C. pengoi [in the Baltic Sea and Lake Ontario] and C. ossiani Mordukhai-Boltovskoi, 1968 [in the Baltic Sea] (Ojaveer & Lumberg 1995, MacIsaac et al. 1999, Simm & Ojaveer 1999). Six other species and intraspecific forms of Cercopagis which are endemic to the Ponto-Azov-Caspian-Aral region, namely C. socialis (Grimm, 1877), C. cylindrata Sars, 1897, C. longicaudata Sars, 1902, C. neonilae Sars, 1902, C. pengoi gracillima Sars, 1902, and C. pengoi aralensis Mordukhai-Boltovskoi, 1974, have also been reported from localities beyond their native ranges (Tseeb et al. 1971, Ivanov & Makartseva 1982, N.V. Aladin, personal observation). However, to date, no evidence has been presented to confirm these distributions.

In this study, we describe morphology, size patterns, fecundity, and demography of Cercopagis pengoi from Lake Ontario, and compare these characteristics with those of populations from the Baltic Sea and Caspian Sea. Furthermore, we examine morphological characters used in the taxonomy of C. pengoi across instar stages, genders, reproductive modes and geographical localities to clarify the extent of ontogenetic (i.e. age) and geographical variability in this species. We also assess weight-length relationships for C. pengoi from Lake Ontario. As well, we review the systematic history of the species and its intraspecific forms, and discuss morphological differences among the forms.

Cercopagis pengoi (Ostroumov, 1892) belongs to the family Cercopagidae (Crustacea: Cladocera: Onychopoda), which comprises two genera, Bythotrephes and Cercopagis. Members of this family are characterized by the following combination of morphological features, many of which reflect their predatory lifestyle (Mordukhai-Boltovskoi 1968a, 1968b): 1) the cephalon, or head, is composed primarily of a compound eye; 2) antennae I, or antennules, are reduced to small appendages positioned beneath the eye; 3) antennae II constitute well developed, biramous appendages used for swimming (Fig. 1); 4) the mandibles possess a sharp cutting edge and terminal denticles and are adapted for biting; 5) the trunk bears four pairs of grasping thoracopods, lacking exopods; 6) thoracopods I are tremendously longer than thoracopods II-IV and differ in shape; and 7) the well-developed abdomen bears an elongate caudal process, extending at least one body length (Mordukhai-Boltovskoi & Rivier 1987).

The number of setae on rami of antennae II is used as a distinguishing feature in separating the genus Cercopagis from Bythotrephes (Sars 1897, Mordukhai-Boltovskoi & Rivier 1987). There are seven natatory setae on each ramus of antennae II in Cercopagis, whereas there are eight setae on the outer (four-segmented) ramus and seven setae on the inner (three-segmented) ramus in Bythotrephes (Sars 1897) (Fig. 2). Members of the genus Cercopagis lack a gnathobasic process on thoracopod I (Sars 1897). The shape and structure of the penis differ between Cercopagis and Bythotrephes (Mordukhai-Boltovskoi & Rivier 1987). In Bythotrephes, the penes are shorter and covered with minute setae, whereas, in Cercopagis, the penes are rather long and smooth (Mordukhai-Boltovskoi & Rivier 1987). In addition, the two genera differ in the shape of the labrum, the condition of the anterior mandibular process, the length of spines on the external face of the thoracopods I-III, and proportion between the length of abdomen and the combined length of the thorax and cephalon (Martin and Cash-Clark 1995).

Systematics in the genus Cercopagis has been revised since its original description in 1897 (Mordukhai-Boltovskoi 1968b, Mordukhai-Boltovskoi & Rivier 1987). Two former genera erected by Sars (1897), Cercopagis and Apagis, were later recognized by Mordukhai-Boltovskoi and Rivier (1987) as subgenera treated within the genus Cercopagis. The two taxa differ in the length and structure of the caudal process and the presence (in Cercopagis) of a loop (Sars 1987, Mordukhai-Boltovskoi 1968a, Mordukhai-Boltovskoi & Rivier 1987). The systematics of Cercopagis and Apagis and relationships between these taxa are poorly understood. It has been suggested that Cercopagis (Apagis) ossiani may represent the first generation of Cercopagis (Cercopagis) pengoi that emerge from resting eggs (Simm & Ojaveer 1999, I. Rivier, unpublished data).

Several species of Cercopagis established by Sars (1897, 1902) were thought to be intraspecific forms of the single polymorphic species C. pengoi (Mordukhai-Boltovskoi 1960, 1965) and were, therefore, combined into the ‘Cercopagis pengoi group’ (Mordukhai-Boltovskoi 1968a). This group comprises four taxa including the typical form (i.e. C. pengoi sensu stricto), C. neonilae, C. pengoi gracillima, and C. pengoi aralensis. Members of this group can be distinguished from all other species of the subgenus Cercopagis based on the following set of characters: 1) the presence of a pointed apex on the brood pouch; 2) the proximal articular spines of the caudal process are well-developed and extend at least 3/4 the distance between the base of these spines on the caudal process and attachment point of the caudal process on the abdomen; 3) the distances between adjacent pairs of articular spines are relatively large and more or less comparable to the abdomen length in females (Mordukhai-Boltovskoi 1968a). It is not entirely clear whether members of the Cercopagis pengoi group represent a number of morphologically similar species or a single, variable species. In addition to morphological differences, members of this group have different geographical distributions and slightly different salinity preferences (Rivier 1967, Mordukhai-Boltovskoi & Rivier 1987).

Cercopagis pengoi was first found in the Azov Sea on 27 July 1876 by Russian scientist K.K. Pengo (Pengo 1879). Owing to similarities in gross morphology, animals were originally identified as marine representatives of Bythotrephes, but no specific name was given. The species B. pengoi was formally established by Ostroumov in 1892 (Ostroumov 1892). Presumably unaware of Ostroumov’s report, Sars described this form as a new species Cercopagis tenera from Caspian Sea material (Sars 1897), and the two species were later recognized to be identical (Zernov 1903, Meisner 1908). Subsequent workers adopted this arrangement (Mordukhai-Boltovskoi 1960, 1968a, Mordukhai-Boltovskoi & Rivier 1987). C. pengoi ranges widely over eastern Europe as native or introduced populations. It occurs over a broad spectrum of salinity, from 0.1 to 14.0‰ (Valkanov 1957, Mordukhai-Boltovskoi 1969, Mordukhai-Boltovskoi & Rivier 1971).

Cercopagis gracillima was originally described by Sars (1902) based on Caspian Sea specimens. It is identified by an extremely extended pointed apex at the tip of the brood pouch, the length of which is greater than the maximum width of the brood pouch, an elongate abdomen that exceeds in length the combined length of the thorax and cephalon, and lengthened, abruptly curved articular spines that far exceed in length the distance between the base of the proximal pair of articular spines and the point of attachment of the caudal process on the abdomen (Mordukhai-Boltovskoi 1968a). Because various transitional phenotypes were observed between the typical forms C. pengoi and C. gracillima in the northern Caspian Sea, the latter is today considered an intraspecific form of C. pengoi (Mordukhai-Boltovskoi & Rivier 1987; Rivier 1998). Relative to C. pengoi, C. gracillima occupies a restricted distribution range in the northern Caspian Sea and tolerates restrained salinity values between 1-12‰ (Rivier 1967).

In erecting C. neonilae from Azov Sea specimens, Sars (1902) described only minor differences between the new species and C. tenera (=pengoi), which caused some doubts with respect to the validity of the former taxon (Meisner 1908, Mordukhai-Boltovskoi 1960, 1965). Meisner (1908) placed C. neonilae in synonymy with C. pengoi. C. neonilae is distinguished from C. pengoi in that it has a shorter pointed apex on the brood pouch, an abdomen which is shorter than the combined length of the thorax and cephalon, and articular spines which are shorter than the distance between the base of the proximal pair of articular spines and the point of attachment of the caudal process on the abdomen (Sars 1902, Mordukhai-Boltovskoi 1968a, Mordukhai-Boltovskoi & Rivier 1987). It remains to be determined whether these features are sufficient to warrant designation of C. neonilae as a separate species, but it has been assigned to the Cercopagis pengoi group (Mordukhai-Boltovskoi 1968a, Mordukhai-Boltovskoi & Rivier 1987). Its distribution encompassing the Caspian Sea, Azov Sea and Black Sea estuaries overlaps with that of C. pengoi (Mordukhai-Boltovskoi & Rivier 1987). In the Caspian Sea, C. neonilae appeared to exhibit a preference to mild salinity of 1-12.5‰ (Rivier 1967).

The subspecies C. pengoi aralensis was first described by Mordukhai-Boltovskoi (1974) based on specimens collected in the Aral Sea. This form is distinguished from typical C. pengoi by its possession of a considerably smaller caudal process (less than 3.5 times the body length) and a variously reduced loop on the caudal process (Mordukhai-Boltovskoi & Rivier 1987). Variation in shape of the loop in this form results in a variety of appearances, from bent, bearing fields of minute spines, to a straight, kinkless process. Owing to a drastic increase in salinity (to levels over 18‰) since the 1980s in the Aral Sea, Cercopagis spp. were likely extirpated from the sea proper and now occur only in adjacent Lake Sudochie in the Amudarya River estuarine region (Aladin 1996, N.V. Aladin, personal observation).

Cercopagis was collected on 27 August 1998 at eight sites [43°N and 79°W] in western Lake Ontario using bottom to surface hauls (£ 27 m) of a 0.5-m-diameter, 253-µm mesh plankton net. Additional C. pengoi were collected from Lake Ontario by trawling offshore in £ 20 m with weighted Trilene fishing line (5.4 kg test). Animals snagged on fishing lines by their caudal process were recovered as a fibrous mass. C. pengoi was collected during July 1997 from the Neva River estuary [60°180 N and 28°430 E], Baltic Sea, Russia, by 10m vertical hauls of a 0.2-m-diameter, 150-µm mesh plankton net. An additional collection was made on the northern Caspian Sea about 75 km from the Volga River delta [45°030 N and 48°170 E], near Zhemchuzhnyi Island, during 22-24 August 1998. The sample was collected using a 0.39-m-diameter, 74-µm mesh plankton net, and is a composite of one horizontal surface tow and one 10 to 3 m vertical tow. All samples were preserved in 70% ethanol or in 4% sugar-formalin. In the laboratory, Cercopagis was separated from other material and transferred to 70% ethanol within a few days after sampling.

All surveys were conducted during relatively calm periods. Water column temperature in all surveyed basins ranged within the limits of 17-24°C. In Lake Ontario, water temperature at the collections sites varied from 23.9°C at surface to 17.4°C at bottom. Less marked temperature variation was recorded during sampling in the Baltic Sea, from 18 to 21°C. Salinity differed greatly among surveyed basins, ranging between 0.14-0.20‰ in Lake Ontario, 2.0-3.4‰ in the Baltic Sea and 6-7‰ in the Caspian Sea.

Only the typical C. pengoi ‘form’ was detected in all localities. We employed both light (12H to 50H ) and scanning electron microscopy (S.E.M.; 60H to 2000H ) to examine morphological characters of C. pengoi. However, because some structures showed evidence of severe distortion after coating with gold, we limit use of S.E.M. images in this study.

We used only intact animals with a complete caudal process to evaluate the relationship between body length and caudal process length. Damaged animals were utilized for body length measurements only if both the entire body and proximal portion of the caudal process were present. In total, we examined 4550 Cercopagis specimens from Lake Ontario, >73 animals from the Baltic Sea and >43 animals from Caspian Sea.

Gender was determined by the presence (in males) or absence (in females) of paired penes on the ventral surface of the thorax, behind the base of the last thoracic limbs. Mature sexual females were discriminated from parthenogenetic females by their possession of resting egg(s) in the brood pouch or ovary (see Results). The term ‘non-ovigerous females’ was used to define females without eggs or embryos in the brood pouch. The caudal process is typically comprised of one articular segment in instar I animals, while two and three articles are present in instars II and III, respectively. According to studies of postembryonic development in Bythotrephes (Zozulya 1977, Yurista 1992), that the caudal process in cercopagids likely grows by addition of a new articular segment from a growth zone, thereby pushing the retentive, older portion of caudal process in the posterior direction. Instars of C. pengoi were distinguished by the number of articular segments and paired spines positioned on the caudal appendage. We recorded the presence of parthenogenetic and sexual females, males, instar stages, and stages of brood development.

Images of Cercopagis were outputted from a Leica Wild M8 dissecting microscope via a Hitachi VK-C370 videocamera and Flashpoint 128 framegrabber, to Optimas 6.2 image analysis software on a personal computer. We measured: 1) body length as the distance from the tip of the eye to the anus, located between the proximal pair of articular spines on the caudal process (after Burkhardt 1994); 2) total length of the caudal process as the distance from the anus to the distal end of the caudal process (after Burkhardt 1994); 3) diameter of the caudal process at the point of its attachment on the abdomen; 4) length of the articular spine III; 5) distances between adjacent articular spine pairs; 6) brood pouch size as the distance from the tip of the brood pouch to the inferior point of insertion of the brood pouch on the trunk; and 7) clutch size. All comparisons among Lake Ontario, Baltic Sea, and Caspian Sea populations were made using alcohol-preserved specimens.

In order to assess weight-length relationships, 334 Cercopagis from Lake Ontario, preserved in sugar-formalin, were sorted according to gender, instar stage and mode of reproduction (parthenogenetic vs. sexual), and categorized into the following body length groups: 0.81-1.00, 1.01-1.20, 1.21-1.40, 1.41-1.60, 1.61-1.80, and 1.80-2.00 mm. Each group included between 5 and 30 individuals, and were dried for 24h at 60°C and held over silica gel prior to weighing. Dry mass was measured with a Cahn 4700 electrobalance to a precision of 1µg. The relationship between body mass and body length was assessed for different parthenogenetic female instars using ANCOVA, with body length as a covariate and instar stage as a categorical variable. The initial model revealed no interaction between mass and length, thus the model presented here lacks an interaction term.

Differences in size dimensions and structural indexes among instars and populations of C. pengoi were examined using one-way ANOVA, followed by Bonferroni’s multiple comparisons test when ANOVA was significant. Two-way ANOVA (instar, basin) test was utilized only for the analysis of differences in body length of instar II and III parthenogenetic females from Lake Ontario, Caspian Sea and Baltic Sea. Differences between genders and among reproductive modes (parthenogenetic vs. sexual) were examined using t-tests. The relationships between caudal process length and body length and between clutch size and body length in all instars of parthenogenetic females from Lake Ontario were examined using linear regression analysis. This analysis was repeated for the population from the Caspian Sea. All data were transformed as log₁₀(x+1) prior to statistical analysis to stabilize variance. All tests were conducted using Systat 8.0.

Morphometric characteristics of C. pengoi varied among localities, instar stages and genders (Table 1). We limit descriptions of the ratio of caudal process length to body size, and the length and structure of the caudal process, largely to instar III parthenogenetic females, as these individuals were numerically dominant in all samples. Body length of instar III females was significantly different among all surveyed basins (ANOVA, P<0.05). The smallest mean body size of instar III parthenogenetic females was recorded in Lake Ontario (1.45 mm), followed by that in Caspian Sea (1.73 mm) and Baltic Sea (1.99 mm). A similar pattern of geographical variation in body size was established for instar II parthenogenetic females (ANOVA, P<0.05).

Caudal process length and body length were positively correlated in instar III females from Lake Ontario (R2=0.19; P<0.001; n=338) (Fig. 3). This relationship was much weaker for comparable individuals from the Caspian Sea (R2=0.23; P=0.096; n=13), possibly owing to small sample size. A significant relationship between caudal process length and body length was also established for in instar II (R2=0.39; P<0.001; n=150) and instar I (R2=0.20; P=0.005; n=36) parthenogenetic females from Lake Ontario.

The Lake Ontario population also differed from other populations in terms of caudal process length (see Table 1). Length of the caudal process in instar III females from Lake Ontario was significantly greater than that in Caspian Sea (t-test, P<0.001). Comparison with animals from the Baltic Sea could not be made because only one instar III female had an intact caudal process. The ratio of caudal process length to body length was distinguishable for each population (Table 2). The highest value of this index for instar III females was reported in Lake Ontario (5.9), followed by the Baltic Sea (4.9) and Caspian Sea (4.3). In the Caspian Sea, C. pengoi populations from northern and central basins were in turn different with respect to relative length of caudal process (I.K. Rivier, unpublished data). For example, the ratio of caudal process length to body length for instar III parthenogenetic females from the central basin of Caspian Sea averaged 6.5 (n=14) (I.K. Rivier, unpublished data) and exceeded the values (4.3) that we reported from the northern basin.

Spacing and relative development of articular spines on the caudal process are additional characters used in Cercopagis taxonomy (Mordukhai-Boltovskoi & Rivier 1987). The ratio of the distance among articular spines II and III to body length in female averaged 0.50 and did not differ significantly between surveyed basins (ANOVA, P=0.16). Length of articular spine III (available in instar III animals) was variable both within (up to 18%) and among (64%; ANOVA, P<0.001) basins. The ratio of size of the articular spine III to diameter of the caudal process typically exceeded 1.0 in males and females (Table 2).

Sexual females of Cercopagis produce only resting eggs. Resting eggs developing in the ovary or brood pouch are visible with light microscopy as a dark-contrast mass shielded by yolk granules (Fig. 4). Instar III sexual females from Lake Ontario possessed significantly larger body size and caudal process length than did parthenogenetic females (t-test, P<0.001) (Table 1). In instar III sexual females, the ratio of the distance between articular spines III and II to body length was approximately 0.50 (Table 2). Mean body length of instar III sexual females in Lake Ontario (1.58 mm) was significantly smaller than that of individuals from the Baltic Sea (1.90 mm; t-test, P=0.019).

Males were rare in the Baltic Sea collection and absent in the Caspian Sea sample. Therefore our treatment of males is limited primarily to individuals from Lake Ontario. Thoracopod I of males is of a grasping type and bears a toothed hook (Mordukhai-Boltovskoi 1968a). Males also possess paired penes behind thoracopods IV. On the dorsal side of the trunk, males bear a tubercle outgrowth, resembling the brood pouch of newborn females (Fig. 5). The loop of the caudal process in males is bent much less than that of females (see Fig. 4C, 5C).

Based on Lake Ontario data, the mean body length of males was significantly smaller than that of parthenogenetic and sexual females for instar II and III animals (ANOVA, P<0.005), but not for instar I individuals (ANOVA, P=0.58). All male instars had significantly shorter caudal process lengths and distances between adjacent articular spines than did comparable females (t-tests, P<0.001). Males also had a lower ratio of caudal process length to body length and ratio of distance between articular spines III and II to body length (Table 2). Males further differed from females in terms of the relative length of articular segments III and II of the caudal process. The mean distance between articular spines I and II (0.33 mm) was significantly greater (t-test, P=0.048) than the distance between articular spines II and III (0.30 mm). These distances respectively correspond with the sizes of articular segments II and III of the caudal process (Fig. 5). The mean distance between articular spines II and III in parthenogenetic and sexual females was greater than that between articular spines I and II (t-test, P=0.049). Trends in size variation of males from Lake Ontario and the Baltic Sea were similar to patterns established for females.

Significant instar-related differences existed with respect to body length and caudal process length for all reproductive groups from Lake Ontario (Table 2). Increases in body length between moults exceeded increases in caudal process length in parthenogenetic females, sexual females and males, resulting in a slightly reduced spiniferous appearance with increasing instar stage. Mean body length increased by 42.2% and the caudal process length increased by 36.9% from instar I to III in parthenogenetic females. In sexual females, comparable increases between instars I to III were 58.0% in body length and 40.5% in caudal process length. There was much less growth in males, as body length and caudal process length increased by only 39.8% and 21.4%, respectively, between these instars.

Shape and size of the brood pouch varied dramatically between instars (Fig. 6). Instar I parthenogenetic females typically possessed a small brood pouch with a rounded to somewhat angular apex (Fig. 6A-B). Between instars I and II, the pouch lengthens dramatically to accommodate the growing embryos, and becomes pointed at the apex. In parthenogenetic females from Lake Ontario, brood pouch length increased by 120% between instars I and II. However, a small number (5 of 171 specimens) of instar II parthenogenetic females examined from Lake Ontario had a divergent brood pouch shape. In these individuals, the brood pouch was rounded and lacked a pointed apex (see Fig. 6C) similar to that described for C. socialis (Sars 1897). The brood pouch of instar III parthenogenetic females was 236% larger than that of instar I individuals and was always distinctly pointed at the apex (Fig. 6E-F). Inter-instar changes in the shape and size of the brood pouch of sexual females were similar to those described for parthenogenetic females (Table 1). Brood pouch size of sexual females from Lake Ontario increased by an average of 337% between instars I and III. All instar II and III sexual females from Lake Ontario always had a pointed apex on the brood pouch.

Instar III parthenogenetic females dominated populations of C. pengoi from all geographical localities (Fig. 7). In the Baltic Sea, there was an apparent modal peak of parthenogenetic females with a body length of 1.8-2.0 mm, comprising 25.0% of the population in July 1997. Instar III parthenogenetic females of 1.6-1.8 mm constituted the predominant size group in Caspian Sea, accounting for 26.2% of the population in August 1998. In Lake Ontario, there was a modal peak of instar III parthenogenetic females with body length of 1.4-1.6 mm, accounting for 20.8% of the population in August 1998. The Lake Ontario population differed from others by the absence of individuals with body length >2.0 mm and by a considerably higher proportion of animals <1.00 mm.

Males made up ca. 9.5 and 3.8% of the total population of C. pengoi in the Baltic Sea and Lake Ontario, respectively. No males were collected from the Caspian Sea.

Fecund individuals of parthenogenetic females were observed at all instars, whereas sexual females were not in a reproductive condition at instar I. Fecundity in sexual females from Lake Ontario was typically lower than that of parthenogenetic females. The mean number of resting eggs carried per ovigerous female in Lake Ontario was somewhat smaller than that in the Baltic Sea. For example, in Lake Ontario, 80.4% of ovigerous sexual females (at instars II and III) carried one-egg broods, and 18% had two-egg broods. In the Baltic Sea, the percentage of ovigerous sexual females with one-egg and two-egg broods was 14.3 and 85.7%, respectively. Of the 98 sexual females examined from Lake Ontario, only one carried three resting eggs. In the Caspian Sea sample, we detected only a single instar II sexual female carrying a one-egg brood.

Mean clutch size of instar III parthenogenetic females from Lake Ontario (3.0; n=292) was significantly lower than that (4.7; n=7) observed for similar individuals from the Baltic Sea (t-test, P=0.047). Furthermore, clutch size of ovigerous parthenogenetic females at instar III averaged 13 in Caspian Sea (Mordukhai-Boltovskoi & Rivier 1987). Differences in fecundity among geographical populations of C. pengoi were comparable to differences in brood pouch length (Table 1). Interestingly, there was no significant correlation between clutch size and body length in all instars of parthenogenetic females from Lake Ontario (R2£ 0.023).

Body mass of C. pengoi from Lake Ontario varied by instar, gender and reproductive mode (Fig. 8A). Differences between instars I and II and between instars II and III were considerably greater for females than for males.

Body mass was closely correlated with body length (F=68; df=1, 7; P<0.001) and instar stage (F=52; df=2, 7; P<0.001) in parthenogenetic females (ANCOVA). As a result of continuous weight increase with increasing length and instar stage, dry mass of parthenogenetic females varied more than 6-fold among instars. Within the same size group of parthenogenetic females, dry mass increased progressively from instar I to III (Fig. 8B). This increase in body mass was partially attributable to differences in brood development between instars. Regression models indicated that inclusion of instar status enhanced prediction of the body mass–body length relationship (Table 3).

Results of our study demonstrated substantial variability in body dimensions, size clutch, and proportions of parthenogenetic and sexual individuals among Cercopagis pengoi populations. Examination of C. pengoi morphology, involving the characterization of instar stages, gender, and reproductive modes from geographically remote localities, revealed a high degree of ontogenetic and geographical variability in the morphological characters that, heretofore, have been used in cercopagid taxonomy. Temperature, food supply, and predation are three factors most often implicated as regulators of body size variation in the related cercopagid Bythotrephes (Burkhardt 1994; Ketelaars et al. 1995; Yan & Pawson 1998). Salinity was also reported to affect species’ phenotypic variability in the Caspian Sea (Mordukhai-Boltovskoi & Rivier 1987). It has been further suggested that variation in body size and other life-history traits in C. pengoi may be the result of genetic differentiation among geographically remote populations (I.K. Rivier, unpublished data).

The present study demonstrated significant differences in body length among populations. We do not know if the geographical location of the waterbodies is an important source of this variation; however, body size of all instars, genders and reproductive modes of C. pengoi was positively related to latitudes, indicating a possible influence of climate on size variability. For example, in Lake Ontario (latitude 43°N) mean body length of instar II and III parthenogenetic females (1.45 and 1.29 mm, respectively) was significantly smaller than that of comparable individuals (1.73 and 1.51 mm) in the Caspian Sea (latitude 45°N). These animals were, in turn, smaller than comparable individuals (1.99 and 1.53 mm, respectively) from the Baltic Sea (latitude 60°N). Likewise, from high to lower latitudes, size-frequency distributions of C. pengoi were characterized by a decline in the proportion of larger individuals (>1.8 mm) and by an increase in the proportion of smaller animals (<1.00 mm). Sun radiation and temperature regimes undoubtedly differed at these latitudes, although water temperature fluctuated within the same range at the collection sites during surveying. Perhaps owing to the species’ euryhalinity, body size showed no consistent relationships to salinity of the surveyed basins. Observed differences in body dimensions and population demography most likely reflected a complex influence of many environmental factors and the relative importance of different factors could vary between systems.

In addition to extensive inter-population variability, mean body length was variable among instars, genders, and between reproductive modes of females within local populations (Table 1). Lake Ontario data revealed a significant positive relationship between caudal process length and body length for instar III parthenogenetic females, though not for instars I and II parthenogenetic females. Moreover, the ratio of caudal process length to body length declined between instars I and III from 6.3 to 5.6 in sexual females, from 6.1 to 5.9 in parthenogenetic females, and from 4.5 to 3.9 in males. This indicates that growth of the body between moults exceeds that of the caudal process. At all instar stages males had significantly shorter caudal process length than did females and were, therefore, characterized by a lower caudal process-to-body ratio. This index also varied markedly between comparable individuals from geographically remote populations (Table 2). Within local populations, spacing of adjacent pairs of articular spines on the caudal process was distinguishable between genders. For example, males typically had shorter distances between adjacent pairs of articular spines. In both parthenogenetic and sexual females the third articular segment exceeded the second one, as evidenced by distances between adjacent pairs of articular spines. Males exhibited the opposite pattern of articular segment lengths (Tables 1, 2). Since the cercopagid caudal process is retained during moulting, its growth could be considered a ‘time axis’ of development (Burkhardt 1994). Consequently, the reduced length of articular segments of the caudal process may indicate faster development of males. However, inasmuch as males and females experience the same environmental conditions, this hypothesis is unlikely. The different spacing of adjacent pairs of articular spines on the caudal process between genders may also be a result of allometric growth.

Previous research identified the adaptive significance of the caudal process for cercopagid onychopods in stabilizing forward motion, steering movement, extinguishing oscillations during functioning of the natatory antennae, buoyancy regulation, and especially as a deterrent against planktivory by small/young fish predators (Zozulya 1978, Barnhisel 1991, Grigorovich et al. 1998). However, it may be questioned whether the caudal process has evolved to function in all these ways.

It has yet to be established whether and to what extent observed geographical variation in the caudal process-to-body length relationships in C. pengoi is related to differences in intensity of zooplanktivory among basins. Research on Daphnia demonstrated that most planktivorous fish prefer large-bodied prey, selectively removing the largest individuals from populations (Tessier at al. 1992). This study has shown that C. pengoi from Lake Ontario had a greater caudal process to body length ratio than animals from either the Caspian Sea or Baltic Sea. This pattern suggests that intensity of planktivory may be greater in Lake Ontario than either of the European basins. The Lake Ontario zooplankton community has been dominated by small-bodied species for many years, apparently due to intense planktivory by alewife (Alosa pseudoharengus) (Mills at al. 1995; Johannsson et al. 1998). Because C. pengoi is now one of the largest zooplankton species in the lake, it should be highly vulnerable to alewife predation. Studies conducted on both the Baltic Sea and Caspian Sea revealed intense positive selection for C. pengoi by a number of planktivorous fishes including Clupea harengus membras, Clupeonella cultriventris, Gasterosteus aculeatus, Pungitius pungitius, Alburnus alburnus, Osmerus eperlanus eperlanus, and Engraulis encrasicolus (Ojaveer & Lumberg 1995; Rivier 1998). Furthermore, larvae of benthivorous fishes including Zosterisessor ophiocephalus, Neogobius melanostomus, Neogobius fluviatilis fluviatilis, and Mesogobius batrachocephalus have also been reported to consume C. pengoi (Yankovskii 1970). A related cercopagid, Bythotrephes, is subject to intense, positive selection by adult lake herring (Coregonus artedi) in Harp Lake, Ontario (Coulas et al. 1998).

In addition to body size and caudal process length variability, we also documented changes in the shape and size of the female brood pouch between instar stages and among populations. These female attributes were so variable as to be almost useless in the species’ determination, except in cases where the pointed apex was well defined in instar III females. During development, embryos increase in size and distend the brood pouch of parthenogenetic females (Fig. 6). Rivier (1974) reported that in C. pengoi females from Caspian Sea, the brood pouch containing fully developed neonates is separated from the maternal body and released. During this process, eggs of the new brood arrive into a new brood chamber, which forms separately under the covering of the old chitin pouch with enclosed embryos, after which the old brood is released (Rivier 1969). At the early stage of brood development (in instar I and II, as well as in instar III immediately after moulting), the brood pouch was widely elliptical and lacked a pointed apex. However, at the late stage of brood development in instar III parthenogenetic females, the brood pouch always terminated with a distinctive point. To accommodate growing embryos, the brood pouch increases in size and becomes distinctly pointed at the tip. A few instar III parthenogenetic females from each basin surveyed had extraordinarily large brood pouches, which equaled or markedly exceeded the body length. Growth of the brood pouch was found to be more independent of growth of the clutch in sexual females than in parthenogenetic females. For example, in sexual females from Lake Ontario, mean size of the brood pouch increased from 0.68 to 0.83 between instars II and III, whereas mean size of resting eggs increased from 246 µm (SD=30 µm, n=4) to 301 µm (SD=40 µm, n=19) (I.A. Grigorovich, unpublished data). Yet in parthenogenetic females from Lake Ontario, mean size of the brood pouch increased from 0.55 to 0.84 between instars II and III, whereas mean length of embryos increased from 152 µm (SD=87 µm, n=133) to 351 µm (SD=122 µm, n=194) (I.A. Grigorovich, unpublished data). Contrary to parthenogenetic broods, resting eggs were carried in smaller number (i.e., one or rarely two eggs per ovigerous female) and shape of the eggs remained invariably globular.

Growth in brood mass was accompanied by marked increases in total body mass of C. pengoi females. Parthenogenetic females exhibited the most variability in total body mass in consequence. Results of this study suggest that total investment in brood development may account for observed instar-related differences in mean body mass within the same size group of parthenogenetic females. Inclusion of instar stage in C. pengoi weight estimation models efficiently improved body mass predictions (Table 3).

Results from this study confirm the importance of the two structural indexes in the identification of C. pengoi – the ratio of the length of articular spine III to diameter of the caudal process at its base and the ratio of the distance between articular spines II and III to body length. The former index typically exceeded 1.0, while the latter averaged nearly 0.5 in females and 0.2 in males. Articular segment III (bearing articular spines III) is present only in instar III animals and varies in size between genders. This pattern poses a problem in identification of instar I and II individuals. We observed considerable individual and instar-specific variation in the shape of the brood pouch, and therefore question its diagnostic utility. Furthermore, the proportion between the mean abdomen length and the mean combined length of the thorax and cephalon varied among populations (I.K. Rivier, unpublished data), thus the utility of this index in the differential diagnosis of C. pengoi is also problematic. For instance, the proportion of abdomen in the total body length in instar III parthenogenetic females was 59.5% (n=28) in Lake Ontario, 47.6% (n=40) in the Baltic Sea and 50.0% (n=14) in the central Caspian Sea (I.K. Rivier, unpublished data).

The present study provides preliminary data on inter-population variability in body dimensions, population demography and fecundity of C. pengoi sensu stricto. The related cercopagid Bythotrephes also exhibits substantial seasonal variation in body size and fecundity (Mordukhai-Boltovskoi & Rivier 1987, Burkhardt 1994, Ketelaars et al. 1995, Yan & Pawson 1998) Consequently, further morphological work is necessary to identify species’ responses to seasonal changes in environmental or biological conditions. Reevaluation of species richness in the genus Cercopagis based on joint morphological and genetic techniques presents another important endeavor for future taxonomic research.

Variability and availability of presumed intraspecific forms indicate that the C. pengoi group is most likely undergoing adaptive radiation (see Mordukhai-Boltovskoi 1965). Mordukhai-Boltovskoi (1965) and Rivier (1967, 1998) hypothesized that speciation in this group in the Caspian Sea might come from C. neonilae via C. pengoi to C. gracillima. This evolutionary flow has led to the disappearance of the angle between the abdomen and the brood pouch, lengthening of the abdomen, increase in size of articular spines and spacing between adjacent articular spine pairs on the caudal process, and elongation of the brood pouch and extension of its pointed apex (Mordukhai-Boltovskoi & Rivier 1987).

C. pengoi was observed to coexist with its congener C. ossiani during the early growth season in the Caspian Sea and Baltic Sea (Simm & Ojaveer 1999, I.K. Rivier, unpublished data). A morphological form resembling C. ossiani was also observed during the early season in Lake Ontario during 1999 (I.A. Grigorovich, personal observation). These observations lead us to believe that C. ossiani may represent a vernal, morphologically distinct generation of C. pengoi. Similar morphological differences have been observed between parthenogenetically and sexually produced individuals of Bythotrephes (Zozulya 1977, Yurista 1992). Rearing of C. pengoi from resting eggs and observation of individuals under controlled conditions should provide evidence as to the taxonomic status of C. ossiani. These and other aspects of C. pengoi biology must be explored in order to comprehend invasion dynamics of the species in new habitats in North America and Europe and quantify the impacts of these invasions on natural food webs.

We gratefully appreciate Dr. Aleksandr Krakhmalnyi (Ukraine) for assistance in electron microscopy. We thank Ronnie Theodory, Mike Weis and Colin van Overdijk for useful suggestions, equipment and technical assistance. This study was supported by NATO and GLIER postdoctoral fellowships to IAG, and by NSERC equipment and research grants to HJM.

References

Aladin, N.V. (1996): Salinity adaptations in Ostracoda and Branchiopoda. – Zoologicheskii institut RAN, St. Petersburg. [Russ.]

Barnhisel, D.R. (1991): The caudal appendage of the cladoceran Bythotrephes cederstroemi as defense against young fish. - J. Plankton Res. 31: 529-537.

Burkhardt, S. (1994): Seasonal size variation in the predatory cladoceran Bythotrephes cederstroemii in Lake Michigan. - Freshwat. Biol. 31: 97-108.

Coulas, R.A., MacIsaac, H.J. & Dunlop, W. (1998): Selective predation on an introduced zooplankter (Bythotrephes cederstroemi) by lake herring (Coregonus artedi) in Harp Lake, Ontario. - Freshwat. Biol. 40: 343-355.

Glamazda, V.V. (1971): On the occurrence of Cercopagis pengoi (Ostr.) in the Tsimlyansk Reservoir. - Gidrobiol. Zhurn. 7(4): 70-71. [Russ.]

Gusynskaya, S.L. (1989): Pelagic zooplankton. – In: Scherbak, G.I. (ed.), Invertebrates and fishes of the River Dnieper and its reservoirs. - Naukova Dumka, Kiev: 21-44. [Russ.]

Gusynskaya, S.L. & Zhdanova, G.A. (1978): Distribution of northern and estuarine-Caspian planktonic crustaceans in the Dnieper reservoirs. - Gidrobiol. Zhurn. 14(6): 25-27. [Russ.]

Grigorovich, I.A., Pashkova, O.V., Gromova, Y.F. & van Overdijk, C.D.A. (1998): Bythotrephes longimanus in the Commonwealth of Independent States: variability, distribution and ecology. - Hydrobiologia 379: 183-198.

Ivanov, N.M. & Makartseva, E.S. (1982): Zooplankton. – In: Smirnova, N.P. & Shnitnikov, A.V. (eds.), Pulsating Lake Chany. - Nauka, Leningrad: 260-272. [Russ.]

Johannsson, O.E., Millard, E.S., Ralph, K.M., Myles, D.D., Graham, D.D. & Taylor, W.D. (1998): The changing pelagia of Lake Ontario (1981-1995): a report of the DFO long-term biomonitoring (Bioindex) program. - Can. Tech. Rep. Fish. Aquat. Sci. 2243: I-IX, 278 pp.

Kaftannikova, O. G. & Basilevich, V.M. (1972): Polyphemidae in the canals of southern Ukraine. – In: The hydrobiology of canals and biological encumbrances in their operation. Inst. of Hydrobiology, Nat. Acad. Sci. Ukraine. - Naukova Dumka, Kiev: 43-44. [Russ.]

Ketelaars, H.A.M., Wagenvoort, A.J., Herbst, R.F., van der Salm, P.A.W. & de Jonge-Pinkster, G.-A.J. (1995): Life history characteristics and distribution of Bythotrephes longimanus Leidig (Crustacea, Onychopoda) in the Biesbosch reservoirs. - Hydrobiologia 307: 239-251.

Krylov, P.I., Bychenkov, D.E., Panov, V.E., Rodionova, N.V. & Telesh, I.V. (1999): Distribution and seasonal dynamics of the Ponto-Caspian invader Cercopagis pengoi (Crustacea, Cladocera) in the Neva Estuary (Gulf of Finland). - Hydrobiologia 393: 227-232.

Krylov, P.I. & Panov, V.E. (1998): Resting eggs in the life cycle of Cercopagis pengoi, a recent invader of the Baltic Sea. - Arch. Hydrobiol./Advanc. Limnol. 52: 382-392.

MacIsaac, H.J. & Grigorovich, I.A. (1999): Ponto-Caspian invaders in the Great Lakes. - J. Great Lakes Res. 25: 1-2.

MacIsaac, H.J., Grigorovich, I.A., Hoyle, J.A., Yan, N.D. & Panov, V.E. (1999): Invasion of Lake Ontario by the Ponto-Caspian predatory cladoceran Cercopagis pengoi. - Can. J. Fish. aquat. Sci. 56: 1-5.

Martin, J. W. & Cash-Clark, C. E. (1995): The external morphology of the onychopod ‘cladoceran’ genus Bythotrephes (Crustacea, Branchiopoda, Onychopoda, Cercopagididae), with notes on the morphology and phylogeny of the order Onychopoda. - Zool. Scr. 24: 61-90.

Meisner, V. (1908): Microscopic representatives of aquatic fauna of the Aral Sea and its affluents in the context of the conditions of their distribution. - Izv. Turkest. Otd. Imp. Russk. Geogr. Obshch. Scientific results of the Aral expedition equipped by the Turkestan Branch of the Imperial Russian Geographical Society 4(8): 1-102, pls. I-IV. [Russ.]

Mills, E.L., Leach, J.H., Carlton, J.T. & Secor, C.L. (1993): Exotic species in the Great Lakes: a history of biotic crises and anthropogenic introductions. - J. Great Lakes Res. 19: 1-54.

Mills, E.L., O’Gorman, R., Roseman, E.F., Adams, C. & Owens R.W. (1995): Planktivory by alewife (Alosa pseudoharengus) and rainbow smelt (Osmerus mordax) on microcrustacean zooplankton and dreissenid (Bivalvia: Dreissenidae) veligers in southern Lake Ontario. - Can. J. Fish. aquat. Sci. 52: 925-935.

Mordukhai-Boltovskoi, F.D. (1960): Caspian fauna in the Azov and Black Sea Basins. - Izdatelstvo Akad. Nauk SSSR, Moscow-Leningrad. [Russ.]

- (1965): Polyphemidae of the Pontocaspian basin. - Hydrobiologia 25: 212-220.

- (1968a): Order Cladocera. – In: Birshtein, Y.A., Vinogradov, L.G., Kondakov, N.N., Astakhova, M.S. & Romanova, N.N. (eds.), Atlas of invertebrates of the Caspian Sea. - Pishchevaya Promyshlennost’, Moscow: 120-160. [Russ.]

- (1968b): On the taxonomy of the Polyphemidae. - Crustaceana 14: 197-209.

- (1969): Order Cladocera. In: Vodianitskii, V.A. (ed.), Guide for identification of the fauna of the Black and Azov Seas. Vol. 2. Freeliving invertebrates. Crustaceans. - Naukova Dumka, Kiev: 12-31. [Russ.]

- (1974): Family Cercopagidae. In: Mordukhai-Boltovskoi, F.D., Kondakov, N.N., Markova, E.L., Romanova, N.N. & Yablonskaya, E.A. (eds.), Atlas of invertebrates of the Aral Sea. - Pishchevaya Promyshlennost’, Moscow: 133-134. [Russ.]

- (1979): Composition and distribution of Caspian fauna in the light of modern data. - Int. Revue ges. Hydrobiol. 64: 1-38.

Mordukhai-Boltovskoi, F. D. & Galinskiy, V. L. (1974): About further spread of caspian polyphemids in the reservoirs of Ponto-Caspian rivers. - Inform. Byul. Biol. Vnutr. Vod. 21: 40-44. [Russ.]

Mordukhai-Boltovskoi, F.D. & Rivier, I.K. (1971): A brief survey of the ecology and biology of the Caspian Polyphemoidea. - Mar. Biol. 8: 160-169.

- - (1987): The predatory cladocerans Podonidae, Polyphemidae, Cercopagidae, and Leptodoridae of the world fauna. - Nauka, Leningrad. [Russ.]

Ojaveer, H. & Lumberg, A. (1995): On the role of Cercopagis (Cercopagis) pengoi (Ostroumov) in Pärnu Bay and the NE part of the Gulf of Riga ecosystem. - Proc. Estonian Acad. Sci. Ecol. 5: 20-25.

Ostroumov, A. (1892): Report on participation in research trip on the Azov Sea on board of transport ship ‘Kazbek’ in summer, 1891. - Mem. Imp. Acad. Sci. St. Petersburg/Suppl. 69(6): 1-19. [Russ.]

Pengo, N. (1879): On Bythotrephes of the Azov Sea. – Tr. Obshch. ispyt. Prirody Imp. Charkov. Univ. 13: 47-67, pl. II. [Russ.]

Rivier, I.K. (1967): Ecology and biology of polyphemids in the Caspian Sea. Candidate Dissertation. - Inst. of Inner Water Biology, Acad. Sci. of the USSR, Borok. [Russ.]

- (1969): Reproduction of cercopagids (Cladocera, Polyphemidae) of the Caspian Sea. - Tr. Inst. Biol. Vnutr. Vod Akad. Nauk SSSR 19(22): 119-127. [Russ.]

- (1974). Brood pouch and egg development in Cladocera. - Tr. Inst. Biol. Vnutr. Vod Akad. Nauk SSSR 25(28): 119-127. [Russ.]

- (1995): The order Polyphemiformes. In: Tsalolikhin, S.Y. (ed.), Key to freshwater invertebrates of Russia and adjacent lands. Vol. 2. Crustaceans. - Zoological Inst., Russian Acad. Sci., St. Petersburg: 69-74. [Russ.]

- (1998): The predatory Cladocera (Onychopoda: Podonidae, Polyphemidae, Cercopagidae) and Leptodorida of the world. - Backhuys, Leiden.

Sars, G.O. (1897): Pelagic Entomostraca of the Caspian Sea. - Ann. Mus. Zool. Imp. Acad. Sci. St. Petersburg 2: 1-73, pls. I-VIII.

- (1902): On the Polyphemidae of the Caspian Sea. - Ann. Mus. Zool. Imp. Acad. Sci. St. Petersburg 7:31-54, pls. I-IV.

Simm, M. & Ojaveer, H. (1999): Occurrence of different morphological forms of Cercopagis in the Baltic Sea. - Proc. Estonian Acad. Sci. Biol. Ecol. 48: 169-172.

Shevtsova, L.V. & Primak, A.B. (1990): Zooplankton. In: Kondratieva, N.V. (ed.), Hydrobiology of the canals in the Ukrainian SSR. - Naukova Dumka, Kiev: 105-118. [Russ.]

Sprules, G.W., Riessen H.P., Jin E.H. (1990): Dynamics of the Bythotrephes invasion of the St. Lawrence Grate Lakes. - J. Great Lakes Res. 16: 346–351.

Tessier, A.J., Young A. & Leibold, M. (1992): Populations dynamic and body-size selection in Daphnia. - Limnol. Oceanogr. 37: 1-13.

Tseeb, Y.Y. (1963): Influence of the flow regulation on the zooplankton in lower reaches of the Dnieper River. - Pr. Inst. Gidrobiol. Akad. Nauk Ukr. RSR 39: 52-65. [Ukr., Russ. summary].

- (1964): Zooplankton. In: Tseeb, Y.Y. (ed.), Kakhovka reservoir. Hydrobiological survey. - Naukova Dumka, Kiev: 60-111. [Ukr., Russ. summary]

Tseeb, Y.Y, Sergeev, A.I. & Grigoriev, B.F. (1971): Zooplankton of the Dnieper-Bug tidal estuary and lower reaches of the Dnieper River under the flow regulation. In: Ivanov, A.I. & Bugai, K.S. (eds.), Dnieper-Bug tidal estuary. - Naukova Dumka, Kiev: 202-228. [Ukr.]

Valkanov, A. (1957): Katalog unserer Schwarzmeerfauna. - Tr. Morsk. biol. St. Varna. Vol. 19. [Bulg., Germ. summary]

Volvich, L.I. (1978): Ponto-Caspian Polyphemoidea in the Veselovsk and Proletarsk Reservoirs. - Gidrobiol. Zhurn. 14(3): 24-25. [Russ.]

Yan, N. D. & Pawson, T. W. (1998): Seasonal variation in the size and abundance of the invading Bythotrephes in Harp Lake, Ontario, Canada. - Hydrobiologia 361: 157-168.

Yankovskii, B.A. (1970). Some features of the structure of digestion organs in Azov-Black Seas gobies in connection with their diet. In: Tseeb, Y.Y. (ed.), Issues for improvement of fish production and sanitary biological conditions of waters in Ukraine. - Naukova Dumka, Kiev: 77-78. [Russ.]

Yurista, P. M. (1992): Embryonic and postembryonic development in Bythotrephes cederstroemii. - Can. J. Fish. aquat. Sci. 49: 1118-1125.

Zernov, S.A. (1903): On the animal plankton of the Aral Sea according to materials collected by L.S. Berg in 1900. - Izv. Turkest. Otd. Imp. Russk. Geogr. Obshch. Scientific results of the Aral expedition equipped by the Turkestan Branch of the Imperial Russian Geographical Society 3: 1-38, pls. I. [Russ.]

Zhuravel, P. A. (1965): On the acclimatization of fauna of estuarine-Caspian type in the reservoirs of Ukraine. - Gidrobiol. Zhurn. 1(3): 59-65. [Russ.]

Zhuravel, P. A. (1967): Formation of the new nidi of the fauna of estuarine-Caspian complex in water bodies of different climatic zones of the USSR. - Zool. Zhurn. 66: 1152-1162. [Russ., Engl. summary]

Zimbalevskaya, L.N. (1989): Littoral zooplankton. In: Scherbak, G.I. (ed.), Invertebrates and fishes of the River Dnieper and its reservoirs. - Naukova Dumka, Kiev, 5-21. [Russ.]

Zozulya, S.S. (1977): Peculiarities of the first generation of Bythotrephes that develops from the latent eggs. - Inform. Byul. Biol. Vnutr. Vod 33: 34-38. [Russ.].

Zozulya, S.S. (1978). On the movement of Bythotrephes longimanus (Leydig). - Gidrobiol. Zhurn. 14(2): 14-17. [Russ.].

 

 

 

 

Figure Legends

Fig. 1. Lateral view of instar III parthenogenetic female of Cercopagis pengoi from Lake Ontario showing body dimensions. Body length is 1.62 mm and caudal process length is 9.59 mm.

Fig. 2. Details of the structure of Cercopagis pengoi parthenogenetic female from Lake Ontario: lateral view of the thoracopods under a dissection microscope (A) and S.E.M. (B) microscopy; cuticle covering of caudal process at high magnification, showing minute spines, S.E.M. (C); loop of caudal process (D); articular spine II, S.E.M. (E); antenna II, showing seven plumose natatory setae on each ramus.

Fig. 3. Relationship between the caudal process length and body length in instar III parthenogenetic females of Cercopagis pengoi from Lake Ontario in August, 1998. Line represents least-squares linear regression (y=2.5x + 5.0; R2=0.19) relating the caudal process length (y in mm) to body length (x in mm).

Fig. 4. Instar III sexual female of Cercopagis pengoi from Lake Ontario. (A) Lateral view of anterior region showing brood pouch containing Nährboden and resting egg. (B) Same animal with complete caudal process. Body length is 1.70 mm and caudal process length is 9.82 mm. (C) Loop of caudal process illustrating extensive curvature.

Fig. 5. Instar III male of Cercopagis pengoi from Lake Ontario. (A) Lateral view of anterior region showing hook, penis and dorsal outgrowth containing heart. (B) Same animal with complete caudal process. Body length is 1.43 mm and caudal process length is 5.49 mm. (C) Loop of caudal process illustrating non-extensive curvature.

Fig. 6. Variation of brood pouch shape in parthenogenetic females of Cercopagis pengoi from Lake Ontario: normal appearance at instars I (A), II (D), III (E, F) and deviating shape of apex at instars I (B) and II (C). Broods are at early (A, B, C, D) and late (E, F) stages of development. Dashed line denotes contour of new brood pouch within currant chamber containing neonates (F). Brood pouch size: A - 0.40 mm; B - 0.28 mm; C - 0.58 mm; and D - 0.77 mm; E - 1.22 mm; F - 1.30 mm.

Fig. 7. Length-frequency histograms of Cercopagis pengoi from the Baltic Sea (n=66), Caspian Sea (n=42) and Lake Ontario (n=765).

Fig. 8. Mean dry body mass and mean body length of sugar-formalin-preserved Cercopagis pengoi from Lake Ontario. Body length-body mass relationships for all animals (A). Instar-specific differences in dry body mass for parthenogenetic females (B).