Nematocysts of the invasive hydroid Cordylophora caspia (Cnidaria: Hydrozoa).
The Ponto-Caspian invasive hydrozoan Cordylophora caspia (Pallas, 1766) is one of only a handful of cnidarians that live in fresh rather than marine waters (Jankowski et al., 2008). It can tolerate salinities ranging from brackish (~15[per thousand]) to freshwater, providing a wide spectrum of suitable habitats (reviewed in Roos, 1979; Folino, 2000). In laboratory settings. C. caspia can survive at salinities as high as 30700 (Kinne, 1958). Colonies are diecious and reproduce sexually, allowing for rapid and broad dispersal within a river, lake, or estuary. Colonies also grow asexually by producing additional hydranths to the upright branches (hydrocauli) or by extending the attached stolons (hydrorhiza) that then produce more hydrocauli and hydranths (Kinne, 1958; Fulton 1962; Jormalainen et al., 1994). C. caspia is often associated with dreissenid mussels (Dreissena polymorpha and D. bugensis), another Ponto-Caspian invasive lineage. Similar to dreissenids. Cordylophora is a biofouler, causing industrial and ecological problems (Folino, 2000; Musko et al., 2008). The spread of C. caspia is difficult to control because colonies are able to regress, forming menonts, which are small pieces of tissue within the perisarc of either the hydrocauli or hydrorhiza. Menonts are ecologically resilient and allow a colony to survive periods of environmental stress because the colony can regenerate from menont tissue when environmental conditions are suitable (Roos, 1979; Folino, 2000). The diversity of growth and reproductive mechanisms observed in Cordylophora in conjunction with a broad salinity tolerance enhance its invasive capacity.
Although its success as an invader comes largely from being clonal and broadly tolerant of salinity and temperature (Roman and Darling, 2007). C. caspia seems to have undergone multiple invasions at some localities (Folino-Rorem et al., 2009). Multiple invasions increase the genetic diversity of the invader species in the new habitat, and thus increase invasion success (Roman and Darling, 2007). C. caspia is inferred to colonize new habitats through ballast water exchange and aquarium release (Olenin, 2006), invasion routes expected to foster multiple introductions (Mac-Isaac et al., 2002) and to promote genetic variation (Stepien et at., 2005). Invasion by C. caspia may have profound effects on an ecosystem: communities may restructure, with decreases in the diversity and abundance of bryozoans and ciliates and increases in the numbers of barnacles, amphi-pods, and polychaetes (Ruiz et al., 1999; Folino, 2000).
Folino-Rorem et at. (2009) identified four genetically distinct lineages of C. caspia within two larger groups through analysis of sequences from nuclear (28S rDNA) and mitochondria! (16S rDNA, cytochrome c oxidase subunit I) markers. Both the larger groups and their constituent clades were highly supported, having posterior probabilities of 100% in Bayesian analyses and bootstrap support of 1.0 in maximum likelihood analyses (Folino-Rorem et at., 2009). The genetic differentiation between the two larger lineages is similar to the interspecific distances between species of other closely related hydrozoan taxa (Folino-Rorem et at., 2009; Ortman et al., 2010; Moura et al., 2011).
Of the two major lineages of Cordylophora identified by Folino-Rorem et al. (2009), one includes only animals from brackish waters; the other includes animals from both freshwater and brackish locales (Fig. 1). Only one subclade is geographically restricted: a brackish lineage represented only by samples from the Pacific coast of the United States. Another subclade is habitat-specific: within the brackish-freshwater group is a small cluster of exclusively freshwater samples, sister to a clade containing both freshwater and brackish samples. Hydroidolina, the subclass to which C. caspia belongs, is composed of mostly marine taxa, and freshwater tolerance is interpreted to be a recent phenomenon in Corclylophora (Jankowski et al., 2008). In several cases, more than one lineage was found at a single location.
Diversity and systematics of Cordylophora caspia
Four valid species are recognized in Cordylophora (see Folino-Rorem, 2009): Cordylophora caspia; Cordylophom japonica (Ito, 1951); Cordylophora mashikoi (Ito, 1952); and Cordylophora solanagiae (Redier, 1967). Only C. caspia is geographically widespread: C. japonica and C. mashikoi are both found in Japan (Ito, 1951, 1952), and C. solangiae is endemic to the Tuamotu atoll (Redier, 1967). C. japonica and C', mashikoi are differentiated by the distinct annulations on the hydrocaulus of the latter (Ito, 1952). C. japonica and C. mashikoi differ from C. caspia by colony form and the structure and number of gonophores (Ito, 1951, 1952).
Populations of C. caspia vary in morphology, but the patterns of variation are difficult to interpret because some attributes can correspond to habitat. Compared to colonies in freshwater, colonies in brackish water generally have taller hydrocauli, more and larger gonozooids, and more and longer tentacles (Kinne, 1958). The branching pattern of the colony, its size, the width of hydranths, the diameter of tentacles, and the shape and number of cells all seem to respond to salinity. Kinne (1958) gradually acclimated clones from a single brackish population of C. caspia to salinities ranging from freshwater to nearly full-strength seawater. Brackish (15[per thousand]) colonies were tallest, had more branching, had the greatest growth rate, and had longer hydranths compared to those in freshwater or saltwater. Freshwater colonies were shortest in height, had the least branching, had slowest growth, and had short, thick hydranths: marine (30[per thousand]) colonies were intermediate in size and had intermediate branching, growth rate, and hydranth size. Cell size and number of cells tracked salinity: cell size was largest and cell number smallest in marine colonies, intermediate in size and number in brackish colonies, and smallest in size and largest in number in freshwater colonies. The variation and plasticity in these features and the lack of overlap with genetic patterns renders previous taxonomic subdivisions based on traditional morphological features unreliable for C. caspia.
Nematocysts: a potential taxonomic indicator for this species complex
Lacking morphological, ecological, distributional, or physiological attributes to differentiate populations, the genetic lineages within C. caspia have not been designated species, despite genetic differentiation characteristic of species or superspecific taxa within Hydrozoa (Folino-Rorem et al., 2009). In addition to the practical problem of defining and recognizing these cryptic units, there is a nomenclatural problem: C. caspia includes several junior synonyms whose application to the genetic lineages cannot be evaluated because those names correspond only to preserved museum material unsuitable for DNA analysis. The problem of correspondence of historical specimens and names to genetically defined lineages is common among invertebrates (e.g., Collin, 2000; Pearse and Francis, 2000: Holland et al., 2004; Barroso et al., 2010; Harmelin et al., 2012) and may be especially problematic for invasive species, for which tracking invasion history through museum material might be especially desirable.
Here we explore the potential of nematocysts to diagnose genetic units within C. caspia. All members of the phylum Cnidaria produce cnidae, complex cellular secretory structures responsible for the stinging capabilities of jellyfish and their relatives. The complement of cnidae distinguishes the anthozoan subclasses Hexacorallia and Octocorallia (reviewed in Daly et al., 2003), and morphology of the nematocyst capsule provides the only morphological evidence for monophyly of the hexacorallian order Actiniaria (Reft and Daly, 2012). Nematocyst size and distribution are integral to species identification in the cnidarian class Anthozoa (reviewed by Fautin, 1988; Ostman, 2000). In Anthozoa, nematocyst size ranges differentiate closely related species (Allcock et al., 1998; Watts et al., 2000) and show congruence with genetic patterns (Hidaka, 1992; Perrin et al., 1999; Abe et al., 2008). Despite their efficacy and widespread use in Anthozoa, these features have been less frequently used in Medusozoa, the clade that includes C. caspia. Nonetheless, their phylogenetic and taxonomic value is clear, at least at higher levels (Papenfuss, 1936; Calder, 1971, 1977). Nematocysts may also be informative taxonomically at lower taxonomic levels: species of Obelia may be differentiated by nematocyst morphology, although ecological and morphological differences also characterize these species (Ostman 1982).
Although nematocysts may have value as specific or higher-level distinguishing features, nematocyst complement (cnidom) may also be shaped by function and ecological circumstance (e.g., Rifkin and Endean, 1983; Purcell, 1984; Purcell and Mills, 1988; Kramer and Francis, 2004; Peach and Pitt, 2005; Colin and Costello, 2007) and so may be subject to ecophenotypic plasticity. Purcell and Mills (1988) found correlations between nematocyst type and prey types, and Purcell (1984) found a positive correlation between the size of copepod prey and the size and quantity of nematocysts. Ecology may affect the size range or the statistical properties of the size range (e.g., mean, variance), condensing the range in species with similar niches so that they are not distinguishable by size even though size range is a characteristic of each species (Fautin, 2009). Geographic and taxonomic scope of sampled individuals may play a role in the accuracy of the account of cnidom and of the size ranges of nematocysts (reviewed in Fautin, 2009).
Here we investigate the sizes of nematocysts from C. caspia from different genetic lineages and from habitats of various salinities. We confirm that C. caspia has two types of nematocysts--euryteles and desmonemes (Fig. 2; reviewed in Mariscal, 1974). In C. caspia, euryteles are larger but desmonemes are more common (Fig. 2B). In documenting the size ranges of the cnidae from individuals of various populations of C. caspia, we discovered that the cnidom does not vary significantly in composition or size with respect to clade or salinity and is thus incapable of differentiating genetic clades: we found the same types in all populations and found no correlation between size range of capsules and clade or salinity. However, we did find significant differences between samples of C. caspia and C. japonica, showing that there is informative interspecific variation in enidom (albeit at a more general level than genetic variation). Our investigation is the first statistical account of nematocyst sizes in hydrozoans, and the first full account of enidom in C. caspia.
Materials and Methods
Cordylophora caspia samples growing on small stones were collected from Lake Erie off the north side of Gibraltar Island (41.658382, -82.821021) at a depth of 0.5-1.0 m. Thirteen additional populations of C, caspia were obtained from cultures maintained by Nadine Folino-Rorem (Table 1); these had been used in the genetic investigations of Folino-Rorem et al. (2009). Colonies were maintained at room temperature (22 [degrees]C) at ambient salinity and fed brine shrimp two to three times a week. Each population was maintained in a 5-gallon (18.9 1) tank with a recirculating filter. Fresh tissue was used for DNA extractions and nematocyst studies; vouchers are deposited at the American Museum of Natural History, New York, NY.
Table 1 Localities, clade identity, and salinity for all populations studied Population locality Population code Clade Salinity Cayuga Lake, NY CL 1A 0 Des Plaines River, Joliet, IL DP 1A 0 Squamscott River, Exeter, NH E 1B 8 Lake Michigan, Chicago. IL FN 1A 0 Illinois River, Henry. IL H 1A 0 Jackson Landing, Durham, NH J 1A 25 Lake Erie, Put-in Bay, OH LE 1A 0 Lake Ontario, Rochester. NY LO 1A 0 La Salle Luke, Marseilles. IL LS 1A 0 Napa River, CA NR 2B 16 Seneca Lake, NY SL 1B 0 Sonoma, CA SO 2B 16 James River. Jamestown, VA V 1B 0.5 Woods Hole, MA WH2 1B 15 See Folino-Rorem et al. (2009) for clade designations. Samples represent several localities from North America, both genetic lineages, and populations found in fresh and brackish water.
For all samples, we removed one haphazardly chosen hydranth from a colony and placed it on a glass slide. For freshwater populations, excess water was removed and Giemsa stain was added to the slide. Excess stain was removed after a minimum of 30 min. For brackish populations, the hydranth was rinsed with deionized water to remove excess salts, excess water was removed, Giemsa stain was added, and the hydranth was stained for a minimum of 4 h (often overnight); even when rinsed well, brackish populations were not as easily stained as the freshwater populations. For both fresh and brackish populations, the stained hydranth was rinsed with acid alcohol to help dissociate tissue and remove excess stain. After a small drop of Permount was added to the stained hydranth, a cover slip was placed atop the tissue; the tissue was dissociated and spread by applying gentle pressure to the cover slip with a fingertip. All measurements were made with a Boeckeler VIA-100 video measurement system (Y/C interface) at 1000X magnification using differential interference contrast microscopy. Measurements were taken by starting at a random location on the slide and moving across the slide in a single direction to avoid measuring any nematocyst twice.
To accurately determine the number and size range of nematocysts per hydranth, 50-75 nematocysts of each type were measured from three hydranths from the Woods Hole population; this population was chosen because it was the largest colony of our study populations. Measurements for each nematocyst type were grouped into blocks of 10 measurements, starting with the first measurement made for the hydranth. The blocks were added to one another in sequence (i.e., first block to second; third to first + second; fourth to first + second + third) and the mean was calculated. The averages calculated for each set (i.e., one block, block 1+2, block 1+2+3) were compared to determine the minimum number of nematocysts needed per hydranth to find a stable and accurate average for that type of nematocyst. For euryteles, mean nematocyst length varied by 0.14 [micro]m or less after 20-30 nematocysts measurements. For desmonemes, mean nematocyst length varied by 0.08 [micro]tm or less after 20 nematocysts. The number of hydranths needed to accurately estimate the population mean was determined in a similar manner, using blocks of 25 nematocysts because the previous protocol indicated that 25 capsules were the minimum required to accurately quantify within-hydranth variation in nematocyst length. For euryteles, mean nematocyst length varied by 0.02 ptin after eight hydranths were measured. More variation in eurytele length was seen when fewer hydranths were measured: variation between the first, second, and third hydranth was 0.24 [micro]m. For desmonemes, mean nematocyst length varied by 0.08 [micro]m or less after only three hydranths were measured. To ensure an accurate quantification of length for both euryteles and desmonemes, we measured both types of nematocysts from eight hydranths per population for each of our 14 study populations.
DNA extraction, amplification, and sequencing
Because at least some localities harbor multiple genetic lineages of C. caspia, we sequenced DNA from all of our study populations to confirm their genetic identity. Two to four hydranths were picked from a single colony and DNA was extracted using the Qiagen DNeasy kit, following the manufacturer's protocol for purification of total DNA. DNA was amplified from the mitochondria] genes COI and 16S and the nuclear gene 28S, using the primers and protocol of Folino-Rorem et at. (2009). Samples that did not amplify were rerun, using lower or higher annealing temperatures; the final annealing temperatures for both PCR cycles varied between 48 [degrees]C and 52 [degrees]C. PCR products were sequenced with the amplification primers in the forward and reverse directions, compiled in Sequencher ver. 5.0 sequence analysis software, and screened against GenBank using BLAST (Johnson et al., 2008). Sequences were compared against those from Folino-Rorem et at. (2009) by creating a data matrix that included our samples and those of Folino-Rorem et at. (2009). This matrix was analyzed following Folino-Rorem et al. (2009). Sequences from the Lake Erie samples are deposited in GenBank as KC489507, KC489508, and KC489509.
Statistical analysis of nematocyst sizes
Hydranths within a population are likely to be related (either as products of asexual or sexual reproduction). Because of this, measurements from hydranths in a population are likely to correlate to one another and are not independent. Within a population, if the means for each type of nematocyst are significantly different between the hydranths, the hydranths of a population constitute pseudoreplicates, and the 25 measurements per nematocyst type per hydranth can be averaged to provide a single mean.
Using Minitab 15, we conducted a one-way ANOVA on the measurements from each population to determine if means differed between populations. We evaluated normality for both desmonemes and euryteles by using a normal quantile plot of residuals because a one-way ANOVA test is not robust to non-normality. We evaluated whether the samples showed constant variance by comparing residuals versus fits: for constant variance, errors must have a mean of zero with constant spread and no patterns. Boxplots were used to determine outliers and demonstrate variance because a one-way ANOVA test is not robust to numerous outliers. Having ascertained that the samples were suited to a one-way ANOVA, we examined full (all populations separate) and reduced (populations with similar means grouped together) models to determine the best model for each nematocyst type. Determining the most reduced model is important for identifying which means differ between populations; the one-way ANOVA only determines that at least one of the means differ. The full model, in which all populations are analyzed separately, is not informative for identifying which of the means differ between populations and which are statistically the same.
Although the overall topology (Fig. 1) matched that of Folino-Rorem et al. (2009), our tree has less resolution, most probably because we consider many fewer samples. All of our samples from cultures used in Folino-Rorem et al. (2009) matched the appropriate published sequences. The Lake Erie samples nested within Clade 1A, as the sister to the Des Plaines River (DP) sample (Fig. 1).
In total, 200 nematocysts of each type were measured from each population, totaling 2800 measurements of each nematocyst and 5600 measurements overall. In all populations there was significant variation in the sizes of euryteles between hydranths (Table 2), but in 3 of 14 populations, the desmonemes did not differ significantly in length between hydranths (Table 3). The small size and smaller differences in size between desmonemes approach the limits of the measurement equipment, and thus the lack of significance in size of desmonemes between populations may reflect these limits rather than similarity among populations. Therefore, means were calculated for each nematocyst type from each hydranth so that each population had 8 means per nematocyst type.
Table 2 Summary of lengths of euryteles from populations studied. Population P Mean Standard Mean Difference value ([mu]m) deviation hydranth SD of hydranth ([mu]m) ([mu]m) SD ([mu]m) CL 0.007 8.44 0.35 0.55 0.17 DP <.001 7.88 0.19 0.40 0.12 E 0.003 8.07 0.22 0.60 0.36 FN <.001 7.94 0.26 0.43 0.32 H <.001 7,96 0.19 0.41 0.14 J < .001 8.15 0.21 0.46 0.26 LE < .001 7.62 0.35 0.74 0.44 LO <.001 7.56 0.27 0.44 0.27 LS <.001 7.69 0.17 0.43 0.18 NR <.001 8.35 0.41 0.74 0.43 SL <.001 7.83 0.18 0.44 0.25 SO <.001 8.04 0.25 0.54 0.25 V <.001 8.07 0.43 0.52 0.24 WH2 0.001 7.73 0.29 0.62 0.16 P values are from a one-way ANOVA between averages of individual hydranth measurements for each population. Mean length (n = 8) for euryteles from each population, standard deviation (n = 8), mean (n = 8) of hydranth standard deviations (SD), and the range difference of hydranth standard deviations. Population codes are presented in Table 1. See Folino-Rorem et at. (2009) for clade designations. Table 3 Summary of lengths of desmonemes from populations studied Population P Mean Standard Mean Difference value ([mu]m) deviation hydranth SD of hydranth ([mu]m) ([mu]m) SD ([mu]m) CL <.001 4.62 0.14 0.40 0.20 DP 0.652 4.36 0.05 0.29 0.09 E 0.048 4.51 0.11 0.38 0.28 FN <.001 4.31 0.21 0.32 0.17 H 0.378 4.35 0.06 0.28 0.18 J <.001 4.58 0.23 0.33 0.22 LE 0.002 4.74 0.18 0.49 0.19 LS 0.001 4.35 0.08 0.26 0.09 LO 0.016 4.14 0.12 0.30 0.16 NR 0.001 4.61 0.18 0.46 0.18 SL 0.020 4.25 0.11 0.31 0.09 SO 0.635 4.42 0.06 0.36 0.23 V 0.001 4,31 0.13 0.43 0.33 WH2 <0.001 4.42 0.34 0.45 0.31 P values are from a one-way ANOVA between averages of individual hydranth measurements for each population. Mean length (n = 8) for desmonemes from each population, standard deviation (n = 8), mean (n = 8) of hydranth standard deviations (SD), and the range difference of hydranth standard deviations. Population codes are presented in Table 1. See Folino-Rorem et al. (2009) for clade designations.
The mean size (Table 2) of euryteles differed between populations (P < 0.001). Length of euryteles was mostly normally distributed (Fig. 3), with some short tails or small standard deviation. Errors had a mean of zero, with constant spread and no pattern. Although treating the samples from a single population as pseudoreplicates reduced the number of outliers, three populations (DP, E, and SO; see Table 1 for full locality information) still contained outlier measurements (Fig. 4). The mean size (Table 3) of desmonemes also differed between populations (P < 0.001). Length of desmonemes was normally distributed (Fig. 5). Errors had a mean of zero, with constant spread and no patterns. Treating the samples of desmonemes from a single population as pseudoreplicates eliminated outliers (Fig 6).
A one-way ANOVA was used to analyze the efficacy of grouping desmoneme or eurytele measurements by clade or by salinity. Although means of clades were statistically different for euryteles (P = 0.014), means were not statistically significant between clades for desmonemes (P = 0.122). Means between groups based on salinity were not statistically different for euryteles (P = 0.122) or des-monemes (P = 0.861). Thus, for euryteles and des-monemes, grouping neither by clade nor by salinity was preferred over considering each population separately (P < 0.001 in all cases).
For both euryteles and desmonemes, a one-way ANOVA was used to group populations with statistically similar means to find a more reduced model. Populations with statistically similar means may have similarities other than salinity or clade to account for similar nematocyst lengths. The best reduced model included four groups, and in each case this was preferred over a more reduced (having fewer groups) model (P = 0.0098 euryteles, P = 0.0105 desmonemes) (Table 4). Both reduced models were preferred over the full models in which all populations were separate (P = 0.9550 euryteles, P = 0.5775 desmonemes). The constituency of the groups differ for euryteles and desmonemes, and in neither case does group membership coincide with clade, geography, or salinity (Table 4): for example, populations SO and V group with E and J for euryteles but with populations DP, FN, N. LS, Si, and WH2 for desmonemes. In none of the best-fit reduced models for either euryteles or desmonemes did we find a correlation between the mean capsule length for a population and clade, salinity, or variation.
Table 4 Populations of Cordylophora caspia grouped by similar mean nematocyst length using a one-way ANOVA Euryteles Group 1 2 3 4 1 Mean 8.40 8.08 7.90 7.65 4.74 ([mu]m) Population Clade Salinity ([per thousand]) CL 1A 0 x NR 2B 16 x E 1B 8 x J 1A 25 x V 1B 0.5 x SO 2B 16 x FN 1A 0 x DP 1A 0 x SL 1B 0 x H 1A 0 x LS 1A 0 x WH2 1B 15 x LO 1A15 x LE 1A 0 x x Desmonemes 2 3 4 4.58 4.35 4.14 Population Clade Salinity ([per thousand]) CL 1A 0 x NR 2B 16 x E 1B 8 x J 1A 25 x V 1B 0.5 x SO 2B 16 x FN 1A 0 x DP 1A 0 x SL 1B 0 x H 1A 0 x LS 1A 0 x WH2 1B 15 x LO 1A15 x LE 1A 0 Mean nematocyst length for each group indicated. The groups indicated were preferred over model with fewer groups for both ewyteles (P = 0.0098) or desmonemes (P < 0.001). Population codes are presented in Table 1. See Folino-Rorem et al. (2009) for clade designations.
Despite their promise in other cnidarian lineages, nematocysts fail to distinguish genetic lineages within Cordylo-phora caspia. Although the nematocysts of C. caspia are small and thus may be difficult to accurately or consistently measure (Gravier-Bonnet, 1987), we found very little variance (Tables 2, 3), suggesting that this factor was minimal in our study of C. caspia.
The examined populations of C. caspia sorted into four groups based on the length of the auryteles and four groups based on the length of the desmonemes. The composition of groups differed, however, depending on the type of nematocyst (Table 4), and none of the groups mirror genetic lineage or salinity (Fig. 1). Although some populations (CL, E. J, NR) had relatively longer euryteles and desmonemes and some (LO, LS, WH2) had relatively smaller euryteles and desmonemes, in LE the mean lengths of euryteles and desmonemes did not correlate (Table 4): LE has the largest mean length for desmonemes and one of the smallest mean lengths for euryteles. LE does not have the greatest (or smallest) variation in capsule lengths for the populations we examined, so this departure from the general pattern is unlikely to be a sampling artifact. This independent variation in size across types suggests that size of each type of nematocyst is controlled independently, rather than both being controlled by a single gene or genetic system.
Although nematocyst length varied between populations, we found no correlation between mean length and genetic clade for either euryteles or desmonemes (Table 4), indicating that for C. caspia, nematocysts are not informative for species or subspecies differentiation. The genetic lineages within C. caspia may be too recently diverged for morphological differences to have been fixed, or niche may be so similar for the lineages within C. caspia that this feature is either not under adaptive selection or is under stabilizing selection to maintain nematocyst sizes within the observed range.
Unlike colony form and polyp morphology, variation in capsule length in C. caspia does not show phenotypic plasticity in response to salinity. We found no correlation between mean length and salinity for either euryteles or desmonemes (Table 4). It is possible that the observed variation may track some other (yet unmeasured) environmental factor, such as prey size. When copepod prey size increases, the size and quantity of nematocysts also may increase (Purcell, 1984). This explanation would account for different trends between nematocyst types, since each type is better adapted for capturing a different prey type: desmonemes adhere to surfaces and so are used to capture hard-bodied prey such as crustaceans (Purcell and Mills, 1988); euryteles are able to penetrate some prey epithelium and are associated with hard- and soft-bodied prey species (Purcell and Mills, 1988). Depending on the prey type, size, and abundance in each location, different environmental pressures could act on the nematocyst types separately.
Although the lengths of desmonemes and euryteles do not correspond to genetic distinctions among populations of C. caspia, nematocyst data may distinguish more distantly related species. For example, comparison of nematocyst lengths supports the distinction between C. caspia and C. japonica. The mean length of euryteles reported from C. japonica is 10-10.2 [micro]m; desmonemes have a mean length of 4.9-5.0 pm (Ito, 1951). The greatest mean length we found for euryteles in C. caspia is 8.44 [micro]m, and the greatest mean length we found for desmonemes is 4.74 [micro]m. Thus, the populations with the largest mean capsule size in C. caspia had smaller nematocysts than are reported for C. japonica. Furthermore, for euryteles, the difference in length between the largest mean length in C. caspia and the smallest mean length in C. japonica is 1.56 [micro]m, nearly 5 times the difference observed between the populations of C. caspia studied here. Another species from Japan, C. mashikoi, has euryteles of mean length 9.1-9.3 [micro]m and desmonemes of mean length 4.3-4.4 [micro]m (Ito 1952). The mean lengths for desmonemes are similar between C. caspia and C. mashikoi, but the mean lengths of euryteles are quite different, still exceeding the maximum distance of 0.32 p.m between groups of C. caspia by about twofold. Note that in these interspecific comparisons, the lengths of euryteles and of desmonemes do not track one another, varying independently as they do within populations of C. caspia.
In light of the lack of correspondence between nematocyst size and clade identity, nematocysts cannot be used as a morphological indicator of genetic lineage for C. caspia. Although populations vary in the mean lengths of both desmonemes and euryteles, the differences do not correlate with lineage. However, differences fail to correlate with ecotype. suggesting that ecophenotypic plasticity has only a minor impact on nematocyst length in this species. Further study of reproductive compatibility and morphometric features (e.g., branching pattern and hydrocaulus length) may provide some means of separating lineages in this cosmopolitan invasive species.
We thank Anthony D'Orazio, Paul Larson, Abby Reft, and Jennifer Yi for assistance with collecting in Lake Erie. Abby Reft provided scanning electron micrographs. This manuscript benefited from comments made by Norman Johnson and John Freudenstein on JW's MSc thesis. Support for this project came from NSF EF-0531763 to MD.
* To whom correspondence should be addressed. E-mail: email@example.com
[dagger] Current address: Center for Learning Innovation, University of Minnesota Rochester, Rochester, Minnesota 55904.
Abe, M., Y. Suzuki, H. Hayakawa, T. Watanabe, and M. Hidaka. 2008. Breeding experiments of the hermatypic coral Galaxea fascicularis: partial reproductive isolation between colonies of different nematocyst types, and enhancement of fertilization success by the presence of parental colonies. Fish Sci. 74: 1342-1344.
Allcock, A. L., P. C. Watts, and J. P. Thorpe. 1998. Divergence of nematocysts in two colour morphs of the intertidal beadlet anemone Actinia equina. J. Mar. Biol. Assoc. UK 78: 821-828.
Barroso, R., M. Klautau, A. M. Soli-Cava, and P. C. Paiva. 2010. Eurythoe complanata (Polychaeta: Amphinomidae), the 'cosmopolitan' fireworm. consists of at least three cryptic species. Mar. Biol. 157: 69-80.
Calder, D. R. 1971. Nematocysts of polyps of Aurelia, Chrysaora, and Cyanea. and their utility in identification. Trans. Am. Microsc. Soc. 90: 269-274.
Calder, D. R. 1977. Nematocysts of the ephyra stages of Aurelia. Chrysaora, Cyanea. and Rhopilema (Cnidaria, Scyphozoa). Trans. Am. Micron., Soc. 96: 13-19.
Colin, S. P., and J. H. Costello. 2007. Functional characteristics of nemalocysts found on the scyphomedusa Cyanea capillata. J. Exp. Mar. Biol. Ecol. 351: 114-120.
Cain, R. 2000. Phylogeny of the Crepidula plana (Gastropoda: Ca-lypnaeidae) cryptic species complex in North America. Can. J. Zool. 78: 1500-1514.
Daly, M., V. A. Cappola, and D. G. Fautin. 2003. Systematics of the Hexacorallia (Cnidaria:Anthozoa). Zool. J. Linn. Soc. 139: 419-437.
Fautin, D. G. 1988. Importance of nematocysts to actinian taxonomy. Pp. 487-500 in The Biology of Nematocysts, D. A. Hessinger and H. M. Lenhoff. eds. Academic Press. San Diego.
Fautin, D. G. 2009. Structural diversity, systematics. and evolution of cnidae. Toxicon 54: 1054-1064.
Folino, N. C. 2000. The freshwater expansion and classification of the colonial hydroid Cordylophora (Phylum Cnidaria, Class Hyclrozoa). Pp. 139-144 in Marine Bioinvasions: Proceedings of the First National Conference, J. Pederson. ed. Massachusetts Institute of Technology Sea Grant College Program, Cambridge, MA.
Folino-Rorem, N. C. 2009. Cordylophora caspia. Invasive Species Compendium. [Online] Available: http://www.cabi.org/isc [2012 April].
Folino-Rorem, N. C., J. A. Darling, and C. A. D'Ausilio. 2009. Genetic analysis reveals multiple cryptic invasive species of the hydrozoan genus Corclylophora. Biol. Invasions 11: 1869-1882.
Fulton, C. 1962. Environmental factors influencing the growth of Cordylophora. Exp. Zool. 151: 61-78.
Gravier-Bonnet, N. 1987. Nematocysts as taxonomic discriminators in thecate hydroids. Pp. 43-55 in Modern Trends in the Systematics, Ecology, and Evolution of Hydroids and Hydrornedusae, J. Bouillion, F. Boero. F. Cicogna, and P. F. S. Cornelius, eds. Clarendon Press, Oxford.
Harman, J. G., L. M. Vieira, A. N. Ostrovsky, J. P. Caceres-Chamizo, and J. Sanner. 2012. Scoipiodinipora costulata (Canu & Bassler, 1929) (Bryozoa. Cheilostomata). a taxonomic and biogeographic dilemma: complex of cryptic species or human-mediated cosmopolitan colonizer? Zoosystema 34: 123-138.
Hidaka, M. 1992. Use of nematocyst morphology for taxonomy of some related species of scleractinian corals. Galaxea 11: 21-28.
Holland, B. S., M. N. Dawson, G. L. Crow, and D. K. Hofmann. 2004. Global phylogeography of Cassiopea (Scyphozoa: Rhizostomcae): molecular evidence for cryptic species and multiple invasions of the Hawaiian Islands. Mar. Biol. 145: 1119-1128.
Ito, T. 1951. A new athecate hydroid, Cordylophora japonica n. sp., from Japan. Mem. Ehime Univ. II 1: 81-86. Available: http://biology.duke.edu/hydrodb/biblionto-%20A%2Onew%20athecate%20hydroid.pdf [2012 April].
Ito, T. 1952. A new species of athecate hydroid Cordylophora from Japan. Pp. 55-57 in A special publication of the Japan Sea Regional Fisheries Research Laboratory on the 31u1 anniversary of its founding, N. K. S. Kenkyrkio, ed. Japan Sea Regional Fisheries Laboratory. Nakao, Japan.
Jankowski, T., A. G. Collins, and R. Campbell. 2008. Global diversity of inland water cnidarians. Hydrobiologia 595: 35-40.
Johnson, M., I. Zaretskaya, Y. Raytselis, Y. Merezhuk, S. McGinnis, and T. L. Madden. 2008. NCBI BLAST: a better web interface. Nucleic Acids Res. 36: W5-W9.
Jormalainen, V.. T. Honkanen, T. Vuorisalo, and P. Laihonen. 1994. Growth and reproduction of an estuarine population of colonial hydroid Cordylophora caspia (Pallas) in the northern Baltic Sea. Helgol. Wiss. Meeresunters. 48: 407-418.
Kinne, O. 1958. Adaptation to salinity variations--some facts and problems. Pp. 92-106 in Physiological Adaptation, C. L. Prosser, ed. American Physiological Society, Washington, DC.
Kramer, A., and L. Francis. 2004. Predation resistance and nematocyst scaling for Metridium senile and M. farcimen. Biol. Bull. 207: 130140.
MacIsaac, H. J., T. C. Robbins, and M. A. Lewis. 2002. Modeling ships ballast water as invasion threats to the Great Lakes. Can. J. Fish. Aquat. Sci. 59: 1245-1256.
Mariscal, R. N. 1974. Nematocysts. Pp. 129-178 in Coelenterate Biology, L. Muscatine and H. M. Lenhoff, eds. Academic Press. New York.
Moura, C. J., M. R. Cunha, F. M. Porteiro, and A. D. Rogers. 2011. The use of the DNA barcode gene 16S mRNA for the clarification of taxonomic problems within the family Sertulariidae (Cnidaria, Hydrozoa). Zool. Scripta 40: 520-537.
Musko, I. B., M. Bence, and C. S. Balogh. 2008. Occurrence of a new Ponto-Caspian invasive species, Cordylophora caspia (Pallas 1771) (Hydrozoa: Clavidae) in Lake Balaton (Hungary). Acta Zool. Acad. Sci. Hung. 54: 169-179.
Olenin, S. 2006. Cordylophora caspia. Delivering alien invasive species inventories for Europe. [Online] Available: http://www.europe-aliens.org/speciesFactsheet.do?speciesId=50150 [2012 April].
Ortman, B. D., A. Buckling F. Pages, and M. Youngbluth. 2010. DNA Barcoding the Medusozoa using mtCOI. Deep Sea Res. II 57: 2148-2156.
Ostman, C. 1982. Nematocysts and taxonomy in Laomedea. Gonothyraea and Melia (Hydrozoa, Campanulariidae). ZooL. Scr. 11: 227-241.
Ostman, C. 2000. A guideline to nematocyst nomenclature and classification. and some notes on the systematic value of nematocysts. Sci. Mar. 64: 31-46.
Papenfuss, E. J. 1936. The Utility of the Nematocysts in the Classification of Certain Scyphomedusae. C. W. K. Gleerup, Lund. Norway.
Peach, M. B., and K. A. Pitt. 2005. Morphology of the nematocysts of the medusae of two scyphozoans, Cato.stylus mosaicus and Phyllorhiza punctata (Rhizostomeae): implications for capture of prey. Invertebr. Biol. 124: 98-108.
Pearse, V. B., and L. Francis. 2000. Anthopleura sola, a new species, solitary sibling species to the aggregating sea anemone, A. elegantis-sima (Cnidaria: Anthozoa: Actiniaria: Actiniidae). Proc. Biol. Soc. Wash. 113: 596-608.
Perrin, M. C., J. P. Thorpe, and A. M. Sole-Cava. 1999. Population structuring, gene dispersal, and reproduction in the Actinia equina species group. Oceanogr. Mar. Biol. Annu. Rev, 37: 129-152.
Purcell, J. E. 1984. The functions of nematocysts in prey capture by epipelagic siphonophores (Coelenterata, Hydro/0a). Biol. Bull. 166: 310-327.
Purcell, J. E., and C. E. Mills. 1988. The correlation or nematocyst types to diets in pelagic Hydrozoa. Pp. 463-485 in The Biology of Nematocysts, D. A. Hessinger and H. M. Lenhoff, eds. Academic Press, San Diego.
Redier, L. 1967. Un nouvel hydraire Cordylophora solangiae n.s. (atoll de Fangatau fa-Tuamotu). Cah. Pac. 11: 117-128.
Reft, A. J., and M. Daly. 2012. Morphology, distribution, and evolution of apical structure of nematocysts in Hexacorallia. J. Morphol. 273: 121-136.
Rifkin. J.. and R. Endean. 1983. The structure and function of the nematocysts of Chironex flecked Southcott, 1956. Cell Tissue Res. 233: 563-577.
Roman, J., and J. A. Darling. 2007. Paradox lost: genetic diversity and the success of aquatic invasions. Trends Ecol. Eval. 22: 454-464.
Roos, P. J. 1979. Two-stage life cycle of a Cordylophora population in the Netherlands. Hydrobiologia 62: 231-239.
Ruiz, G. M., P. Fofonoff, A. H. Hines, and E. D. Grosholz. 1999. Non-indigenous species as stressors in estuarine and marine communities: assessing invasion impacts and interactions. Limnol. Oceanogr. 44: 950-972.
Stepien, C. A., J. E. Brown, M. E. Neilson, and M. A. Tumeo. 2005. Genetic diversity of invasive species in the Great Lakes versus their Eurasian source populations: insights for risk analysis. Risk Anal. 25: 1043-1060.
Watts, P. C., A. L. Allcock, S. M. Lynch, and J. P. Thorpe. 2000. An analysis of the nematocysts of the beadlet anemone Actinia equina and the green sea anemone Actinia prctsina. J. Mar Biol. Assoc. UK 80: 719--724.
JENNIFER WOLLSCHLAGER (1) * [dagger], NADINE FOLINO-ROREM (2), AND MARYMEGAN DALY (1)
(1.) Department of Evolution. Ecology, and Organismcd Biology, the Ohio State University, Columbus, Ohio 43210; and (2.) Department of Biology, Wheaton College, Wheaton. Illinois 60187
|Printer friendly Cite/link Email Feedback|
|Author:||Wollschlager, Jennifer; Folino-Rorem, Nadine; Daly, Marymegan|
|Publication:||The Biological Bulletin|
|Date:||Apr 1, 2013|
|Previous Article:||Latitudinal diversity of sea anemones (Cnidaria: Actiniaria).|
|Next Article:||Vertical visual features have a strong influence on cuttlefish camouflage.|