Insights & Perspective Why coelacanths are not ‘living fossils’ Think again A review of molecular and morphological data Didier Casane1)2) and Patrick Laurenti1)2)Ã A series of recent studies on extant coelacanths has emphasised the slow certain extent orders now widely rate of molecular and morphological evolution in these species. These stud- separated in the natural scale. These anomalous forms may almost ies were based on the assumption that a coelacanth is a ‘living fossil’ that be called living fossils; they have has shown little morphological change since the Devonian, and they pro- endured to the present day, from posed a causal link between low molecular evolutionary rate and morpho- having inhabited a confined area, logical stasis. Here, we have examined the available molecular and and from having thus been exposed morphological data and show that: (i) low intra-specific molecular diversity to less severe competition. does not imply low mutation rate, (ii) studies not showing low substitution Since then, the term ‘living fossil’ has rates in coelacanth are often neglected, (iii) the morphological stability of a double meaning: it is a species which coelacanths is not supported by paleontological evidence. We recall that has no close living relatives, and which intra-species levels of molecular diversity, inter-species genome divergence underwent very few changes during the rates and morphological divergence rates are under different constraints and course of evolution. However, in his they are not necessarily correlated. Finally, we emphasise that concepts founding book, Darwin used ‘living fossil’ in two passing comments only. The term such as ‘living fossil’, ‘basal lineage’, or ‘primitive extant species’ do not was actually popularised eighty years make sense from a tree-thinking perspective. later after the discovery of Latimeria chalumnae [2], an extant species belong- Keywords: ing to the infraclass Actinistia that was, at .coelacanth; evolutionary rate; Latimeria; living fossil; slow evolution; substitution that time, exclusively known from the rate; tree-thinking fossil record (see [3] for review). This unexpected discovery made the extant coelacanths (L. chalumnae,andthemore recently discovered L. menadoensis [4]) Introduction been less severe than elsewhere; the paradigmatic example of a ‘living new forms will have been more slowly fossil’, thus defining them as an evolutio- The term ‘living fossil’, was coined by formed, and old forms more slowly narily conserved group of species that Charles Darwin in the first edition of exterminated. And it is in fresh water has evolved little over a geological time On the origin of species...[1], as follows: that we find seven genera of Ganoid scale. fishes, remnants of a once preponder- Transposing the concept of ‘living All fresh-water basins, taken together, ant order: and in fresh water we find fossil’ to the genomic level has led to make a small area compared with that some of the most anomalous forms the hypothesis of genetic stasis (or at of the sea or of the land; and, con- now known in the world, as the least to the idea of a reduced molecular sequently, the competition between Ornithorhynchus and Lepidosiren, evolutionary rate) that is in sharp con- fresh-water productions will have which, like fossils, connect to a trast with the principles of evolutionary genetics [5]. Genomes change continu- ously under the combined effects of DOI 10.1002/bies.201200145 various mutational processes, that pro- duce new variants, and genetic drift 1) Laboratoire Evolution, Ge´ nomes et Spe´ ciation, *Corresponding author: and selection, that eliminates or fixes UPR 9034 CNRS, Gif sur Yvette, France Patrick Laurenti 2) Universite´ Paris Diderot, Sorbonne Paris Cite´ , E-mail: [email protected] them in populations [6]. In other terms, France the only possibility for genomes to rep- 332 www.bioessays-journal.com Bioessays 35: 332–338,ß 2013 WILEY Periodicals, Inc. .....Insights & Perspective D. Casane and P. Laurenti licate without change implies at least canth population is situated, the popu- mechanisms that are under selective one of the two following conditions: lation size is about 300–400 adult pressure. The mutation rate is the result (i) new variants do not appear (i.e. no individuals [13]. Even if larger popu- of a trade-off between the need for Think again mutations), and (ii) new variants are lations of coelacanths existed in the new variants to adapt to environmental systematically eliminated by selection past, the small sizes of extant popu- changes and selection for a reduced (i.e. no genetic drift and very powerful lations are sufficient to explain low mutation rate because mutations are selection against new variants). Of course genetic diversity in this species. The more often deleterious than advan- we can consider a less extreme case, geographic differentiation of popu- tageous [15]. The substitution rate i.e. a reduced evolutionary rate of the lations depends on abiotic factors such depends chiefly on fixation of neutral genome, but this still implies a lower as geographic barriers (e.g. land masses and slightly deleterious mutations [16] mutation rate and/or stronger selection and ocean currents for marine fish) and and it is thus positively correlated with against new variants than observed in biotic factors such as migration (e.g. the mutation rate and negatively correlated other species. existence of planktonic larvae and with population size because selection However, the long held idea that adult migration). L. chalumnae and is more efficient in large populations coelacanths are ‘living fossils’ has been L. menadoensis are viviparous species [17]. An unexpectedly low substitution recently revitalised following several that live in the deep ocean and give birth rate could thus be the result of a molecular studies that concluded that to juveniles about 30 cm in length [3]. peculiarly low mutation rate and/or there is low molecular diversity within The low genetic differentiation found in an especially strong selection against populations of L. chalumnae [7] and low Latimera could thus be the result of the deleterious mutations [16, 17]. substitution rates [8–10] in the lineage absence of geographic barriers, and the Molecular support for the hypothesis leading to the two extant coelacanths. absence of small pelagic larvae could that L. chalumnae and L. menadoensis A causal link between low molecular explain the low impact of ocean cur- have slowly evolving genomes comes evolutionary rate and morphological rents on population isolation. chiefly from two studies on HOX gene stasis was systematically hypothesised. Interestingly, although the two clusters [8, 9]. This is supported by an In order to re-evaluate the pace of available analyses of L. chalumnae analysis of the L. menadoensis sonic genomic and anatomical evolution in natural population genetics both hedgehog (shh) locus that has low substi- coelacanths, here we review the avail- detected low allelic diversity and low tution rates [10]. Assuming that the Hox able molecular and morphological data levels of geographic differentiation, gene clusters and shh are evolving for this group. We question whether the they came to different conclusions. particularly slowly, does this indicate data actually support the hypotheses of The first study [11] concluded that the a general trend for the whole nuclear a slowly evolving genome and morpho- low allelic diversity is the result of a genome? Analyses of three nuclear logical stability in Actinistia. small population size and the absence genes (protocadherin clusters [18], vitel- of strong geographic isolation, as has logenin genes [19], and nuclear-encoded been concluded for similar findings in recombination activating genes [19, 20]) Low molecular diversity other species [14]. In contrast, the more and two of the whole mitochondrial and low geographic recent study, Lampert et al. [7], rejected genome [21, 22] are often quoted in sup- these straightforward hypotheses. On port of the hypothesis of slow evolution. differentiation in the assumption that coelacanths are However, a closer look at the data chal- coelacanths ‘living fossils’, the authors concluded lenges this interpretation: depending on that their data ‘confirm the assumed the analysed sequence, the coelacanth Two population genetics studies in slow rate of molecular evolution in branch is not systematically shorter Latimeria chalumnae are available and coelacanths’. As we found this con- than the branches leading to other both show low allelic diversity and low clusion surprising, we questioned to species. In addition, most phylogenetic levels of geographic differentiation what extent low mutation rates have analyses – including analysis of Hox [7, 11]. For neutral alleles, diversity been detected in Latimeria. sequences [23–26] – do not support depends on two opposite forces, the rate the hypothesis that the Latimeria of mutation that is the source of new genome is slow evolving, i.e. they do alleles and the genetic drift that elimin- Is the genome of the not place coelacanth sequences on short ates alleles. The effect of genetic drift branches nor do they detect low substi- depends on the population size, i.e. ran- coelacanth slowly tution rates [24, 27–31]. The clearest dom changes in allele frequencies are evolving? example, which involves the largest more important in small populations number of genes, is a phylogeny based and rare alleles are lost more rapidly The rate of molecular evolution in a study of forty-four nuclear genes that [6] (see also [12] for recent review). lineage, the substitution rate, depends does not show a dramatic decrease, if Thus, for a given mutation rate, the on the mutation rate and the rate of any, in the rate of molecular evolution in genetic diversity is positively correlated fixation of new variants, which itself the coelacanth lineage [32]. What we with the population size: smaller popu- depends on the selective values of these know about the biology of coelacanths lations are expected to show a lower new variants and genetic drift [6].