• allopatry;
  • echinoderms;
  • ecological speciation;
  • hybridization;
  • phylogeography


  1. Top of page
  2. Abstract
  3. References

Where do new species arise? When do they form and how do they diverge from a common ancestor? A new comprehensive study of Arbacia sea urchins provides surprising answers to these questions. By combining mtDNA phylogeographic markers with a nuclear locus (encoding the sperm acrosomal protein bindin) known to be susceptible to high rates of adaptive codon evolution, Lessios et al. (2012) show that new species and lineages arose relatively recently, most often in association with latitudinal shifts between the temperate zones and the tropics, and in one case, in association with a significant geological barrier to gene flow (the rise of the Isthmus of Panama). In addition to the ‘where’ and ‘when’ of Arbacia speciation, these new data resolve an important question about ‘who’Arbacia species are by revealing extensive allele sharing at both loci between a pair of broadly sympatric nominal species (that should perhaps be considered a single taxon). ‘How’Arbacia diverge from each other is less easily resolved: there is no evidence for reinforcement (via selection on bindin) as an important source of divergence between nominal species, and there are few other data to decide among the alternative hypotheses to explain Arbacia speciation.

In marine communities, the conventional answer to the first question posed earlier has been ‘In the tropics’. Shallow-water epifaunal sea urchins are a puzzle in this respect: most sea urchin genera include only tropical or only temperate species, but tropical warm-water genera include species in several different oceans that have been cut-off from each other by Pliocene, Miocene or older oceanographic and geological barriers (like the Isthmus of Panama). More recent migration routes pass through cold Arctic and Antarctic waters that also seem like barriers to dispersal of tropical sea urchins and their planktonic larvae. Sorting out the history of these switches between latitudes and ocean basins has been hard to do, because comparisons among close relatives (congeners) were limited to just tropical or just temperate latitudes. Do tropical sea urchins give rise to temperate sister groups that can tolerate cold water and invade new ocean basins? Do those temperate lineages then give rise to new tropical species? Or are temperate species all recent descendants of various old tropical lineages that diverged from each other long before the formation of late Cenozoic barriers between tropical seas? Arbacia species provide an unusual testing ground for these ideas, because (unlike other sea urchin genera) members of this genus occur in both temperate and tropical waters and in three different ocean basins (Pacific, Atlantic, Mediterranean). Arbacia could be an evolutionary oddball consisting of very old species (Mayr 1954), or a taxonomic artefact (consisting of a nonmonophyletic mixture of tropical and temperate species) or the exception that proves a rule (and disproves some others). So, although Arbacia is a small genus of just six nominal species, a study of its phylogeographic history would return a disproportionate reward.

Carrying out such a study depends on thorough sampling of the geographic ranges of each species, including some that live in highly inaccessible habitats in the southern Atlantic and sub-Antarctic regions. Lessios and his many coauthors cooperatively assembled just such a thorough collection. By sampling all living species in all oceans for multiple parts of the mitochondrial genome, and rooting that gene tree with a closely related eastern Pacific species (Tetrapygus niger), Lessios et al. (2012) show that the earliest speciation event among extant Arbacia lineages did not occur in the distant past nor in the tropics (as previously proposed), but relatively recently (<5 Mya) in the temperate southeastern Pacific. This speciation gave rise to two lineages, both of which have temperate living descendants, including A. dufresni found in both the southern Pacific and Atlantic basins, and the lineage that gave rise to A. spatuligera and its closest relatives. Mapping latitudinal range onto the species tree (Fig. 1) suggests a series of evolutionary transitions from the temperate southern Pacific into the tropics (A. stellata) and then back to temperate latitudes (A. lixula, A. punctulata in the Atlantic), with some of the largest latitudinal differences between sister groups. Thus, Arbacia species are not evolutionary oddballs, and they form a good clade of close relatives, but their biogeographic history is not rooted in the tropics. Instead, tropical living evolved from a temperate ancestor (at least, that is the most parsimonious accounting of the species differences on the gene tree) and gave rise to new temperate species in turn (in a different ocean basin).


Figure 1.  A cladogram of Arbacia sea urchin species, which arose in temperate high latitudes (blue) in the southeast Pacific (A. dufresni) and diversified into subtropical (A. spatuligera, purple), tropical (A. stellata, red) and then new temperate lineages in the Atlantic (A. lixula, A. punctulata).

Download figure to PowerPoint

Understanding the ‘where’ and ‘when’ of Arbacia speciation leads the reader to a tempting series of speculative hypotheses regarding ‘how’ they diverged from each other. The large latitudinal differences between some sister groups (Fig. 1) might imply large ecological differences in their characteristic habitats (temperature and other types of physiological stress, food organisms, competitors, predators) and a source of selection for adaptations to different habitats. Such adaptive ecological divergence could occur in allopatry (without gene flow), while separated by geological or oceanographic barriers to dispersal, or in sympatry (and despite ongoing gene flow). Alternatively, allopatric divergence could involve no selection and merely reflect the effects of genetic drift and fixation of alternative alleles and phenotypes.

As readers of this journal know, the standard of evidence for distinguishing among these hypotheses—and for ascribing particular speciation events to ecological causes in sympatry—is high (Schluter 2009; Johannesson 2010). In the case of Arbacia, the current geographic distributions of most species (in allopatry) leave few opportunities for testing any of those alternative ‘how’ processes against each other. Lessios et al. (2012) conservatively (perhaps wisely) limit themselves to considering just one possible mechanism: reinforcement against hybrids in two zones of sympatry. They found a few strongly supported conflicts between the mitochondrial and nuclear gene trees that imply occasional hybridization between members of one species pair (A. dufresni and A. spatuligera) that have a narrow zone of sympatry in southern Chile. They also found extensive sharing of alleles at both loci between two other species (A. dufresni and A. ‘crassispina’) that are broadly sympatric in the southern Atlantic. This latter case could reflect either poor taxonomy in what is really a single species (A. dufresni) or poor barriers to hybridization between two species that are diverging, despite gene flow between them. Because the phenotypic differences between the two species are very slight, Lessios et al. (2012) favour the former interpretation (that the shared alleles and haplotypes are ancestral polymorphisms in a single species), but different methods are needed to rule out the possibility of more recent gene flow by hybridization as the source of allele sharing between two divergent species (Pinho & Hey 2010).

In neither of these cases, although, was there evidence for reinforcement. Bindin molecular evolution in sympatric hybridizing species is not characterized by higher rates of amino acid substitutions (relative to silent nucleotide substitutions) compared with other Arbacia lineages (which are allopatric and not evidently hybridizing). This is relatively weak evidence against reinforcement: sea urchin fertilization depends on other sperm–egg recognition molecules in addition to bindin (including an egg peptide pheromone that acts as a sperm chemoattractant and an egg surface glycoprotein that induces sperm activation; Lessios 2011), and evidence of reinforcement (positive selection in sympatry) at these other gamete recognition loci would be consistent with a hypothesis of secondary contact and reinforcement between A. dufresni and its sister group (including A. spatuligera).

So, in the end, we now understand where and when Arbacia speciation occurred, but not the cause(s) of those events. The patterns discovered by Lessios et al. (2012) suggest a series of ecological and reproductive comparisons between temperate, subtropical and tropical species (e.g. A. dufresni, A. spatuligera, A. stellata) that might shed light on possible sources of adaptive divergence between them (Oliver et al. 2010) and might explain apparent variation in localized hybridization between A. spatuligera and the two species with which it is partially sympatric at its southern (A. dufresni) and northern range ends (A. stellata) (McCartney & Lessios 2002). Keen readers will have to await the sequel to find out.


  1. Top of page
  2. Abstract
  3. References