Biol. Bull.
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (10)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Collin, R.
Right arrow Articles by Véliz, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Collin, R.
Right arrow Articles by Véliz, D.
Biol. Bull. 212: 83-92. (April 2007)
© 2007 Marine Biological Laboratory

Molecular Phylogenetic and Embryological Evidence That Feeding Larvae Have Been Reacquired in a Marine Gastropod

Rachel Collin1,*, Oscar R. Chaparro2, Federico Winkler3 and David Véliz4

1 Smithsonian Tropical Research Institute, Apartado Postal 0843-03092, Balboa, Ancon, Republic of Panama
2 Instituto de Biologia Marina Dr. Jurgen Winter, Universidad Austral de Chile, Casilla 567, Valdivia, Chile
3 Dept. Biologia Marina, Universidad de Catolica Norte, Sede Coquimbo, Casilla 117, Coquimbo, Chile
4 Departamento de Ciencias Ecológicas, Facultad de Ciencias, Universidad de Chile, Las Palmeras 3425, Ñuñoa, Santiago, Chile

* To whom correspondence should be addressed. E-mail: collinr{at}si.edu


   Abstract
TOP
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Conclusion
 LITERATURE CITED
 
Evolutionary transitions between different modes of development in marine invertebrates are thought to be biased toward the loss of feeding larvae. Because the morphology of feeding larvae is complex and nonfeeding larvae or encapsulated embryos with benthic development often have simplified morphologies, it is presumed to be easier to lose a larval stage than to reacquire it. Some authors have gone so far as to suggest that feeding larvae, morphologically similar to the ancestral feeding larvae, cannot be reacquired. However, the larval structures of some groups, most notably gastropods, are often retained in the encapsulated embryos of species that hatch as benthic juveniles. Therefore the re-evolution of feeding larvae using the same structures may be possible in these groups. Here we present the first well-substantiated case for the recent re-evolution of feeding larvae within a clade of direct-developers. DNA sequence data show that Crepipatella fecunda, a species of calyptraeid gastropod with planktotrophic development, is nested within a clade of species with direct development, and that Crepipatella dilatata, a species with direct development, appears to be paraphyletic with respect to C. fecunda. Observation of the embryos of C. dilatata shows that the features necessary for larval feeding and swimming are retained in the encapsulated veligers, suggesting that heterochronic shifts in hatching time and changes in nurse-egg allotment could have resulted in the re-evolution of feeding larvae in this species.


   Introduction
TOP
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Conclusion
 LITERATURE CITED
 
The mode of development of marine invertebrates (direct or indirect, brooded, encapsulated, or planktonic) has far-reaching evolutionary and ecological consequences that are mediated via differences in dispersal ability. Species with planktotrophic (swimming and feeding) development often have higher rates of dispersal than do those with benthic, nonfeeding development (referred to here as "direct development" for simplicity). These differences in dispersal are thought to result in higher rates of gene flow and increased colonization ability (e.g., for gastropods: Hoskin, 1997; Kyle and Boulding, 2000; Wilke and Davis, 2000; Collin, 2001) for planktotrophs, as well as larger geographic ranges (Scheltema, 1989; Emlet, 1995; but see O’Foighil, 1989) and lower speciation and extinction rates when compared to species with direct development (e.g., Hansen, 1978, 1980; Jablonski, 1986; Gili and Martinell, 1994). Since mode of development has such profound evolutionary consequences, and since it is a feature that varies widely among species within most phyla of marine invertebrates, understanding the factors that shape evolution in mode of development is a major goal of marine invertebrate biologists (e.g., Strathmann, 1974, 1978a, b, 1985, 1990; Raff, 1987; Havenhand, 1995; Smith, 1997; Duda and Palumbi, 1999; McEdward, 2000; Jeffery et al., 2003; Collin, 2004).

The evolution of mode of development is presumed to be biased toward the loss of feeding larvae (Strathmann, 1974, 1978a; Hart, 1996). The idea is that species lacking planktonic feeding stages do not experience the selective pressure to retain the complex larval structures for feeding and swimming in the plankton. This should eventually result in the loss or reduction of such structures, and the likely degeneration of the genetic machinery that supports their development. Indeed, loss of such structures is commonly observed in nonfeeding larvae or embryos of echinoderms and molluscs (e.g., Strathmann, 1978a,b; Hadfield and Iaea, 1989; Emlet, 1994; Wray, 1996). Once lost, the difficulty of re-evolving a set of structures that function effectively for feeding and swimming makes it unlikely that feeding larvae will be reacquired from an ancestor with direct development.

If planktotrophic larvae have been reacquired de novo, the morphological structures used for feeding and swimming are expected to differ noticeably from those structures in the ancestral larval type (Strathmann, 1978a,b). The hypothesized ancient acquisition of feeding larvae in lingulid brachiopods through co-option of juvenile structures for function in a feeding larva is an example of this scenario (Strathmann, 1978a). In addition, some groups, most notably polychaete families (Rouse, 2000), have evolved multiple ways of feeding in the plankton that are associated with distinct larval morphologies. The cases identified by Strathmann (1978a) as possible independent acquisitions of feeding larvae were all ancient. Although more recent reacquisitions of feeding larvae are possible, there are no well-documented cases of such an event. Applications of phylogenetic methods to the comparative study of invertebrate development has further supported the idea that losses of feeding larvae are common, while little evidence of the possible re-evolution of feeding larvae has been recovered (Wray, 1996; Hart et al., 1997; Smith, 1997; Duda and Palumbi, 1999; Jeffery et al., 2003; Jeffery and Emlet, 2003).

Most of our ideas about the loss and reacquisition of feeding larvae are based on the extensive study of echinoid echinoderms. Groups such as gastropods, in which larval structures are not so extremely modified in direct-developing species, may be more permissive for the reacquisition of feeding larvae (Strathmann, 1978a). Many direct-developing gastropod species do indeed lose any identifiable larval structures. For example, the direct-developing brooded calyptraeids Crepidula adunca and Crepidula philippiana lose almost all vestiges of the larval velum and operculum (Gallardo, 1977b; Collin, 2000a). Although the loss of larval structures is also well documented for muricids (e.g., Fioroni, 1966), it is also common for species without a free-living larval stage to retain larval features during development. For example, Moran (1999) demonstrated that the embryonic velum in Littorina species with encapsulated development is not simplified or reduced compared to that of species with planktotrophic larvae. Likewise, Pernet (2003) showed that nonfeeding sabellid polychaete larvae retained the ability to capture particles by using the opposed-band ciliary system. The reduction and loss of larval structures appears to be correlated with a proxy for the time since the loss of feeding larvae (Collin, 2004). This suggests that after the evolution of direct development there is a window when larval structures are retained and during which the re-evolution of feeding larvae would be possible (Collin, 2004).

A recent phylogenetic analysis of calyptraeid gastropods did not support the irreversibility of the loss of feeding larvae (Collin, 2004). Maximum likelihood reconstruction of ancestral states did not demonstrate a statistically significant difference between the rate of gains versus losses of feeding larvae in calyptraeid gastropods (Collin, 2004). Furthermore, the rate of gains was significantly different from 0. Likewise, parsimony reconstructions suggested that feeding larvae could have been regained if the probability of loss was twice that of gaining feeding larvae (Collin, 2004). This study identified three areas in the phylogeny where a planktotrophic species appeared to be nested within a clade of direct-developing species and were, therefore, likely candidates for the discovery of secondarily evolved feeding larvae. Here we examine one of these clades, the genus Crepipatella, in detail and provide more evidence that C. fecunda is a species in which planktotrophic development may have been recently reacquired.


   Materials and Methods
TOP
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Conclusion
 LITERATURE CITED
 
Biology and natural history of Crepipatella
Calyptraeid gastropods comprise a family of sedentary, filter-feeding, protandrous marine limpets (Fig. 1 A, B). They are tolerant of widely varying ecological conditions but generally occur in intertidal or shallow subtidal habitats. Species in this family have been the focus of many studies of life histories, reproduction, and development. The genus Crepipatella, first described by Lesson in 1830, was synonymized with Crepidula until phylogenetic analysis of anatomical and DNA sequence data showed that Crepipatella groups more closely with other calyptraeid genera, and most closely with Crucibulum and Bostrycapulus, than with Crepidula (Collin, 2003a, b). About 15 Crepipatella species have been described since 1800, but recent studies indicate that there are only six distinct species. Three of these co-occur along the coast of Chile. These species are morphologically indistinguishable as adults, but they can be distinguished on the basis of development. Two of these species, C. dilatata and C. fecunda, are among the most well-studied species of gastropods from the southern hemisphere (Gallardo, 1976, 1977a, 1979; Gallardo and Garrido, 1987; Chaparro and Paschke, 1990; Chaparro et al., 2002, 2005; Veliz et al., 2003).


Figure 1
View larger version (143K):
[in this window]
[in a new window]

  		Figure 1. Adults, embryos, and larvae of Crepipatella species. (A) Adult C. dilatata in the intertidal near Coquimbo, Chile. (B) Adult C. lingulata attached to the glass of an aquarium in Friday Harbor, Washington. (C) Planktotrophic larva of C. lingulata. (D) Encapsulated development of C. dilatata. The uncleaved nurse eggs are about 190 µm, and the extended velum is clearly visible on several of the embryos. (E) Encapsulated embryo (above) and nurse eggs (below) of Crepipatella sp. This early in development, the embryos are not much larger than the nurse eggs, but the clearly developed mouth, embryonic kidneys, head-vesicle, and beginnings of the velum are all clearly visible, whereas the nurse eggs are irregular ciliated balls of yolk. (F) Embryos of C. capensis after consumption of the nurse eggs. The foot and tentacles are visible, but the embryonic morphology is dominated by the large yolk-filled visceral mass. (G) An embryo of C. capensis that has not yet consumed all of the nurse eggs. The irregular outlines of the nurse eggs are just visible at the edges of the visceral mass.

 
Gastropod larval morphology
Like larvae of most groups of marine invertebrates, gastropod larvae use bands of compound cilia arranged on a supporting structure to both propel them through the water and, at the same time, capture particles and transport them to the mouth. Gastropod larvae swim and feed with the velum, a pair of lobate flaps that extend from the head and are edged by two bands of compound cilia. These ciliary bands flank a ciliated food groove along which food is transported to the mouth. In direct-developing marine gastropods these structures are often reduced. The velum can either be small or reduced to a raised ciliated ridge, and the cilia are shortened and are often rearranged into a uniform ciliary field covering the head. Other modifications of direct-developing embryos include (1) enlargement or elaboration of the "embryonic kidneys," a pair of round cells located laterally behind the velum that are used for uptake of intracapsular albumin; (2) enlargement of the head vesicle, a ciliated, fluid-filled extension of the head with unknown function; (3) loss of the operculum in species that lack an operculum as adults; and (4) reduction of the coiling and sculpture of the shell (Collin, 2004).

Field collections and observations
Individuals of six species of Crepipatella were collected, with their brooded embryos, in the intertidal and shallow subtidal zones. For all animals that were preserved in 95% ethanol for subsequent DNA sequencing, mode of development was scored by examining living embryos with a stereomicroscope. Multiple stages of development were observed and described for all species, and special note was taken of the features that are known to often be modified in direct-developing species (see above). In most cases many additional animals were collected for other ongoing studies by the principal investigators (e.g., Chaparro et al., 2002; Véliz et al., 2003; Collin, 2004, 2006). Locality information for the live-collected specimens that were examined or sequenced for this study are listed in Table 1. Vouchers are deposited at the Field Museum of Natural History, Chicago, USA; The Natural History Museum, London, England; and the Academy of Natural Sciences, Philadelphia, USA (Table 1). We sampled all apparently valid species of the genus Crepipatella: C. lingulata (Gould 1846), C. dorsata (Broderip 1834), C. capensis (Quoy & Gaimard 1832–1833), C. fecunda (Gallardo 1979), C. dilatata (Lamarck 1822), and a currently undescribed species we refer to as Crepipatella sp. The only described species that is not likely to be a synonym of the species included here is Crepipatella fluctuosa Taki 1938 from Japan, which is reported from shells only.


View this table:
[in this window]
[in a new window]

  		Table 1 Material collected, location, and haplotype designations

 
DNA sequencing and analysis
A 564-base-pair fragment of the mitochondrial cytochrome oxidase I (COI) gene was sequenced for individuals from each locality (Table 1), and sequences have been deposited in GenBank (AF546051–AF546054; AF550491; DQ811116–DQ811133). DNA was extracted, amplified, and sequenced following the methods of Collin (2003b). DNA was extracted from ethanol-preserved tissue with Puregene (Gentra Systems) or DNeasy (Quiagen) extraction kits, amplified using Ready-To-Go PCR beads (Pharmacia Biotech), and primers and PCR profile of Folmer et al. (1994). PCR products were sequenced in both directions with dRhodamine (Perkin Elmer) or Big Dyes cycle sequencing dye terminator kits using the amplification primers and an ABI 377 automated sequencer.

Phylogenetic analyses were conducted using PAUP* 4b02 (Swofford, 1998). An equal-weighted parsimony analysis was performed, using a heuristic search with TBR branch swapping and 100 random additions. Bootstrap support for each clade was assessed on the basis of 1000 bootstrap replicates with TBR branch swapping and five random additions. Likelihood analyses were also conducted with PAUP, using TBR branch swapping and the model of evolution identified as the most appropriate by ModelTest (see below). Crucibulum auricula and Crucibulum cf. quiriquinae were used to root these phylogenies.

DNA sequence data were also analyzed by a Bayesian approach using MrBayes 2.01 (Huelsenbeck and Ronquist, 2001). The appropriate model and starting parameters for Bayesian analysis were chosen for each of the data sets by using the likelihood ratio test implemented in ModelTest 3.06 (Posada and Crandall, 1998, 2001) with the default settings and an {alpha}-level of 0.01. The Bayesian analysis, using one cold and three incrementally heated chains, started from a random tree with a uniform prior for branch lengths and for the gamma shape parameter. Invariant sites were retained in the sequences, and their frequency was estimated using the "invgamma" setting. The Metropolis-coupled Markov chain Monte Carlo (MCMCMC) analysis was run five times for 4,000,000 generations, and the number of trees to be discarded as representing a "burn-in" period was determined graphically. Majority-rule consensus trees for every 50th tree after the "burn-in" period were created using PAUP*, and consensus phylograms were created in MrBayes.

Likelihood reconstructions of character evolution were performed using Discrete (Pagel, 1997, 1999a,b). The maximum likelihood rates of character change were used to map the evolution of development onto the Bayesian topology using the "calculate fossil likelihood" command. Those nodes where the likelihoods of the two states differed by more than 2 log units were considered to provide significant support for one state at that node in preference to the other state (Pagel, 1997, 1999a,b). For comparison, ancestral states were also reconstructed using equal-weighted parsimony.


   Results
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Conclusion
 LITERATURE CITED
 
Observations of development
The gastropod genus Crepipatella (Calyptraeidae) contains two species that have typical planktotrophic development (Crepipatella lingulata and Crepipatella fecunda; Fig. 1) and three species that have direct development with nurse eggs (Crepipatella capensis, Crepipatella sp., and Crepipatella dilatata). Mode of development was documented for samples from each location where animals were collected. Development did not vary among populations of any of the species.

The development of the planktotrophs has been described in detail previously (Collin, 2000a; Chaparro et al., 2002, 2005). Neither the development nor the larval morphology of either species differs in any substantive way from that described for the well-known larvae of Crepidula fornicata (Conklin, 1897; Werner, 1955; Fig. 1C). We observed no modifications to the development of the velum, head vesicle, embryonic kidneys, or operculum that were similar to any features of the embryos of the direct-developing calyptraeid species. There were no intra-specific differences in larval morphology between C. lingulata from Washington State and California, or between C. fecunda from different populations in Chile.

All three direct-developing species produced small eggs that developed into large, crawling juveniles by ingesting nurse eggs (other eggs deposited in the same egg capsule). The development of C. dilatata has been described in some detail previously (Gallardo, 1976, 1977a, 1979; Chaparro et al., 2002). To summarize, a small number of the 220–240-µm eggs in each capsule develop into an intracapsular veliger with a coiled shell, a single embryonic kidney on each side, an operculum, and a distinct velum with a food groove (Fig. 1D). In fact, embryos of C. dilatata retain the ability to capture particles and transport them to the mouth via the food groove when removed from the capsule (Chaparro et al., 2002). However, they cannot swim (Chaparro et al., 2002). At this stage, the embryos of C. dilatata consume the uncleaved nurse eggs that were deposited in the capsule with them. The embryos lose all larval features before they hatch as crawling juveniles.

The development of Crepipatella sp. is similar to that of C. dilatata. The embryos develop into intracapsular veligers complete with a coiled shell, an operculum, and a velum with a distinct food groove. Like C. dilatata, the embryos do not have any obvious modifications for encapsulated development. However, unlike C. dilatata, the nurse eggs cleave in the same way as the embryos, and the two can only be distinguished from each other after gastrulation when the development of the nurse eggs becomes abnormal. Unlike the situation in C. dilatata where the nurse eggs are consumed whole, the embryos of Crepipatella sp. appear to somehow suck the yolk from the nurse eggs. Early in development the nurse eggs are yolk-filled ciliated balls, while later the yolk is restricted to the center of the ball (Fig. 1E). The nurse eggs are all finally consumed, and the larval features are lost before the embryos hatch as crawling juveniles.

The embryos of C. capensis also have direct development with nurse eggs. However, they lack a distinct intracapsular veliger stage (Fig. 1F, G). The embryos begin to consume the uncleaved nurse eggs much earlier in development, somewhat after gastrulation. The shell develops around the yolk-filled embryo and therefore does not develop a distinct coil like that of the other species of Crepipatella. The velum is a vestigial ridge with some ciliation but lacks a well-organized food groove. There is no operculum, and the head vesicle is large, although the larval kidneys remain simple.

Crepipatella dorsata is a very rare, subtidal species, about which little is known. None of the four individuals we collected were brooding egg capsules, and they failed to produce eggs when maintained in the laboratory for several months. Therefore, the development of this species remains unknown.

Phylogenetic relationships
Sequences of 564 base pairs of COI representing 33 haplotypes were obtained from 55 individuals (Table 1). Phylogenetic analyses of these sequences using Bayesian, parsimony, and maximum likelihood methods all resulted in congruent tree topologies (Fig. 2). In all cases the genus Crepipatella was rooted on the branch between C. lingulata + C. dorsata and the rest of the clade. Crepipatella capensis and Crepipatella sp. were the successive sisters to a clade comprising C. fecunda and C. dilatata. All species except the fecunda-dilatata group appeared as distinct and well-supported monophyletic clades (bootstrap > 70%, posterior probability > 95%; Fig. 2), which were separated by 10%–18% Kimura two-parameter distances.


Figure 2
View larger version (25K):
[in this window]
[in a new window]

  		Figure 2. The maximum likelihood topology of the COI haplotypes of Crepipatella. Redundant sequences are not shown, but the number of individuals with each haplotype is indicated after each species name. Asterisks above the branches indicate Bayesian support >95%, and those below the branches indicate support of >70% from a nonparametric bootstrap of the maximum likelihood analysis. In Bayesian, parsimony, and likelihood analyses the best trees all showed C. fecunda haplotypes nested within C. dilatata. Colored branches represent the parsimony state reconstruction of mode of development, and colored circles at the nodes indicate the maximum likelihood reconstruction of the ancestral condition reconstructed by Discrete (Pagel, 1999a, b). Solid blue circles represent nodes where the support for one state over the other was statistically significant and where that state was reconstructed more than 95% of the time. Blue = direct development; red = planktotrophy; black = unknown.

 
Together the haplotypes of C. fecunda and C. dilatata formed a well-supported (bootstrap > 70%, posterior probability > 95%) monophyletic clade, but the haplotypes from the two species were not reciprocally monophyletic. The haplotypes of C. fecunda formed a monophyletic clade in all analyses, and this clade was consistently placed within a paraphyletic C. dilatata by all three phylogenetic methods. Crepipatella dilatata comprised two haplotype clades, one made up predominantly of haplotypes from southern Chile and the other mostly of haplotypes from Argentina. This indication of population structure is what would be expected from a direct-developing species. However, both clades also contained haplotypes from northern Chile. The average Kimura two-parameter distance between sequences from C. fecunda and C. dilatata was 1.66% (1.26%–1.99%).

Both parsimony and maximum likelihood trait reconstructions support the conclusion that the planktotrophy of C. fecunda has re-evolved from direct development. Parsimony reconstructions of mode of development on the phylogeny (Fig. 2) give the ancestral condition at the base of the tree as ambiguous, because the first branching leads to one clade with planktotrophic development and one with direct development, and because the condition in the outgroups is variable (Collin, 2004). Within Crepipatella, the ancestral condition for C. fecunda is reconstructed as planktotrophy, whereas the four nodes between C. fecunda and the base of the tree, including the base of the fecunda-dilatata clade, are all reconstructed as direct development. Maximum likelihood reconstructions, of the traits using Discrete recovered a similar pattern (Fig. 2). The ancestral condition for C. fecunda is reconstructed as significantly more likely to be planktotrophy than direct development (likelihood ratio, P < 0.05), whereas the base of the fecunda-dilatata clade is significantly more likely to be direct development (likelihood ratio, P < 0.05). The mode of development for the node below the base of the fecunda-dilatata clade is marginally significant for direct development (P < 0.10; Fig. 2).


   Discussion
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Conclusion
 LITERATURE CITED
 
In the evaluation of the two alternate hypotheses—secondary planktotrophy in Crepipatella fecunda versus ancestral planktotrophy in C. fecunda—two lines of evidence are consistent with the idea that planktotrophy has been reacquired: the reconstruction of mode of development on the molecular phylogeny of the genus, and comparative observations of the embryology and larval morphology. The reconstruction of mode of development on the phylogeny shows that the planktotrophic C. fecunda has four successive sister clades with direct development (two clades of C. dilatata, C. capensis, and the Crepipatella sp.). The nesting of C. fecunda deeply within a paraphyletic clade of C. dilatata suggests that C. dilatata is the extant paraphyletic ancestor of C. fecunda. Thus, the direct-developer is reconstructed as the extant ancestor of the planktotrophic C. fecunda, giving support to the idea that feeding larvae have been reacquired. Finally, nodes ancestral to the most recent common ancestor of C. fecunda are reconstructed by Discrete as statistically significantly more likely to have direct development. If further sampling leads to a re-rooting of this clade such that C. fecunda is sister to C. dilatata or that there is an unresolved polytomy at the base of the fecunda-dilatata clade, the argument would be weakened somewhat, but C. fecunda would still be nested within three successive sister clades of direct-developers. Such a topology would still support the reacquisition of feeding larvae under parsimony analysis and is likely to be supported by maximum likelihood, although such a change could also result in an ambiguous reconstruction. Either way, the alternate hypothesis—that feeding larvae are the retained ancestral state—would not receive support from the topology.

The reacquisition of planktotrophy in C. fecunda is also consistent with the dynamics of evolution of development across the entire family. If planktotrophy is assumed for the ancestral state at the base of the Crepipatella clade, the hypothesis of reacquisition of planktotrophy requires a single loss of feeding larvae and one subsequent gain in C. fecunda. The alternate hypothesis—that planktotrophy is the ancestral condition retained in C. fecunda—would require that direct development with nurse eggs evolved three times independently (assuming that direct development evolved only once in C. dilatata), and that feeding larvae never re-evolve. Global maximum likelihood analysis of mode of development for 72 species of calyptraeids found no statistically significant differences in rates of loss and gain of feeding larvae across the family (Collin, 2004). This study also found that the loss of nurse eggs was almost 5 times more likely than the gain of nurse eggs. Therefore, the global analysis suggests that (1) there is no strong bias toward the loss of feeding larvae and (2) nurse eggs are more likely to be lost than gained. These results are more consistent with the pattern of character-state changes generated by the secondary gain of feeding larvae in C. fecunda than by the retention of ancestral planktotrophy. If the ancestral state for Crepipatella is reconstructed as direct development or if C. dorsata (development currently unknown) has direct development, the case for the reacquisition of feeding larvae becomes stronger.

Observations of embryonic and larval morphologies, an independent line of evidence, are also consistent with the idea that feeding larvae could have re-evolved in C. fecunda. The closest direct-developing sister of C. fecunda and one of the other sisters both retain complete larval morphologies during intracapsular development. The closest sister, C. dilatata, also retains the ability to capture particles from suspension (using the opposed-band ciliary mechanism), and its inability to swim is apparently due only to its large size (Chaparro et al., 2002). If this kind of development with an intracapsular veliger were ancestral at the base of the C. fecunda clade, feeding larvae could have been gained by the simple reduction of nurse egg production, as suggested by Chaparro et al. (2002). This is not as unexpected as the reacquisition of larvae de novo.

How common is the reacquisition of feeding larvae?
As described in the introduction of this paper, there is ample evidence from echinoids and asteroids to support the scenario that losses of feeding larvae are relatively common and that feeding larvae have not been reacquired (e.g., Emlet, 1994; Wray, 1996). In contrast, vast variation in mode of development among many families or genera of gastropods, polychaetes (Kupriyanova, 2003; Pernet, 2003), and ophiuroids, and phylogenetic analyses of such groups show that transitions in mode of development have occurred frequently and recently (e.g., Hart et al., 1997; Duda and Palumbi, 1999; Jeffery et al., 2003; Collin, 2004; Ellingson and Krug, 2006). In fact, changes can appear so frequently that our ability to reconstruct the likely pattern of evolution is limited (Collin, 2004). The same groups that display such high levels of variation in mode of development are also the groups in which the retention of structures specialized for planktotrophy is common among direct-developers. This makes it likely that such clades include other examples of secondary larval feeding or swimming. Although independent acquisitions of feeding larvae were predicted by Strathmann (1978a) to have occurred at the levels of order or family in these groups, it is somewhat surprising that well-supported cases of the re-evolution of larvae among closely related species have not come to light before now. There is a single documented case in which a species with swimming larvae, the starfish Pteraster tessulatus, has evolved from a species with completely brooded development (McEdward, 1995). However, these ciliated, yolky swimming larvae do not express any of the morphological features typical of feeding larvae and cannot feed. We are not aware of any similar cases that report the recent re-evolution of feeding larvae.

Are C. fecunda and C. dilatata distinct species?
The COI sequence data show that C. fecunda is as differentiated from the two clades of C. dilatata as they are from each other. In addition, these low levels of divergence are similar to intra-specific divergences among other species of calyptraeids (Collin, 2000b, 2001, 2005). These small divergences and the fact that the mitochondrial haplotypes have not yet coalesced suggest that C. fecunda and C. dilatata diverged very recently relative to other calyptraeid species. A similar conclusion was reached by Véliz et al. (2003) on the basis of allozyme variation.

Despite these very similar COI sequences, it is unlikely that C. fecunda and C. dilatata are conspecific. The variation in development between these otherwise morphologically indistinguishable species is not due to poecilogony (intraspecific variation in mode of development), as evidenced by the fact that many years of research in southern Chile (Gallardo, 1976, 1977a, 1979; Gallardo and Garrido, 1987; Chaparro and Paschke, 1990; Chaparro et al., 2002, 2005) have failed to produce any evidence of intra-specific variation in development among either C. dilatata or C. fecunda. Observations of embryos and larvae of all populations included here, and therefore species identifications, were standardized by a single worker (R.C.). This cross-validation of developmental characteristics in northern and southern Chile, Peru, and Argentina, as well as many years of research on these species by all of the authors, ensures that development was consistently characterized. Furthermore, poecilogony is unknown for any species of caenogastropod (Hoagland and Robertson, 1988; Bouchet, 1989) and thus is unlikey to occur in these species.

The status of C. dilatata and C. fecunda as distinct species is supported by several lines of evidence in addition to the consistent differences in development. The geographic ranges of both species, although largely sympatic, do not overlap completely. Crepipatella fecunda extends as far north as Lima, Peru, where C. dilatata is undocumented (although, C. fecunda has been misidentified as C. dilatata in this region), and C. dilatata occurs commonly along the southern Argentine coast where C. fecunda is unknown. Allozyme data for the three sympatrically occurring species from Northern Chile show that there are fixed differences between C. fecunda and C. dilatata in ACP, EST-2, and PEPB-1 (Véliz et al., 2003) and frequency differences or private alleles in three other enzymes. In addition, cytogenetic data also support the distinct status of the two species. Contreras (1999) found that Crepipatella sp. has half the chromosome number of C. fecunda and C. dilatata, although the large number of chromosomes in these last species (2N = 120–160) mades it difficult to obtain a precise count. Despite having a similar number of chromosomes as C. fecunda, C. dilatata has significantly more DNA (almost 4 times more) than the other two species (Contreras, 1999; F. Winkler, unpubl. data).


   Conclusion
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Conclusion
LITERATURE CITED
 
Here we show that a clade of direct-developing marine snails, which retains the morphological features necessary for planktotrophy within the egg capsule, appears to have recently given rise to a species with planktotropic development. As larval characteristics are often retained by direct-developing gastropods, the recent re-evolution of feeding larvae may be more common than previously expected among some groups.


   Acknowledgments
 
We thank C. Gallardo, N. Ciocco, G. Branch, A. Indacochea, and S. Valle for help in the field, and M. Venegas for sequencing some of the samples included in this analysis. Financial support was provided by the Smithsonian Tropical Research Institute (to R.C.) and by Fondecyt—Chile 1980984 to (O.R.C.),


   Footnotes
 
Received 24 April 2006; accepted 29 November 2006.


   LITERATURE CITED
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Conclusion
 LITERATURE CITED
 

    Bouchet, P. 1989. A review of poecilogony in gastropods. J. Molluscan Stud. 55:67–78.[Abstract/Free Full Text]
    Chaparro, O. R., and K. A. Paschke. 1990. Nurse egg feeding and energy balance in embryos of Crepidula dilatata Lamarck during intracapsular development. Mar. Ecol. Prog. Ser. 65:183–191.
    Chaparro, O. R., J. L. Charpentier, and R. Collin. 2002. Embryonic velar structure and function of two sibling species of Crepidula with different modes of development. Biol. Bull. 203:80–86.[Abstract/Free Full Text]
    Chaparro, O. R., C. L. Saldivia, S. V. Pereda, C. J. Segura, Y. A. Montiel, and R. Collin. 2005. The reproductive cycle and development of Crepipatella fecunda (Gastropoda: Calyptraeidae) from southern Chile. J. Mar. Biol. Assoc. UK 85:157–161.
    Collin, R. 2000a. Sex change, reproduction and development of Crepidula adunca and C. lingulata (Gastropoda: Calyptraeidae) Veliger 43:24–33.
    Collin, R. 2000b. Phylogeny of the Crepidula plana (Gastropoda: Calyptraeidae) cryptic species complex in North America. Can. J. Zool. 78:1500–1514.
    Collin, R. 2001. The effects of mode of development on phylogeography and population structure of North Atlantic Crepidula (Gastropoda: Calyptraeidae). Mol. Ecol. 10:2249–2262.[Medline]
    Collin R. 2003a. The utility of morphological characters in gastropod phylogenetics: an example from the Calyptraeidae. Biol. J. Linn. Soc. 78:541–593.[Web of Science]
    Collin R. 2003b. Phylogenetic relationships among calyptraeid gastropods and their implications for the biogeography of speciation. Syst. Biol. 52(5):618–640.
    Collin, R. 2004. Phylogenetic effects, the loss of complex characters, and the evolution of development in calyptraeid gastropods. Evolution 58:1488–1502.[Web of Science][Medline]
    Collin, R. 2005. Development, phylogeny, and taxonomy of Bostrycapulus (Caenogastropoda: Calyptraeidae), an ancient cryptic radiation. Zool. J. Linn. Soc. 144:75–101.
    Conklin, E. G. 1897. The embryology of Crepidula. J. Morphol. 13:3–226.
    Contreras, P. 1999. Diferenciación citgenética en especies de Crepidula (Mollusca: Mesogastropoda) del norte Chileno: evidencias de una possible poliploidía. Licenciado thesis in marine biology, Faculty of marine sciences, Universidad Católica del Norte, Chile.
    Duda, T. F., and S. R. Palumbi. 1999. Developmental shifts and species selection in gastropods. Proc. Natl. Acad. Sci. 96:10272–10277.[Abstract/Free Full Text]
    Ellingson, R. A., and P. J. Krug. 2006. Evolution of poecilogony from planktotrophy: Cryptic speciation, phylogeography, and larval development in the gastropod genus Alderia. Evolution 60(11):2293–2310.
    Emlet, R. B. 1994. Body form and patterns of ciliation in non-feeding larvae of echinoderms—functional solutions to swimming in the plankton. Am. Zool. 34:570–585.[Web of Science]
    Emlet, R. B. 1995. Developmental mode and species geographic range in regular sea urchins (Echinodermata: Echinoidea). Evolution 49:476–489.[Web of Science]
    Fioroni, P. 1966. Zur Morphologie und Embryogeneses des Darmtraktes und der transitorischen Organe bei Prosobranchiern (Mollusca, Gastropoda). Rev. Suisse Zool. 73:621–876.[Medline]
    Folmer, O., M. Black, W. Hoeh, R. Lutz, and R. Vrijenhoek. 1994. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol. Mar. Biol. Biotechnol. 3:294–299.[Medline]
    Gallardo, C. S. 1976. Historia natural y reproducción de Crepidula dilatata Lamarck en una población de Bahía Mehuín (Prov. Valdivia, Chile). Medio Ambiente 2:44–50.
    Gallardo, C. S. 1977a. Two modes of development in the morphospecies Crepidula dilatata (Gastropoda: Calyptraeidae) from Southern Chile. Mar. Biol. 39:241–251.
    Gallardo, C. S. 1977b. Crepidula philippiana n. sp. nuevo gastropodo Calyptraeidae de Chile con especial referencia al patron de desarrollo. Stud. Neotrop. Fauna Environ. 12:177–185.
    Gallardo, C. S. 1979. Especies gemelas del genero Crepidula (Gastropoda, Calyptraeidae) en la costa de Chile: una redescripción de C. dilatata Lamarck y descripción de C. fecunda n. sp. Stud. Neotrop. Fauna Environ. 14:215–226.
    Gallardo, C. S., and O. A. Garrido. 1987. Nutritive egg formation in the marine snails Crepidula dilatata and Nucella crassilabrum. Int. J. Invertebr. Reprod. Dev. 11:239–254.
    Gili, C., and J. Martinell. 1994. Relationship between species longevity and larval ecology in nassariid gastropods. Lethaia 27:291–299.[Web of Science]
    Hadfield, M. G., and D. K. Iaea. 1989. Velum of encapsulated veligers of Petaloconchus (Gastropoda), and the problem of re-evolution of planktotrophic larvae. Bull. Mar. Sci. 45:377–386.
    Hansen, T. A. 1978. Larval dispersal and species longevity in Lower Tertiary gastropods. Science 199:886–887.
    Hansen, T. A. 1980. Influence of larval dispersal and geographic distribution on species longevity in neogastropods. Paleobiology 6:193–207.[Abstract]
    Hart, M. W. 1996. Evolutionary loss of larval feeding: development, form, and function in a facultatively feeding larva, Brisaster latifrons. Evolution 50:174–187.[Web of Science]
    Hart, M.W., M. Byrne, and M. J. Smith. 1997. Molecular phylogenetic analysis of life-history evolution in asterinid starfish. Evolution 51:1848–1861.[Web of Science]
    Havenhand, J. N. 1995. Evolutionary ecology of larval types. Pp. 79–122 in Ecology of Marine Invertebrate Larvae, L. R. McEdward, ed. CRC Press, Boca Raton, FL.
    Hoagland, K. E., and R. Robertson. 1988. An assessment of poecilogony in marine invertebrates: phenomenon or fantasy? Biol. Bull. 174:109–125.[Abstract/Free Full Text]
    Hoskin, M. G. 1997. Effects of contrasting modes of larval development on the genetic structure of three species of prosobranch gastropods. Mar. Biol. 127:647–656.
    Huelsenbeck, J. P., and F. Ronquist. 2001. MrBayes 2.0: Bayesian Inference of Phylogeny. Distributed by the author. Department of Biology, University of Rochester and Department of Biology, University of Uppsala.
    Jablonski, D. 1986. Larval ecology and macroevolution of marine invertebrates. Bull. Mar. Sci. 39: 565–587.
    Jeffery, C. H., and R. B. Emlet. 2003. Macroevolutionary consequences of developmental mode in temnopleurid echinoids from the Tertiary of southern Australia. Evolution 57:1031–1048.[Web of Science][Medline]
    Jeffery, C. H., R. B. Emlet, and D. T. J. Littlewood. 2003. Phylogeny and evolution of development mode in temnopleurid echinoids. Mol. Phylogenet. Evol. 28:99–118.[Web of Science][Medline]
    Kupriyanova, E. K. 2003. Live history evolution in Serpulimorph polychaetes: a phylogenetic analysis. Hydrobiologia 496:105–114.[Web of Science]
    Kyle, C. J., and E. G. Boulding. 2000. Comparative population genetic structure of marine gastropods (Littorina spp.) with and without pelagic larval dispersal. Mar. Biol. 137:835–845.
    McEdward, L. 2000. Adaptive evolution of larvae and life cycles. Semin. Cell Dev. Biol. 11:403–409.[Web of Science][Medline]
    McEdward, L. R. 1995. Evolution of pelagic direct development in the starfish Pteraster tesselatus (Asteroidea: Velatida). Biol. J. Linn. Soc. 54:299–327.[Web of Science]
    Moran, A. L. 1999. Intracapsular feeding by embryos of the gastropod genus Littorina. Biol. Bull. 196:229–244.[Abstract]
    Ó Foighil, D. 1989. Planktotrophic larval development is associated with a restricted geographic range in Lasaea, a genus of brooding hermaphrodite bivalves. Mar. Biol. 103:349–358.
    Pagel, M. 1997. Inferring evolutionary processes from phylogenies. Zool. Scr. 26:331–348.[Web of Science]
    Pagel, M. 1999a. Inferring the historical patterns of biological evolution. Nature 401:877–884.
    Pagel, M. 1999b. The maximum likelihood approach to reconstructing ancestral character states of discrete characters on phylogenies. Syst. Biol. 48:612–622.[Free Full Text]
    Pernet, B. 2003. Persistent ancestral feeding structures in nonfeeding annelid larvae. Biol. Bull. 205:295–307.[Abstract/Free Full Text]
    Posada, D., and K. A. Crandall. 1998. Modeltest: testing the model of DNA substitution. Bioinformatics 14:817–818.[Abstract/Free Full Text]
    Posada, D., and K. A. Crandall. 2001. Selecting the best-fit model of nucleotide substitution. Syst. Biol. 50:580–601.[Abstract/Free Full Text]
    Raff, R. A. 1987. Constraint, flexibility, and phylogenetic history in the evolution of direct development in sea urchins. Dev. Biol. 119:6–19.[Web of Science][Medline]
    Rouse, G. W. 2000. Polychaetes have evolved feeding larvae numerous times. Bull. Mar. Sci. 67: 391–409.
    Scheltema, R. S. 1989. Planktonic and non-planktonic development among prosobranch gastropods and its relationships to the geographic ranges of species. Pp. 183–188 in Reproduction, Genetics, and Distributions of Marine Organisms, J. S. Ryland and P. A. Tyler, eds. Olsen and Olsen, Fredensborg, Denmark.
    Smith, A. B. 1997. Echinoderm larvae and phylogeny. Annu. Rev. Ecol. Syst. 28:219–241.[Web of Science]
    Strathmann, R. R. 1974. Introduction to function and adaptation in echinoderm larvae. Thalassia Jugosl. 10:321–339.
    Strathmann, R. R. 1978a. The evolution and loss of feeding larval stages in marine invertebrates. Evolution 32:899–906.
    Strathmann, R. R. 1978b. Progressive vacating of adaptive types during the Phanerozoic. Evolution 32:907–914.[Web of Science]
    Strathmann, R. R. 1985. Feeding and nonfeeding larval development and life history evolution in marine invertebrates. Annu. Rev. Ecol Syst. 16:339–361.[Web of Science]
    Strathmann, R. R. 1990. Why life histories evolve differently in the sea. Am. Zool. 30:197–207.[Web of Science]
    Swofford, D. L. 1998. PAUP*. Phylogenetic Analysis Using Parsimony (* and Other Methods). Ver. 4. Sinauer Associates, Sunderland, MA.
    Taki, I. 1938. Systematic study of Japanese species of Calyptraeidae. Venus 8:136–147.
    Véliz, D., F. Winkler, and C. Guisado. 2003. Developmental and genetic evidence for the existence of three morphologically cryptic species of Crepidula in northern Chile. Mar. Biol. 143:131–142.
    Werner, B. 1955. Über die Anatomie, die Entwicklung und Biologie des Veligers und der Veliconcha von Crepidula fornicata L. (Gastropoda, Prosobranchia). Helgol. Wiss. Meeresunters. 5:169–217.
    Wilke, T., and G. M. Davis. 2000. Infraspecific mitochondrial sequence diversity in Hydrobia ulvae and Hydrobia ventrosa (Hydrobiidae: Rissooidea: Gastropoda): Do their different life histories affect biogeographic patterns and gene flow? Biol. J. Linn. Soc. 70:89–105.[Web of Science]
    Wray, G. 1996. Parallel evolution of nonfeeding larvae in echinoids. Syst. Biol. 45:308–322.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Biol. Bull.Home page
P. J. Krug
Not My "Type": Larval Dispersal Dimorphisms and Bet-Hedging in Opisthobranch Life Histories
Biol. Bull., June 1, 2009; 216(3): 355 - 372.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
K. Domes, R. A. Norton, M. Maraun, and S. Scheu
Reevolution of sexuality breaks Dollo's law
PNAS, April 24, 2007; 104(17): 7139 - 7144.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (10)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Collin, R.
Right arrow Articles by Véliz, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Collin, R.
Right arrow Articles by Véliz, D.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS