Molecular species delimitation methods and population genetics data reveal extensive lineage diversity and cryptic species in Aglaopheniidae (Hydrozoa)
Graphical abstract
Introduction
The global biodiversity crisis requires focusing conservation efforts on key areas that ensure the long-term persistence of the greatest fraction of global biodiversity (Barnosky et al., 2011, Myers et al., 2000). The identification of such regions requires reliable assessments of alpha diversity, i.e. the number of species present in the area of interest (Margules and Pressey, 2000). Historically, species have been described and identified based on morphological characters. Even if traditional taxonomic work continues to be important for species inventories and conservation (Daugherty et al., 1990), ignoring cryptic diversity leads to incomplete taxon sampling and erroneous assessments of biodiversity, biogeographic patterns and speciation processes (Heath et al., 2008). Indeed, traditional taxonomy does not integrate genetic diversity and might ignore potential cryptic species [i.e. two distinct sympatric species classified under the same taxonomic name based on morphological characters (see box 1 in Bickford et al., 2007)]. Hence, protecting only the ‘visible biodiversity’ will negatively affect conservation and management efforts of biodiversity and evolutionary processes (Briggs, 2005, Moritz, 2002, Moritz, 1999).
Most evolutionary biologists concur that species and higher taxonomic levels form independent genealogical lineages of organisms (Mayden, 2002, Samadi and Barberousse, 2006) and the broad use of molecular markers during the past decades has revealed the prevalence of cryptic lineage diversity among marine organisms, especially in marine invertebrates (Boissin et al., 2008, Duda et al., 2008, Hoareau et al., 2013, Huelsken et al., 2013, Knowlton, 1993, Lindner et al., 2011, Niemiller et al., 2011, Palumbi, 1994, Pfenninger and Schwenk, 2007, Prada et al., 2014). The main issue in species delimitation is not the definition of what a species is, but resides in the issue of choosing a criterion (e.g. morphology, ecology, genetic distances, etc.) to identify lineages. No consensus exists (De Queiroz, 2007, De Queiroz, 2005, De Queiroz, 1992) and the criterion used can possibly lead to a false representation of a taxon’s evolution and diversity (Agapow et al., 2004, De Queiroz, 2005, Mayden, 2002, Samadi and Barberousse, 2006). Thus, when describing alpha-diversity, identification and description of species using morphological characters generate primary species hypotheses (PSHs) that need to be tested and confronted to other sources of information in order to delineate robust secondary species hypotheses (SSHs) and potentially reveal cryptic species [see Puillandre et al. (2012b) for an example in gastropods].
Among marine invertebrates, cnidarians in particular can be expected to present cryptic diversity due to the paucity of morphological characters useful for species description and systematics (e.g. Addamo et al., 2012, McFadden et al., 2014, Stampar et al., 2012). These morphological clues do not necessarily represent phylogenetic relationships since evolution of reproductive, ecological and physiological traits and eventually speciation do not always have morphological outcomes. Among cnidarians, the class Hydrozoa is particularly subject to morphological plasticity and taxonomic incertitude (Bavestrello et al., 2000, Bouillon et al., 2006, Leclère et al., 2009, Meroz-Fine et al., 2003, Miglietta et al., 2009). Hydrozoans are found in almost all aquatic ecosystems: polar to tropical regions, shallow waters to abyssal plains, freshwater and marine ecosystems (Bouillon et al., 2006). Recent phylogenies showed that morphological characters thought to be taxonomically significant in this class were actually highly labile and plastic (e.g. Leclère et al., 2007, Miglietta et al., 2009, Moura et al., 2012, Postaire et al., 2015c). Several studies investigated cryptic diversity in hydrozoans (e.g. Schuchert, 2014, Folino-Rorem et al., 2009, Moura et al., 2008, Govindarajan et al., 2005), and the existence of “true” cryptic species (sensu Bickford et al., 2007) in cosmopolitan morpho-species (i.e. species delimited using morphological characters) has already been revealed or hypothesized (Lindner et al., 2011, Miglietta et al., 2007, Schuchert, 2005). Unfortunately, recent studies on hydrozoan systematics are often limited to a DNA barcoding approach, a method which was originally developed to help species identification by associating DNA sequences to type specimens (Hebert et al., 2003, Puillandre et al., 2011, Vernooy et al., 2010). While DNA barcoding and genetic data in general were not initially proposed to be employed as species delimitation tools, it can help to uncover species diversity in complex taxa (Hebert et al., 2004) and several novel methods have been developed to use genetic data as a first step to delimit putative species, especially in taxa with limited comprehensive information (Carstens et al., 2013). These methods are particularly useful in taxa lacking clear synapomorphies and species boundaries, such as hydrozoans [see Castelin and Lambourdière (2010) for an example in gastropods].
Here we used three of these methods on the Aglaopheniidae (Agassiz, 1862), a highly specious family of mostly brooding hydrozoans with unresolved taxonomy and phylogeny (Bouillon et al., 2006, Moura et al., 2012, Postaire et al., 2015c). First, the Automatic Barcode Gap Discovery (ABGD) method (Puillandre et al., 2012a) uses a genetic distance based approach to detect a barcode gap dividing candidate species in the dataset by assuring that intra- and inter-specific genetic distances do not overlap. This method is independent of tree topology. ABGD calculates all pairwise distances and orders them as ranked values. A sliding window is then applied to calculate a local slope function across these values: the barcode gap is the first significant peak (increase of slope) that allows inferring primary partitions of the dataset. Each sub-partition is then analyzed using the same approach until no further significant gaps are found (Puillandre et al., 2012a). Secondly, species delimitation methods based on the Generalized mixed Yule-coalescent (GMYC) model (Fontaneto et al., 2010, Fujita et al., 2012, Pons et al., 2006) are based on tree topologies to infer species hypotheses. Using a likelihood function modeling evolutionary processes, this model states that each node of a phylogenetic tree corresponds to one of two possible events: divergence between species following a strict Yule process [no extinction; (Yule, 1925)] or neutral coalescent events between lineages forming a species (Kingman, 1982). As coalescent events are assumed to occur at higher rates than speciation, it is thus possible to identify a limit on a phylogenetic tree between inter- and intra-specific divergence, delimiting clusters of leaves. Such clusters represent genetically isolated, independently evolving lineages, in which selection and genetic drift operate (Fujita et al., 2012), i.e. species hypotheses. Finally, the Poisson tree processes (PTP) species delimitation method is based on the differences between sequences (number of substitutions), but contrary to the GMYC models, it does not use a calibrated tree (Zhang et al., 2013). This method makes the assumptions that each mutation event has a non-null probability of forming a new species and, as a consequence, that the number of substitutions between species is significantly higher than the number of substitutions within species. We used a combination of the three methods (i.e. ABGD, GMYC and PTP) to delineate SSHs.
Although considered efficient in identifying species limits (Puillandre et al., 2012b), even when singletons (i.e. a single haplotype per species hypothesis, PSH and/or SSH) represent an important part of the dataset (Talavera et al., 2013), some studies highlighted the tendency of these methods to overestimate the number of species (Hamilton et al., 2014, Lohse, 2009, Puillandre et al., 2012a, Zhang et al., 2013). Several studies further underlined that basing species delimitation solely on genetic data, a fortiori on a single mitochondrial marker, must be made cautiously (Dellicour and Flot, 2015, Hamilton et al., 2014, Jörger et al., 2012, Lohse, 2009). Indeed, deeply divergent mitochondrial lineages do not always imply distinct species: divergence might result from ancestral polymorphism, genetic introgression or hybridization (Ladner and Palumbi, 2012). Furthermore, gene trees do not always reflect species evolution (Hoelzer, 1997) and phylogenies must be discussed in an integrative framework that includes all available information (Dayrat, 2005, Padial et al., 2010, Puillandre et al., 2009, Schlick-Steiner et al., 2010).
We tested SSHs for two PSHs, Lytocarpia brevirostris (Busk, 1852) and Macrorhynchia phoenicea (Busk, 1852), using newly developed microsatellite markers to assess the congruence between the ABGD method, GMYC models, the PTP method and population genetics data. Microsatellite data have already been efficiently used for this purpose (Hausdorf et al., 2011, Hausdorf and Hennig, 2010, Turini et al., 2014). Furthermore, microsatellite markers present the advantage of being co-dominant, bi-parentally inherited and neutral, unlike mitochondrial markers. Finally, we compared species delimitation outputs to phylogenies produced with two nuclear markers: the first exon of the calmodulin and a sequence comprising ITS1 and ITS2.
This study is the first to apply species delimitation methods based on DNA sequences in Aglaopheniidae, a highly diversified family lacking clear synapomorphies and/or species limits (Moura et al., 2012, Postaire et al., 2015c). It aims to clarify taxonomic issues associated with cryptic diversity by using a combination of population genetics data, molecular phylogenies and DNA based species delimitation methods.
Section snippets
Sampling of PSHs
The samples used for phylogenetic analyses are the same as in Postaire et al. (2015c). Between 2007 and 2014, we explored three marine provinces as defined by Spalding et al. (2007), comprising six localities in two ecoregions of the Western Indian Ocean (WIO) province (western/northern Madagascar and Mascarene Islands), one locality in the South-East Polynesian province (SEP; Moorea, Society Islands) and one in the tropical South-Western Pacific province (SWP; New Caledonia) (Fig. 1a). We
16S variability in Aglaopheniidae and phylogenetic reconstruction
A total of 37 PSHs were identified in this study. A set of 340 sequences (207 unique haplotypes) of 521 base pairs (bp) was generated and analyzed; this dataset comprised 396 polymorphic sites, 98 identical sites (19.0%), 81.9% pairwise identity and a GC content of 24.6% (base composition: A = 42.5%, C = 12.3%, G = 14.6%, T = 30.6%). Six PSHs were singletons (i.e. PSHs represented by only one haplotype): Streptocaulus multiseptatus (Bale, 1915), S. dolfusi (Billard, 1924), Cladocarpus integer (Sars,
Discussion
We used the mitochondrial marker 16S to study lineage diversity in Aglaopheniidae and detect potential cryptic species. Based on our protocol (i.e. excluding singletons), the combination of several molecular based species delimitation methods identified 35 SSHs among 37 PSHs (Fig. 2, Appendix C). Several SSHs did not correspond to PSHs, revealing potential cases of synonymy and the presence of independent lineages within PSHs. Nevertheless, the majority of PSHs were monophyletic and SSHs were
Conclusions
In this study we employed species delimitation tools to assess the congruence between genetic data and morphological classification in hydrozoans from the Aglaopheniidae family. We identified several independent lineages in nominal morpho-species, representing true cryptic species. These results, based on the congruence of sequence and microsatellite data, reveal part of the hidden diversity within the Aglaopheniidae family. Our results underline the relevance of integrative taxonomy in
Acknowledgements
This work was supported by the Laboratoire d’Excellence CORAIL. Hydrozoan sampling in New Caledonia (HM) was carried out during Cobelo (doi: http://dx.doi.org/10.17600/14003700) and Bibelot (doi: http://dx.doi.org/10.17600/13100100) oceanographic campaigns on board of RV Alis (IRD). Sampling in Reunion Island (HM, BP, HB, CAFB) was supported by program HYDROSOOI (Labex CORAIL fund); in Madagascar (HM) supported by project Biodiversity (POCT FEDER fund) and in Juan de Nova (HM) by program
References (136)
- et al.
Cryptic species as a window on diversity and conservation
Trends Ecol. Evol.
(2007) - et al.
Zoogeography and life cycle patterns of Mediterranean hydromedusae (Cnidaria)
Biol. J. Linn. Soc.
(1993) Coral reefs: conserving the evolutionary sources
Biol. Conserv.
(2005)ITS2 is a double-edged tool for eukaryote evolutionary comparisons
Trends Genet.
(2003)- et al.
Molecular evidence for multiple episodes of paedomorphosis in the family Hydractiniidae
Biochem. Syst. Ecol.
(1993) - et al.
Coalescent-based species delimitation in an integrative taxonomy
Trends Ecol. Evol.
(2012) - et al.
An evaluation of sampling effects on multiple DNA barcoding methods leads to an integrative approach for delimiting species: a case study of the North American tarantula genus Aphonopelma (Araneae, Mygalomorphae, Theraphosidae)
Mol. Phylogenet. Evol.
(2014) - et al.
Population genetic patterns revealed by microsatellite data challenge the mitochondrial DNA based taxonomy of Astyanax in Mexico (Characidae, Teleostei)
Mol. Phylogenet. Evol.
(2011) - et al.
Phylogeny and biogeography of Hydra (Cnidaria: Hydridae) using mitochondrial and nuclear DNA sequences
Mol. Phylogenet. Evol.
(2010) - et al.
Desmophyllum dianthus (Esper, 1794) in the scleractinian phylogeny and its intraspecific diversity
PLoS ONE
(2012)