Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

Syst. Biol. 69(3):413–430, 2020

© The Author(s) 2019. Published by Oxford University Press on behalf of the Society of Systematic Biologists. This is an Open Access article distributed under the terms of the Creative
Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium,
provided the original work is properly cited. For commercial re-use, please [email protected]
DOI:10.1093/sysbio/syz059
Advance Access publication September 5, 2019

Macroevolutionary Analyses Suggest That Environmental Factors, Not Venom Apparatus,

Play Key Role in Terebridae Marine Snail Diversification
MARIA VITTORIA MODICA1,2,† , JULIETTE GORSON3,4,5,6,† , ALEXANDER E. FEDOSOV7 , GAVIN MALCOLM8 , YVES TERRYN9 ,
NICOLAS PUILLANDRE9 , AND MANDË HOLFORD3,4,5,6,∗
1 Department of Biology and Evolution of Marine Organisms, Stazione Zoologica Anton Dohrn, Villa Comunale, 80121 Naples, Italy; 2 UMR5247,

Université de Montpellier CC 1703, Place Eugène Bataillon 34095 Montpellier, France; 3 Department of Chemistry, Hunter College Belfer Research Center,
413 E. 69th Street, BRB 424, New York, NY 10021, USA; 4 Department of Biochemistry, Weill Cornell Medical College, Cornell University, New York, NY
10021, USA; 5 Division of Invertebrate Zoology, The American Museum of Natural History, New York, NY 10024, USA; 6 Programs in Biology,
Biochemistry, and Chemistry, The Graduate Center, City University of New York, New York, NY 10016, USA; 7 Institute of Ecology and Evolution of
Russian Academy of Sciences, Leninskiy Prospect, 33, Moscow 119071, Russia; 8 Bird Hill, Barnes Lane, Milford on Sea, Hampshire, UK; and 9 Institut
Systématique Evolution Biodiversité (ISYEB), Muséum national d’Histoire naturelle, CNRS, Sorbonne Université, EPHE, Université des Antillles, 57 rue
Cuvier, CP 26, 75005 Paris, France
† Maria Vittoria Modica and Juliette Gorson contributed equally to this article.
*Correspondence to be sent to: Department of Chemistry, Hunter College Belfer Research Center, 413 E. 69th Street, BRB 424, New York, NY 10021, USA;
E-mail: [email protected].

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

Received 09 November 2018; reviews returned 18 August 2019; accepted 19 August 2019
Associate Editor: Jeanne Serb

Abstract.—How species diversification occurs remains an unanswered question in predatory marine invertebrates, such as
sea snails of the family Terebridae. However, the anatomical disparity found throughput the Terebridae provides a unique
perspective for investigating diversification patterns in venomous predators. In this study, a new dated molecular phylogeny
of the Terebridae is used as a framework for investigating diversification of the family through time, and for testing the
putative role of intrinsic and extrinsic traits, such as shell size, larval ecology, bathymetric distribution, and anatomical
features of the venom apparatus, as drivers of terebrid species diversification. Macroevolutionary analysis revealed that
when diversification rates do not vary across Terebridae clades, the whole family has been increasing its global diversification
rate since 25 Ma. We recovered evidence for a concurrent increase in diversification of depth ranges, while shell size appeared
to have undergone a fast divergence early in terebrid evolutionary history. Our data also confirm that planktotrophy is the
ancestral larval ecology in terebrids, and evolutionary modeling highlighted that shell size is linked to larval ecology of the
Terebridae, with species with long-living pelagic larvae tending to be larger and have a broader size range than lecithotrophic
species. Although we recovered patterns of size and depth trait diversification through time and across clades, the presence
or absence of a venom gland (VG) did not appear to have impacted Terebridae diversification. Terebrids have lost their venom
apparatus several times and we confirm that the loss of a VG happened in phylogenetically clustered terminal taxa and that
reversal is extremely unlikely. Our findings suggest that environmental factors, and not venom, have had more influence
on terebrid evolution. [Terebridae; macroevolution; phylogenetic comparative methods; venom; Conidae; diversification.]

Explaining the amazing biodiversity of species that clades (Rabosky et al. 2013; Rainford et al. 2014; Sánchez-
inhabit our planet remains a significant challenge. García and Matheny 2017). In addition, environmental
With the exception of a few well-known taxa, such as modifications may create new ecological opportunities
vertebrates or angiosperms, current hypotheses about for specific clades, through the availability of new
diversity patterns remain largely untested across the habitats or the extinction of predators or competitors
majority of Earth’s biodiversity (Jetz et al. 2012; Pyron (Harmon et al. 2008; Parent and Crespi 2009; Des Roches
and Burbrink 2012; Rainford et al. 2014; Legendre et al. 2011).
and Condamine 2018). This is especially true for Many marine organisms rely on the production of
marine invertebrates, in which their basic biology, venomous secretions to deter predators or subdue
diversification patterns, and evolutionary dynamics preys. The onset of a venom system, made up of
remain largely unknown. Several hypotheses proposed specialized glands and delivery structures such as
to explain diversity patterns focus on key innovations beaks, fangs, harpoons, spines, or pincers, is considered
that affect the adaptation of organisms to their an opportunistic innovation that favors speciation of
predators by enabling the exploitation of new ecological
environment. The innovations can be derived from
niches characterized by different potential prey species
intrinsic factors like morphology, physiology, behavior, (Vidal and Hedges 2005; Fry et al. 2006; Castelin et al.
ecology, or from extrinsic environmental factors, such 2012). Venom plays a crucial role in prey capture and
as depth and temperature (Benton and Harper 2009; survival, which makes it a potential key innovation, as
Yoder et al. 2010; Ng and Smith 2014; Wiens 2017). The also suggested by its convergent evolution in multiple
acquisition of key innovations is proposed to lead to lineages (Barlow et al. 2009; Casewell et al. 2013). The
faster diversification rates either by increasing speciation components of venom are often encoded by rapidly
rates or by decreasing extinction rates, which may evolving gene families (Kordis and Gubensek 2000; Fry
account for differences in species richness between et al. 2009; Casewell et al. 2013), suggesting a strong

413

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 413 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

414 SYSTEMATIC BIOLOGY VOL. 69

diversifying selective pressure on venom composition. (lecithotrophy; Thorson 1950). Although planktotrophic
However, the hypothesis that venom production may larvae can spend a considerable time in the water column
affect diversification has only been examined in a few (typically weeks or months), lecithotrophic larvae have
cases, mostly in vertebrates or terrestrial invertebrates a shorter pelagic phase due to the limited yolk reserve
and is generally targeted at the species level using and consequently tend to have reduced dispersal kernels
indirect evidence (Daltry et al. 1996; Fry et al. 2008; (Shanks 2009). The duration of pelagic larval phase has
Duda et al. 2009). For example, in snakes, which exhibit been demonstrated to influence genetic connectivity in
exceptional species richness, it is proposed that the gastropods (Collin 2001; Wright 2002; Modica et al. 2017),
majority of the diversity stems from an early radiation through dispersal ability, suggesting that the acquisition
within the superfamily Colubroidea, possibly due to the of lecithotrophy may lead to increased speciation rates by
evolution of venom delivery systems that allowed the reducing gene flow between populations (Harvey et al.
colonization of new areas (Pyron and Burbrink 2012). 2017).
Marine snails belonging to the superfamily A relationship has been proposed between
Conoidea are among the most prominent marine diversification and abiotic factors such as habitat
venomous lineages. To date extensive toxinological complexity, sea temperature, sea level, ocean
and phylogenetic investigations have focused almost productivity, and oxygen content, for different lineages
exclusively on Conus species, neglecting other related of marine organisms (Figueirido et al. 2011; Stein et al.

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

lineages, including the Terebridae or auger snails 2014; Davis et al. 2016; Costello and Chaudhary 2017;
(Holford et al. 2009a; Puillandre et al. 2011; Castelin Stigall 2017; Lewitus et al. 2018; Rabosky et al. 2018).
et al. 2012). Terebrids demonstrate a high level of Indeed, depth has been identified as a diversification
morphological disparity in feeding-related traits, in shell driver in several lineages of marine fish (Ingram 2011;
size range, and ecological diversity, providing a basis Sorenson et al. 2014; Gaither et al. 2016). Given terebrids
for investigating the role of such traits as diversification have a broad span of bathymetric distribution globally
drivers. The more than 400 described terebrid species in subtropical and tropical oceans, where they have
display anatomical disparity in the foregut comparable been found on the shore line as well as at depths >700
with the entire Conoidea superfamily (Miller 1971; Mills m (Taylor 1990; Terryn 2007; Terryn and Holford 2008),
1979; Castelin et al. 2012). The terebrid foregut has been depth is another important factor to investigate for
shaped by multiple losses of key anatomical structures influence on terebrid diversification.
such as the venom gland (VG) and proboscis, as well as In this study, we reconstruct the first dated terebrid
by the convergent evolution of the main venom delivery phylogeny with a three-fold increase in number of
structure, the hypodermic radula, in three lineages specimens analyzed from prior efforts and use this
(Castelin et al. 2012). Given this remarkable variation, tree to carry out a phylogenetic comparative analysis
we examined if morphological traits pertaining to the of morphological and life-history traits, along with
use of venom may have affected terebrid evolution. bathymetric distribution, and their association to
Recognizing that not all terebrids have a venom diversification regimes in terebrid marine snails (Fig. 1).
apparatus, we also examined the role of additional We separately evaluate support for the hypothesis that
biotic and abiotic traits pertaining to shell size, larval the venom apparatus, shell size, larval development, and
ecology, and depth in driving diversification of the depth have facilitated diversification in marine snails.
Terebridae. Similar to foregut anatomy, shell size
displays a high level of diversification in Terebridae,
which in adult specimens can range from 15 to 230 mm
(Taylor 1990; Terryn 2007; Terryn and Holford 2008). MATERIALS AND METHODS
Body size influences multiple aspects of organismal
morphology, physiology, life history, and ecology, and Sample Collection
may dramatically affect behavior and extinction rates. All of the materials used in this study were
The relationship between body size and diversification collected during several expeditions conducted by
rates is mostly unresolved and has been confirmed only the Museum National d’Histoire Naturelle of Paris
in a few cases (Knouft and Page 2003; Fontanillas et al. (MNHN—www.expeditions.mnhn.fr) and the Holford
2007; Rabosky et al. 2013). However, most studies failed Laboratory. The data set includes 1275 specimens
to identify a clear effect of size on lineage diversification collected from 25 localities with a focus on the Indo-
(Gittleman and Purvis 1998; Owens et al. 1999; Rainford Pacific province (Supplementary Table S1 available
et al. 2014; Feldman et al. 2016; Lee et al. 2016). In on Dryad at http://dx.doi.org/10.5061/dryad.j008j5h).
terebrids, diversification of shell size might both affect Samples were collected from 0 m to approximately 800
speciation rates allowing access to multiple trophic m in depth and specifically fixed for molecular analysis
niches and influence the extinction risk through a in the field. Live specimens were anesthetized using
balance between the higher metabolic expenditure and magnesium chloride (MgCl2 ) isotonic with seawater,
the differential susceptibility to predation. and a piece of tissue was cut from the foot and fixed
As in other marine gastropods, terebrids can produce in 95% ethanol. Specimens collected after 2012 were
pelagic larvae that either actively feed on phytoplankton processed with a microwave oven to facilitate removal
(planktotrophy) or rely exclusively on yolk reserves of soft tissue from the shell (Galindo et al. 2014). The

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 414 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

2020 MODICA ET AL.—DIVERSIFICATION AND EVOLUTION OF THE TEREBRIDAE 415

model. Robustness of the nodes was assessed using

the thorough bootstrapping algorithm (Felsenstein 1985)
with 1000 replicates. Phylogenies were jointly estimated
using the Bayesian Markov Chain Monte Carlo method
implemented in BEAST version 1.8.4 (Drummond and
Rambaut 2007). The program BEAUti version 1.8.4
(Drummond and Rambaut 2007) was used to generate
the file used in BEAST. A birth–death process speciation
prior was implemented and the substitution models
identified in jModelTest2 version 2.1.6 were applied to
each gene independently. An uncorrelated lognormal
clock was applied to estimate the relaxed molecular
clock. The analysis ran for 75 million generations and
sampled every 1000 generations. The oldest known
Terebridae, Mirula plicata (Lamarck, 1803) from the
lower Eocene (56.0–47.7 Ma) was used to constrain the
stem node of Terebridae with a normal distribution

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

mean of 50.7 Ma and a standard deviation (SD) of
FIGURE 1. Examination of environmental, ecological, and 1.48 (Abdelkrim et al. 2018). A burn-in of 10% was
morphological traits to determine factors driving evolution and removed after convergence analysis was evaluated using
diversification in the Terebridae. One thousand seven hundred and Tracer version 1.7 (Drummond and Rambaut 2007)
sixty-one (1761) specimens of Terebrids were collected globally and
sequenced using a multigene strategy to reconstruct a phylogenetic
to check that all effective sample size (ESS) values
hypothesis that was dated using input from the fossil record, and were > 200. Analyses were performed on the Cipres
subsequently used to infer diversification patterns for the family. Science Gateway (http://www.hylo.org/portal2), using
Disparities in size, larval ecology, depth, and presence or absence of the RAxML-HPC2 on XSEDE tool for ML and the BEAST
the venon gland were evaluated to determine their impact on terebrid on XSEDE tool for BA.
diversification rates.

majority of shells was kept intact for identification Shell Size Measurements
and deposited as vouchers in MNHN and the Holford Shell sizes were determined for 325 intact adult
laboratory. The taxonomy of the family Terebridae specimens representing 137 species of our data
was reworked based on the new phylogeny provided set. Reliability and species-level representativeness of
in this study. The nomenclature for new taxa and these measurements were checked against size ranges
revised classification of Terebridae based on the recently published by Bratcher and Cernohorsky (1987) or in the
portrayed relationships is followed (Fedosov et al. 2019). original descriptions of the shells. For trait-dependent
diversification analyses, shell size was converted into
a binary categorical trait with the states “small” and
DNA Sequencing and Molecular Phylogenetic Analyses “large,” following a shell size partitioning that was
obtained as follows: From the species present in our DNA
Total genomic DNA was extracted from foot tissue
data set, we calculated the lowest 25% quartile for species
using NucleoSpin® 96 Tissues (Macherey-Nagel) or
size and adopted 25 mm, which accommodated 30% of
the Epmotion 5075 robot (Eppendorf), following
the species, as the boundary for the categorical size trait
the manufacturer’s protocol. Fragments of three
with the states of “small” or “large” for each species.
mitochondrial genes [Cytochrome Oxidase I (COI), 16S
Each measurement was confirmed against published
rRNA, and 12S rRNA] and one nuclear gene (28S
information regarding shell size to ensure that the
rRNA) were amplified. PCR reactions were performed
allocation to the small or large species category was
as described in Holford et al. (2009a). Successfully
reasonably valid.
amplified products were sent to Genewiz (South
Plainfield, NJ, USA) or to the Eurofins sequencing facility
(France) for bidirectional Sanger sequencing.
Sequences were aligned for each gene independently Larval Ecology
using MUSCLE version 3.2 (Edgar 2004). The accuracy In Terebridae, as in many other families of marine
of these alignments was manually inspected using gastropods, larval ecology can be easily inferred from
BioEdit version 7.0.0.0 (Hall 1999). Best-fit substitution the appearance of protoconch, the larval shell that is
models were identified for each gene separately using often maintained at the tip of adult shell (Jablonski and
jModelTest2 version 2.1.6 (Posada 2008). Best-scoring Lutz 1983; Lima and Lutz 1990; Eldredge et al. 2005).
maximum likelihood (ML) trees were estimated using Depending on the protoconch appearance, species are
RAxML (Stamatakis 2006, 2014). Each gene, and each defined as planktotrophic, that is possessing a pelagic
codon position within the COI gene, was considered free swimming stage during which the veliger larva can
as independent, each following its best-fit substitution actively collect phytoplankton, when the protoconch is

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 415 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

416 SYSTEMATIC BIOLOGY VOL. 69

multispiral, or lecithotrophic, relying on yolk reserves for samples due to technical limitations (Bouchet et al. 2008,
survival until metamorphosis (Thorson 1950), when the 2009).
protoconch is paucispiral. The protoconchs of 638 intact
terebrid shells were examined under a microscope and
categorized as multi- or paucispiral, and the number of Species Delimitation and Species Diversity Estimations
whorls present was counted to the nearest quarter whorl
(Bouchet and Kantor 2004). All samples were first identified morphologically.
Then, independent gene trees were used to confirm
that conspecific samples were all included in a single
clade, separated by genetic distances compatible with
Foregut Anatomy intraspecific distances (i.e. inferior to genetic distances
The anatomy of the terebrids was studied by manual among species).
dissections—when possible, on the same specimens To estimate total Terebridae diversity, we used the
sequenced for phylogeny. As most informative Chao1 estimator (Gotelli and Chao 2013):
morphological characters in Conoidea are related  
to feeding, we specifically focused on the anterior SChao1 = Sobs +f12 / 2f2
alimentary channel structures to infer ability of the where Sobs is the observed species richness, and f1 and

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

Terebridae lineages to envenomate their preys. Manual f2 the number of, respectively, singletons (species found
dissections were complemented by SEM studies of only once in the study area) and doubletons (species
radular morphology, known to be extremely diverse found twice).
in the Terebridae. When present, radular sacs were As the overall sampling effort has been uneven with
isolated, and soft tissues immersed in a 3–5% solution respect to the worldwide distribution of Terebridae, we
of commercially available bleach. The radulae were then used a two-steps strategy to estimate global Terebridae
rinsed several times in distilled water, mounted on a biodiversity. First, we calculated the SChao1 for the Indo-
12-mm SEM stub, air-dried, gold-coated, and examined Pacific subset of our Terebridae data set, because it
using a TeScan TS5130MM microscope at the Joint corresponds both to a biodiversity hotspot for molluscan
Usage Center “Instrumental methods in ecology” at the fauna and to the most densely sampled area, obtaining
Institute of Ecology and Evolution of Russian Academy the estimated Indo-Pacific diversity. We then calculated
of Sciences (IEE RAS). the ratio of the estimated Indo-Pacific diversity to the
sampled Indo-Pacific diversity, a measure of how well
our sampling reflects the real diversity for that specific
Bathymetric Distributions area. Assuming that the effectiveness of our sample is
the same worldwide (which is reasonable given that
To calculate the bathymetric range for each species, all both diversity and sampling effort are lower outside
the individual specimens had a depth range recorded the Indo-Pacific), we applied the same ratio to the
at the time of collection giving the maximum and total number of Terebridae species described in WoRMS
minimum depth of the dredge/dive at its collection (WoRMS Editorial Board 2018). Finally, we added to
station. If a station was sampled at a constant depth, the the estimate the number of newly delimited species
same depth value was adopted as both the maximum from this study, to derive the total estimated Terebridae
and minimum depth for the specimen. For each biodiversity. The same approach was applied to estimate
species with multiple specimens recorded, we adopted the number of Terebridae species presenting alternate
a minimum depth for the species based on the lowest character state for depth, size, and larval ecology, except
maximum depth at any collecting station for a specimen that the ratio was calculated between the number of
of that species. This approach allowed us to be certain Indo-Pacific species presenting, for example state 0 and
that at least one specimen of the species was found at the total number of Indo-Pacific species for which we
that depth or shallower. Likewise a maximum depth for had available information (state 0 + state 1). The ratio
the species was adopted based on the highest minimum relative to state 0 and state 1 was then applied to the
depth of all the specimens of the species. The resulting total Terebridae diversity estimated as described above.
range of depth can, therefore, be considered as a reliable
but a minimal value. This algorithm was implemented
in an in-house Python script to quickly analyze large
data sets of species occurrences (Supplementary Fig. S1 Diversification Rates through Time and Across Clades
available on Dryad). For trait-dependent diversification Macroevolutionary dynamics of diversification
analyses, depth was converted into a binary categorical were modeled across the Terebridae phylogeny (after
trait with the two states “shallow” and “deep” using a outgroup removal) using the software Bayesian
100-m threshold. The use of this depth threshold value Analysis of Macroevolutionary Mixtures (BAMM)
roughly corresponds on average to the end of the photic v.2.5.0 (Rabosky et al. 2013; Rabosky et al. 2014) on the
zone and is in agreement with previous publications on Maximum Clade Credibility tree obtained in BEAST.
marine gastropods, and represents a zone for which it BAMM explores models of lineage diversification
is generally observed a drop in the number of collected implementing a Metropolis Coupled Markov Chain

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 416 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

2020 MODICA ET AL.—DIVERSIFICATION AND EVOLUTION OF THE TEREBRIDAE 417

Monte Carlo (MC3) to improve the efficiency in species were defined as shallow (0) when found above
simulating the posterior probability distribution. Ten 100 m or deep water (1) below 100 m; and 4) size,
million generations of reversible jump Markov Chain where species were identified as either small (0) for
Monte Carlo sampling were run, drawing samples shell length lower than 25 mm or large (1) for lengths
from the posterior every 10,000 generations. Priors exceeding 25 mm. Continuous traits were transformed
were chosen using the setBAMMpriors command into categorical two-state traits using appropriate
in the R package BAMMtools (Rabosky et al. 2014), thresholds as described above. We applied the Binary
except for the prior probability of rate shift, which State Speciation and Extinction model (BiSSE; Maddison
has been shown to affect BAMM results (Moore et al. et al. 2007) for the four two-states data sets, accounting
2016; Rabosky et al. 2017). For this prior, we tested for state-specific incomplete taxon sampling, estimated
values ranging from 0.1 to 50 and we chose the value based on our data as detailed in the Supplementary
leading to the highest ESS values for LogLikelihood and Materials available on Dryad. The BiSSE model has six
NumberOfShifts (Supplementary Table S2 available on distinct parameters: two speciation rates, two extinction
Dryad). We accounted for incomplete taxon sampling rates, and two transition rates (i.e. anagenetic change)
using a sampling fraction of 26%, estimated using a between the trait states. Analyses were performed using
total Terebridae diversity value obtained as described the R package diversitree (Fitzjohn 2012) on the MCC tree
above. We processed the output data using BAMMtools obtained from BEAST, using the functions make.bisse to

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

to obtain summary statistics after removing a 10% construct the likelihood functions for each model based
burn-in, and to plot diversification rate through time. on the data, and the functions constrain and find.mle to
BAMM was used both to estimate diversification rates apply different diversification scenarios (Supplementary
through time and among/within clades, and to define Table S4 available on Dryad). We used AIC to select
diversification rates for continuous traits (depth and among different models: the scenario supported with
size) using the same parameters. the lowest AIC was considered the best when AIC> 2
To corroborate BAMM results we used the time- and AIC ω > 0.5 against other models.
dependent diversification approach implemented in
the R package RPANDA (Morlon et al. 2016). This
approach enables both speciation and extinction to
change through time, whereas in BAMM the extinction Phylogenetic Signal and Phylogenetic Diversity
rates are assumed to be constant, thus allowing scenarios We compared the phylogenetic signal of the
in which diversification rates are negative (Morlon phenotypic traits taken into consideration (venom
et al. 2011). For the whole Terebridae tree (with a apparatus, shell size, larval development, and depth)
26% sampling fraction), we tested with RPANDA six using different metrics for the different type of
nested diversification models: 1) a Yule model, with characters. For continuous traits (size and depth), we
a constant speciation rate and null extinction, 2) a calculated Pagel’s  using the function phylosig in the
constant birth–death model, with constant speciation R package Phytools: a  = 0 indicates a trait is random
and extinction rates, 3) a variable speciation rate with respect to phylogeny (i.e., there is no phylogenetic
model without extinction, 4) a variable speciation signal), whereas a  = 1 is consistent with a trait that
rate model with constant extinction, 5) a rate-constant has evolved according to the Brownian motion model
speciation and variable extinction rate model, and (Freckleton et al. 2002). For binary discrete traits (VG
6) a model in which both speciation and extinction and larval development), we applied the D statistic
rates vary (Legendre and Condamine 2018). To select proposed by Fritz and Purvis (2010), using the function
the best-fitting model, ML score of each model and the phylo.d in the R package caper: D = 1 indicates that the
resulting corrected Akaike information criterion (AICc) trait has a phylogenetically random distribution across
were compared (Supplementary Table S3 available on the tips of the phylogeny (i.e., lack of phylogenetic
Dryad). signal), whereas D = 0 if the observed trait is as clumped
as if it had evolved according to a Brownian motion
model. Values of D can also fall outside this range: D < 0
suggests a highly clustered trait whereas D > 1 suggests
Trait-Dependent Diversification phylogenetic overdispersion.
To model simultaneously the evolution of discrete We used a phylogenetic diversity (PD) approach to
traits and their impact on diversification, we used measure how functional and ecological discrete traits
trait-dependent diversification models, in which species are distributed along Terebridae phylogeny. As defined
are characterized by an evolving trait and their by Faith (1992), PD can be measured as “the minimum
diversification follows a birth–death process in which total length of all the phylogenetic branches required to
speciation and extinction rates may depend on the trait span a given set of taxa on the phylogenetic tree.” In
state. We used four characters: 1) Larval ecology, where this particular context, this approach depicts how the
species were defined by having either a planktotrophic distribution of a trait state among taxa is influenced by
(0) or nonplanktotrophic (1) ecology; 2) VG, where the underlying evolutionary processes, or in other words
species were defined according to either the presence how each trait state contribute to the phylogenetic signal
(0) or the absence (1) of this structure; 3) depth, where for that particular discrete trait.

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 417 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

418 SYSTEMATIC BIOLOGY VOL. 69

PD was calculated for two subsets of taxa simplest OU model (OU1) has a single optimum ()
corresponding: 1) the planktotrophic vs. lecithotrophic applied to all branches. The remaining four OU models
developers and 2) the species with VG vs. species differ in how the rate parameters are allowed to vary in
that had lost it. In both cases, PD was calculated the model. In the first (OUM model) phenotypic optima
using different metrics, standardized for unequal means (x ) are different whereas both the strengths of
richness sampling, using the R package picante (Kembel selection (x ) and the rate of stochastic motion around
et al. 2010, 2013). First, we calculated Faith’s PD, the optima (2x ) acting on all selective regimes are
corresponding to the sum of the total phylogenetic identical. We also fitted a model that only allowed
branch length for one or multiple samples (Faith strengths of selection to vary among selective regimes
1992). Then, we measured beta diversity in each (1 , 2 : OUMA model), as well as one that only allowed
subset both as the mean nearest taxon distance the rates of stochastic evolution away from the optimum
(MNTD) separating taxa with alternative trait states, to vary (2A , 2B : OUMV model). Eventually, we fitted
corresponding to the average phylogenetic distance a model (OUMVA) that allowed all three parameters
to the most similar taxon in the other cluster, and as (,,) to vary among the different selective regimes.
the mean pairwise distance (MPD) separating taxa To choose the best-fitting model, we used a model-
in two clusters (Gotelli and Colwell 2001; Webb et al. averaging approach, where we calculated the Akaike
2002; Helmus et al. 2007). All metrics were calculated weights for each model, that is the relative likelihood

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

as SES (standardized effect size) values (Warren et al. of each model (Burnham and Anderson 2002) by means
2008). As MPD and MNTD have different sensitivity, of the second-order AICc that includes a correction
being more sensitive, respectively, to tree-wide vs. for reduced sample sizes (Hurvich and Tsai 1989). We
tips-accumulating patterns of phylogenetic clustering. ensured that the eigenvalues of the Hessian matrix
Positive values (mpd.obs.z ≥ 0) and high quantiles calculated in our OUwie analysis were positive, because
(mpd.obs.p > 0.95) indicate phylogenetic evenness, or this is an indication of the reliability of parameters
a greater phylogenetic distance among species sharing estimation (Beaulieu et al. 2012).
a same character state than expected. Conversely,
negative values and low quantiles (mpd.obs.p
< 0.05) indicate phylogenetic clustering, or small
phylogenetic distances among species sharing a same RESULTS
character state than expected (Gotelli and Colwell 2001;
Webb et al. 2002; Helmus et al. 2007). Species Diversity Identifies Potential Cryptic Lineages
A data set of 1275 samples was used to reconstruct
the molecular phylogeny of the Terebridae family (Fig. 2
and Supplementary Table S1 available on Dryad).
Evolutionary Modeling Among them, 130 species were confidently identified
To test whether shifts in larval development are because their shell matched a described species and
associated with selective constraints on the evolution corresponded to a unique lineage in the independent
of shell size and bathymetric distribution, and if depth gene trees. Some names previously synonymized with
shifts are associated with selective constraint on shell others were elevated at the species level (marked with
size evolution, we fitted two Brownian Motion (BM) an * in Fig. 2; Fedosov et al. 2019). In addition, 69
models and five different Ornstein-Uhlenbeck (OU) new species were identified based on morphological
models using the R package OUwie (Beaulieu 2016) grounds and/or correspondence to divergent lineages
to 100 trees reconstructed with stochastic character in the independent gene trees, with genetic distances
mapping of the trait “larval development” and the trait among species equivalent or even higher to genetic
“depth” (coded as discrete) using the make.simmap distances recovered among already described species
function available in the R package phytools. For (K2P genetic distances > 2.5%). For example, the name
the parametrization of make.simmap, we used the Punctoterebra textilis was originally applied to eight
estimated ancestral state, and a transition matrix with lineages recognized in the COI tree. After re-examination
equal rates estimated from our empirical data with a of the shells, we applied the name P. textilis to one of
MCMC search, and we performed 100 replicates then them, the names Punctoterebra roseata and Punctoterebra
summarized in a consensus tree, to account for the soulyeti, previously considered as synonyms of P. textilis,
inherent stochasticity of the process. BM models are to two others, and the remaining five lineages are
processes where phenotypic variation accumulates with considered new. In all but one case taxa belonging to
time, as is the case with random variation, neutral genetic these species complexes fall within one major Terebridae
drift, or drift-mutation equilibrium (Felsenstein 2001; clade consistent with one genus. The single exception
Beaulieu et al. 2012). Here, we fitted BM1 and BMS is the Profunditerebra orientalis complex, in which two
models, respectively, with a single rate and different rate lineages cluster within the genus Profunditerebra (E3)
parameters for each state in the tree. The OU models, and a morphologically strikingly similar form is found
add to the stochastic displacement described by BM in Maculauger (E5A; Fig. 2). In most of these species
models an optimal trait value and a tendency toward complexes, a thorough re-examination of the shells
that optimum (Hansen 1997; Beaulieu et al. 2012). The revealed morphological differences, suggesting they

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 418 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

2020 MODICA ET AL.—DIVERSIFICATION AND EVOLUTION OF THE TEREBRIDAE 419

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

FIGURE 2. Dated phylogenetic reconstruction of the Terebridae using a multigene approach confirms terebrids are monophyletic and
comprised of six major clades. A Bayesian phylogenetic terebrid tree was constructed using 12S, 16S, 28S, and COI DNA sequences. Six major
Clades (A–F) were recovered, which are shown with a unique color shaded block in the tree. Each shade represents a different genera listed
A–F with subheadings such as A, E1–E5, and F1, F2, within the main clades. Character traits larval ecology and anatomy types are mapped
onto the tree. Blue dots with the letter “M” represent a multispiral protoconch, whereas red dots with the letter “P” represent a paucispiral
protoconch. Roman numerals represent newly defined anatomy types. Shells represent 12 of the 17 cryptic species complexes identified. Posterior
Probabilities (pp) are marked with dots on the nodes, where black darkened dots represent a pp of 1 and gray dots represent a pp between 0.9
and 1.0.

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 419 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

420 SYSTEMATIC BIOLOGY VOL. 69

comprise pseudo-cryptic species. Our findings suggest in Fedosov et al. (2019), this clade has been further
that a considerable fraction of the Terebridae diversity divided into F1 and F2, which correspond to the revised
still requires formal description. genus Duplicaria and the genus Partecosta, respectively
Three species complexes comprised pairs of lineages (Fedosov et al. 2019).
with allopatric distribution, and in three clusters We used the current fossil record of the Terebridae to
comprising three or more divergent lineages (P. textilis, produce a calibrated tree. The origin of the Terebridae
Tabellaria fenestrata and Punctoterebra trismacaria) where is estimated at 50.6 Ma with 95% highest probability
at least one of them does not overlap in distribution density: 44.1–51.2, matching the well-documented
with others. In addition, our data suggest difference Terebridae fossils found in the Early Eocene period
in bathymetric distribution in at least four putative (stage Ypresian: 47.8–56 Ma). The six main lineages of
species complexes: Terebra cumingii, Myurella burchi, P. terebrids all appeared before the end of the Eocene. The
trismacaria, and P. orientalis. However, such differences diversifications of each of the main lineages, including
do not exist between sister-lineages, suggesting that the the subgroups within the Clades A, E, and F, all started
lineages within a species complex actually correspond concomitantly, between the mid-Oligocene (30 Ma) and
to different species (Puillandre et al. 2012). Confirming the early Miocene (20 Ma).
whether these lineages correspond to different species
or to populations within a single species would require

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

further study, including more samples per lineages that Terebrid Diversification Rates Increase Over Time
are currently represented in most cases by less than
We examined terebrid diversification rates as a
five specimens each. For the subsequent analysis, we
function of time and across the six individual Clades A–
considered that our data set includes 199 species.
F delineated in our phylogenic reconstruction (Fig. 2).
Using a realistic sampling fraction of 26%, BAMM
analysis supported a model that indicated a steady
Dated Terebridae Molecular Phylogeny Recovers New Sister rate of terebrid diversification over time, with a 0.97
Clade posterior probability. Both posterior probabilities and
Bayes factors were remarkably lower for alternative
A multigene approach was applied using COI (1161
models with one or two rate shifts (Supplementary
samples), 16S (717 samples), 12S (817 samples), and
Table S2 available on Dryad). The credible shifts plot
28S (263 samples) genes. Analyses of each individual
depicts a single evolutionary regime for the Terebridae
gene were performed using RAxML and no supported
regardless of the value attributed to the prior probability
conflicts were found between the four separately
of a rate shift (Fig. 3A and Supplementary Table S2
generated gene trees (Supplementary Figs. S2–S5
available on Dryad). The rate-through-time BAMM plot
available on Dryad). The four genes were combined
supports a scenario of a slow increase of diversification
to produce a consensus tree (Fig. 2). Only samples
for the whole Terebridae (Fig. 3B). This scenario is further
with ≥ 2 genes successfully sequenced were used in
corroborated by the results of RPANDA analysis, which
the combined gene data set, a total of 898 samples.
recovered a rate-constant speciation (lambda = 0.134
Even though the species representation doubled and the
lineages/myr) and rate-variable extinction model as best
number of samples tripled from the previous reported
to describe the evolutionary pattern of the Terebridae.
terebrid molecular phylogenies, the overall topology of
More specifically, the extinction rate has decreased
the terebrid tree is largely consistent with the previous
over time and the diversification rate has plateaued,
study and the family has remained monophyletic as
according to the best-fit RPANDA model (Fig. 3C). From
described in the first molecular phylogeny of the group
these analyses, the decrease in terebrid extinction rate
(Holford et al. 2009b).
can explain an increase in global diversification rate
Our new terebrid phylogenetic reconstruction divides
beginning around 25 Ma, as has been observed in other
the family into six major clades as found in previously
marine taxa (Alfaro et al. 2007; Williams and Duda 2008).
published reports (Castelin et al. 2012). In this study, we
use the same naming system for clades (A–F). However,
Clade A (Pellifronia) is no longer a sister group to all other
terebrids and two lineages were recognized, Pellifronia Evolution Rate Shifts in Depth and Shell Size
jungi and Bathyterebra coriolisi (Fedosov et al. 2019; Fig. 2). Despite the absence of across-clade heterogeneity in
The genera represented by Clade B (Oxymeris), Clade C diversification rates, the most supported configurations
(Terebra), and Clade D (Hastula) are consistent with their recovered by BAMM analysis for continuous traits
previous placement (Holford et al. 2009a; Castelin et al. displayed evidence of shifts in evolutionary rates of
2012). The largest Clade E is subdivided into subclades terebrid traits. Specifically, for shell size, we recovered
E1–E5, with the corresponding genera E1 (Myurella), E2 two likely evolutionary rate shifts: one for the single
(Punctoterebra), E3 (Profunditerebra), E4 (Neoterebra), and species Myurella pertusa belonging to clade E1 and the
E5 (Maculauger and Myurellopsis). In addition, Clade F, other for Clades B and C, corresponding to the Terebra
consisting of 11 species in our data set, is now the sister and Oxymeris genera (Supplementary Fig. S6 available
group to all other terebrids with a posterior probability on Dryad). Shell size appeared to have undergone
of 1. Based on morphological findings summarized a fast divergence at the beginning of the Terebridae

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 420 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

2020 MODICA ET AL.—DIVERSIFICATION AND EVOLUTION OF THE TEREBRIDAE 421

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

FIGURE 3. Terebridae Diversification rates vary across clades and time. a) The single BAMM credible shifts plot representing the rate shift
configuration and a posteriori probability shift configuration corresponding to 0.97. b) BAMM plot depicting the net diversification rates-through-
time trajectory as analyzed by BAMM. c) RPANDA plot showing the estimated speciation (blue, straight line), extinction (red, descending dashed
line), and net diversification (purple, ascending dashed line) rates through time for the Terebridae phylogeny. d) RPANDA plot showing the
estimated accumulation of species richness through time for the Terebridae phylogeny.

evolutionary history, followed by several oscillations were both shifts to deep waters (Supplementary Fig. S8
between 35 and 15 Ma, with the evolutionary rate available on Dryad).
still increasing toward the present (Supplementary Fig. The rate-through-time plot for depth distribution
S7 available on Dryad). Our sample ranged in length emphasizes a constant, very low evolutionary rate at the
from 10 mm (Partecosta trilineata) to 274 mm (Oxymeris beginning of Terebridae evolutionary history, followed
maculata), with an average length of 61 mm, 104 species by a steep increase at ca. 40 Ma, a marked decrease after
were classified as being >25 mm and 27 species ≤25 mm. 30 Ma, and a second rapid increase from ca. 25 Ma to the
Similarly, depth apparently underwent seven shifts present (Fig. 4). From the specimens used in our data set,
in evolutionary rates that are summarized in the four certain species, such as Pellifronia jungi, which was found
groups outlined (Supplementary Fig. S8 available on 400–780 m over a range of widespread localities, remain
Dryad): 1) One shift for a subset of clade C including in deep waters, whereas other species, such as Hastula
Terebra n. sp. aff. cumingii 1 (shallow), Terebra n. sp. hectica, remain in shallow waters exhibiting a minimum
aff. cumingii 2 (deep), Terebra n. sp. 27 (shallow) depth of 0 m and maximum depth of 3 m. One hundred
and Terebra cumingii (deep). 2) One shift for the and forty eight species were classified as deep water
subset of Clade E1, which is a shift to deep waters being found below 100 m and 64 species classified as
shared by Myurella brunneobandata, M. pseudofortunei shallow were found above 100 m. Although most species
and M. n. sp. aff. Fortune. 3) Three shifts from have a narrow depth range, certain terebrid species have
shallow to deep for subsets of Clade E2, including, a broad depth range, such as Myurella nebulosa, which
respectively, Punctoterebra teramachii and Punctoterebra has a minimum depth of 1 m and maximum depth of
baileyi, Punctoterebra polygyrata, P. trismacaria and P. 762 m, or M. joserosadoi with a minimum depth of 5 m
textilis, P. sp. aff. textilis 1, and P. n. sp. aff. trismacaria and maximum depth of 287 m.
1. 4. The last two shifts are in the E5B clade for the According to the values retrieved for Pagel’s  (0.91
species Myurellopsis joserosadoi and Myurellopsis guphilae for both traits), both depth and shell size have a strong

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 421 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

422 SYSTEMATIC BIOLOGY VOL. 69

with both a venom apparatus and APS, suggesting this

morphology could be an intermediate between terebrids
that have a venom apparatus and those that lack it. The
APS is usually found in terebrid and other conoidean
species that have lost radula and VG, and even on those
occasions it is a seldom occurrence in these families
(Fedosov 2007; Fedosov and Kantor 2008). The APS
was suggested to have enabled novel feeding strategies
which did not involve prey envenomation, or enhanced
switch to different prey taxa (Fedosov and Kantor 2008;
Holford et al. 2009b). Anatomy Type XI represents the
traditional conoidean venom features and is found in
terebrids, cone snails, and most other Conoidea lineages.
Summarily, the 12 anatomy types identified reflect the
substantial degree of plasticity in terebrid foregut.
Phylogenetic signal and PD analysis with regard to
the presence or absence of a VG were carried out

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

on a subset of 51 species. The strong phylogenetic
signal (D = t1.08) obtained for the VG indicates that
the trait is phylogenetically conserved, indicating that
members of a same clade tend to share same trait state.
Through a PD analysis, negative SES values and low
quantiles were obtained both for the MNTD and for
the MPD of the species without a VG, indicating that
their phylogenetic distance is smaller than expected
(Supplementary Table S6 available on Dryad). These
results confirm the conservatism of the trait identified
FIGURE 4. Terebrid depth diversification rate varies over time. Rate by the phylogenetic signal, and highlight that the loss of
vs. time plot from the depth trait BAMM analysis, where “trait rate”
is given as depth change per million years, and “time before present”
a VG happened in phylogenetically clustered terminal
is in millions of years. At the start of terebrid evolution depth trait has taxa, and that when the VG is lost in the ancestor, the
a constant diversification rate, then in the Oligocene there is a sharp reversal is extremely unlikely.
increase, followed by a decline until approximately 25 Ma, when the
depth trait appears to steadily increase continuing into present day.

phylogenetic signal, indicating that close relatives are Distribution and PD of Terebrid Larval Ecology
more similar to each other for what concerns these We examined the protoconch in a total of 638 intact
traits than to distant relatives (Supplementary Table S5 terebrid adult specimens belonging to 116 species. In
available on Dryad). our data set, multispiral (m) protoconchs had between
three and five whorls, and paucispiral (p) protoconchs
had a maximum of 2.25 whorls. A number of specimens
displayed an intermediate protoconch, with 2.5 whorls
Redefinition and PD of Terebrid Foregut Anatomy Involving
and a general appearance compatible either with a
Predation-Related Traits lecithotrophic larva with a longer dispersive stage, or a
The presence or absence of a proboscis (PR), VG, short-lived planktotrophic larva. In those cases, instead
odontophore, accessory proboscis structure (APS), and of using only whorl numbers, the shell was attributed to
salivary glands (SG), and ranked the type of marginal one of the two developmental types based on protoconch
teeth (RadT; absent, duplex, solid recurved, flat, semi- characteristics, where a small nucleus and an evident
enrolled, or hypodermic) were evaluated to redefine the boundary between protoconch and teleoconch were
feeding types present in 51 of the 199 terebrid species considered indicative of a planktotrophic development.
used in this study. We identified 12 unique foregut Of the 199 species examined in the study, 72% are
anatomies (Types I–XII) defined by unique combinations planktotrophic and 28% are lecithotrophic (Fig. 2 and
of the six studied characters (Fig. 2 and Table 1). It is Supplementary Fig. S9 available on Dryad).
important to note our anatomy Types I–XII are distinct Phylogenetic signal was quite strong for larval
from Miller Types I–III (Miller 1971). In our analyses, development (D = t0.21), whereas PD analysis recovered
certain anatomy types are clade specific, such as Type negative SES values and low quantiles for MNTD
XII, which is only found in the genus Terebra (Clade C), of the lecithotrophic community only. The values
whereas other anatomy types can be found in multiple obtained for MPD were negative with low quantiles
clades, such as Type I, which can be found in Oxymeris for the planktotrophic community, and positive with
clade B and in the Myurella, Punctoterebra, Neoterebra, high quantiles for the lecithotrophic community
and Maculauger E subclades. Type XII represents species (Supplementary Table S6 available on Dryad).

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 422 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

2020 MODICA ET AL.—DIVERSIFICATION AND EVOLUTION OF THE TEREBRIDAE 423

The negative MNTD values for the lecithotrophic the optima across our Terebridae data set. The optimal
community indicate that the phylogenetic distance shell size value itself () has a value of 70 (±18) mm
among lecithotrophic species is smaller than expected, for planktotrophic and 21 (±7) mm for lecitotrophic
and that this clustering can be detected closer to the tips species. Our results suggest that species with long-living
of the phylogeny. Therefore, lecithotrophy appears to pelagic larvae not only tend to be generally larger, but
be a trait shared by closely related species, indicating also have a wider shell size range than lecitotrophic
that it has evolved before separation of the species-level species. The best-fitting model for depth distribution was
lineages and supporting the current view that reversal a simple Brownian model (BM), which did not support
to planktotrophy is an unlikely event. Conversely, any correlation between depth and larval development.
the obtained MPD values suggest that PD is high for Likewise, when coded as a discrete trait, there was no
planktotrophic developers, and indicates a more ancient support for a correlation between shell size and depth
origin of phylogenetic clustering. distribution.

Evolutionary Modeling of Traits Establishes Larval

No Clear Drivers of Terebrid Diversification
Development and Shell Size Relationship
Potential key innovations such as venom apparatus,
We identified an evolutionary link between larval

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

larval development, shell size, and depth distribution
ecology and shell size in the Terebridae using OUwie were examined in BiSSE using several models of trait
analyses. Specifically, the rate of shell size evolution evolution to determine potential drivers of terebrid
is more than five times higher in planktotrophic diversification (Supplementary Table S4 available on
species (2 = 83.15±0.23) than in lecithotrophic species Dryad). Contrary to our expectations, for presence
(2 = 15.1±0.51), whereas the strength of pull toward or absence of VG, the best-fit model had irreversible
a shell size optimum is about three times higher transition rates and equal speciation and extinction rates,
for lecithotrophic species ( = 0.30±0.043) than for suggesting the presence or absence of the VG does not
planktotrophic ones ( = 0.67 ±0.01). This finding is impact the rate of diversification in the terebrids. For
based on the best-fitting model for the Terebridae larval development, shell size, and depth, we recovered
adult shell size across the species included in our data the same speciation, extinction, and transition rates
set, which is the OUMVA, according to the Akaike for the two trait states considered, thus detecting no
weights, with a delta AICc >5 with respect to the significant departure from the null model. These results
second best-fitting model OUMA (Supplementary Table do not identify any of the tested traits as drivers of
S7 available on Dryad). This model allows the larval diversification, suggesting that either additional traits
ecology to influence the optimal shell size, the rate of and/or sampling of species is required, or terebrid
shell size evolution and the strength of pull toward diversification is not driven by a key innovation but

TABLE 1. Twelve newly defined terebrid anatomy types

Defining characteristics
Anatomy type Species representatives Clade Proboscis VG SG APS Marginal teeth
I Oxyermis areolata, Myurella B, E1, E2, E4, E5A Absent
amoena, Punctoterebra solangeae,
Neoterebra armillata, Maculauger
minipulchra
II Myurella affinis, Myurellopsis E1, E5B  Absent
parkinsoni
III Neoterebra variegata, Maculauger B, E4  Absent
pseudopertusa
IV Myurellopsis nebulosa, Myurellopsis E1, E5B   Absent
undulata
V Partecosta fuscolutea, Duplicaria F1, F2  Solid recurved
bernadii
VI Punctoterebra succincta E2    Absent
VII Neoterebra puncturosa E4  ?  Absent
VIII Profunditerebra poppei E3    Duplex
IX Punctoterebra lineaperlata E2    Flat
X Hastula stylata D    Semienrolled
XI Terebra subulata, Hastula hectica, C, D, E5B    Hypodermic
Myurellopsis kilburni
XII Terebra quoygaimardi C     Hypodermic
Notes: Twelve anatomy types were defined by looking at the presence or absence of a proboscis, VG, SG, or APS, as well as looking at the type of
marginal tooth. Species listed do not encompass all species with the anatomy type, but rather a subset, while clades represent all of the clades
that contain each of the anatomy types.

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 423 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

424 SYSTEMATIC BIOLOGY VOL. 69

rather by ecological opportunity due to environmental still increasing. These results were corroborated by
conditions. RPANDA analysis that also highlighted that the increase
in diversification rates can be attributed to a decrease in
extinction rate starting about 25 Ma (Fig. 3C).
The lack of clade-specific diversification rate shifts
DISCUSSION was unexpected given the uneven species richness and
A robust dated phylogenetic reconstruction of anatomical disparity observed in different clades. The
predatory terebrid marine snails was used as a relationship between species richness and diversification
framework for investigating the influence of several rates has been intensely debated, and it is presently
potential factors, such as anatomical features linked to generally accepted (McPeek and Brown 2007; Rabosky
the venom apparatus, larval development, shell size, and 2009; Wiens 2011; Rabosky et al. 2012). The strength
depth distribution, on terebrid species diversification. of this relationship has been demonstrated to be
The molecular phylogeny presented here is based on a rather variable, and may be lowered by several factors
significant increase in the taxonomic coverage compared including negative age–diversification relationships in
with previously published phylogenies for the group, which younger clades tend to evolve faster than older
tripling the number of specimens used and almost clades (Kozak and Wiens 2016; Scholl and Wiens 2016).
doubling the number of species. Specifically, the number This pattern may be explained by density-dependence

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

of specimens sequenced increased from 406 used in the slowing diversification rates over time, or by the younger
previous terebrid phylogeny (Castelin et al. 2012) to 1275 clades having novel traits that can help explain their
in the current study. This sampling increase corresponds rapid diversification (Rabosky 2009; Wiens 2011; Scholl
to about 40% of the >400 described species, which is 26% and Wiens 2016). In addition, homogeneity of across-
of the estimated species diversity, and further confirms clade diversification has been explained in some cases by
the monophyly of the family Terebridae and the existence the acquisition of a group-wide key innovation, which
of six major clades (Clades A–F; Holford et al. 2009b; lead to a constant diversification rates, as is the case with
Castelin et al. 2012). pharyngeal jaws in labrid fishes, climbing behavior in
In our molecular phylogenetic analysis, Clade F woodcreepers, and locomotion strategies in triggerfish
(including genera Duplicaria and Partecosta) has a new (Alfaro et al. 2009; Dornburg et al. 2011; Claramunt et al.
position and is recovered as a sister group to all other 2012). In some circumstances, diversification rates have
terebrids. In prior publications, Pellifronia clade A was been even shown to decrease after the acquisition of such
found to be the sister group to all other terebrids key innovations, as evidenced by the development of
(Castelin et al. 2012; Modica et al. 2014; Fig. 2). This shift foregut fermentation in colobine monkeys (Tran 2014).
could be explained be the addition of Bathyterebra coriolisi
in Clade A, which is a very different species compared
with P. jungi, the only species in the Castelin et al. (2012)
analyses. We also identified a number of pseudo-cryptic Foregut Anatomy and Ecological Traits Are Not Drivers for
species within species complexes, which suggest that a Terebrid Diversification
considerable fraction of the diversity in the Terebridae Our results suggest that trait evolution in
still needs formal description (Fedosov et al. 2019). morphological and ecological traits is not linked
Although the overall topology of terebrid phylogeny did to terebrid diversification. Using a BiSSE analysis,
not change significantly, adding more samples helped to none of the traits examined, venom apparatus, larval
reveal species complexes and to increase the overall node development, bathymetric distribution, and shell
support, illustrating the importance of dense species size, was identified as key innovations able to affect
sampling to more accurately reconstruct phylogenies. Terebridae diversification rates. The finding that foregut
anatomy did not have any effect on diversification rates
was surprising given the uneven species richness
observed across lineages with different foregut
Diversification Is Constant Across Clades, and Slowly anatomies. This is particularly relevant for the VG
Increasing Across Time in Terebridae in the foregut as the production of venom has been
proposed as a key innovation driving diversification in
The results obtained by BAMM analysis of terebrid Conoidea (Castelin et al. 2012) and in other venomous
diversification rates across clades outlined the absence taxa such as snakes (Vidal and Hedges 2005; Fry et al.
of any clade-specific shift in diversification rates. 2006), as it can allow the exploitation of new prey types
In contrast, the diversification rate-through-time plot and thus the colonization of novel niches. Our results,
obtained in BAMM suggests that the diversification however, agree with a recent work, which reported the
rate is slowly increasing in the Terebridae, when using presence of a VG had no effect on diversification rates
a sampling fraction of 26% of total extant terebrid across the conoidean tree (Abdelkrim et al. 2018).
diversity (Fig. 3B). The shape of the rate-through-time It should be noted that the venom apparatus, which
plot suggests that diversification rates were increasing consists of a VG, hyperdermic radular teeth, and
faster at the roots of the Terebridae phylogenetic tree, proboscis, is a shared evolutionary novelty of most
and tend to slow down closer to the present while lineages of the Conoidea and is lacking in some

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 424 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

2020 MODICA ET AL.—DIVERSIFICATION AND EVOLUTION OF THE TEREBRIDAE 425

terebrids. In other words, in some clades of the originated in shallow waters, increased their depth range
Terebridae the loss of the venom apparatus and not by moving with a set of adaptions that progressively
its acquisition is observed, for example, in the entire allowed them to reposition at deeper zones when sea
Oxymeris clade (Clade B). BiSSE best-fit model supported levels began to fall. This led them to colonize new
the hypothesis, already proposed on anatomical basis, niches, where selective pressure due to competition
that the loss of the VG is irreversible and this was and predation were weaker, which enabled a slow, but
also corroborated by the PD results. It is unclear how steady increase of diversification due to the reduction of
these species can effectively predate, but the evidence extinction rate. The conquest of deep-sea habitats may
of increased abundance of terebrid species with no have been triggered by an increase of sea level, which
VG, compared with those retaining a venom apparatus reached a maximum in the early Eocene (∼50 Ma; Miller
within a given area or locality seems to suggest that this et al. 2005; Kominz et al. 2008). At the same time, the
loss does not imply any selective disadvantage (Kantor increase in sea levels may have contributed to lower the
et al. 2012; Fedosov et al. 2014). This finding is confirmed extinction rates through the increase of the continental
by a recent stable isotope study investigating feeding shelf surface and, therefore, an increase in habitats
habits of the Terebridae in which the ranges of trophic (Orzechowski et al. 2015). Both the time estimates for
niches were indistinguishable between lineages with a main increase of depth diversification rate retrieved from
venom apparatus and those without (Fedosov et al. 2014). BAMM and the paleontological dating of Eocene sea-

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

In addition, venom components were reported in foregut level rise match with the time corresponding to lowest
glands such as the SG, which are not considered as part estimate of the extinction rate found in RPANDA analysis
of the venom apparatus, suggesting that, as in other (Fig. 3C). As sea levels began to fall, extinction rates in
venomous gastropods, even those Terebridae lineages the Terebridae began to level off (∼30 Ma). Therefore,
that lack a venom apparatus may still produce bioactive similar to the “colonization of deep waters” hypothesis,
compounds that can be released into the water to subdue the availability of more habitats created by the increased
prey (Modica et al. 2015; Gerdol et al. 2018). These sea level would have contributed to an ecological release
observations, along with the finding that neither the through a decrease in competition for resources on the
loss nor the acquisition of a venom apparatus influence shelf. The mosaic of habitat types in the Indo-West
diversification rates in Terebridae, imply that venom Pacific, a diversity hotspot for Terebridae as well as
apparatus is not, by itself, a good indicator of selective for other marine invertebrates, might have contributed
advantages linked to trophic ecology. Other feeding- to ecological release, as already suggested for other
related traits such as SG or general biochemical venom gastropod taxa (Williams and Duda 2008).
diversification may reveal better proxies of trophic The lack of statistical support for this hypothesis from
adaptation. BiSSE modeling may be due to insufficient taxonomic
coverage. In fact, simulation studies suggested that
BiSSE modeling performs best with >300 terminal taxa
(Davis et al. 2013; Gamisch 2016). Despite the three-fold
Colonization of Deep Waters May Have Affected Overall
increase with respect to previous phylogenies, our data
Terebridae Diversification set still represents merely 26% of estimated Terebridae
The observed lack of support for clade-specific diversity. In addition, our sampling effort has been
terebrid diversification rate shifts, suggests the overall mostly concentrated on less known deep-water habitats,
increase in diversification rate affecting the family may leading to a potential overrepresentation of deep-water
be due to a group-wide factor, rather than to traits species in our data set. We recognize that our deep water
displaying a high level of lineage-specific disparity. sampling bias may not reflect the actual distribution of
A potential hypothesis to explain the generalized Terebridae diversity, and may have affected the results
increase in diversification rates across the entire of trait evolution modeling.
terebrid family is an ecological release initiated by the
colonization of deep waters. A constant increase in
diversification rates was identified in bird genus Grebes
and was hypothesized to be caused by fragmentation Larval Development Affects Terebrid Adult Shell Size
of habitat, a factor that affected the entire family Notably, for the first time we demonstrate that
(Ogawa et al. 2015). Similarly, a study focused on lecithotrophy has evolved at least 18 times in the
freshwater snails showed an increase in speciation Terebridae and there is a link between adult shell
rates after experiencing ecological opportunity through size and larval development. We corroborate in the
dispersal to new locations (Delicado et al. 2018). For terebrids, as in previous studies on other gastropod
Terebridae, we retrieved a BAMM rate-through-time taxa, that larval development evolution trends are
plot of depth distribution characterized by a constant generally unidirectional, moving from planktotrophy to
and low diversification rate at the beginning, followed lecithotrophy (Gould and Eldredge 1986; Rouse 2000;
by a steep increase at about 40 Ma, a decrease at 30 Collin et al. 2007). In the Terebridae, shell size appears
Ma, and a second rapid increase in diversification rates to follow a complex history of diversification. Across
from about 25 Ma to the present (Fig. 4). Therefore, our entire data set, the best-fitting model estimates
we propose a scenario wherein terebrids, after having with strong support, according to Akaike weights,

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 425 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

426 SYSTEMATIC BIOLOGY VOL. 69

a different optimal size for the two divergent larval Burbrink 2012) and venomous lizard lineages (Fry et al.
ecologies but with a higher strength of pull toward a 2006). However, our results are in agreement with recent
size optimum in the lecithotrophic species. In detail, findings that the presence of a VG does not significantly
this model consistently estimates that adult size in affect diversification rates across the conoidean tree of
lecithotrophic species is significantly smaller, and more life (Abdelkrim et al. 2018). Although larval development
strictly size-constrained, than in planktotrophic species, did not appear to play a role in the diversification of
despite a larger egg size, which in turns determines Terebridae, evolutionary modeling identified a strong
the appearance of the protoconch. This implies that link between larval ecology and variability of adult
size in later stages of life is mostly linked to the shell size, highlighting larval ecology as an indirect
length of the larval stage (Levin et al. 1987; Miller factor shaping the Terebridae evolutionary trajectory.
and Hadfield 1990; Havenhand 1993). The increased Our results indicate that environmental factors linked
shell size in planktotrophic terebrids could be derived to past sea level increase and depth range may have
from longer generation times, which has been discussed played a key role in terebrid diversification, acting
in the settlement-timing hypothesis (Todd and Doyle as major evolutionary drivers on the whole family.
1981). A pelagic larval development is displayed by The terebrids are a microcosm for understanding
the vast majority (ca 70%) of marine invertebrate diversification within marine mollusks, and our findings
species, and is considered the ancestral larval ecology are an important milestone in disentangling true

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

in gastropods (Thorson 1950; Nielsen 2009), including drivers of evolutionary success that lead to the
most lineages of Caenogastropoda (Haszprunar 1988). astounding speciation of this group and in the family
The dichotomy between the two contrasting larval Conoidea.
ecologies has been well studied in marine invertebrates:
planktotrophic species have smaller egg sizes and high
female fecundity and lecithotrophic species possess
lower female fecundity and larger egg sizes, and they SUPPLEMENTARY MATERIAL
can, therefore, be placed at the two edges of an r-
Data available from the Dryad Digital Repository:
K continuum (Thorson 1950; Vance 1973; Strathmann
http://dx.doi.org/10.5061/dryad.j008j5h.
1977; Todd and Doyle 1981). As lecithotrophic terebrid
species rely on yolk reserves that are fixed at the
moment of egg production, their shell size at the time
of hatching is fixed, whereas in planktotrophic species
it may vary according to the length of larval stage FUNDING
and food intake. Thus, it may be argued that the shift This work was supported by funding from the
to lecithotrophy, with its stronger constraint on adult Camille and Henry Dreyfus Teacher-Scholar Award,
shell size, may reduce the plasticity of species and National Academy of Sciences Arab American Frontiers
their ability to adapt to new niches, partly explaining Fellowship Grant, The Alfred P. Sloan Foundation and
why the acquisition of lecithotrophy, despite leading NSF awards [CHE-1247550 and CHE-1228921 to M.H.],
to a reduction of interpopulational gene flow, does not and in part by the CONOTAX project funded by
increase speciation rate. The link between adult shell size the French National Research Agency [grant number
and larval development we have identified in terebrids ANR-13-JSV7-0013-01], and by the Russian Science
may upon examination also be present in other families Foundation grant [RSF 16-14-10118 to A.E.F]. Access to
of marine gastropods. ship time on the French Oceanographic Fleet is gratefully
acknowledged; the Taiwan and South China Sea
cruises were supported by bilateral cooperation research
funding from the Taiwan Ministry of Science and
CONCLUSIONS Technology (MOST 102-2923-B-002-001-MY3, PI Wei-Jen
Identifying the factors that influence predator–prey Chen) and the French National Research Agency (ANR
interactions and macroevolutionary patterns that lead to 12-ISV7-0005-01, PI Sarah Samadi). Additional materials
species diversification remains a challenge in neglected were used from fieldwork expeditions organized in
marine invertebrates. In this study, we examined the collaboration with the Smithsonian Tropical Research
Terebridae, an understudied group of predatory sea Institute in Panama in 2006–2008 with funding to M.H.
snails that possess a notable range of foregut anatomical (NSF-CHE 0610202). All expeditions operated under the
features and a complexity of venom arsenals comparable regulations then in force in the countries in question
with other groups of the Conoidea (Imperial et al. and satisfy the conditions set by the Nagoya Protocol for
2007; Kendel et al. 2013; Anand et al. 2014; Gorson access to genetic resources. The participation of A.E.F.
et al. 2015; Eriksson et al. 2018). Despite a long-standing was supported by the grant No. 16-14-10118 from the
hypothesis that venom can be a driver for diversification, Russian Science Foundation (principal investigator Yu I.
we did not find a correlation between possession of Kantor). M.V.M. was supported by a European Union’s
a venom apparatus and terebrid diversification. This Horizon 2020 research and innovation program under
is a remarkable difference from what is reported in grant agreement No. 748902. J.G. was supported partially
advanced snakes (Vidal 2002; Fry et al. 2008; Pyron and by CUNY Graduate Research Fellowship.

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 426 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

2020 MODICA ET AL.—DIVERSIFICATION AND EVOLUTION OF THE TEREBRIDAE 427

ACKNOWLEDGMENTS Benton M., Harper D.A.T. 2009. Introduction to paleobiology and the
fossil record. Chichester: Wiley.
The authors thank Philippe Bouchet, Philippe Bouchet P., Héros V., Lozouet P., Maestrati P. 2008. A quarter-century
Maestrati, Virginie Héros, Barbara Buge, and Julien of deep-sea malacological exploration in the South and West
Brisset for their help in curating the vouchers. The Pacific: where do we stand? How far to go? In: Héros V., Cowie
authors are grateful to Yuri Kantor and John Taylor R.H., BOUCHET P., editors. Tropical Deep-Sea Benthos 25. Paris:
Mémoires du Muséum national d’Histoire naturelle. 196:9–40.
for generously sharing their data on anatomy, Paolo Bouchet P., Kantor Y.I. 2004. New Caledonia: the major centre of
Colangelo (CNR, Italy) and Iacopo Bertocci (SZN, Italy) biodiversity for volutomitrid molluscs (Mollusca: Neogastropoda:
for helpful discussions on PD and diversity estimates, Volutomitridae). Syst. Biodivers. 1:467–502.
Fabien Condamine (CNRS, France) for advice on Bouchet P., Ng P.K.L., Largo D., Tan S.H. 2009. PANGLAO 2004:
macroevolutionary analyses. The material in this article investigations of the marine species richness in the Philippines.
Raffles Bull. Zool. 20:1–19.
originates from several shore-based expeditions and Bratcher T., Cernohorsky W. 1987. Living terebras of the world: a
deep-sea cruises, conducted, respectively, by MNHN monograph of the recent Terebridae of the world. Melbourne,
and Pro-Natura International (PNI) as part of the Florida & Burlington (MA): American Malacologists.
Our Planet Reviewed program (SANTO 2006, INHACA Burnham K.P., Anderson D. 2002. Model selection and multimodel
inference: a practical information-theoretic approach. New York:
2011, ATIMO VATAE and KAVIENG 2014; PI Philippe Springer-Verlag.
Bouchet) and/or by MNHN, Institut de Recherche Casewell N.R., Wüster W., Vonk F.J., Harrison R.A., Fry B.G. 2013.
pour le Développement (IRD) and other partners Complex cocktails: the evolutionary novelty of venoms. Trends Ecol.

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

as part of the Tropical Deep-Sea Benthos program Evol. 28:219–229.
(SALOMON 2, EXBODI, TERRASSES, MAINBAZA, Castelin M., Puillandre N., Kantor Y.I., Modica M. V, Terryn Y., Cruaud
C., Bouchet P., Holford M. 2012. Macroevolution of venom apparatus
MIRIKY, KARUBENTHOS 2, MADEEP, NANHAI 2014, innovations in auger snails (Gastropoda; Conoidea; Terebridae).
ZHONGSHA 2015; PIs Philippe Bouchet, Sarah Samadi, Mol. Phylogenet. Evol. 64:21–44.
Wei-Jen Chen, Tin-Yam Chan). In-country scientific Claramunt S., Derryberry E.P., Brumfield R.T., Remsen J.V. 2012.
partners included the University of Papua New Ecological opportunity and diversification in a continental radiation
of birds: climbing adaptations and cladogenesis in the Furnariidae.
Guinea (UPNG); National Fisheries College (Kavieng, Am. Nat. 179:649–666.
Papua New Guinea); Universidade Eduardo Mondlane Collin R. 2001. The effects of mode of development on phylogeography
(Maputo); Parc National de la Guadeloupe, and and population structure of North Atlantic Crepidula (Gastropoda:
Université des Antilles; University of Taipei and National Calyptraeidae). Mol. Ecol. 10:2249–2262.
Taiwan Ocean University (Keelung, Taiwan). Funders Collin R., Chaparro O.R., Winkler F., Véliz D. 2007. Molecular
phylogenetic and embryological evidence that feeding larvae have
and sponsors included the Total Foundation, Stavros been reacquired in a marine gastropod. Biol. Bull. 212:83–92.
Niarchos Foundation, European Regional Development Costello M.J., Chaudhary C. 2017. Marine biodiversity, biogeography,
Fund (ERDF), the French Ministry of Foreign Affairs, and deep-sea gradients, and conservation. Curr. Biol. 27:511–527.
Fonds Pacifique. Materials from expedition PANGLAO Daltry J.C., Wuster W., Thorpe R.S. 1996. Diet and snake venom
evolution. Nature 379:537–540.
2005 (a joint project of MNHN, University of San Carlos, Davis K.E., Hill J., Astrop T.I., Wills M.A. 2016. Global cooling as a
Cebu City, and the Philippines Bureau of Fisheries driver of diversification in a major marine clade. Nat. Commun.
and Aquatic Research); sampling in Congo arranged by 7:13003.
Bernard Thomassin. Davis M.P., Midford P.E., Maddison W. 2013. Exploring power and
parameter estimation of the BiSSE method for analyzing species
diversification. BMC Evol. Biol. 13:38.
Delicado D., Hauffe T., Wilke T. 2018. Ecological opportunity may
facilitate diversification in Palearctic freshwater organisms: a case
REFERENCES study on hydrobiid gastropods. BMC Evol. Biol. 18:55.
Des Roches S., Robertson J.M., Harmon L.J., Rosenblum E.B. 2011.
Abdelkrim J., Aznar-Cormano L., Fedosov A., Kantor Y., Lozouet Ecological release in white sands lizards. Ecol. Evol. 1:571–578.
P., Phuong M., Zaharias P., Puillandre N. 2018. Exon-capture Dornburg A., Sidlauskas B., Santini F., Sorenson L., Near T.J., Alfaro
based phylogeny and diversification of the venomous gastropods M.E. 2011. The influence of an innovative locomotor strategy on
(Neogastropoda, Conoidea). Mol. Biol. Evol. 35:2355–2374. the phenotypic diversification of triggerfish (family: Balistidae).
Alfaro M.E., Brock C.D., Banbury B.L., Wainwright P.C. 2009. Does Evolution 65:1912–1926.
evolutionary innovation in pharyngeal jaws lead to rapid lineage Drummond A.J., Rambaut A. 2007. BEAST: Bayesian evolutionary
diversification in labrid fishes? BMC Evol. Biol. 9:1–14. analysis by sampling trees. BMC Evol Biol. 7:214.
Alfaro M.E., Santini F., Brock C.D. 2007. Do reefs drive diversification in Duda T.F., Chang D., Lewis B.D., Lee T. 2009. Geographic variation in
marine teleosts? Evidence from the pufferfish and their allies (order venom allelic composition and diets of the widespread predatory
tetraodontiformes). Evolution 61:2104–2126. marine gastropod Conus ebraeus. PLoS One. 4:e6245.
Anand P., Grigoryan A., Bhuiyan M.H., Ueberheide B., Russell V., Edgar R.C. 2004. MUSCLE: multiple sequence alignment with high
Quinoñez J., Moy P., Chait B.T., Poget S.F., Holford M., Quinonez accuracy and high throughput. Nucleic Acids Res. 32:1792–1797.
J., Moy P., Chait B.T., Poget S.F., Holford M. 2014. Sample limited Eldredge N., Thompson J.N., Brakefield P.M., Gavrilets S., Jablonski D.,
characterization of a novel disulfide-rich venom peptide toxin from Jackson J.B.C., Lenski R.E., Lieberman B.S., Mcpeek M.A., Iii W.M.
terebrid marine snail Terebra variegata. PLoS One 9:e94122. 2005. The dynamics of evolutionary stasis. Paleobiology 31:133–145.
Barlow A., Pook C.E., Harrison R.A., Wuster W. 2009. Coevolution of Eriksson A., Anand P., Gorson J., Grijuc C., Hadelia E., Stewart J.C.,
diet and prey-specific venom activity supports the role of selection Holford M., Claridge-Chang A. 2018. Using Drosophila behavioral
in snake venom evolution. Proc. R. Soc. B Biol. Sci. 276:2443–2449. assays to characterize terebrid venom-peptide bioactivity. Sci. Rep.
Beaulieu J.M., Jhwueng D.C., Boettiger C., O’Meara B.C. 2012. 8:15276.
Modeling stabilizing selection: expanding the Ornstein-Uhlenbeck Faith D.P. 1992. Conservation evaluation and phylogentic diversity.
model of adaptive evolution. Evolution 66:2369–2383. Biol. Conserv. 61:1–10.
Beaulieu M.J., O’Meara B. 2016. OUwie: analysis of evolutionary rates Fedosov A. 2007. Anatomy of accessory rhynchodeal organs of
in an OU framework. R package version 1.5. Veprecula vepratica and Tritonoturris subrissoides: new types of

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 427 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

428 SYSTEMATIC BIOLOGY VOL. 69

foregut morphology in Raphitominae (Conoidea). Ruthenica Gotelli N.J., Colwell R.K. 2001. Quantifying biodiversity: procedures
17:33–41. and pitfalls in the measurement and comparison of species richness.
Fedosov A.E., Kantor Y.I. 2008. Toxoglossan gastropods of the Ecol. Lett. 4:379–391.
subfamily Crassispirinae (Turridae) lacking a radula, and a Gould S.J., Eldredge N. 1986. Punctuated equilibrium at the third stage.
discussion of the status of the subfamily Zemaciinae. J. Molluscan Syst. Biol. 35:143–148.
Stud. 74:27–35. Hall T. 1999. BioEdit: a user-friendly biological sequence alignment
Fedosov A.E., Malcolm G., Terryn Y., Gorson J., Modica M.M., Holford editor and analysis program for Windows 95/98/NT. Nucleic Acids
M., Puillandre N. 2019. Phylogenetic classification of the family Symp. Ser. 41:95–98.
Terebridae (Gastropoda: Mollusca). J. Molluscan Stud. in press. Hansen T.F. 1997. Stabilizing selection and the comparative analysis of
Fedosov A.E., Tiunov A. V., Kiyashko S.I., Kantor Y.I. 2014. Trophic adaptation. Evolution 51:1341–1351.
diversification in the evolution of predatory marine gastropods of Harmon L.J., Melville J., Larson A., Losos J.B. 2008. The role of
the family Terebridae as inferred from stable isotope data. Mar. Ecol. geography and ecological opportunity in the diversification of day
Prog. Ser. 497:143–156. geckos (Phelsuma). Syst. Biol. 57:562–573.
Feldman A., Sabath N., Pyron R.A., Mayrose I., Meiri S. 2016. Body sizes Harvey M.G., Seeholzer G.F., Smith B.T., Rabosky D.L., Cuervo A.M.,
and diversification rates of lizards, snakes, amphisbaenians and the Brumfield R.T. 2017. Positive association between population genetic
tuatara. Glob. Ecol. Biogeogr. 25:187–197. differentiation and speciation rates in New World birds. Proc. Natl.
Felsenstein J. 1985. Confidence limits on phylogenies: an approach Acad. Sci. USA 114:6328–6333.
using the bootstrap. Evolution 39:783–791. Haszprunar G. 1988. On the origin and evolution of major gastropods
Felsenstein J. 2001. Taking variation of evolutionary rates between group, with special reference to the Streptoneura. J. Molluscan Stud.
sites into account in inferring phylogenies. J. Mol. Evol. 53: 54:367–441.
447–455. Havenhand J.N. 1993. Egg to juvenile period, generation time, and the

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

Figueirido B., Janis C.M., Perez-Claros J.A., De Renzi M., Palmqvist P. evolution of larval type in marine invertebrates. Mar. Ecol. Prog. Ser.
2011. Cenozoic climate change influences mammalian evolutionary 97:247–260.
dynamics. Proc. Natl. Acad. Sci. USA 109:722–727. Helmus M.R., Bland T.J., Williams C.K., Ives A.R. 2007. Phylogenetic
Fitzjohn R.G. 2012. Diversitree: comparative phylogenetic analyses of measures of biodiversity. Am. Nat. 169:E68–E83.
diversification in R. Methods Ecol. Evol. 3:1084–1092. Holford M., Puillandre N., Modica M.V., Watkins M., Collin
Fontanillas E., Welch J.J., Thomas J.A., Bromham L. 2007. The influence R., Bermingham E., Olivera B.M. 2009a. Correlating molecular
of body size and net diversification rate on molecular evolution phylogeny with venom apparatus occurrence in panamic auger
during the radiation of animal phyla. BMC Evol. Biol. 7:95. snails (Terebridae). PLoS One 4:e7667.
Freckleton R.P., Harvey P.H., Pagel M. 2002. Phylogenetic analysis and Holford M., Puillandre N., Terryn Y., Cruaud C., Olivera B.M.,
comparative data: a test and review of evidence. Am. Nat. 160:712– Bouchet P. 2009b. Evolution of the toxoglossa venom apparatus as
726. inferred by molecular phylogeny of the Terebridae. Mol. Biol. Evol.
Fritz S.A., Purvis A. 2010. Selectivity in mammalian extinction risk 26:15–25.
and threat types: a new measure of phylogenetic signal strength in Hurvich C.M., Tsai C.-L. 1989. Regression and time series model
binary traits. Conserv. Biol. 24:1042–1051. selection in small samples. Biometrika 76:297.
Fry B.G., Roelants K., Champagne D.E., Scheib H., Tyndall J.D.A., Imperial J.S., Kantor Y., Watkins M., Iii F.M.H., Stevenson B., Chen
King G.F., Nevalainen T.J., Norman J.A., Lewis R.J., Norton R.S., P., Hansson K., Stenflo J., Ownby J., Bouchet P., Baldomero B.M.
Renjifo C., de la Vega R.C.R. 2009. The toxicogenomic multiverse: 2007. Venomous auger snail hastula (impages) hectica (Linnaeus, 1758):
convergent recruitment of proteins into animal venoms. Annu. Rev. molecular phylogeny, foregut anatomy and comparative toxinology.
Genomics Hum. Genet. 10:483–511. J. Exp. Zool. 756:744–756.
Fry B.G., Scheib H., Van der Weerd L., Young B., McNaughtan J., Ingram T. 2011. Speciation along a depth gradient in a marine adaptive
Ramjan S.F.R., Vidal N., Poelmann R.E., Norman J. 2008. Evolution radiation. Proc. R. Soc. B Biol. Sci. 278:613–618.
of an arsenal: structural and functional diversification of the venom Jablonski D., Lutz R.A. 1983. Larval ecology of marine benthic
system in the advanced snakes (Caenophidia). Mol. Cell. Proteomics invertebrates: paleobiological implications. Biol. Rev. 58:
7:215–246. 21–89.
Fry B.G., Vidal N., Norman J.A., Vonk F.J., Scheib H., Ramjan S.F.R., Jetz W., Thomas G.H., Joy J.B., Hartmann K., Mooers A.O. 2012.
Kuruppu S., Fung K., Hedges S.B., Richardson M.K., Hodgson W.C., The global diversity of birds in space and time. Nature 491:
Ignjatovic V., Summerhayes R., Kochva E. 2006. Early evolution of 444–448.
the venom system in lizards and snakes. Nature 439:584–588. Kantor Y.I., Fedosov A.E., Marin I.N. 2012. An unusually high
Gaither M.R., Violi B., Gray H.W.I., Neat F., Drazen J.C., Grubbs R.D., abundance and diversity of the Terebridae (Gastropoda: Conoidea)
Roa-Varón A., Sutton T., Hoelzel A.R. 2016. Depth as a driver of in Nha Trang Bay, Vietnam. Zool. Stud. 51:663–670.
evolution in the deep sea: insights from grenadiers (Gadiformes: Kembel A.S.W., Ackerly D.D., Blomberg S.P., Cornwell W.K., Cowan
Macrouridae) of the genus Coryphaenoides. Mol. Phylogenet. Evol. P.D., Hel M.R., Morlon H., Webb C.O. 2013. Picante: R tools for
104:73–82. integrating phylogenies and ecology. R package version 1.6.
Galindo L.A., Puillandre N., Strong E.E., Bouchet P. 2014. Using Kembel S.W., Cowan P.D., Helmus M.R., Cornwell W.K., Morlon H.,
microwaves to prepare gastropods for DNA barcoding. Mol. Ecol. Ackerly D.D., Blomberg S.P., Webb C.O. 2010. Picante: R tools for
Resour. 14:700–705. integrating phylogenies and ecology. Bioinformatics 26:1463–1464.
Gamisch A. 2016. Notes on the statistical power of the binary state Kendel Y., Melaun C., Kurz A., Nicke A., Peigneur S., Tytgat J., Wunder
speciation and extinction (BiSSE) model. Evol. Bioinform. 12:165– C., Mebs D., Kauferstein S. 2013. Venomous secretions from marine
174. snails of the Terebridae family target acetylcholine receptors. Toxins
Gerdol M., Cervelli M., Oliverio M., Modica M.V. 2018. Piercing fishes: (Basel) 5:1043–1050.
porin expansion and adaptation to hematophagy in the vampire Knouft J.H., Page L.M. 2003. The evolution of body size in extant groups
snail cumia reticulata. Mol. Biol. Evol. 35:2654–2668. of North American freshwater fishes: speciation, size distributions,
Gittleman J.L., Purvis A. 1998. Body size and species-richness in and Cope’s rule. Am. Nat. 161:413–421.
carnivores and primates. Proc. R. Soc. B Biol. Sci. 265:113–119. Kominz M.A., Browning J. V., Miller K.G., Sugarman P.J., Mizintseva
Gorson J., Ramrattan G., Verdes A., Wright E.M., Kantor Y., Srinivasan S., Scotese C.R. 2008. Late Cretaceous to Miocene sea-level estimates
R.R., Musunuri R., Packer D., Albano G., Qiu W.G., Holford M. from the New Jersey and Delaware coastal plain coreholes: an error
2015. Molecular diversity and gene evolution of the venom arsenal of analysis. Basin Res. 20:211–226.
Terebridae predatory marine snails. Genome Biol. Evol. 7:1761–1778. Kordis D., Gubensek F. 2000. Adaptive evolution of animal toxin
Gotelli N.J., Chao A. 2013. Measuring and estimating species richness, multigene families. Gene 261:43–52.
species diversity, and biotic similarity from sampling data. In: Levin Kozak K.H., Wiens J.J. 2016. Testing the relationships between
S.A., editor. Encyclopedia of Biodiversity. 2nd ed. Vol. 5. Waltham diversification, species richness, and trait evolution. Syst. Biol.
(MA): Academic Press. p. 195–211. 65:975–988.

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 428 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

2020 MODICA ET AL.—DIVERSIFICATION AND EVOLUTION OF THE TEREBRIDAE 429

Lee M.S.Y., Sanders K.L., King B., Palci A. 2016. Diversification rates Puillandre N., Kantor Y.I., Sysoev A., Couloux A., Meyer C., Rawlings
and phenotypic evolution in venomous snakes (Elapidae). R. Soc. T., Todd J.A., Bouchet P. 2011. The dragon tamed? A molecular
Open Sci. 3:150277. phylogeny of the Conoidea (Gastropoda). J. Molluscan Stud. 77:259–
Legendre F., Condamine F.L. 2018. When Darwin’s special difficulty 272.
promotes diversification in insects. Syst. Biol. 67:873–887. Puillandre N., Modica M. V., Zhang Y., Sirovich L., Boisselier M.C.,
Levin L.A., Caswell H., Depatra K.D., Creed E.L. 1987. Demographic Cruaud C., Holford M., Samadi S. 2012. Large-scale species
consequences of larval development mode: planktotrophy vs. delimitation method for hyperdiverse groups. Mol. Ecol. 21:2671–
lecithotrophy in Streblospio benedicti. Ecology 68:1877–1886. 2691.
Lewitus E., Bittner L., Malviya S., Bowler C., Morlon H. 2018. Clade- Pyron R.A., Burbrink F.T. 2012. Extinction, ecological opportunity, and
specific diversification dynamics of marine diatoms since the the origins of global snake diversity. Evolution 66:163–178.
Jurassic. Nat. Ecol. Evol. 2:1715–1723. Rabosky D.L. 2009. Ecological limits and diversification rate:
Lima G.M., Lutz R.A. 1990. The relationship of larval shell morphology alternative paradigms to explain the variation in species richness
to mode of development in marine prosobranch gastropods. J. Mar. among clades and regions. Ecol. Lett. 12:735–743.
Biol. Assoc. UK 70:611–637. Rabosky D.L. 2014. Automatic detection of key innovations, rate
Maddison W.P., Midford P.E., Otto S.P. 2007. Estimating a binary shifts, and diversity-dependence on phylogenetic trees. PLoS One
character’s effect on speciation and extinction. Syst. Biol. 56:701–710. 9:e89543.
McPeek M.A., Brown J.M. 2007. Clade age and not diversification rate Rabosky D.L., Chang J., Title P.O., Cowman P.F., Sallan L., Friedman
explains species richness among animal taxa. Am. Nat. 169:E97– M., Kaschner K., Garilao C., Near T.J., Coll M., Alfaro M.E. 2018.
E106. An inverse latitudinal gradient in speciation rate for marine fishes.
Miller B.A. 1971. Feeding mechanisms of the family Terebridae. Ann. Nature 559:392–395.
Rep. Am. Mal. Union 1970:72–74. Rabosky D.L., Grundler M., Anderson C., Title P., Shi J.J., Brown J.W.,

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

Miller K.G., Kominz M.A., Browning J. V., Wright J.D., Mountain G.S., Huang H., Larson J.G. 2014. BAMMtools: an R package for the
Katz M.E., Sugarman P.J., Cramer B.S., Christie-Blick N., Pekar S.F. analysis of evolutionary dynamics on phylogenetic trees. Methods
2005. The phanerozoic record of sea-level change. Science 310:11293– Ecol. Evol. 5:701–707.
11298. Rabosky D.L., Mitchell J.S., Chang J. 2017. Is BAMM Flawed?
Miller S.E., Hadfield M.G. 1990. Developmental arrest during Theoretical and practical concerns in the analysis of multi-rate
larval life and life-span extension in a marine mollusc. Science diversification models. Syst. Biol. 66:477–498.
248:356–358. Rabosky D.L., Santini F., Eastman J., Smith S.A., Sidlauskas B., Chang J.,
Mills P. 1979. Radular tooth structure of three species of Terebridae Alfaro M.E. 2013. Rates of speciation and morphological evolution
(Mollusca: Toxoglossa). Veliger 19:259–265. are correlated across the largest vertebrate radiation. Nat. Commun.
Modica M.V., Lombardo F., Franchini P., Oliverio M. 2015. The 4:1958.
venomous cocktail of the vampire snail Colubraria reticulata Rabosky D.L., Slater G.J., Alfaro M.E. 2012. Clade age and species
(Mollusca, Gastropoda). BMC Genomics 9:441. richness are decoupled across the eukaryotic tree of life. PLoS Biol.
Modica M.V., Puillandre N., Castelin M., Zhang Y., Holford M. 2014. A 10:e1001381.
good compromise: rapid and robust species proxies for inventorying Rainford J.L., Hofreiter M., Nicholson D.B., Mayhew P.J.
biodiversity hotspots using the Terebridae (Gastropoda: Conoidea). 2014. Phylogenetic distribution of extant richness suggests
PLoS One. 9:e102160. metamorphosis is a key innovation driving diversification in
Modica M.V., Russini V., Fassio G., Oliverio M. 2017. Do larval insects. PLoS One 9:e109085.
types affect genetic connectivity at sea? Testing hypothesis in two Rouse G.W. 2000. The epitome of hand waving? Larval feeding and
sibling marine gastropods with contrasting larval development. hypotheses of metazoan phylogeny. Evol. Dev. 2:222–233.
Mar. Environ. Res. 127:92–101. Sánchez-García M., Matheny P.B. 2017. Is the switch to an
Moore B.R., Höhna S., May M.R., Rannala B., Huelsenbeck J.P. 2016. ectomycorrhizal state an evolutionary key innovation in mushroom-
Critically evaluating the theory and performance of Bayesian forming fungi? A case study in the Tricholomatineae (Agaricales).
analysis of macroevolutionary mixtures. Proc. Natl. Acad. Sci. USA Evolution 71:51–65.
113:9569–9574. Scholl J.P., Wiens J.J. 2016. Diversification rates and species richness
Morlon H., Lewitus E., Condamine F.L., Manceau M., Clavel J., Drury across the Tree of Life. Proc. R. Soc. B Biol. Sci. 283:20161334.
J. 2016. RPANDA: an R package for macroevolutionary analyses on Shanks A.L. 2009. Pelagic larval duration and dispersal distance
phylogenetic trees. Methods Ecol. Evol. 7:589–597. revisited. Biol. Bull. 216:373–385.
Morlon H., Parsons T.L., Plotkin J.B. 2011. Reconciling molecular Sorenson L., Santini F., Alfaro M.E. 2014. The effect of habitat on
phylogenies with the fossil record. Proc. Natl. Acad. Sci. USA modern shark diversification. J. Evol. Biol. 27:1536–1548.
108:16327–16332. Stamatakis A. 2006. RAxML-VI-HPC: maximum likelihood-based
Ng J., Smith S.D. 2014. How traits shape trees: New approaches for phylogenetic analyses with thousands of taxa and mixed models.
detecting character state-dependent lineage diversification. J. Evol. Bioinformatics 22:2688–2690.
Biol. 27:2035–2045. Stamatakis A. 2014. RAxML version 8: a tool for phylogenetic
Nielsen C. 2009. How did indirect development with planktotrophic analysis and post-analysis of large phylogenies. Bioinformatics 30:
larvae evolve? Biol. Bull. 216:203–215. 1312–1313.
Ogawa L.M., Pulgarin P.C., Vance D.A., Fjeldså J., van Tuinen M. 2015. Stein A., Gerstner K., Kreft H. 2014. Environmental heterogeneity as a
Opposing demographic histories reveal rapid evolution in grebes universal driver of species richness across taxa, biomes and spatial
(Aves: Podicipedidae). Auk 132:771–786. scales. Ecol. Lett. 17:866–880.
Orzechowski E.A., Lockwood R., Byrnes J.E.K., Anderson S.C., Stigall A.L. 2017. Ordovician oxygen and biodiversity. Nat. Geosci.
Finnegan S., Finkel Z. V., Harnik P.G., Lindberg D.R., Liow L.H., 10:887–888.
Lotze H.K., Mcclain C.R., Mcguire J.L., O’Dea A., Pandolfi J.M., Strathmann R.R. 1977. Egg size, larval development, and juvenile size
Simpson C., Tittensor D.P. 2015. Marine extinction risk shaped by in benthic marine invertebrates. Am. Nat. 111:373–376.
trait-environment interactions over 500 million years. Glob. Chang. Taylor J.D. 1990. The anatomy of the foregut and relationships in the
Biol. 21:3595–3607. Terebridae. Malacologia 32:19–34.
Owens I.P.F., Bennett P.M., Harvey P.H. 1999. Species richness among Terryn Y. 2007. A collectors guide to recent Terebridae (Mollusca:
birds: body size, life history, sexual selection or ecology? Proc. R. Neogastropoda). Hackenheim, Germany: Conchbooks &
Soc. B Biol. Sci. 266:933. NaturalArt.
Parent C.E., Crespi B.J. 2009. Ecological opportunity in adaptive Terryn Y., Holford M. 2008. The Terebridae of the Vanuatu archipelago
radiation of galápagos endemic land snails. Am. Nat. 174:898–905. with a revision of the genus Granuliterebra Oyama 1961, the
Posada D. 2008. jModelTest: phylogenetic model averaging. Mol. Biol. description of a new genus and a three new species. Visaya 3:
Evol. 25:1253–1256. 1–96.

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 429 413–430

 Copyedited by: YS MANUSCRIPT CATEGORY: Systematic Biology

430 SYSTEMATIC BIOLOGY VOL. 69

Thorson G. 1950. Reproductive and larval ecology of marine bottom R.C., Nixon B., Dacheux J.L., Niwa H., Sekita Y., Huang X., Stark A.,
invertebrates. Biol. Rev. 25:1–45. Kheradpour P., Kellis M., Flicek P., Chen Y., Webber C., Hardison
Todd C.D., Doyle R.W. 1981. Reproductive strategies of marine benthic R., Nelson J., Hallsworth-Pepin K., Delehaunty K., Markovic C.,
invertebrates: a settlement-timing hypothesis. Mar. Ecol. Prog. Ser. Minx P., Feng Y., Kremitzki C., Mitreva M., Glasscock J., Wylie T.,
4:75–83. Wohldmann P., Thiru P., Nhan M.N., Pohl C.S., Smith S.M., Hou
Tran L.A.P. 2014. The role of ecological opportunity in shaping S., Renfree M.B., Mardis E.R., Wilson R.K. 2008. Genome analysis
disparate diversification trajectories in a bicontinental primate of the platypus reveals unique signatures of evolution. Nature 453:
radiation. Proc. R. Soc. B Biol. Sci. 281:20131979. 175–183.
Vance R. 1973. More on reproductive strategies in marine benthic Webb C.O., Ackerly D.D., McPeek M.A., Donoghue M.J. 2002.
invertebrates. Am. Nat. 107:353–361. Phylogenies and community ecology. Annu. Rev. Ecol. Syst. 33:475–
Vidal N. 2002. Colubroid systematics: evidence for an early appearance 505.
of the venom apparatus followed by extensive evolutionary Wiens J.J. 2011. The causes of species richness patterns across space,
tinkering. J. Toxicol. Toxin Rev. 21:21–41. time, and clades and the role of “ecological limits.” Q. Rev. Biol.
Vidal N., Hedges S.B. 2005. The phylogeny of squamate reptiles 86:75–96.
(lizards, snakes, and amphisbaenians) inferred from nine nuclear Wiens J.J. 2017. What explains patterns of biodiversity across the
protein-coding genes. C R Biol. 328:1000–1008. Tree of Life? New research is revealing the causes of the dramatic
Warren W.C., Hillier L.D.W., Marshall Graves J.A., Birney E., Ponting variation in species numbers across branches of the Tree of Life.
C.P., Grützner F., Belov K., Miller W., Clarke L., Chinwalla A.T., BioEssays. 39:1600128.
Yang S.P., Heger A., Locke D.P., Miethke P., Waters P.D., Veyrunes Williams S.T., Duda T.F. 2008. Did tectonic activity stimulate Oligo-
F., Fulton L., Fulton B., Graves T., Wallis J., Puente X.S., López- Miocene speciation in the Indo-West Pacific? Evolution 62:1618–
Otín C., Ordóñez G.R., Eichler E.E., Chen L., Cheng Z., Deakin 1634.

Downloaded from https://academic.oup.com/sysbio/article/69/3/413/5561106 by guest on 03 March 2025

J.E., Alsop A., Thompson K., Kirby P., Papenfuss A.T., Wakefield WoRMS Editorial Board. 2018. World Register of Marine Species
M.J., Olender T., Lancet D., Huttley G.A., Smit A.F.A., Pask A., (WoRMS). Available from http://www.marinespecies.org.
Temple-Smith P., Batzer M.A., Walker J.A., Konkel M.K., Harris Wright S. 2002. Breeding structure of populations in relation to
R.S., Whittington C.M., Wong E.S.W., Gemmell N.J., Buschiazzo E., speciation. Am. Nat. 74:232–248.
Vargas Jentzsch I.M., Merkel A., Schmitz J., Zemann A., Churakov Yoder J.B., Clancey E., Des Roches S., Eastman J.M., Gentry L., Godsoe
G., Ole Kriegs J., Brosius J., Murchison E.P., Sachidanandam R., W., Hagey T.J., Jochimsen D., Oswald B.P., Robertson J., Sarver B.A.J.,
Smith C., Hannon G.J., Tsend-Ayush E., McMillan D., Attenborough Schenk J.J., Spear S.F., Harmon L.J. 2010. Ecological opportunity
R., Rens W., Ferguson-Smith M., Lefèvre C.M., Sharp J.A., Nicholas and the origin of adaptive radiations. J. Evol. Biol. 23:1581–
K.R., Ray D.A., Kube M., Reinhardt R., Pringle T.H., Taylor J., Jones 1596.

[17:17 19/3/2020 Sysbio-OP-SYSB190058.tex] Page: 430 413–430