- Research article
- Open Access
The evolution of euhermaphroditism in caridean shrimps: a molecular perspective of sexual systems and systematics
BMC Evolutionary Biologyvolume 10, Article number: 297 (2010)
The hippolytid genus Lysmata is characterized by simultaneous hermaphroditism, a very rare sexual system among Decapoda. Specialized cleaning behavior is reported in a few pair-living species; these life history traits vary within the genus. Unfortunately, the systematics of Lysmata and the Hippolytidae itself are in contention, making it difficult to examine these taxa for trends in life history traits. A phylogeny of Lysmata and related taxa is needed, to clarify their evolutionary relationships and the origin of their unique sexual pattern. In this study, we present a molecular phylogenetic analysis among species of Lysmata, related genera, and several putative hippolytids. The analysis is based upon DNA sequences of two genes, 16S mtDNA and nuclear 28S rRNA. Phylogenetic trees were estimated using Bayesian Inference, Maximum Likelihood, and Maximum Parsimony.
Phylogenetic analysis of 29 species of Lysmata, eight genera of Hippolytidae and two genera of Barbouriidae based on a single (16S, 28S) and combined gene approach (16S+28S) indicates that three groups of Lysmata differentiate according to antennular morphology: (1) Lysmata, having a multi-segmented accessory branch, (2) Hippolysmata (prior to Chace 1972), with a one-segmented accessory branch, and (3) a third group of Lysmata outliers, with one-segmented unguiform accessory branch, and close affinity to the genera Exhippolysmata and Lysmatella. The monophyly of the clade bearing a multi-segmented accessory branch is robust. Within the short accessory branch clade, species with specialized cleaning behaviors form a monophyletic clade, however, the integrity of the clade was sensitive to alignment criteria. Other hippolytid and barbouriid genera used in the analysis are basal to these three groups, including one displaying simultaneous hermaphroditism (Parhippolyte). The two barbouriid species occur in a separate clade, but among hippolytid taxa.
The data support the historical morphological division of Lysmata into clades based on accessory branch morphology. The position of the "cleaner" shrimps, indicates that specialized cleaning behavior is a derived trait. The topologies of the cladograms support the monophyly of the barbouriids, but do not support their elevation to familial status. Taxa ancestral to the genus Lysmata display simultaneous hermaphroditism, suggesting that this life history trait evolved outside the genus Lysmata.
The hippolytid shrimp genus, Lysmata (Risso, 1816), has attracted the attention of biologists for several decades. Members of this genus are small caridean shrimp and occur in tropical to warm temperate marine coastal waters worldwide. They are popular marine aquarium pets, with some species engaging in cleaning behavior of fishes. One species (L. seticaudata) was used as a model organism for ground breaking studies on sexual differentiation in Crustacea [1–3], with the mistaken impression that this species underwent male-to-female sex change, or protandric hermaphroditism (PH). For many years, PH was thought to be the only form of hermaphroditism in the decapod crustacea, albeit uncommon. This perception changed in the last decade with the discovery that the reproductive system of two Lysmata species was a form of simultaneous hermaphroditism, or euhermaphroditism [4, 5]. This system has been confirmed in every Lysmata species examined (e.g. L. amboinensis , L. wurdemanni , L. nilita , L. seticaudata , L. californica , L. bahia , L. intermedia , L. rafa  and L. holthuisi ). Among confirmed euhermaphroditic Lysmata species, all individuals pass through a functional male phase early in life [4–7]. This is the impetus of the term "protandric simultaneous hermaphroditism" or "PSH" (e.g. ) to describe the system. The early male phase also contributed to the mistaken impression that Lysmata species were protandric hermaphrodites [1–3].
Bauer [12, 13] postulated that the evolution of PSH in Lysmata was related to social systems and/or behavioral characteristics among members of the genus. He divided Lysmata into two informal, non-taxonomic ecological groupings: 1) low density, pair living, specialized "cleaner shrimps", with bright and contrasting coloration, including yellow and red colors and long white antenna, and famous for their ability to actively "clean" fish (e.g. L. amboinensis, L. grabhami, L. debelius, and L. splendida); 2) high density, group living, "peppermint shrimps", with color patterns consisting of semi-translucent bodies with longitudinal and lateral red bands (e.g. L. wurdemanni, L. californica, and L. seticaudata). Bauer [12, 13] hypothesized that PSH must have evolved from a paired, cleaning ancestor species living at low densities with few opportunities to find mates and further suggested that group-living species diverged once or multiple times from these paired species. However, this explanation was made without phylogenetic inference for the genus Lysmata and its socioethological patterns. Furthermore, there are indications that PSH may have evolved outside the genus. Recent studies have shown that PSH occurs in the hippolytid genera Exhippolysmata Stebbing, 1916 [14, 15], Lysmatella Borradaile, 1915 (Rhyne unpub.), and one barbouriid genus (previously a hippolytid) Parhippolyte Borradaile, 1899 (Onaga & Fiedler, unpub.). The presence of PSH within these few taxa suggests an opportunity to examine the evolution of this unique system via a molecular phylogenetic approach.
Unfortunately, the systematics of both genus Lysmata and the family Hippolytidae are still unsettled. Recent revisions in the caridean genus Lysmata have increased the number of species to nearly 40, an expansion of 33% over the last 10 years, and this has not abated [9, 10, 16, 17]. Members of the genus Lysmata were originally split into two genera: Hippolysmata Stimpson, 1860 and Lysmata. These two genera were previously differentiated by the presence of a multi-segmented accessory antennal branch in Lysmata species, and the lack thereof in Hippolysmata species (see  for an example). Chace  placed Hippolysmata in synonymy of Lysmata, based upon a perceived wide intraspecific variation in the accessory branch morphology. However, Chace may have failed to properly delineate several species based on this character, which directly led to his misinterpretation (c.f. ). Furthermore, both generic names were in use two decades later by Holthuis .
The family Hippolytidae has also been under recent scrutiny. Christoffersen  concluded that the Hippolytidae are a polyphyletic group, based upon a detailed manual cladistic analysis of morphological characters. He went so far as to rearrange member genera between the superfamilies Alpheoidea Rafinesque, 1815 and Crangonoidea Haworth, 1825. He placed the genus Lysmata with the closely related Lysmatella and Exhippolysmata in its own family, the Lysmatidae Dana, 1852. Chace  did not agree with Christoffersen's rearrangement of taxa into new superfamilies. He performed a non-cladistic analysis of the 40 genera originally assigned to the Hippolytidae, examining 107 separate characters . He concluded that the family was "reasonably homogenous", but agreed with Christoffersen's  suggestion to move the genera Barbouria Rathbun, 1912, Janicea Manning and Hart, 1984, and Parhippolyte from the Hippolytidae to a new family, Barbouriidae Christoffersen, 1987. Martin and Davis  recognized some of the inconsistencies detailed by Christoffersen , and use Barbouriidae in their classification of recent Crustacea. However, they kept the Barbouriidae within the superfamily Alpheoidea, because of similarities to hippolytids. Furthermore, Martin and Davis  kept the rest of the hippolytids intact, not recognizing any of Christoffersen's  other new families. More recently, in a phylogenetic analysis of the Infraorder Caridea based on 16S and 18S sequence data, the genus Lysmata formed a distinct clade, well separated from the other hippolytids , supporting Christoffersen's view of a paraphyletic Hippolytidae. Hence, the accepted phylogeny of the Hippolytidae and related taxa is as yet unresolved.
In this paper, we present a phylogeny of 29 Lysmata species and eight genera of related hippolytids and two barbouriids, based upon sequences from both mitochondrial and nuclear ribosomal gene sequences. Our use of two genes from independently evolving genomes, a thorough taxonomic coverage of the Lysmata and related genera, and a robust analysis in terms of alignment strategies improves upon a very recent preliminary phylogeny of the genus Lysmata . We demonstrate that PSH evolved outside the genus Lysmata, as it is present in at least one ancestral taxon. Our phylogenetic analyses support the past division of Lysmata and Hippolysmata species based on the morphology of the antennular accessory flagellum, and the need for revision of both past and present Hippolytidae.
We obtained specimens from Lysmata and other hippolytid genera, from all over the world (Table 1). Hereafter, when discussing phylogenetic relationships we refer to the historical Hippolysmata/Lysmata taxonomic nomenclature (prior to ) based on the presence or absence of a multi-segmented accessory branch on the dorsolateral flagellum of the antennule. We differentiate Lysmata as ornamented with a short, one-segmented accessory branch, a long multi-segmented accessory branch, or unguiform accessory branch (newly described here). Most of the Indo-Pacific specimens were collected by the first author in Hawaii, Japan, and other Pacific locations; the majority of West Atlantic specimens were provided by AR. Other specimens were kindly provided by individuals from a variety of locations, including Indonesia, the Mediterranean, and Brazil. Many specimens were photographed prior to fixation, as color information is critical in the ultimate determination of species identity . Species identities were determined using published descriptions (e.g. ), the most recent morphological keys (e.g. [16, 22]), and descriptions of several new species  Specimens or portions of specimens were fixed in 80-100% ethanol by their respective sources. A small number of specimens were frozen for mitochondrial separation procedures (see below). Where possible, we included replicate specimens for each species, including confirmed specimens from different geographical regions. For example, Lysmata wurdemanni was sampled from two locations in Florida and one location in Texas.
We have also included representatives of the hippolytid genera Alope, Exhippolysmata, Heptacarpus, Tozeuma, and Thor, as well as two barbouriids (Barbouria, Parhippolyte) to explore their phylogenetic relationship with Lysmata. The snapping shrimp Synalpheus brevicarpus from the closely related Alpheidae  was selected as the designated outgroup (Table 1). The final data sets consist of a combination of our novel sequences with published sequences obtained from GenBank. The sources of the GenBank sequences are recent papers by Porter et al. , Baeza et al. , and Rhyne et al. . Samples including taxonomic authority, location, and GenBank accession numbers are given in Table 1.
DNA was isolated from individual specimens using one (or more) of three techniques, dependent upon sample condition, fixation method, and laboratory location. Total DNA extractions from EtOH-fixed specimens were performed in one of two ways: a) using the PureGene DNA isolation kit (Gentra) for fixed-tissue or b) via SDS & phenol/chloroform extraction [29, 30]. When available, frozen samples were also subjected to preferential mtDNA extraction using the alkaline lysis procedure . This procedure was used because of concerns that mtDNA sequences (i.e., 16S) were confounded by the presence of putative mitochondrial pseudogenes (numt) in several species [32, 33].
The 16S region was amplified with the 1471-1472 primers . The 28S region was amplified with "28S01" 5'-GACTACCCCCTGAATTTAAGCAT-3' and "28SP19F" 5'-GAGATTACCCGCCTAATTTAAGCAT-3' as forward primers paired with the reverse primer "28SR-02" 5'-CTCCTTGGTCCGTGTTTC-3'. PCR conditions were optimized for each gene-species combination via gradient PCR procedures. PCR products were assessed via electrophoresis of 2.5-5 μl of amplicon on a 0.7-1% agarose gel. Amplified bands were visualized under UV light and stored digitally. PCR products were cleaned of excess dNTPs, primers, and other impurities with one of two methods: a) enzymatic treatment with EXOSAP or b) silica gel extraction and wash . All successful PCR products were processed for sequencing using the Big Dye 3.1 Terminator Cycle Sequencing Kit and the ethanol precipitate products were loaded into either an ABI 3130xl 16-capillary Genetic Analyzer or an ABI 377 DNA sequencer. DNA products were sequenced from both directions. Sequence traces were viewed and processed with Phrap/Phred/Consed software [36–38] or 4Peaks software  and the chromatographs were cross-checked during contig building. Identical sequences were collapsed in MacClade  and represented as one taxon in the analysis. We reconstructed phylogenies based on both the 16S and 28S data sets separately, and combined. Preliminary analysis of the 28S region in several species showed that there was either no variation or very little variation among closely related species, so representative species of each of the 3 main clades of the 16S tree were chosen for sequencing. DNA sequences were aligned in ClustalX  using the default parameters. The resulting alignments of 16S and 28S consisted of conserved and highly variable regions. Some of highly variable regions could not be aligned unequivocally and those regions were removed by Gblocks v 0.91b . The Gblocks parameters for the 16S and 28S data sets were: minimum number of sequences for a conserved position (11/32), minimum number of sequences for a flanking position (11/32), maximum number of contiguous non-conserved positions (8/8), minimum length of a block (5/5), and allowed gap positions (with half/with half). We explored further the robustness of the phylogenetic signal of the datasets against a) the alignment deriving after highly variable regions were removed with even more stringent criteria by Gblocks and b) the alignment deriving from the default settings in ClustalX. All alignments are available as supplementary data (Additional File 1: Table S1).
We analyzed the phylogenetic relationships of the sequences by using MCMC-based Bayesian inference (BI) as implemented in MrBayes v. 3.2  and maximum parsimony (MP) and maximum likelihood (ML) in PAUP* . Data specific models of nucleotide evolution were evaluated with ModelTest  by the AIC criterion. In the BI of the combined data set (16S+28 S), each data partition was assigned a different model of substitution. The conditions for the Bayesian analysis were: three million generations, four simultaneous independent runs, and tree sampling every 1000th generation. For the Bayesian analysis of the concatenated data, different nucleotide substitution models were applied to each data partition. Graphs of ln(L) against number of generations were inspected to determine the burn-in factor. A consensus tree was calculated after discarding the first 10% trees as burn-in, which ensured that non-optimal trees were not included. Searches for the MP tree(s) run using the full heuristic option with 10 random replicates and for ML trees the fast stepwise addition option was used. The robustness of each clade was assessed with 100 replicates for ML and 1000 for MP of the non-parametric bootstrap procedure . For each bootstrap replicate, in MP a heuristic search was performed with 10 random taxon addition sequences and TBR branch swapping and in ML, a heuristic search was performed with the stepwise-addition option and TBR branch swapping. The Bayesian trees are presented and important topological discrepancies among the three phylogenetic methods are discussed. Posterior probabilities (pP) and bootstrap support (bp) values are used to indicate clade support.
We obtained 16S sequences from more than 100 specimens belonging to 29 species of Lysmata (27 named, 2 new), in addition to 16 species belonging to 8 other hippolytid genera and two genera of Barbouriidae, from all over the world (Table 1). In addition, we obtained partial sequences of the 28S ribosomal gene from 11 species of Lysmata and 9 species (eight genera) of other hippolytids and barbouriids (Table 1). The TrN + I + Γ and the GTR + Γ models of substitution were selected as the appropriate models for the 16S and 28S data sets, respectively.
Phylogenetic analyses of 16S (Figure 1), 28S (Figure 2) and the concatenated data sets (16S+28 S; Figure 3) overall supported the historical division of the Lysmata based on antennular accessory branch morphology. The short accessory branch group includes L. bahia, L. ankeri, L. pederseni, L. bogessi, L. rafa, L. wurdemanni, L. gracilorostris, L. nayaritensis, L. amboinensis, L. grabhami, L. debelius, L. californica, and L. olavoi (Figure 1, pP = 63). The low clade values (pP < 50) supporting the basal position of L. olavoi with respect to species possessing a short accessory branch reflect the uncertainty of its placement in the phylogenetic tree (Figure 1). The Lysmata group ornamented with a long accessory branch consists of L. galapagensis, L. moorei, L. nilita, L. intermedia, L. seticaudata, L ternatensis, L. trisetacea, L. acicula and was recovered as a highly supported monophyletic group (Figure 1, pP = 100). The three species (Lysmata hochi, L. cf. anchisteus, and L. lipkei) that are ornamented with an unguiform branch are recovered outside Lysmata and clustered with Exhippolysmata and Lysmatella (Figure 1), but this topological arrangement is not consistent with all alignment strategies.
The analyses also support a behavioral split within the short accessory branch clade - the so-called "cleaner" vs. "peppermint" shrimps. The specialized cleaner shrimps are defined as species with white antennae and legs, and bright body coloration, where peppermint shrimps lack white antennae and legs, and bright body coloration . The "peppermint" shrimps, which are represented in the 16S tree by L. wurdemanni, L. boggessi, L. pederseni, L. ankeri, L. rafa, L. bahia, L. gracilirostris, L. nayaritensis are differentiated from the "cleaners" L. debelius, L. amboinensis and L. grabhami as separate clades (pP = 88 and pP = 71, respectively). The placement of L. californica, which is considered a peppermint shrimp, is contingent to the alignment strategy of the 16S data. Different alignments placed this species ancestral to cleaners or ancestral to non-cleaners (Additional File 2: Figure S1) or nested within the cleaner clade, sister taxon to L. debelius (Figure 1). The phylogenetic divisions between short and long accessory branch clades and between behavioral groups within the short accessory branch clade are supported mainly by BI and not by the ML or the MP method.
The clade including Exhippolysmata oplophoroides, Lysmatella prima, and the three unguiform Lysmata is positioned outside of Lysmata (short and long accessory branch) but the placement is either weakly supported (pP = 62; Figure 1), or in different positions in alternative alignments (Additional File 2: Figure S1). This group of species, along with Parhippolyte mistica, Barbouria cubensis, and Merguia Kemp 1914, are basal to all Lysmata, except those ornamented with an unguiform accessory branch (Figure 1). Finally, the 16S topology indicates two pairs of sister taxa, the genera Hippolyte with Tozeuma Stimpson, 1860, and Heptacarpus Holmes, 1900 with Thor Kingsley, 1878.
The resulting phylogeny from the 28S data alone (Figure 2) displays a general topology similar to that observed from the 16S (Figure 1) and the concatenated 16S+28S data set (Figure 3). The same major clades are apparent among the Lysmata, Lysmatella, and Exhippolysmata taxa. The only exception is the relative ancestral/derived positions of these clades, which are not resolved. This loss of resolution may simply be due to the relatively more conserved 28S region. The concatenated data strongly support the behavioral division within the shrimps possessing a short accessory branch: cleaners (pP = 100, bp = 87) vs. peppermint shrimps (pP and bp = 100). These data also support the historical division (prior to ) between Lysmata (pP = 100, bp = 99) and Hippolysmata (pP = 100, bp = 96), though neither L californica, L. olavoi or L. nayaritensis) are included. Lysmata cf. anchisteus and Lysmata lipkei are clustered outside Lysmata and Hippolysmata forming a clade with Exhippolysmata and Lysmatella, an association observed in the 16S analysis. Similar to the 16S tree, Merguia, Parhippolyte, and Barbouria are basal to Lysmata.
I. Lysmatataxonomy & phylogeny
A. Historical division between Lysmata & Hippolysmata
Our data generally support the historical division of Lysmata based on accessory flagellum morphology. It also partly supports Rhyne's  further division of Lysmata according to morphology and/or color pattern: (1) Lysmata, having a long accessory branch, (2) Hippolysmata (prior to ), having a short accessory branch and displaying typical peppermint color patterns, and (3) cleaner shrimps, within Hippolysmata, with a short accessory branch and displaying bright coloration with white antenna. However, support for these groupings is contingent upon analysis method and alignment strategy (Figure 1, Additional File 2: Figure S1). Specifically, the positions of three peppermint shrimp species are problematic. The inconsistent placement of L. californica and L. nayaritensis challenges the monophyly of the peppermint shrimps. Furthermore, the support for the monophyly of the species with short accessory branch is weakened by the variable topological position of L. olavoi. Regardless, the discovery of a putative third group with unguiform branch (see below), renders the genus Lysmata paraphyletic. The BI recovers the different groups more consistently than both ML and MP, especially in the combined dataset (Figure 3). However the absence of L. californica, L. nayaritensis and L. olavoi from the combined data set weakens the comparison between the 16S and the 16S/28S trees. For comparison, the Baeza et al.  analysis recovers Exhippolysmata and L. hochi within the clade with the short accessory branch. Another problematic taxon is L. olavoi which is placed (pP = 63) in a basal position in the group with short accessory branch (Figure 1). Unlike L. californica, there is no behavioral data for L. olavoi, which has only been collected with traps from >125 m depth in Azores . Lysmata olavoi is placed ancestrally to all other Lysmata in . Any interpretation of the current results and those of Baeza et al.  should be made cautiously, as the evolutionary nature of the ribosomal datasets (i.e. excessive indel events) may limit the phylogenetic information they can convey.
The "cleaners" (L. amboinensis, L. debelius, and L. grabhami) may form a monophyletic group , except for the inclusion of L. californica within the cleaner clade. The placement of L. californica is unresolved, because it is strongly influenced by the alignment strategy. Lysmata californica has "peppermint shrimp" characteristics, lacking white legs and antenna, having translucent body with red stripes and living in groups. The monophyly of the remaining peppermint shrimps (i.e., L. wurdemanni, L. rafa, L. boggessi, L. pederseni, L. ankeri, L. bahia, L. gracilirostris, L. nayaritensis) was strongly supported (pP = 88; Figure 1). In contrast, the support for a monophyletic clade with species bearing a long accessory branch was very robust (pP = 100) and insensitive to the alignment conditions and the dataset used.
The topologies based on ribosomal data are also sensitive to the inclusion of particular taxa. By including Lysmatella prima in the analysis, L. hochi is no longer the sister taxon of Exhippolysmata, as indicated in . Rather, Lysmatella is the "new" sister taxon, whereas L. hochi is consistently grouped with L. anchisteus (Figure 1). There are arguments supporting that a denser phylogenetic sampling of taxa will generally improve the phylogenetic accuracy [50, 51], but others highlight the importance of longer sequences rather than denser taxon sampling . However, the addition of more sequence data without concomitantly increasing the sampled taxa can lead to strong systematic biases, producing highly supported, but incorrect or misleading topologies . Without a doubt, more species and more genes will be added in the future and should better resolve the systematic inconsistencies of Lysmata and related genera. Besides the potential problem of limited taxa and gene sampling, the tree topology may be more influenced by the final alignment itself than by the phylogenetic reconstruction method [54, 55]. There are several possible ways to resolve the problem of alignment uncertainty: 1) explore the effect of different alignment strategies, 2) removal of uncertain regions and/or 3) include protein coding genes where homologous alignment may be more objective by using the more conserved amino acid sequences. The alignment uncertainty of ribosomal data sets caused by the excessive indel events of the Lysmata phylogeny will be ameliorated when nuclear protein-coding genes are included in the analysis.
B. A possible third clade of Lysmata
Three Lysmata species (L. anchisteus, L. hochi, L. lipkei) are robustly placed outside the Lysmata + Hippolysmata clade in the 16S phylogeny. Similarly, the 16S/28S concatenated phylogeny places L. cf. anchisteus and L. lipkei in the same clade with Lysmatella and Exhippolysmata outside of their traditional taxonomic boundaries (Figure 3); this grouping suggests that an additional clade might be formed by species with a highly reduced antennular accessory branch. Lysmata anchisteus, L. hochi, and L. lipkei possess a vestigial antennal flagella, at most one segment in length with an unguiform shape [[19, 20, 26, 49], respectively]. The position of these three species suggests they are basal to the other clades of Lysmata. Data from morphologically similar species (e.g. L. uncicornis and L. kuekenthali) may help clarify the occurrence of this clade. These "outlier" Lysmata also present a challenge to any revision of the nomenclature of the genus. If one proposes to resurrect Hippolysmata and Lysmata to their previous status based upon phylogenetic data, the outliers could not be placed into either genus. A new genus may have to be erected, once their relationship with Exhippolysmata and Lysmatella is clarified. Alernatively, the presence of Exhippolysmata and Lysmatella in the putative third clade of Lysmata, may indicate that the generic definitions of these two genera based on raised basal crest (Exhippolysmata) and lack of epipods (Lysmatella) may be insufficient to raise these species to the genus level. Unlike the Lysmata species of the third clade that are ornamented with an unguiform antennal flagella, Exhippolysmata and Lysmatella have a short blunt flagella.
C. Exhippolysmata & Lysmatella
One of the surprising findings in  is the position of the genus Exhippolysmata within Lysmata, rendering the genus Lysmata paraphyletic. We have shown that the position of Exhippolysmata depends on the alignment strategy for 16S and the taxon sampling. Additionally, single gene approaches of closely related species should be interpreted cautiously as they often represent the phylogeny of the genes [56–58] or the organelles  and not the "true" organismal phylogeny. It is obvious that more genes and additional taxon sampling are needed to resolve the phylogenetic issues of Lysmata and other closely related genera. When the morphology of the two genera is taken in to consideration (raised basal crest in Exhippolysmata; lack of epipods on the first four pereiopods in Lysmatella) it seems highly improbable that species with vastly different morphological characters would be nested within a clade of Lysmata. Even though we present a phylogeny based on a denser taxon sampling and an additional gene from the independently evolved nuclear genome, our approach is still limited. We have proceeded by concatenating the two gene sequences prior to the phylogenetic analysis (i.e. total evidence approach), because it has been demonstrated empirically that concatenation of multiple genes often results in a single well-supported topology . Theoretical work, however, has shown that especially when the coalescent process is highly variable from gene-to-gene , concatenation of data sets can produce inconsistent phylogenetic estimates .
II. Hermaphroditism & Life History Patterns
Our data do not support any relationship between cleaning symbiosis or social system and the origin of PSH or the genus itself. Results from our phylogenetic analyses suggest that fish cleaning is a derived behavior within the short accessory branch clade. Lysmata californica, a peppermint shrimp that commonly associates with moray eels is placed within the clade that includes species living in pairs and bright coloration indicating strong specialized behavior (Figure 1). For comparison, L. californica is basal to the cleaner clade in the study of Baeza et al. . The different placement of this taxon is an alignment artifact as both studies used different alignment criteria. There is also evidence of moray eel interactions with species bearing long accessory branch . Since well-developed cleaning behavior evolved once within Lysmata, there seems to be no obvious connection of the so-called "paired cleaning species" with the origin of PSH; PSH is ubiquitous within Lysmata and likely evolved ancestrally to the genus. Furthermore, most of the Lysmata species examined would be classified as group living species, including taxa basal to the "paired cleaner" clade. The assumption that group size = mating system in nature should be substantiated with supporting observations of behavior under natural conditions.
PSH has been recently been reported in Exhippolysmata , Lysmatella prima (Rhyne, unpublished) and Parhippolyte (Onaga & Fiedler, unpub.), a genus placed ancestrally to Lysmata, regardless of the alignment strategy. Clearly, studies attempting to determine the origins of PSH must focus on related genera that are ancestral to Lysmata, a point that has been highlighted also in . Christoffersen  subdivided the hippolytids into superfamily and families based on morphological comparisons. The placement of Parhippolyte and Lysmata in different families (Barbouriidae and Lysmatidae, respectively) would further support that PSH evolved well outside of Lysmata and could be far more common than previously considered. The cave dwelling, group-living shrimp genus Parhippolyte possesses PSH and all phylogenetic analyses support the ancestral position of this group relative to all Lysmata and Exhippolysmata. When Bauer [12, 13, 47] postulated the evolution of PSH within Lysmata and why there are two distinct ecological clades, he was unaware that PSH is secondary to the divisions within the genera. The evolution of PSH likely predates the diversification of Lysmata and may have little or no bearing on the evolution of different ecological groups within the genus. For Lysmata, the question is not how PSH is related to socio-ecological systems, but rather why pair living and specialized fish cleaning behavior evolved from a group living ancestor.
III. Phylogenetic issues in the Hippolytidae and related taxa
Based on cladistic analysis, Christoffersen  split the hippolytid genera into several different families. Notably, Lysmata, Calliasmata Holthuis 1973, Exhippolysmata, and Mimocaris Vereshchaka 1997 were assigned to the family Lysmatidae, while Barbouria, Parhippolyte, among other genera were assigned to the Barbouriidae. These two families were included with the Processidae Ortmann, 1890 and Crangonidae, and the genera Merguia and Glyphocrangon Milne-Edwards 1881, in the superfamily Crangonoidea. The rest of the hippolytids are assigned by Christoffersen to various families within the superfamily Alpheoidea. Chace  rejected this wholesale rearrangement, though agreed with the erection of the Barbouriidae. This assertion was re-affirmed by Martin and Davis . We consistently recover the branch of Barbouria + Parhippolyte in all trees, therefore we cannot reject Christoffersen's suggestion for separating the Barbouriidae. However, it is not clear whether or not the level of differentiation from Lysmata, Exhippolysmata, Lysmatella, and Merguia is sufficient to propose a separate family. Although the relative placement of Merguia and the two barbouriids is susceptible to alignment strategies, they are clearly more closely related to Lysmata than the other hippolytid genera in our phylogenies. The ancestral relationship of Merguia to Lysmata and its basal position to the barbouriids could invalidate the Crangonoidea sensu Christoffersen. A denser sampling of taxa from Christoffersen's proposed groups may help to clarify the level of differentiation present among these taxa and others that have traditionally been a part of the Hippolytidae. Until more convincing conclusions can be drawn, the current delineations proposed by Martin and Davis  should be maintained.
Our mitochondrial and nuclear ribosomal data generally support the historic morphological division of Lysmata based on accessory branch morphology. Shrimps within the short accessory branch clade differentiate according to behavior and color pattern. The monophyly of the Lysmata group which is bearing a multi-segmented accessory branch is strongly supported, underlying the taxonomic importance of this character. Lysmata with an unguiform accessory branch are part of a third clade which includes Lysmatella and Exhippolysmata. The third clade does not conform to the historic division between Lysmata and Hippolysmata. PSH is ubiquitous within Lysmata and occurs in Barbouriidae, suggesting that this rare reproductive system that evolved ancestrally to the genera Lysmata, Exhippolysmata and Lysmatella. The two representative species of barbouriids form a monophyletic group and are consistently placed within the Hyppolytidae, therefore not providing support for the family Barbouriidae. The ribosomal data provides a unique view of the phylogeny of Lysmata and life history traits, however, the position of some taxa is sensitive to alignment strategies.
Charniaux-Cotton H: Masculinisation des femelles de la Crevette à hermaphrodisme protérandrique Lysmata seticaudata, par greffe de glandes androgènes. Interprétation de l'hermaphrodisme chez les Décapodes. Note préliminaire. Comptes Rendus de l' Academie des Sciences. 1959, 249: 1580-1582.
Charniaux-Cotton H: Physiologie de l'inversion sexuelle chez la Crevette à hermaphrodisme protérandrique fonctionnel, Lysmata seticaudata. Comptes Rendus de l' Academie des Sciences. 1960, 250: 4046-4048.
Dohrn PFR: Studi sulla Lysmata seticaudata Risso (Hippolytidae). Pubblicazioni della Stazione zoologica di Napoli. 1950, 22: 257-272.
Bauer RT, Holt GJ: Simultaneous hermaphroditism in the marine shrimp Lysmata wurdemanni (Caridea: Hippolytidae): an undescribed sexual system in the decapod Crustacea. Marine Biology. 1998, 132: 223-235. 10.1007/s002270050388.
Fiedler GC: Functional, simultaneous hermaphroditism in female phase Lysmata amboinensis (Decapoda: Caridea: Hippolytidae). Pacific Science. 1998, 52: 161-169.
D'Udekem D'Acoz C: Lysmata seticaudata (Risso, 1816) and L. nilita Dohrn & Holthuis, 1950 are protandrous simultaneous hermaphrodites (Decapoda, Caridea, Hippolytidae). Crustaceana. 2002, 75: 1149-1152. 10.1163/156854002763270545.
Bauer RT, Newman WA: Protandric simultaneous hermaphroditism in the marine shrimp Lysmata californica (Caridea: Hippolytidae). Journal of Crustacean Biology. 2004, 24: 131-139. 10.1651/C-2426.
Baeza JA: Protandric simultaneous hermaphroditism in the shrimps Lysmata bahia and Lysmata intermedia. Invertebrate Biology. 2008, 127: 181-188. 10.1111/j.1744-7410.2007.00122.x.
Rhyne AL, Anker A: Lysmata rafa, a new species of peppermint shrimp (Crustacea, Caridea, Hippolytidae) from the subtropical western Atlantic. Helgoland Marine Research. 2007, 61: 291-296. 10.1007/s10152-007-0077-4.
Anker A, Baeza JA, De Grave S: A new species of Lysmata (Crustacea, Decapoda, Hippolytidae) from the Pacific Coast of Panama, with observations of its reproductive biology. Zoological Studies. 2009, 48: 682-692.
Bauer RT: Reproductive ecology of a protandric simultaneous hermaphrodite, the shrimp Lysmata wurdemanni (Decapoda: Caridea: Hippolytidae). Journal of Crustacean Biology. 2002, 22: 742-749. 10.1651/0278-0372(2002)022[0742:REOAPS]2.0.CO;2.
Bauer RT: Simultaneous hermaphroditism in caridean shrimps: a unique and puzzling sexual system in the Decapoda. Journal of Crustacean Biology. 2000, 20: 116-128.
Bauer RT: Same sexual system but variable sociobiology: evolution of protandric simultaneous hermaphroditism in Lysmata shrimps. Integrative and Comparative Biology. 2006, 46: 430-438. 10.1093/icb/icj036.
Braga AA, López Greco LS, Santos DC, Fransozo A: Morphological evidence for protandric simultaneous hermaphroditism in the caridean Exhippolysmata oplophoroides. Journal of Crustacean Biology. 2009, 29: 34-41. 10.1651/08.3015.1.
Laubenheimer H, Rhyne AL: Experimental confirmation of protandric simultaneous hermaphroditism in a Caridean shrimp outside of the genus Lysmata. Journal of the Marine Biological Association of the UK. 2008, 88: 301-305. 10.1017/S0025315408000702.
Rhyne AL, Lin J: A western Atlantic peppermint shrimp complex: redescription of Lysmata wurdemanni, description of four new species, and remarks on Lysmata rathbunae (Crustacea: Decapoda: Hippolytidae). Bulletin of Marine Science. 2006, 79: 165-204.
Baeza JA, Anker A: Lysmata hochi n. sp., a new species of hermaphroditic shrimp from the southern Caribbean. Journal of Crustacean Biology. 2008, 28: 148-155. 10.1651/07-2839R.1.
Holthuis LB: The recent genera of the caridean and stenopodidean shrimps (Crustacea, Decapoda): With an appendix on the order Amphionidacea. 1993, Leiden: Nationaal Natuurhistorisch Museum Leiden
Chace F: The shrimps of the Smithsonian-Bredin Caribbean Expeditions with a summary of the West Indian shallow-water species (Crustacea: Decapoda: Natantia). Smithsonian Contributions to Zoology. 1972, 98: 1-179.
Udekem d'Acoz Cd: Redescription of Lysmata intermedia (Kingsley, 1879) based on topotypical specimens, with remarks on Lysmata seticaudata (Risso, 1816) (Decapoda, Caridea, Hippolytidae). Crustaceana. 2000, 73: 719-735. 10.1163/156854000504750.
Christoffersen ML: Phylogenetic relationships of hippolytid genera, with an assignment of new families for the Crangonoidea and Alpheoidea (Crustacea, Decapoda, Caridea). Cladistics. 1987, 3: 348-362.
Chace FJ: The Caridean Shrimps (Crustacea: Decapoda) of the Albatross Philippine Expedition, 1907-1910, Part 7: families Atyidae, Eugonatonotidae, Rhynchocinetidae, Bathypalaemonellidae, Processidae and Hippolytidae. Smithsonian Contributions to Zoology. 1997, 587: 1-106.
Martin JW, Davis GE: An updated classification of the Recent Crustacea. Natural History Museum of Los Angeles County, Science Series. 2001, 39: 1-124.
Bracken HD, de Grave S, Felder DL: Phylogeny of the Infraorder Caridea based on mitochondrial and nuclear genes (Crustacea: Decapoda). Decapod Crustacean Phylogenetics. Edited by: Martin JW, Crandall KA, Felder DL. 2009, Boca Raton, FL: CRC Press, Taylor & Francis Group, 281-305.
Baeza JA, Schubart CD, Zillner P, Fuentes S, Bauer RT: Molecular phylogeny of shrimps from the genus Lysmata (Caridea: Hippolytidae): the evolutionary origins of protandric simultaneous hermaphroditism and social monogamy. Biological Journal of the Linnean Society. 2009, 96: 415-424. 10.1111/j.1095-8312.2008.01133.x.
Okuno J, Fiedler GC: Lysmata lipkei, a new species of peppermint shrimp (Decapoda, Hippolytidae) from the warm temperate and subtropical waters of Japan. Crustaceana Monograph. Edited by: Fransen CHJM, De Grave S, Ng PKL. 2010, Studies on Malacostraca: Lipke Bijdeley Holthuis Memorial Volume,
Porter ML, Pérez-Losada M, Crandall KA: Model-based multi-locus estimation of decapod phylogeny and divergence times. Molecular Phylogenetics and Evolution. 2005, 37: 355-369. 10.1016/j.ympev.2005.06.021.
Rhyne AL, Zhang D, Lin J, Schizas NV: Not any two will do: DNA divergence and interpopulation reproductive compatibility in a simultaneous hermaphroditic shrimp, Lysmata wurdemanni. Marine Ecology Progress Series. 2009, 388: 185-195. 10.3354/meps08062.
Sambrook E, Fritsch F, Maniatis T: Molecular Cloning. 1989, Cold Spring Harbor, New York: Cold Spring Harbor Press
Segawa RD, Aotsuka T: The mitochondrial genome of the Japanese freshwater crab, Geothelphusa dehaani (Crustacea: Brachyura): evidence for its evolution via gene duplication. Gene. 2005, 355: 28-39. 10.1016/j.gene.2005.05.020.
Tamura K, Aotsuka T: Rapid isolation method of animal mitochondrial DNA by the alkaline lysis procedure. Biochemical Genetics. 1988, 26: 815-819.
Williams ST, Knowlton N: Mitochondrial pseudogenes are pervasive and often insidious in the snapping shrimp genus Alpheus. Molecular Biology and Evolution. 2001, 18: 1484-1493.
Schubart CD: Mitochondrial DNA and decapod phylogenies: the importance of pseudogenes and primer optimization. Decapod Crustacean Phylogenetics. Edited by: Martin JW, Crandall KA, Felder DL. 2009, Boca Raton, FL: CRC Press, Taylor & Francis Group, 47-65.
Crandall KA, Fitzpatrick JFJ: Crayfish molecular systematics: Using a combination of procedures to estimate phylogeny. Systematic Biology. 1996, 45: 1-26.
Boom R, Sol CJA, Salimans MM, Jansen CL, Wertheim-van Dillen PME, van der Noordaa J: Rapid and simple method for purification of nucleic acids. Journal of Clinical Microbiology. 1990, 28: 495-503.
Ewing B, Green P: Basecalling of automated sequencer traces using phred. II. Error probabilities. Genome Research. 1998, 8: 186-194.
Ewing B, Hillier L, Wendl M, Green P: Basecalling of automated sequencer traces using phred. I. Accuracy assessment. Genome Research. 1998, 8: 175-185.
Gordon D: Viewing and editing assembled sequences using consed. Current Protocols in Bioinformatics. Edited by: Baxevanis AD, Davison DB. 2004, New York: John Wiley & Co, 11.12.11-11.12.43.
Griekspoor A, Groothuis T: 4Peaks. 2006, [http://mekentosj.com]
Maddison D, Maddison W: MacClade 4: Analysis of Phylogeny and Character Evolution. Book MacClade 4: Analysis of Phylogeny and Character Evolution. 2000, Sinauer
Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins DG: The ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Research. 1997, 24: 4876-4882. 10.1093/nar/25.24.4876.
Castresana J: Selection of conserved blocks from multiple alignments for their use in phylogenetic analysis. Molecular Biology and Evolution. 2000, 17: 540-552.
Ronquist F, Huelsenbeck JP: MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics. 2003, 19: 1572-1574. 10.1093/bioinformatics/btg180.
Swofford DL: PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods), Version 4.0b10 edition. 2002, Sinauer
Posada D, Crandall KA: Modeltest: testing the model of DNA substitution. Bioinformatics. 1998, 14: 817-818. 10.1093/bioinformatics/14.9.817.
Felsenstein J: Confidence limits on phylogenies: an approach using the bootstrap. Evolution. 1985, 39: 783-791. 10.2307/2408678.
Bauer RT: Remarkable shrimps: natural history and adaptations of the carideans. 2004, Norman, Oklahoma: University of Oklahoma Press
Rhyne AL: Biology and systematics of Western Atlantic peppermint shrimp, Lysmata spp. (Decapoda: Caridea: Hippolytidae). 2006, Florida Institute of Technology, Department of Biological Sciences
Fransen CHJM: Lysmata olavoi, a new shrimp of the family Hippolytidae (Decapoda, Caridea) from the eastern Atlantic Ocean. Arquipélago, Life and Earth Sciences. 1991, 9: 63-73.
Zwickl DJ, Hillis DM: Increased taxon sampling greatly reduces phylogenetic error. Systematic Biology. 2002, 51: 588-598. 10.1080/10635150290102339.
Debry RW: The systematic component of phylogenetic error as a function of taxonomic sampling under parsimony. Systematic Biology. 2005, 54: 432-440. 10.1080/10635150590946745.
Rosenberg MS, Kumar S: Incomplete taxon sampling is not a problem for phylogenetic inference. Proceedings of the National Academy of Sciences of the United States of America. 2001, 98: 10751-10756. 10.1073/pnas.191248498.
Heath TA, Hedtke SM, Hillis DM: Taxon sampling and the accuracy of phylogenetic analyses. Journal of Systematics and Evolution. 2008, 46: 239-257.
Morrison DA, Ellis JT: Effects of nucleotide sequence alignment on phylogeny estimation: a case study of 18S rDNAs of apicomplexa. Molecular Biology and Evolution. 1997, 14: 428-441.
Wong KM, Suchard MA, Huelsenbeck JP: Alignment uncertainty and genomic analysis. Science. 2008, 319: 473-476. 10.1126/science.1151532.
Pamilo P, Nei M: Relationships between gene trees and species trees. Molecular Biology and Evolution. 1988, 5: 568-583.
Nichols R: Gene trees and species trees are not the same. Trends in Ecology and Evolution. 2001, 16: 358-364. 10.1016/S0169-5347(01)02203-0.
Hudson RR: Testing the constant-rate neutral allele model with protein sequence data. Evolution. 1983, 37: 203-217. 10.2307/2408186.
Ballard JW, Chernoff B, James AC: Divergence of mitochondrial DNA is not corroborated by nuclear DNA, morphology or behavior in Drosophila simulans. Evolution. 2002, 56: 527-545.
Rokas A, Williams BL, King N, Carroll SB: Genome-scale approaches to resolving incongruence in molecular phylogenies. Nature. 2003, 425: 798-804. 10.1038/nature02053.
Hudson RR, Turelli M: Stochasticity overrules the ''threetimes rule'': genetic drift, genetic draft, and coalescence times for nuclear loci versus mitochondrial DNA. Evolution. 2003, 57: 182-190.
Kubatko LS, Degnan JH: Inconsistency of phylogenetic estimates from concatenated data under coalescence. Systematic Biology. 2007, 56: 17-24. 10.1080/10635150601146041.
Wicksten MK: Interactions with fishes of five species of Lysmata (Decapoda, Caridae, Lysmatidae). Crustaceana. 2009, 82: 1213-1223. 10.1163/156854009X448899.
De Grave S, et al: A classification of living and fossil genera of decapod crustaceans. Raffles Bulletin of Zoology. 2009, 21: 1-109.
Chiba S: A review of ecological and evolutionary studies on hermaphroditic decapod crustaceans. Plankton & Benthos Research. 2007, 2: 107-119.
AR is grateful to the Bahamian Department of Fisheries and to Gerace Marine Laboratory, College of the Bahamas (Dr. Thomas, A. Rothfusand and Rochelle Hanna), and Dr. Robert Knowlton for their logistical support and/or permitting of collections on San Salvador Island. GCF thanks the Hawaii Institute of Marine Biology, University of Hawaii (Dr. Ernst Reese), Sesoko Tropical Biosphere Research Center, University of the Ryukus (Dr. Kazuhiko Sakai), Akajima Marine Science Laboratory, and Shimoda Marine Research Center, University of Tsukuba, for their logistical support of specimen collection. AR and GCF are grateful to Helio Laubenheimer, Massimo Boyer, Hitoshi Onaga, Drs. Yoshihisa Fujita, Sammy De Grave, Tohru Naruse, and Peter Wirtz for sample collections. We thank Drs. Arthur Anker and Sammy De Grave for their contribution to the background discussion relating to the generic and familiar status of Lysmata and surrounding genera. Lastly, we are deeply indebted to the two anonymous reviewers for their time and helpful comments on earlier versions of the manuscript. A large portion of GCF's contribution was supported by a Japan Society for the Promotion of Science Postdoctoral Fellowship. Partial funding for this project was provided by the 2007 Arts and Sciences Seed Money of University of Puerto Rico, Mayagüez awarded to NVS. A portion of DNA sequencing was performed in the Sequencing and Genotyping facility, University of Puerto Rico, Río Piedras which is supported in part by NCRR AABRE Grant #P20 RR16470, NIH-SCORE Grant #S06GM08102, University of Puerto Rico Biology Department, NSF-CREST Grant #0206200.
GCF and AR conceived the project and collected the specimens. RS, TA and NVS provided monetary support, facilities and contributed to the manuscript. GCF, RS and NVS generated the molecular data. GCF, AR and NVS carried out the analyses and wrote the manuscript. All authors read and approved the final manuscript.