Molecular Phylogenetics and Evolution 120 (2018) 212–217 Contents lists available at ScienceDirect Molecular Phylogenetics and Evolution journal homepage: www.elsevier.com/locate/ympev Short Communication Updated molecular phylogeny of the squid family Ommastrephidae: Insights T into the evolution of spawning strategies ⁎ M. Cecilia Pardo-Gandarillasa, Felipe I. Torresa,b, Dirk Fuchsc, Christian M. Ibáñezb, a Departamento de Ciencias Ecológicas, Facultad de Ciencias, Universidad de Chile, Las Palmeras 3425, Ñuñoa, Santiago, Chile b Departamento de Ecología y Biodiversidad, Facultad de Ecología y Recursos Naturales, Universidad Andres Bello, República 440, Santiago, Chile c Department of Natural History Sciences, Hokkaido University, Sapporo, Hokkaido, Japan ARTICLE INFO ABSTRACT Keywords: Two types of spawning strategy have been described for ommastrephid squids: coastal and oceanic. It has been Cephalopoda suggested that ancestral ommastrephids inhabited coastal waters and expanded their distribution into the open Evolution ocean during global changes in ocean circulation in the Oligocene. This hypothesis could explain the different Reproduction reproductive strategies in oceanic squids, but has never been tested in a phylogenetic context. In the present Ancestral states study, we assess the coastal-to-open-ocean hypothesis through inferring the evolution of reproductive traits Divergence times (spawning type) of ommastrephid squids using the phylogenetic comparative method to estimate ancestral states and divergence times. This analysis was performed using a robust molecular phylogeny with three mitochondrial genes (COI, CYTB and 16S) and two nuclear genes (RHO and 18S) for nearly all species of ommastrephid squid. Our results support dividing the Ommastrephidae into the three traditional subfamilies, plus the monotypic subfamily Todaropsinae as proposed previously. Divergence times were found to be older than those suggested. Our analyses strongly suggest that early ommastrephid squids spawned in coastal areas, with some species subsequently switching to spawn in oceanic areas, supporting previous non-tested hypotheses. We found evi- dence of gradual evolution change of spawning type in ommastrephid squids estimated to have occurred since the Cretaceous. 1. Introduction species would be characterized by long periods of stasis followed by short punctuated bursts of evolution associated with speciation Environmental changes in geological and ecological dimensions (Eldridge and Gould, 1972; Pagel et al., 2006). Phylogenetic evidence have been proposed to promote the evolution of species across a broad attributes great importance to punctuated models of evolution in pro- spectrum of biological aspects (e.g., life histories, distribution, specia- moting rapid evolutionary divergence in different organisms (Pagel lization, speciation, extinction) (Dynesius and Jansson, 2000; Schoener, et al., 2006). To assess the effects of gradual evolution on certain traits, 2011). In the marine environment, the evolution of life forms from some sophisticated phylogenetic approaches have been developed to benthic to nektonic reflects a progressive evolution related to sea level scale the relationship between individual phylogenetic branch lengths changes in different modern taxonomic groups (e.g., fishes, cephalo- and trait evolution (e.g., Pagel et al., 2006). pods, reptiles and mammals; Nesis, 1978). In modern cephalopods, Squids of the family Ommastrephidae are the most abundant, Lindgren et al. (2012) reported convergent evolution of multiple phe- widely distributed, ecologically active cephalopods (Wormuth, 1998). notypic traits correlated with habitat type. Nigmatullin (1979, 2007) Traditionally the family has been divided into three subfamilies (Om- proposed the hypothesis that oceanic and coastal spawning in om- mastrephinae, Todarodinae and Illicinae) based on morphological mastrephid squids evolved in relation to environmental changes during traits, such as the structures of the funnel groove, morphology of the the Oligocene – Miocene transition and the radiation of scombrid fishes. club suckers and hectocotylus, and presence and placement of photo- These evolutionary patterns suggest the influence of gradual change as phores (Roeleveld, 1988; Wormuth, 1998). Although due to peculia- predicted by phyletic gradualism theory, but in some cases the observed rities in morphological traits it has been proposed creating two addi- rate of evolution is faster than expected according to this theory tional subfamilies Ornithoteuthinae (Nigmatullin, 1979) and (Eldridge and Gould, 1972). Rapid evolution is related to punctuated Todaropsinae (Nigmatullin, 2000) to place Ornithoteuthis and Todaropsis equilibrium, a theory that suggests evolutionary divergence among respectively, the three-subfamily classification prevails (Wormuth, ⁎ Corresponding author. E-mail addresses: [email protected], [email protected] (C.M. Ibáñez). https://doi.org/10.1016/j.ympev.2017.12.014 Received 16 February 2017; Received in revised form 3 November 2017; Accepted 11 December 2017 Available online 14 December 2017 1055-7903/ © 2017 Elsevier Inc. All rights reserved. M.C. Pardo-Gandarillas et al. Molecular Phylogenetics and Evolution 120 (2018) 212–217 1998; Wakabayashi et al., 2012). According to their spawning location, the substitution model GTR+G+I. The clock rate was estimated using ommastrephid squids can be classified into coastal and oceanic. Coastal a relaxed clock with an uncorrelated lognormal distribution. The Birth- spawning (on the continental shelf) is common among the subfamilies Death model served as the tree prior, and parameters were logged every Illicinae and Todarodinae, it is characterized by a long maturation 1000 iterations, to a sample total of 20,000,000 generations. Burn-in period without somatic growth, and intermittent spawning with de- was determined by Tracer, accounted for the initial 10% of the first creasing intensity (with potential fecundity of 0.04–2.5 million, estimation, and was discarded from calculations. The best evolution Nigmatullin and Laptikhovsky, 1994; Nigmatullin, 2007). Notable ex- model of the molecular clock (strict or relaxed) was evaluated by ceptions however include the two species of Ornithoteuthis (according to comparing Bayes Factors, adjusting the molecular clock using four ca- Wakabayashi et al., 2012 a member of the Todarodinae), these are libration points. The first one was the split between Decapodiformes known to spawn far from the shore, as O. antillarum which has been and Octopodiformes (Carninan, 236 Mya; Kröger et al., 2011) with a recorded spawning around seamounts (Nigmatullin and Laptikhovsky, gamma distribution and the following parameters: offset = 236, 1994). Oceanic spawning is characteristic of Ommastrephinae, which shape = 2.0, scale = 3.0. The second calibration point was the diver- display quick maturation with somatic growth and multiple spawning gence of Vampyromorpha and Octopodiformes using genus Teudopsis over several months with relatively stable intensity (potential fecundity from the Lower Jurassic (Toarcian, 182 Mya; Fuchs and Weis, 2008), 0.1–32 million, Nigmatullin and Laptikhovsky, 1994; Nigmatullin, The third calibration point used was the origin of Myopsida using lo- 2007). Nigmatullin (1979, 2007) suggesting that ancestral ommas- liginid statoliths from the mid-Eocene (48 Mya; Neige et al., 2016) with trephids used to inhabit coastal waters and then expanded their ranges gamma distribution and the following parameters: offset = 48, to open ocean during the Oligocene due to global ocean circulation shape = 2.0, scale = 1.0, and the final one was the split of northern and changes. This historical change of habitat could explain the origin of southern Atlantic (USA vs Brazil) genetic lineages of Doryteuthis pleii: different spawning strategies for oceanic and coastal squids offset = 16 Mya (Sales et al., 2017), shape = 2.0, scale = 1.0. (Nigmatullin, 2007), implicating that there may have been a gradual The evolution of the reproductive type (oceanic and coastal evolution of reproductive strategies on a par with slow changes in ha- spawning) was evaluated using the multistate model on BayesTraits v.3 bitat. Although insightful, the ideas proposed by Nigmatullin (1979, (http://www.evolution.rdg.ac.uk/BayesTraitsV3/BayesTraitsV3.html). 2007) were never statistically tested. The evolution of a binary trait (0 = coastal and 1 = oceanic) was The main purpose of the present study is to test previous ideas on modelled estimating two parameters, a rate of change from 0 to 1 (q01) the evolution of ommastrephid spawning by using quantitative tech- and a rate of change from 1 to 0 (q10), assuming these two rates are niques for the first time. We apply the phylogenetic comparative equal. In order to estimate these rate of change, we used the informa- method which is based on a robust Bayesian molecular phylogeny with tion about the distribution of 0 s and 1 s at the tips of the trees (i.e. for mitochondrial and nuclear genes for 18 out of the 22 extant ommas- each species), and the branch lengths of these trees. Through a Bayesian trephid species. framework the rate of gain (q01) and loss (q10) of the spawning type was estimated in each branch of the 1000 phylogenetic random trees 2. Materials and methods selected from a sample of trees, using BayesTrees v.1.3 (http://www. evolution.reading.ac.uk/BayesTrees.html). A beta hyperprior was used Through a literature review we acquired information