Printer Friendly

Contrasting Metatrochal Behavior of Mollusc and Annelid Larvae and the Regulation of Feeding While Swimming.


Veliger larvae of molluscs and some annelid larvae capture food with opposed bands of cilia (Hatschek, 1878, 1880a, b, 1885; Nielsen, 1987; Hansen, 1993; Pernet et al., 2015; Pernet, 2018). A preoral band of cilia, the prototroch, encircles the larva and beats with an effective stroke from anterior to posterior. The prototroch produces the current for swimming and feeding. In feeding, the prototroch overtakes particles and removes them from the current that passes the swimming larva (Strathmann et al., 1972; Strathmann and Leise, 1979; Gallager, 1988; Romero et al., 2010; Pernet and Strathmann, 2011; Fig. 1), a process also occurring in other animals that capture particles on the downstream side of a ciliary band (Strathmann et al., 1972; Riisgard et al., 2000). Parallel to the prototroch, a postoral band of shorter cilia, the metatroch, encircles the body and beats with an effective stroke from posterior to anterior, opposing the beat of the prototroch. Algal cells and other particles are caught between these opposed ciliary bands. Also, some particles that are not intercepted by prototrochal cilia are intercepted by metatrochal cilia, and some particles that initially pass by the food groove then enter the food groove across the metatrochal band (Strathmann et al., 1972; Romero et al., 2010). Cilia in the food groove that is between the prototroch and the metatroch transport captured particles to the mouth.

The larvae reduce their ingestion of particles when they are satiated at high concentrations of food or when particles are noxious or non-nutritious (Pechenik and Fisher, 1979; Sprung, 1984;Gallager, 1988; Hansen and Ockelmann, 1991; Hansen, 1993; Juhl et al., 2008). Because the current produced by the beating prototroch serves for both swimming and particle capture, the means of dissociating swimming from the capture of food are not obvious. Rates of capture of particles by swimming mollusc larvae can be reduced by reducing extension of the velum (Chan et al., 2013), but reducing length of the exposed velar edge does not by itself entirely eliminate capture of particles. The larvae also reject captured particles. Particles that have reached the mouth can be rejected in a strand of mucus that passes posteriorly, along the developing foot in molluscs or the neurotroch of annelids, but this means of rejection, if prolonged, involves loss of mucus, and also the drag from a trailing mucous strand could increase capture of particles (Emlet, 1990; Fenchel and Ockelmann, 2002; Strathmann and Grunbaum, 2006).

Arrest of metatrochal cilia could be one way of regulating rate of particle capture during swimming. Trochophore larvae of serpulid annelids sometimes arrest beat of the metatroch while the prototroch beats (Strathmann et al., 1972; Lacalli, 1981, 1984). The same occurs in larvae of sabellid annelids that, although non-feeding larvae, capture particles between opposed prototroch and metatroch (Pernet, 2003). Although some particles can be captured and transported during metatrochal arrests of serpulid larvae (Lacalli, 1984), absence of opposed beat by the metatroch is expected to reduce capture rates by swimming larvae. The beat of metatrochal cilia is inferred to increase capture of particles by prototrochal cilia by increasing the difference between speed of effective strokes of the prototrochal cilia and speed of the water passing through this ciliary band (Emlet, 1990; Strathmann and Grunbaum, 2006) and also by aiding retention of particles in the food groove (Strathmann et al., 1972; Romero et al., 2010). However, the veliger larvae of mollu scs may not use metatrochal arrest as a way of reducing captures during swimming. Strathmann and Leise (1979) did not observe metatrochal arrests during prototrochal beat for larvae of a caenogastropod, an opisthobranch gastropod, and a bivalve. These observations suggested that molluscan veligers do not arrest metatrochal beat while swimming and do not regulate particle capture in this way.

Our first objective was to examine the generality of metatrochal arrest in larval annelids and non-arrest in larval molluscs during prototrochal beat. We recorded activity of metatrochal cilia of larvae of a larger sample of taxa during longer periods than in previous observations. Although our primary interest was in regulation of capture of particles while swimming, a difference between phyla in metatrochal behavior could also provide new data for discussions of evolution of the opposed-band feeding mechanism (Haszprunar et al., 1995; Rouse, 2000; Page, 2009; Nielsen, 2012; Pernet et al., 2015; Page and Hookham, 2017).

Our second objective was to observe the extent to which mollusc larvae could reduce particle captures while both prototroch and metatroch beat. Our hypothesis was that veligers can reduce captures per encounter between particle and cilium, thereby dissociating swimming and particle capture, even while both prototroch and metatroch beat. The ability is suggested by differences in the zone in which algal cells passing through the band of beating prototrochal cilia are caught. Some authors observed capture of only those algae passing the proximal part of the prototrochal cilia in their effective strokes (Strathmann and Leise, 1979; Gallager, 1988), but Romero et al. (2010) observed captures almost to the tips of the prototrochal cilia. This difference could indicate that the larvae can reduce the length of a prototrochal cilium that captures particles and thereby reduce captures per encounter. Gallager (1988) observed that for a bivalve veliger, the ratio of the number of cells reaching the mouth to the number of encounters was lower at high concentrations of algae. Gallager's estimate of encounters was concentration of algae times volume flow through the zone of captures of algae passing the velar edge. Gallager's observation could indicate either regulation of captures per encounter or losses from the food groove after capture. To test for reduced captures per encounter, we first confirmed that capture by interception of algae near the tips of prototrochal cilia occurs in veligers of a bivalve, as in the previous report for a gastropod; we then observed captures per encounter directly, at the velar edge, when algal cells were offered at a high concentration. Previous studies of this feeding mechanism by video microscopy have focused on capture of particles. We endeavored to examine larvae that produced a feeding current while not capturing particles, or capturing very few.

Materials and Methods


A video clip of a larva of Urechis caupo Fisher & MacGinitie, 1928 had been obtained at the Bodega Marine Laboratory for a different study (Miner et al., 1999). Other annelid larvae were obtained in the Salish Sea, NE Pacific, and maintained at the Friday Harbor Laboratories (FHL). Adults of Serpula columbiana Johnson, 1901 that were removed from their tubes released sperm and eggs that were then combined for fertilization. Adults of Notomastus sp. were picked by hand from shoveled mud, allowed to crawl free of sediment and mucus, then cut on posterior segments to release gametes, as in Pernet et al (2015). Sperm were added to eggs for fertilization. Larvae of polygordiid, sabellariid, and oweniid annelids were obtained by towing a plankton net near the dock of the FHL. A larva of Armandia brevis (Moore, 1906) was reared from an embryo in water containing spawning epitokes, attracted to the FHL dock with a submerged light. Identification of the larva of A. brevis, reared through metamorphosis, was confirmed by correspondence with figures in Hermans (1964).

Egg capsules of the gastropod Concholepas concholepas (Bruguiere, 1789) that had been deposited in a continuous-flow aquarium at the Estacion de Biologia Marina Abate Juan Ignacio Molina, of the Universidad Catolica de la Santisima Conception, Chile, were kept in jars in aerated seawater at 14 [degrees]C, at the Campus San Andres of the same university, where veligers hatched over several days. Other veligers were obtained from molluscs in or near the Salish Sea and maintained at the FHL. Egg capsules of Fusitriton oregonensis (Redfield, 1846) were laid in aquaria. Veligers of Crepidula fornicata (Linnaeus, 1758) were from broods of animals from southern Puget Sound, maintained in quarantine at the FHL. Egg capsules of Littorina sp. (Littorina scutulata Gould, 1849 or Littorina plena Gould, 1849) were obtained from adults that were collected from intertidal rocks and placed in mesh-sided chambers overnight in an aquarium, where they shed egg capsules. The preceding gastropod species are caenogastropods. Egg masses of the opisthobranch gastropods Melanochlamys diomedea (Bergh, 1894) and Haminoea vesicula (Gould, 1855) were collected intertidally and kept at room temperature until hatching. For bivalves, veligers of Ostrea lurida Carpenter, 1864 were released from broods of adults obtained from Totten Inlet, Puget Sound, and maintained at the FHL. Veligers of Crassostrea gigas (Thunberg, 1793) were at eyed stages and had been reared for a different experiment at the FHL. Unidentified bivalve veligers were obtained by plankton tows near the FHL dock.

Methods of maintaining larvae were as described by Strathmann (1987). Bivalve larvae from the plankton, larvae of C. concholepas, F. oregonensis, and the annelid larvae were maintained at 12-14 [degrees]C. The other larvae were maintained at room temperature, which ranged from 14 to 21 [degrees]C. The larvae were kept in seawater that had passed through either a 0.45-[micro]m membrane filter or a 30-[micro]m mesh.

Algae used as particles

Algae used for observations of feeding were Rhodomonas sp., Isochrysis galbana Parke, 1949 (T-ISO), Dunaliella tertiolecta Butcher, 1959, and Skeletonema sp. These algae were offered at high concentrations because an objective was to observe larvae that produced a current for swimming while reducing capture of particles. Some larvae were exposed to satiating concentrations of I. galbana for at least 12 h prior to observation. The concentrations exceeded 40,000 cells of I. galbana per milliliter, far exceeding concentrations that Sprung (1984) found to be satiating for veligers of Mytilus edulis. Others were maintained on slides long enough for satiation. The time to satiation is not long. Gallager (1988) observed a decline in ingestion in 6-9 minutes after a bivalve veliger was introduced to a suspension of algal cells. We added seawater with a suspension of algal cells to one edge of the cover glass and withdrew seawater and algal cells from the opposite edge.

Metatrochal activity

Observations were with a 20 x objective and bright-field illumination, with the condenser iris constricted to increase contrast. Analog video signals from the camera were imported into the program iMovie (Apple, Cupertino, CA) through an analog-to-digital converter. Selected clips were exported to QuickTime (Apple), then de-interlaced when converted into MPEG4 format. Video recordings were at 30 Hz. This low frequency permitted longer video records of behavior than in previous high-frequency recordings for studies of capture mechanisms. Longer records were necessary for a search for infrequent events. That advantage outweighed the disadvantage of inability to follow a single cilium through a beat cycle.

For video recordings, the annelid larvae and opisthobranch veligers were in a confined space between a slide and a cover glass supported at the corners with modeling clay. Some, with more vertical space, swam against an obstruction. Larvae of U. caupo swam within the space of a nylon mesh placed on a slide topped with a cover glass, so that the larvae were free to rotate and change orientation but not to move forward continuously.

Most veligers (C. concholepas, F. oregonensis, Littorina sp., O. lurida, and C. gigas) were glued to glass needles, which allowed a larger space, about 1.5-2 mm, between slide and cover glass. Glass needles were pulled from glass rods or Pasteur pipettes held in a flame of burning propane or alcohol. The needles were mounted on a small ball of modeling clay that allowed attachment of the needles to the rim of a beaker for immersion of veligers until use and then allowed positioning of each needle on a microscope slide for observations. The modeling clay allowed rotation of the needle to adjust the view of a veliger. To attach a veliger to a needle, we placed the veliger on a glass slide, withdrew most of the seawater, moved the veliger out of the drop with a glass needle, dipped the tip of a glass needle in a drop of cyanoacrylate glue, touched the needle's tip to the veliger's shell, lifted the veliger from the slide, and placed the tip of the needle with the veliger in a beaker with ~60 mL seawater. After several veligers on needles were in the beaker, we changed the water.

Arrests of metatrochal cilia of molluscs were inferred when cilia appeared in the same position for more than 6 successive frames (>0.2 s). Recorded arrests were much longer for the annelids. Activity of a long row of metatrochal cilia could be recorded in focus in cases where light passed through the larval body and an extensive part of the metatroch was near the focal plane. For other cases, the number of metatrochal cilia that could be observed simultaneously was limited by opacity of the larval body or convolution of the ciliary band. Because cilia near the mouth have a special role in rejection and ingestion of particles, cilia near the mouth were not included in observations of metatrochal arrest during prototrochal beat.

Captures per encounter

Direct observations of individual larvae are necessary to distinguish reduced rate of capture from later rejection of captured particles. Before observation, veligers of C. concholepas and C. gigas were glued to glass needles. For estimates of captures per encounter, velar edges were viewed as in Figure 2. Captures per encounter were also estimated for a trochophore of S. columbiana that had swum against the cover glass. Captures were easily observed and were defined as the number of encountered cells that entered the food groove and were then transported toward the mouth (Fig. 2). Cells that were lost from the food groove after initial transport along the food groove were counted as captures. The measure of encounters was the number of cells passing the beating prototrochal cilia within 60 [micro]m of the velar edge for C. concholepas, 50 [micro]m for C. gigas, and 30 [micro]m for S. columbiana. These distances were about 2/3 of the visible maximum reach of prototrochal cilia of each larva in its effective strokes. Two previous studies of veligers (Strathmann and Leise, 1979; Gallager, 1988) observed that captures of algal cells were frequent in the proximal part of the zone swept by prototrochal cilia, although captures can occur nearly to the ciliary tips (Romero et al., 2010). Our criterion for encounters therefore underestimated the number of passing cells that would have been captured at maximal clearance rates. Also, we may have missed cells that were far from the focal plane in successive frames, and some cells that passed close to recovery strokes may have been obscured. A possible opposite bias for the ratio of captures per encounter is that tethering could increase captures by increasing the velocity of prototrochal cilia relative to algal cells (Emlet, 1990). Overall, our estimate of captures per encounter is a conservative measure of how far captures per encounter can be reduced. Estimates of encounters and captures that are based on direct observations of algae passing through the prototrochal band are unaffected by changes in concentrations of algal cells as cells settle, adhere to the cover glass or slide, or are removed by the larva.

Encounters and captures were counted along 165 [micro]m of the velar edge for C. concholepas, 120 [micro]m for C. gigas, and 120 [micro]m for S. columbiana. Any 5-second interval with a velar contraction was skipped so that estimated captures per encounter were for veligers with velum extended and prototrochal cilia beating. Veligers' metatrochal cilia were in focus along part of the band much of the time and were beating whenever seen sufficiently in focus. The prototroch and metatroch of the trochophore were visible and beating throughout the periods of observation. The veliger of C. concholepas with cells of Rhodomonas sp. was observed continuously for about 13 minutes with no renewal of algal cells beneath the cover glass and a decreasing number of algal cells encountered. The same veliger was subsequently observed for four minutes with cells of D. tertiolecta. Total encounters were 784 with Rhodomonas sp. and 222 with D. tertiolecta. Encounters during observations with Rhodomonas sp. decreased from about 120 to 40 per minute. Encounters during observations with D. tertiolecta were 51-62 encounters each minute. A veliger of C. gigas was observed with cells of Rhodomonas sp. at intervals over 70 minutes with periodic additions of algal cells and subsequently observed for 1 minute with cells of Skeletonema sp. Total encounters were 3080 with Rhodomonas sp. and 173 with Skeletonema sp. The trochophore of S. columbiana was observed with J. galbana for 40 s with 171 encounters. For the veligers, counts of captures and encounters were averaged over 1-minute intervals for estimates of captures per encounter and reported as mean [+ or -] SD. For the trochophore the estimate was for the 40 s of observation.


Annelid metatrochal arrests

The metatrochs of annelid larvae of six families (Polygordiidae, Sabellariidae, Serpulidae, Opheliidae, Capitellidae, and Oweniidae) frequently arrested beat while the prototroch beat. Similar metatrochal arrests during prototrochal beat were visible at times in a brief video clip of a larva of the Echiuridae. In the polygordiid, sabellariid, serpulid, and opheliid, rows of metatrochal cilia arrested in alignment at or near the beginning of the effective stroke, pointing away from the food groove and positioned near the surface of the larval body (Fig. 3). Opacity of the body of the capitellid larva prevented observation of rows of metatrochal cilia because the metatrochal cilia were clearly visible only at the edge of the larval body, with the plane of focus an optical section across prototroch, food groove, and metatroch. The position of arrested metatrochal cilia was, however, similarly near the larval body with the cilia pointing away from the food groove. Also, simultaneous arrests of metatrochal cilia on both sides of the capitellid larva's body (Fig. 3E), as well as successive views of different arrested metatrochal cilia as the larva rotated, indicated that metatrochal arrests were extensive. The arrested metatrochal cilia of the oweniid were usually aligned and pointed away from the food groove, but they pointed outward away from the body's surface instead of arresting near the surface of the larval body (Fig. 3F-H). Because the oweniid larva's parallel bands of prototroch, food groove, and metatroch are on the edges of lobes of the larval body, arrest of metatrochal cilia next to the larval body would require a greater rotation away from the food groove than in the other annelid larvae observed here. The arrested metatrochal cilia of the oweniid were usually nearly straight (Fig. 3F) but sometimes curved (Fig. 3G, H). In the oweniid, metatrochal cilia frequently arrested throughout the length of ciliary band that was in focus, but the lengths of the band in focus were short because of the curvature of the ciliary bands. During a video clip of 13.8 s of a rotating echiurid larva, arrested metatrochal cilia were visible for 1.7 s and beating metatrochal cilia for 5.5 s while prototrochal cilia beat continuously.

Incidental observations of the polygordiid, serpulid, sabellariid, capitellid, and oweniid larvae showed that during prototrochal beat, the metatrochal cilia can arrest along one part of the band while continuing to beat elsewhere. In the video records, the metatroch usually arrested when the prototroch arrested, but in the polygordiid larva, in which prototrochal arrests were frequent, the metatrochal cilia were sometimes seen to move while the opposing prototroch cilia were arrested. Beat of cilia of the food groove was sufficient for transport within the food groove without beat of the metatroch, as seen with larvae of the oweniid, polygordiid, and serpulid. In one instance in the polygordiid, an algal cell was transported in the food groove past arrested cilia of both prototroch and metatroch. Occasionally, an algal cell was captured where part of the metatroch was arrested, with subsequent transport in the food groove, as was seen with larvae of the capitellid (Fig. 4) and also with the oweniid and the serpulid, confirming Lacalli's (1984) observations on a serpulid larva. The particle concentrations during these observations were often very high, however, and captured algal cells were a small fraction of those passing within reach of the beating prototrochal cilia.

Molluscan metatrochal arrests

Metatrochal cilia of veligers infrequently arrested while the prototrochal cilia beat. Finding a few metatrochal arrests during prototrochal beat required a total of 703 minutes of video recordings for the molluscan larvae. In contrast, numerous metatrochal arrests occurred during 285 minutes of video recordings for annelid larvae. Metatrochal arrests during prototrochal beat were detected for larvae of Concholepas concholepas and Crassostrea gigas but were not easily seen, in part because arrested metatrochal cilia often remained curved, a position resembling the recovery strokes (Fig. 5). Arrested metatrochal cilia of these larvae extended away from the larval body, often in a variety of positions. The indication that metatrochal cilia had arrested beat was that cilia remained in the same position in successive frames (Fig. 5). An instance of a more extensive arrest of metatrochal cilia during prototrochal beat occurred with a larva of Crepidula fornicata, viewed with the velar edge in the plane of focus; groups of metatrochal cilia appeared to be aligned near the beginning of the effective stroke but with some movement (Fig. 6). Metatrochal arrests during prototrochal beat were too infrequent to account for the low rates of captures per encounter observed when veligers were exposed to high concentrations of algal cells (see Low captures per encounter while swimming).

Incidental observations showed continued movement of metatrochal cilia during prototroch arrest in nearly all of the veligers. including larvae of caenogastropods, opisthobranchs, and bivalves. The one exception was with larvae of C. fornicata, for which there were no recordings of prototrochal arrests while metatrochal cilia were in focus. Larvae of C. concholepas, Fusitriton oregonensis, and Littorina sp. were seen to sometimes arrest metatrochal cilia during prototrochal arrests.

Low captures per encounter while swimming

Captures of algae by veligers of Ostrea lurida included captures near the tips of prototrochal cilia, confirming that particles can be captured through all or nearly all of the zone swept by prototrochal cilia. Captures near the tips of prototrochal cilia included cells of the flagellate Isochrysis galbana and the diatom Skeletonema sp. (Fig. 1) and also cells of Rhodomonas sp. Contact with the end of the diatom chain was sufficient for capture. Thus, in bivalve veligers, as well as with gastropod veligers (Romero et al., 2010), algae can be captured along nearly the entire length of prototrochal cilia beyond the zone of recovery strokes.

However, rates of capture of algal cells that passed through the zone swept by prototrochal cilia were low when there were large numbers of algal cells. The orientation of larvae and focus were adequate to show that captures per encounter were low while both the prototroch and the metatroch were beating in video recordings of veligers of the bivalves C. gigas and O. lurida, unidentified bivalve veligers from the plankton, veligers of the gastropods C. concholepas, C. fornicata, F. oregonensis, and Littorina sp., and larvae of the serpulid and oweniid annelids. The ability to greatly reduce captures while both prototroch and metatroch beat is widespread.

Rates of capture per encounter were estimated for veligers of the gastropod C. concholepas and the bivalve C. gigas from video recordings of algal cells passing the zone of effective strokes of prototrochal cilia (Fig. 2). Algal cells seen within the proximal 2/3 of effective strokes of prototrochal cilia were counted as encounters, and most of these cells passed the prototrochal band without retention while the metatroch and food groove cilia also beat. The veligers did, however, continue to capture algal cells between the prototrochal and metatrochal bands at a low rate while exposed to satiating concentrations of algal cells. Capture rates varied, but there was no prolonged period with no captures.

For the veliger of C. concholepas, exposed to high concentrations of cells of Rhodomonas sp., estimates of captures per encounter varied from minute to minute (Fig. 7), with mean [+ or -] SD of 0.088 [+ or -] 0.027 (n = 12). Although capture rates were low, captures continued (Fig. 7). The same veliger was observed 4 days later after a previous exposure of about 15 minutes to a satiating concentration of cells of Dunaliella tertiolecta. The veliger captured cells of D. tertiolecta at a rate of 0.056 [+ or -] 0.047 captures per encounter in each minute (n = 4). During one minute, there were no captures of D. tertiolecta in the observed area, but during this interval there were captures elsewhere on the velum.

The veliger of C. gigas was exposed to high concentrations of cells of Rhodomonas sp., with additional cells added between video recordings, so that numbers of cells passing through the observed zone were sustained over a longer interval (Fig. 7). Captures per encounter during each 1-minute interval of observation were 0.072 [+ or -] 0.023 (n = 11). Again, the capture rate did not decline to zero. Five minutes later, for the same veliger but with short chains of cells of Skeletonema sp., the ratio of captures per encounter was 0.092 during a minute of observation.

During the quantitative observations of C. concholepas and C. gigas, cilia of the metatroch were often in focus and beating while there were few captures per encounter. The low rates of capture per encounter that were observed for C. concholepas and C. gigas occurred with three kinds of algae. Cells of Rhodomonas sp. and D. tertiolecta were similar in size, between 10 and 15 [micro]m long and less than 10 [micro].m wide. Sizes for Skeletonema sp. varied greatly because of varying chain length and clumping of chains.

For a trochophore of S. columbiana, exposed to high concentrations of algal cells, captures per encounter were similarly low: 0.064. This annelid trochophore, like the molluscan veligers, was capable of a low rate of capture per encounter while prototroch and metatroch beat. There could be a differing bias in the quantitative estimates for the trochophore and the veligers because the trochophore was swimming against the cover glass, whereas the veligers were tethered farther away from this wall, presumably diminishing its influence on the current produced by the prototroch.

The initial paths of captured and non-captured algal cells were sometimes similar as they passed through the band of beating prototrochal cilia. To illustrate, captured and non-captured cells of Rhodomonas sp. passed the prototroch of C. gigas, with initially similar paths, within an interval of three seconds (Fig. 8). Passage on flow lines close to the recovery strokes did not assure capture.

Other means of reducing captures or ingestion

Reducing captures per encounter was not the only means by which the larvae reduced captures or ingestion while the prototroch produced a current for swimming and the metatroch beat. An unidentified bivalve veliger and veligers of C. gigas and Littorina sp. sometimes produced a current with the velum only partly extended, so that along part of the velar edge, the bases of the prototrochal cilia were at the edge of the shell. In this posture, algal cells passed through the zone of prototrochal effective strokes, indicating a current for swimming, but the algal cells were not captured at that part of the velar edge (Fig. 9). Keeping the food groove and metatrochal cilia within the shell could reduce captures along part of the velar edge while permitting swimming, but the entire velar edge cannot be placed at the shell aperture so as to expose prototrochal cilia while sequestering all of the opposing metatrochal cilia.

Larvae of both annelids and molluscs lost captured particles from the food groove. The processes governing retention or loss during transport along the food groove are obscure. We did not assess the effect of the infrequent velar contractions that narrowed the food groove or the effect of other contractions that change shape or position of the food groove.

Particles are sometimes rejected at the mouth even while others are being captured, as reported previously by Gallager (1988). As an example, a veliger of O. lurida captured a cell of I. galbana on the dorsal side of its velum while another cell arrived at its mouth via the food groove and was rejected (Fig. 10).


Differences between annelids and molluscs

The annelid and mollusc larvae differed in metatrochal behavior. Arrests of the metatroch during beat of the prototroch were frequent in the annelid larvae but infrequent to unobserved in the mollusc larvae. In the annelid larvae, arrested metatrochal cilia were commonly aligned in long rows, at or near the position of the start of an effective stroke. In the molluscs, when metatrochal cilia arrested, they were usually in a variety of positions, often with other metatrochal cilia moving nearby.

In previous observations, arrests of metatrochal cilia during prototrochal beat were reported for serpulid larvae (Strathmann et al., 1972; Lacalli, 1981, 1984) and sabellid larvae (Pernet, 2003) in the annelids but were not seen in larvae of two gastropods and a bivalve (Strathmann and Leise, 1979). Our observations extended the sample of taxa and observations of metatrochal behavior. Opposed prototrochal and metatrochal bands are known from 10 families of annelids (Pernet et al., 2015). Our observations, together with the sabellids observed by Pernet (2003), include larvae in eight of these families. In these eight families, metatrochs arrested beat during prototrochal beat. In seven of the eight families, the arrested metatrochal cilia were positioned near the larval body. Arrested metatrochal cilia of the oweniid often pointed outward from, rather than lying against, the larval body, but nevertheless frequently arrested while the prototroch beat. In contrast, in the larvae of the gastropod and bivalve molluscs, the metatroch rarely arrested when the prototroch was beating. In these molluscs, metatrochal arrests during prototrochal beat were infrequent, usually brief, and usually with the arrested metatrochal cilia in a variety of positions rather than aligned near the larval body.

Further observations will test the adequacy of our sample of taxa and expand the known behavioral repertoire of opposed ciliary bands. Our observations did not permit accurate quantitative comparisons of frequency and duration of metatrochal arrests, because arrests of the molluscs' metatrochs were not only infrequent but also so limited as to be easily missed when they did occur. Indeed, our initial impression was that the veligers of molluscs did not arrest metatrochal cilia while the prototroch beat. It was through repeated viewing of video recordings of the veligers that we discovered metatrochal arrests during prototrochal beat. In contrast, for the annelid larvae, metatrochal arrests during prototrochal beat were obvious.

Mollusc larvae were glued to glass needles, and annelid larvae were not, but we cannot account for the interphylum differences by this difference in methods. Some annelid larvae were held between cover glass and slide, some had space to move, and some had swum upward against the cover glass; yet in all of these conditions, metatrochal arrests during prototrochal beat were frequent, of long duration, and with metatrochal cilia aligned and positioned near the beginning of an effective stroke. Tethering did not appear to restrict other aspects of the ciliary behavior of the molluscan veligers. They arrested their cilia and moved velar lobes in varied ways but did not arrest metatrochal cilia during prototrochal beat in the manner of the annelid larvae.

One hypothesis for this difference between phyla is that some difference between annelid and mollusc larvae favors the differences in metatrochal arrest during prototrochal beat. One difference is that the velum of molluscan larvae is capable of a greater range of movements, but it is as yet unclear how that capability would eliminate advantages of prolonged and orderly arrests of the metatroch during prototrochal beat.

Another hypothesis is that the difference in metatrochal behavior is an accident of ancestry rather than arising from different functional requirements associated with such traits as a shell or velar lobes. Evidence of ancestral presence or absence of a metatroch in larvae is indirect. The hypotheses involve uncertainties about homology, phylogeny, bias in evolutionary transitions, and interpretation of fossils. If opposed prototrochal and metatrochal bands evolved several times within each phylum, as suggested by Haszprunar et al. (1995) and Rouse (1999, 2000), then consistent differences between the phyla as an accident of ancestry would be improbable. If opposed prototrochal and metatrochal bands evolved before divergence of annelids and molluscs, as suggested by Jagersten (1972) and Nielsen (2009, 2012), then a single divergence could have occurred after these phyla diverged, yet early in the history of each phylum. An intermediate hypothesis, apparently without advocates, is a single origin within each phylum.

Cell lineages appear to be unreliable indicators of homology or homoplasy of metatrochs. An argument for multiple origins of the metatroch was its formation by cells of the second quartet in the caenogastropod Crepidula fomicata (Hejnol et al., 2007; Henry et al., 2007), in contrast to its formation by cells of the third quartet in the annelid Polygordius (Woltereck, 1904); but the metatroch of the caenogastropod Ilyanassa obsoleta is formed by cells of both the second and third quartet (Gharbiah et al., 2013).

The fossil record suggests absence of feeding by larval molluscs of the Cambrian, which in turn suggests absence of a metatroch in ancestral gastropods and bivalves. Absence of larval growth has been inferred from larger shell apices and therefore larger initial sizes of Cambrian molluscs than of Ordovician molluscs (Niitzel et al., 2006). Freeman and Lundelius (2007) contested the inference of no feeding larvae in Cambrian molluscs, in part because the measurements were from internal casts of shells; and it is the external surface of the shell that can indicate boundaries of the larval shell. Nevertheless, there appears to be no evidence for feeding molluscan larvae in the Cambrian (Runnegar, 2007). If larval feeding had preceded shell formation in ancestral molluscs, as it does today in disciniscid brachiopods (Chuang, 1977; Freeman and Lundelius, 1999), then shell apices would not reflect initial size of larvae; but if the common ancestor of bivalves and gastropods had a shell at the earliest larval stages (Kristof et al., 2015), that would preclude larval feeding before shell formation.

Inferences of presence or absence of an ancestral metatroch have also been based on the distribution of metatrochs in presently existing annelids and molluscs, inferred relationships within these taxa, and the assumption of weak bias in evolutionary loss and gain of the metatroch and opposed-band feeding (Haszprunar et al., 1995; Rouse, 1999). However, a model of strength of bias in the evolutionary transitions is necessary for such inferences (Keever and Hart, 2008), and the strength of bias in evolutionary origins and losses of metatrochs has not been established. Hypothetical evolutionary routes to independent acquisition of opposed-band feeding have been proposed for molluscs (Hadfield et al., 1997) and annelids (Miner et al., 1999), but these hypotheses do not indicate probabilities of evolutionary gains relative to losses.

A complication for assessing gain and loss of opposed-band feeding is that evolutionary loss of a feeding planktonic larval stage does not necessarily imply loss of a functional metatroch and food groove. Some annelid larvae, as in the Sabellidae, have a metatroch and food groove and capture and transport particles, but they lack a mouth and do not ingest particles (Pernet, 2003). Encapsulated veligers of many gastropods retain prototroch, metatroch, and food groove, with which they feed on particles from nurse eggs; and an evolutionary return to planktotrophy is inferred for a calyptraeid whose ancestor lacked a planktotrophic stage but fed with opposed prototroch and metatroch during its encapsulated development (Collin et al., 2007).

Changes in inferred relationships of annelid families (Weigert et al., 2014), together with the existence of metatrochs in more families than assumed by Rouse (1999), change the number of implied losses or gains of a metatroch (Pernet and Strathmann, 2011; Pernet et al., 2015); but the hypothesis that the metatroch evolved once in the annelids still implies numerous losses. An inferred phylogeny for molluscan classes that puts gastropods, bivalves, and scaphopods in a monophyletic group (Kocot, 2013) reduces the implied number of losses of a metatroch if a metatroch originated in the ancestor of that clade; but implied losses still outnumber gains because numerous clades within the Bivalvia and Gastropoda lack metatrochs. Within each class, some clades without metatrochs diverged early: protobranch bivalves (Gonzalez et al., 2015), patellogastropods, and vetigastropods (Zapata et al., 2014); and many also diverged later. Inferences about ancestral metatrochs that are based on parsimony or likelihood models remain uncertain, primarily because of uncertainty about probabilities of gains relative to losses (Strathmann and Eernisse, 1994).

It is unclear to us what differences in neural anatomy and inputs underlie the observed difference in metatrochal behaviors of annelids and molluscs. In our observations, beat and arrest of prototroch and metatroch were, to some degree, independent in molluscan larvae, even though the metatroch seldom arrested while the prototroch beat. Braubach et al. (2006) found different responses of prototroch and metatroch to catecholamines, an indication that these ciliary bands are differently affected by neural inputs.

Swimming without particle capture

Previous direct observations of capture of particles by molluscan larvae aimed for high rates of captures per encounter while larvae were confined or tethered. We had the easier task of aiming for low rates of capture per encounter. With high concentrations of algal cells, captures per encounter were less than 0.1 for algal cells passing the proximal 2/3 of the zone swept by prototrochal cilia of larvae of a gastropod, bivalve, and annelid. Low rates of capture per encounter were also indicated by qualitative observations of other larvae in this study and of a tethered larva of a sabellariid annelid (Pernet and Strathmann, 2011). One indication that captures per encounter less than 0.1 are much lower than with maximal feeding is that captures per encounter of 0.15 to 0.44 were calculated for veligers that were capturing algal cells at less than maximal rates (Strathmann and Leise, 1979). In that calculation, encounters were estimated from the entire zone swept by prototrochal cilia, but the veligers were capturing only those particles encountering the proximal part of prototrochal cilia. An ability to capture at a higher rate is indicated by captures toward the tips of prototrochal cilia by veligers of both gastropods (Romero et al., 2010) and bivalves (this study).

The processes for regulating captures per encounter are unknown. One possibility is changes in secretion of materials that enhance or prevent adhesion of cilia to particles (Hermans, 1983). At least some captures are by direct interception, in which a prototrochal cilium overtakes a particle and adheres to it (Romero et al., 2010). Secretory-like structures appear near the prototroch in transmission electron micrographs of veligers of gastropods (L. Page, University of Victoria, pers. comm., 2013). Changes in adhesion between cilium and particle could regulate capture. Another possibility for regulating captures per encounter is differences in ciliary behavior that are subtler than beat and arrest. Some aspects of form and behavior of cilia were outside the scope of our study. Prototrochs have more than one row of cilia. In some larvae, there appeared to be more than one row of metatrochal cilia capable of differences in beat, but observations were too few for confirmation.

As far as sampling of taxa and behaviors permitted, our observations demonstrated that annelid and mollusc larvae differ in metatrochal arrest during prototrochal beat; but in both phyla, larvae with opposed bands can decouple swimming from particle capture almost completely, while the opposed prototrochal and metatrochal ciliary bands are fully deployed and beating.


This research was supported by Fondo Nacional de Dessarollo Cientifico y Tecnologico (FONDECYT 1170598), the National Science Foundation (OCE0623102), funding for Special Academic Activities from the Universidad Catolica de la Santfsima Conception (UCSC), the Faculty of Science of the UCSC, the Estacion de Biologfa Marina Abate Juan Ignacio Molina of UCSC, and the Friday Harbor Laboratories (FHL) of the University of Washington. Bruno Pernet provided the capitellid. The FHL Marine Zoology class provided the net tow with the polygordiid larvae. We are grateful for the advice and help of Daniel Gillon, B. Holthuis, Ricardo Otaiza O'Ryan, Nicole Padilla, Louise Page, Bruno Pernet, and staff at UCSC and FHL.

Literature Cited

Braubach, O. R., A. J. G. Dickinson, C. C. E. Evans, and R. P. Croll. 2006. Neural control of the velum in larvae of the gastropod, Ilyanassa obsoleta. J. Exp. Biol. 209: 4676-1689.

Chan, K. Y. K., H. Jiang, and D. K. Padilla. 2013. Swimming speed of larval snail does not correlate with size and ciliary beat frequency. PLoS One 8: e82764.

Chuang, S.-H. 1977. Larval development in Discinisca (inarticulate brachiopod). Am. Zool. 17: 39-53.

Collin, R., O. R. Chaparro, F. Winkler, and D. Veliz. 2007. Molecular phylogenetic and embryological evidence that feeding larvae have been reacquired in a marine gastropod. Biol. Bull. 212: 83-92.

Emlet, R. B. 1990. Flow fields around ciliated larvae: the effects of natural and artificial tethers. Mar. Ecol. Prog. Ser. 63: 211-225.

Fenchel, T., and K. W. Ockelmann. 2002. Larva on a string. Ophelia 56: 171-178.

Freeman, G., and J. W. Lundelius. 1999. Changes in the timing of mantle formation and larval life history traits in linguliform and craniiform brachiopods. Lethaia 32: 197-217.

Freeman, G., and J. W. Lundelius. 2007. Origin of planktotrophy--evidence from early molluscs: a response to Niitzel et al. (2006). Evol. Dev. 9: 307-310.

Gallager, S. M. 1988. Visual observations of particle manipulation during feeding in larvae of a bivalve mollusc. Bull. Mar. Sci. 43: 344-365.

Gharbiah, M., A. Nakamoto, and L. M. Nagy. 2013. Analysis of ciliary band formation in the mollusc Ilyanassa obsoleta. Dev. Genes Evol. 223: 225-235.

Gonzalez, V. L., S. C. Andrade, R. Bieler, T. M. Collins, C. W. Dunn, P. M. Mikkelsen, J. D. Taylor, and G. Giribet. 2015. A phylogenetic backbone for Bivalvia: an RNA-seq approach. Proc. R. Soc. Biol. Sci. B 282: 20142332.

Hadfleld, M. G., M. F. Strathmann, and R. R. Strathmann. 1997. Ciliary currents of non-feeding veligers in putative basal clades of gastropods. lnvertebr. Biol. 116: 313-321.

Hansen, B. 1993. Aspects of feeding, growth and stage development by trochophora larvae of the boreal polychaete Mediomastus fragile (Rasmuissen) (Capitellidae). J. Exp. Mar. Biol. Ecol. 166: 273-288.

Hansen, B., and K. W. Ockelmann. 1991. Feeding behaviour in larvae of the opisthobranch Philine aperta. I. Growth and functional response at different developmental stages. Mar. Biol. 111: 255-261.

Haszprunar, G., L. V. Salvini-Plaweri, and R. M. Rieger. 1995. Larval planktotrophy--a primitive trait in the Bilateria? Acta Zool. 76: 141-154.

Hatschek, B. 1878. Studien uber Entwicklungsgeschichte der Anneliden. Arb. Zool. Inst. Univ. Wien 1: 277-404.

Hatschek, B. 1880a. Ueber Entwicklungsgeschichte von Echiurus. Arb. Zool. Inst. Univ. Wien 3: 45-78.

Hatschek, B. 1880b. Ueber Entwicklungsgeschichte von Teredo. Arb. Zool. Inst. Univ. Wien 3: 1-44.

Hatschek, B. 1885. Entwicklung der Trochophora von Eupomatus uncinatus, Philippi (Serpula uncinata). Arb. Zool. Inst. Univ. Wien 6: 121-148.

Hejnol, A., M. Q. Martindale, and J. Q. Henry. 2007. High-resolution fate map of the snail Crepidula fomicata: the origins of ciliary bands, nervous system, and muscular elements. Dev. Biol. 305: 63-76.

Henry, J. Q., A. Hejnol, K. J. Perry, and M. Q. Martindale. 2007. Homology of ciliary bands in spiralian trochophores. Integr. Cotnp. Biol. 47: 865-871.

Hermans, C. O. 1964. The reproductive and developmental biology of the opheliid polychaete, Armandia brevis. M.S. thesis, University of Washington. Seattle.

Hermans, C. O. 1983. The duo gland adhesive system. Oceanogr. Mar. Biol. 21: 283-339.

Jagersten, G. 1972. Evolution of the Metazoan Life Cycle. Academic Press. London.

Juhl, A. R., C. A. Martins, and D. M. Anderson. 2008. Toxicity of Alexandrium lusitanicum to gastropod larvae is not caused by paralytic-shellfish-poisoning toxins. Harmful Algae 7: 567-573.

Keever, C. C., and M. W. Hart 2008. Something for nothing? Reconstruction of ancestral character states in asterinid sea star development. Evol. Dev. 10: 62-73.

Kocot, K. M. 2013. Recent advances and unanswered questions in deep molluscan phylogenetics. Am. Malacol. Bull. 31: 195-208.

Kristof, A., A. E. Luiz, A. L. De Oliveira, K. G. Kolbin, and A. Wanninger. 2015. Neuromuscular development in Patellogastropoda (Mollusca: Gastropoda) and its importance for reconstructing ancestral gastropod bodyplan features. J. Zool. Syst. Evol. Res. 54: 22-39.

Lacalli, T. C. 1981. Structure and development of the apical organ in trochophores of Spirobranchus polycerus, Phyllodoce maculata, and Phyllodocemucosa (Polychaeta).Proc. R. Soc. Biol. Sci. B212: 381-402.

Lacalli, T. C. 1984. Structure and organization of the nervous system in the trochophore larva of Spirobranchus. Philos. Trans. R. Soc. B Biol. Sci. 306: 79-135.

Miner, B. G., E. Sanford, R. R. Strathmann, B. Pernet, and R. B. Emlet. 1999. Functional and evolutionary implications of opposed bands, big mouths, and extensive oral ciliation in larval opheliids and echiurids (Annelida). Biol. Bull. 197: 14-25.

Nielsen, C. 1987. Structure and function of metazoan ciliary bands and their phylogenetic significance. Acta Zool. 68: 205-262.

Nielsen, C. 2009. How did indirect development with planktotrophic larvae evolve? Biol. Bull. 216: 203-215.

Nielsen, C. 2012. How to make a protostome. Invertebr. Syst. 26: 25-40.

Niitzel, A., O. Lehnert, and J. Fryda. 2006. Origin of planktotrophy--evidence from early molluscs. Evol. Dev. 8: 325-330.

Page, L. R. 2009. Molluscan larvae: pelagic juveniles or slowly metamorphosing larvae? Biol. Bull. 216: 216-225.

Page, L. R., and B. Hookham. 2017. The gastropod foregut--evolution viewed through a developmental lens. Can. J. Zool. 95: 227-238.

Pechenik, J. A., and N. S. Fisher. 1979. Feeding, assimilation, and growth of mud snail larvae, Nassarius obsoletus (Say), on three different algal diets. J. Exp. Mar. Biol. Ecol. 38: 57-80.

Pernet, B. 2003. Persistent ancestral feeding structures in nonfeeding annelid larvae. Biol. Bull. 205: 295-307.

Pernet, B. 2018. Larval feeding: mechanisms, rates, and performance in nature. Pp. 87-102 in Evolutionary Ecology of Marine Invertebrate Larvae. T. J. Carrier, A. M. Reitzel, and A. Heyland, eds. Oxford University Press, Oxford.

Pernet, B., and R. R. Strathmann. 2011. Opposed ciliary bands in the feeding larvae of sabellariid annelids. Biol. Bull. 220: 186-198.

Pernet, B., L. H. Harris, and P. Schroeder. 2015. Development and larval feeding in the capitellid annelid Notomastus cf. tenuis. Biol. Bull. 228: 25-38.

Riisgard, H. U., C. Nielsen, and P. S. Larsen. 2000. Downstream collecting in ciliary suspension-feeders: the catch-up principle. Mar. Ecol. Prog. Ser. 207: 33-51.

Romero, M. R., H. C. P. Kelstrup, and R. R. Strathmann. 2010. Capture of particles by direct interception by cilia during feeding of a gastropod veliger. Biol. Bull. 218: 145-159.

Rouse, G. W. 1999. Trochophore concepts: ciliary bands and the evolution of larvae in spiralian Metazoa. Biol. J. Linn. Soc. 66: 411-464.

Rouse, G. W. 2000. Polychaetes have evolved feeding larvae numerous times. Bull. Mar. Sci. 67: 391-409.

Runnegar, B. 2007. No evidence for planktotrophy in Cambrian molluscs. Evol. Dev. 9: 311-312.

Sprung, M. 1984. Physiological energetics of mussel larvae (Mytilus edulis). II. Food uptake. Mar. Ecol. Prog. Ser. 17: 295-305.

Strathmann, M. F. 1987. Reproduction and Development of Marine Invertebrates of the Northern Pacific Coast. University of Washington Press. Seattle.

Strathmann, R. R., and D. J. Eernisse. 1994. What molecular phylogenies tell us about the evolution of larval forms. Am. Zool. 34: 502-512.

Strathmann, R. R., and D. Grunbaum. 2006. Good eaters, poor swimmers: compromises in larval form. Integr. Comp. Biol. 46: 312-322.

Strathmann, R. R., and E. Leise. 1979. On feeding mechanisms and clearance rates of molluscan veligers. Biol. Bull. 157: 524-535.

Strathmann, R. R., T. L. Jahn, and J. R. C. Fonseca. 1972. Suspension feeding by marine invertebrate larvae: clearance of particles by ciliated bands of a rotifer, pluteus. and trochophore. Biol. Bull. 142: 505-519.

Weigert, A., C. Helm, M. Meyer, B. Nickel, D. Arendt, B. Hausdorf, S. R. Santos, K. M. Halanych, G. Purschke, C. Bleidorn, and T. H. Struck. 2014. Illuminating the base of the annelid tree using transcriptomics. Mol. Biol. Evol. 31: 1391-1401.

Woltereck, R. 1904. Beitriige zur praktischen Analyse der Polygordius-Entwicklung nach dem "Nordsee-" und dem "Mittelmeer-Typus." Arch. Entwickl. Org. 18: 377-403.

Zapata, F., N. G. Wilson, M. Howison, S. C. Andrade, K. M. Jorger, M. Schrodl, F. Goetz, G. Giribet, and C. W. Dunn. 2014. Phylogenomic analyses of deep gastropod relationships reject Orthogastropoda. Proc. R. Soc. Biol. Sci. B 281: 20141739.


(1) Friday Harbor Laboratories and Department of Biology, University of Washington, Friday Harbor, Washington 98250; (2) Departamento de Ecologia, Facultad de Ciencias, Universidad Catolica de la Santisima Concepcion, Casilla 297, Conception, Chile; (3) Centro de Investigacion en Biodiversidad y Ambientes Sustentables (CIBAS), Universidad Catolica de la Santisima Concepcion, Casilla 297, Concepcion, Chile; and Centro i~mar, Universidad de Los Lagos, Camino Chinquihue Km 6, Puerto Montt, Chile

Received 30 August 2018; Accepted 14 November 2018; Published online 13 February 2019.

(*) To whom correspondence should be addressed. Email:

Abbreviation: FHL, Friday Harbor Laboratories.

DOI: 10.1086/701730
COPYRIGHT 2019 University of Chicago Press
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2019 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Strathmann, Richard R.; Brante, Antonio; Oyarzun, Fernanda X.
Publication:The Biological Bulletin
Date:Apr 1, 2019
Previous Article:Larvae of Caribbean Echinoids Have Small Warming Tolerances for Chronic Stress in Panama.
Next Article:GABA as a Neurotransmitter in Gastropod Molluscs.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters