Printer Friendly

The distribution of marine invertebrate larvae near vertical surfaces in the Rocky subtidal zone.

INTRODUCTION

The question of how far larvae disperse away from their point of release is fundamental to understanding community dynamics in marine benthic systems (Connell 1985, Sebens 1985). Three questions are relevant: (1) how long does a larva remain in the plankton? (2) how much influence, by swimming, does the larva have on where it goes? and (3) what is the water flow at the time of release and how will it influence the larva's ultimate destiny? The distance larvae disperse depends on the speed and direction of currents and the length of time a larva can spend in the water column. This dispersal distance is important for a number of reasons: (1) it increases the potential to genetically link separated populations of adults (Scheltema 1971, 1986), (2) it allows recolonization of distant areas subjected to disturbance (Crisp 1974, Day and McEdward 1984), and (3) it plays a major role in determining the likelihood of speciation and extinction events (Scheltema 1977, 1986, Jablonski and Lutz 1983, Jablonski 1986). However, planktonic dispersal has a negative side, i.e., the longer a larva spends in the plankton, the lower its probability of survival (Thorson 1950, Strathmann 1974). There must be a balance between increasing the probability of finding a good settlement site and decreasing the probability of mortality in the water column. There is evidence that planktonic larvae have some control over the length of their planktonic phase via the ability to delay metamorphosis (for review, see Pechenik 1990).

The small size of most invertebrate larvae and the extreme difficulty of in situ study make direct observations of the planktonic stage laborious, and frequently impossible. Historically, the larval stage of invertebrates has been a "black box." Despite these difficulties, advances in sampling and observing planktonic marine invertebrate larvae have been made. Attempts to stain or mark larvae and track them in the field (radioactive isotopes or genetic markers) have produced few successful studies of larval movements in marine invertebrates (European oyster larvae, Millar 1961). The difficulties associated with dilution of marked larvae, the inability to sample synoptically, and the enormous amount of labor involved in finding marked larvae in plankton samples have discouraged the application of release and recovery techniques (Levin 1990). One obvious method to study planktonic larvae in situ is to directly observe their behavior and travel distance. This is feasible only if the larvae are very large and spend a short time in the plankton before settling. Olson (1985) and Young (1986), for example, both used scuba to directly observe ascidian larvae throughout their planktonic stage.

Based on settling plate experiments, Keough (1983) inferred three possible distributions of larvae in the plankton: (1) no patchiness, i.e., a mixed larval "soup" transported by currents, (2) small-scale patchiness on a scale of centimetres to metres resulting in larval swarms, and (3) large-scale patchiness on a scale of kilometres. Cyprids of some barnacle species maintain a precise vertical level relative to the water surface prior to settlement. This vertical distribution pattern is reflected in the settlement pattern for species on adjacent intertidal rocks in very protected habitats (Grosberg 1982). Gaines et al. (1985) found that spatial variation in larval concentrations of Balanus glandula was a cause of spatial settlement variation. The presence of such planktonic patterns of larvae clearly indicates that settlement surfaces are not being washed by a uniform bath of larvae. Each experimental site and each larval type must be examined individually before generalizations can be made.

It is often difficult to reconcile data on potential dispersal of larvae, based on swimming speeds and length of time they are competent, with their realized dispersal. Grosberg and Quinn (1986) found that, in a quiet harbor, the dispersal of Botryllus schlosseri was within centimetres of adult colonies. Olson and McPherson (1987) found that the realized dispersal of the ascidian Lissoclinum patella on the Great Barrier Reef was an order of magnitude less than its potential dispersal, and Svane and Havenhand (1993) described very localized dispersal of larvae for the ascidian Ciona intestinalis. Conversely, J. Witman and C. Arnold (unpublished manuscript) found that larvae of some bryozoans and sponges could be carried many kilometres from their origin on one or two tidal cycles.

The purpose of this project was to examine realized dispersion for a range of larval types. This was accomplished by quantifying patterns of larval distribution at various distances from large subtidal rock walls covered with encrusting invertebrates and coralline algae. Three walls were studied over a 2-yr period. Plankton were sampled at different distances from the rock walls using a plankton pump that collected from a 1-5 cm layer of water (Sebens and Maney 1992). Three hypotheses form the basis for this study: (1) differences in the time larvae must spend in the plankton result in distinct patterns of larval distribution near substrata, (2) larvae are found in similar densities over crustose coralline algae-covered rock surfaces as over nearby (metres or less) invertebrate-covered areas. Concerning this hypothesis, Breitburg (1984) found that certain larvae were less likely to settle on crustose-coralline algae, presumably because of algal deterrents. (3) "Layering" of larvae near rock surfaces is less evident on days of high flow compared to days of low flow; increased flow and mixing could facilitate larvae moving away from the walls and into dispersive currents. Animals with larvae that spend a very short time (seconds, minutes) in the water may not get far off the substratum into dispersive currents and, therefore, would be collected in higher densities a few centimetres from the rock surface. There may also be a similar pattern for larvae that spend a long time in the plankton, such as mussel and barnacle larvae, and then enter a search phase. These late-stage (e.g., pediveliger or cyprid) larvae may also be found at high densities close to rock surfaces regardless of their place of origin because they accumulate as they search for an appropriate settlement site.

METHODS

Sampling was conducted from September 1989 to November 1991 on three subtidal rock walls at the Shag Rocks, East Point, Nahant, Massachusetts. The exposed side of the Shag Rocks faces incoming swells during most wind conditions. Wave-induced oscillatory flow is generally strong here (Sebens 1985) with water movement parallel to subtidal rock walls. The sampling locations were large vertical to slightly undercut rock walls perpendicular to shore (not facing the incoming swells) up to 3 m high and 10 m long. These walls are covered with encrusting invertebrates including the following common species: anemones (Metridium senile), octocorals (Alcyonium siderium), ascidians (Aplidium glabrum, Molgula citrina, M. manhattensis, Didemnum albidum, and Dendrodoa carnea), sponges (Halichondria panicea, H. bowerbankii, Halisarca dujardini, Cliona celata, Leucosolenia cancellata, Isodictya spp., and Haliclona oculata), barnacles (Balanus balanus), tube worms (Spirorbis spp.), encrusting bryozoans and mussels (Modiolus modiolus and Mytilus edulis) (Sebens 1985). There are also patches ranging in size up to [greater than]1 [m.sup.2] area dominated by three species of crustose coralline algae: Lithothamnion glaciale, L. lemoinae, and Phymatolithon rugulosum (Sebens 1986).

The community dynamics of the benthos have been studied extensively at this site (Sebens 1986); there is a competitive dominance hierarchy, with patches of open space created by small-scale disturbances, typically by predators. These communities are characterized by intense competition for space. Encrusting species are constantly being overgrown, removed by predators, recruiting into cleared space, and losing or gaining space from competitors. Despite this intense space competition, the diversity of species is high. One exception to this diversity is the production of large patches of rock grazed by herds of sea urchins (Sebens 1985, 1986); these "barrens" are dominated by crustose coralline algae.

All three study sites are at [approximately equal to]8-10 m depth. Wall 1 was a long regular wall 2-5 m high and 10 m long, covered with encrusting invertebrates and with large ([greater than]2 [m.sup.2]) patches of crustose coralline algae. Most of the algal crust patches were on the 2-m tall end of the wall, but some were interspersed with the invertebrate-covered patches. Wall 2 was somewhat concave, 6 m long, and 2-3 m tall. Patches of coralline algae were smaller than on wall 1 ([less than]2 [m.sup.2]) and were interspersed with invertebrate-covered areas. Wall 3 was flat and regular but also somewhat concave on one end, 5 m long, and 3-7 m high. Patches of crust were also small here ([less than]2 [m.sup.2]) and mixed with invertebrate-covered areas.

Three walls, and three seasons, were included to avoid problems of pseudoreplication, whereby all conclusions would relate only to a single site (wall) or period of time. Samples were taken during times when a settling plate experiment was in progress on the same walls (K.P. Sebens, unpublished data). Approximately 5 d of samples were taken during each 3-wk experimental period, with sampling at various times of day (0900 to 1700). There were 13 3-wk long experimental periods between September 1989 and November 1991. It was logistically impossible to run experiments on more than one wall at a time, because current meters, racks, and panels had to be set up at one site at a time; therefore, sampling periods do not overlap for the three walls.

Plankton samples were collected at 8-10 m depth by scuba divers and over deeper water by lowering the same pump to 10-m depth from a boat, [approximately equal to]200 m east of Shags Rocks over water deeper than 25 m. The pump used was designed by K. Sebens (Sebens and Maney 1992) and consists of a sampling wand with three in-current heads connected to a length of 12.5 cm diameter polyvinyl chloride (PVC) pipe with a 3500 Rule bilge pump on the end [ILLUSTRATION FOR FIGURE 1 OMITTED]. Plankton were captured in a 40-[[micro]meter] mesh Nitex plankton net inside the PVC pipe, upstream of the pump. The incurrent heads were designed so that water was drawn from a defined layer. The heads consisted of two 6-cm diameter plates separated by a 0.6-cm gap with the inlet between the plates. The pump heads were analyzed qualitatively using a flow tank and video camera. In still water, the pump sampled a layer [approximately equal to]3-5 cm above and below the plates. In moving water, however, this layer became much narrower; the pump sampled a 2-3 cm wide layer at 5 cm/s and [less than]1 cm at speeds over 10 cm/s. Flow in situ would rarely be less than 5 cm/s on these walls (Patterson 1984, Sebens 1984). Water flow through the pump was measured by a General Oceanics propeller flow meter mounted inside the PVC pipe. The pump was powered by a 12-V battery connected in the boat via a 30-m cable. For the last set of samples (wall 3), the pump had a 12-V battery pack attached underwater by a 2-m cable. The pump filtered water at [approximately equal to]1 L/s and was calibrated regularly for each pump used.

This sampling device was held by a diver so that the incurrent heads were at a specified distance from the wall surface. The heads were moved over the wall surface slowly without touching the wall. Divers made every effort not to stir up sediments or interfere with the flow. Four or five samples were taken during each dive: (1) 1-5 cm distance away from the rock face over invertebrates (5 min), (2) 1-5 cm distance away from the rock over crustose coralline algae (1990 and 1991 only; 5 min), (3) 10-15 cm distance away from the rock over invertebrate-dominated areas (1989 and 1990 only; 5 min), (4) 1 m away from the rock (5 min), and (5) offshore over deep water ([greater than]200 m distance) by dropping the pump to 10 m depth (10 min). At least four successful sample collections were made during each 3-wk experimental period.

To provide a statistically rigorous analysis of variation among samples during one time period, four replicate groups of samples were taken on wall 1 in summer 1990. Replicate samples were taken on wall 1 by two teams of divers with two identical plankton pumps; these pumps were calibrated separately to determine the amount of water filtered per rotation of the flow meter and data were treated accordingly. Locations for each replicate sample were derived by random choice along a transect next to the wall. Three replicate samples were taken at the same distances described above.

For all collections, live samples were taken to Northeastern University's Marine Science Center in the plankton bags floating in sea water. Each plankton sample was rinsed with filtered sea water into jars (400 mL) and fixed in 5% buffered formalin in sea water, then stained with Rose Bengal, all within 1 h of collection. Plankton samples were mixed thoroughly and subsamples (dipped during shaking) were transferred into plastic petri dishes with 0.5-[cm.sup.2] grids engraved on the bottom. All larvae of sessile invertebrates in the samples were identified and counted using a Wild dissecting microscope (16x, 25x magnification). The entire contents of each sample were analyzed, except in cases where larvae and holoplankton were extremely numerous. In these cases, subsamples were counted and the results were calculated for the entire volume of the sample; at least half of the original sample was analyzed and data were adjusted to full sample volume when calculating densities.

All computational and graphic analyses were done using a Macintosh II; statistics were generated using the Statview II program. Replicate samples were analyzed using the Kruskal-Wallis nonparametric test because the assumptions for analysis of variance could not be met, even with transformation. This statistic was applied comparing the data for samples close to the wall (over both invertebrates and crustose algae) to far from the wall (both 1 m and [greater than]200 m away) (Table 2). The same statistic was also applied comparing the data for samples close to the wall over invertebrates, close to the wall over crust, 1 m away from the wall, and [greater than]200 m away from the wall (Table 3).

All individual sample days were grouped by walls and by seasons and were examined visually using the cross reference graphs [ILLUSTRATION FOR FIGURES 2-8 OMITTED]. Statistical analysis was performed by using Wilcoxon's signed-ranks test (paired) because the assumptions of the analysis of variance could not be met. Samples were analyzed separately for each wall and season (spring/summer and fall). The Wilcoxon's signed-ranks test was used to test the significance of differences among densities of larvae found 1-5 cm over invertebrate-covered walls vs. densities of larvae found in all the other sample areas: (1) 1-5 cm over crustose coralline algae-covered surfaces, (2) 1 m from invertebrate-covered surfaces, and (3) far ([greater than]200 m) away from rock surfaces.

To determine if there were differences in densities of larvae between days of high and low flow, data from all three walls were grouped and separated by flow characteristics. Flow data were taken from dive notes and from a Marine Science Center data file on wave heights observed daily over the Shags Rocks. The full data set was divided equally to form the low- and high-flow groups (low flow: waves [less than]0.7m and tidal currents [less than] 5 cm/s). The Wilcoxon's signed-ranks test was applied to test for differences in densities among 1-5 cm over invertebrates, 1-5 cm over crust-covered surfaces, 1 m, and [greater than]200 m from walls. This analysis was repeated using flow data from Interocean S4 current meters placed 1 m from each rock wall, which were available for many, but not all, of the days on which samples were taken.

RESULTS

Plankton samples contained a variety of larvae (meroplankton) and many other zooplankton (holoplankton). The most common larvae were small straight-hinge stage bivalve veligers, which were abundant in nearly all samples. These larvae were omitted from analysis because they are the larvae of clam species not occurring on rock substrata and because their numbers usually swamped the larvae of interest. Barnacle nauplii and cyprids were common (most likely Balanus balanus at this depth) during spring and summer. Veliconcha-stage veligers of a variety of bivalves (mostly Mytilus edulis and Modiolus modiolus) were also common during spring, summer, and fall and Modiolus modiolus pediveligers (and possibly Mytilus edulis) were also abundant. A variety of bryozoan cyphonautes larvae and polychaete setigers were found irregularly. Ascidian tadpole larvae and hydroid actinula larvae were found regularly in samples during spring and summer.

[TABULAR DATA FOR TABLE 1 OMITTED]

[TABULAR DATA FOR TABLE 2 OMITTED]

[TABULAR DATA FOR TABLE 3 OMITTED]

The two most common ascidians on the rock walls were Aplidium glabrum (a colonial ascidian) and Molgula citrina (a solitary ascidian). M. citrina is an early successional species, while A. glabrum is a more aggressive and long-lived competitor that overgrows most other invertebrates (Sebens 1986). Two different forms of ascidian larvae were found in samples near the walls, which are probably the larvae of A. glabrum and M. citrina, but have not all been identified individually.

The actinula larvae are believed to be those of Tubularia spp. because (1) the actinula closely resembled illustrations of Tubularia larvae, and were similar in color to the polyps, and (2) Tubularia was by far the most abundant hydroid on these walls. Pediveligers collected on settling plates at the same time these samples were taken were cultured and identified as Modiolus modiolus. Anomia veligers were found only in the summer and fall; the species commonly found on these walls is Anomia simplex. This bivalve was frequently found on settling plates at this site but was not very abundant on the walls, perhaps due to postsettlement predation.

All graphs comparing larval densities at two or more distances from the rock walls, or at two locations, are presented as scatter plots with a y=x line (equal larval density) for comparison. This method allows direct visual comparison of samples taken on days of different overall larval density. Samples are grouped by seasons during which each larval type was most abundant. Seasons with few larvae present are omitted from the graphs and from statistical analysis.

Distance from wall vs. length and planktonic dispersal period

Ascidian larvae. - Overall, densities of ascidian tadpole larvae were uniformly highest in samples nearest the rock surfaces. On all three walls [ILLUSTRATION FOR FIGURE 2 OMITTED] there was a significant difference (Wilcoxon's signed-ranks test, P [less than] 0.05) between density of tadpole larvae collected 1-5 cm over invertebrate-covered surfaces and [greater than]200 m from wall; more larvae were found 1-5 cm from the substratum (Table 1). On wall 1 (spring/summer) there were no differences in densities 1-5 cm from the wall and 1 m from wall. On wall 1 (fall) walls 2 and 3 (all seasons) there were significantly more larvae at 1-5 cm from the rock surface than at 1 m from the rock surface. During the replicate sampling on 30 July and 29 August 1990 there were also significantly more ascidian larvae close to the rock surfaces than farther away (Table 2).

Mussel pediveligers. - Densities of pediveligers were generally highest in the layer 1-5 cm from the rock surfaces. On all three walls [ILLUSTRATION FOR FIGURE 3 OMITTED], (spring, summer, fall) there were significant differences (Wilcoxon's signed-ranks test, P [less than] 0.05) between the density of pediveligers found 1-5 cm over invertebrate-covered surfaces vs. [greater than]200 m (Table 1). On wall 1 [ILLUSTRATION FOR FIGURE 3 OMITTED], spring/summer) there were higher densities of pediveligers at 1 m from the wall than at 1-5 cm from the wall. On wall 1 (fall) there were higher densities of pediveligers at 1-5 cm from the wall than at 1 m from the wall. On walls 2 and 3, there were no differences in densities between 1-5 cm and 1 m from the walls. On 24 July and 30 July 1990 there was a significant difference between densities of pediveligers 1-5 cm from the wall vs. densities of pediveligers at all distances farther away from the wall (Table 2). No significant differences were found on the other replicate sampling days.

Mussel veliger larvae. - Veliconcha-stage mussel veliger larvae were usually not differentially distributed among samples, occurring at equal densities at all distances from walls. However, when differences were present, the samples [greater than]200 m from walls had higher abundances. They were found in significantly higher densities in samples [greater than]200 m from wall 2 (spring/summer) and wall 3 (fall) (Wilcoxon's signed-ranks test, P [less than] 0.05; Table 1).

Hydroid actinula larvae. - Actinula larvae were most common in samples taken close to the rock surfaces. No actinula larvae were found in fall samples. On walls 1 and 2 [ILLUSTRATION FOR FIGURE 4 OMITTED], spring/summer) there was a significant (Wilcoxon's signed-ranks test, P [less than] 0.05) difference between samples collected 1-5 cm over invertebrate-covered surfaces vs. all other distances. More actinula were found close to the rock surface and at 1 m than [greater than]200 m away (Table I). On 29 August 1990 (Table 2), there were significantly more actinula larvae found close to the rock walls. No significant differences were found on the other replicate sampling days.

Anomia veligers. - There were no significant differences in densities of Anomia veligers at 1 m and [greater than]200 m away from rock surfaces vs. 1-5 cm from wall 1 during all seasons, and wall 2 during spring/summer [ILLUSTRATION FOR FIGURE 5 OMITTED], Table 1). Wall 3 had a higher density of Anomia veligers 1-5 cm from the wall than at 1 m but also a higher density [greater than]200 m from the wall than at 1-5 cm [ILLUSTRATION FOR FIGURE 4 OMITTED]. On 24 July and 30 July 1990 there was a significant difference between densities of veligers 15 cm from the rock surfaces vs. densities of veligers farther away from the rock surfaces; there were more Anomia offshore ([greater than]200 m) than close to wall surfaces in those samples (Table 2).

Other larvae. - Barnacle nauplii and cyprids were found only in spring and summer. There were significantly more nauplii found [greater than]200 m away from wall 1 than close to this wall during spring and summer (Table 1), but there were no other differences for any sets of samples; this was due primarily to the low total number of samples containing these larvae. Polychaete larvae, spionid setigers, and various postsetigers also showed no significant differences. Larvae of the abundant polychaete Spirorbis spp. and the octocoral Alcyonium siderium were exceedingly rare or absent in plankton samples in any season.

Density differences over crustose coralline algae and invertebrate areas

Ascidian larvae. - On wall 1 during spring/summer, there was a significant difference (Wilcoxon's signed-ranks test, P [less than] 0.05) between density of ascidian tadpole larvae 1-5 cm over invertebrate-covered surfaces vs. 1-5 cm over crustose coralline algae-covered surfaces (Table 1, [ILLUSTRATION FOR FIGURE 6 OMITTED]). More larvae were found over the invertebrate-covered surfaces. On wall 2 [ILLUSTRATION FOR FIGURE 6 OMITTED] there was a significant difference also; more larvae were found over invertebrate-covered surfaces than over crust-covered surfaces. On wall 3 (fall), no statistically significant differences in densities were found between invertebrate-covered surfaces and crust-covered surfaces; however, the few data points available clustered under the x=y line, also showing a trend of greater density over invertebrate-covered surfaces. The lack of significance is probably because the patches of crustose algae on wall 3 were smaller and more interspersed with invertebrate patches; the samples contained very few larvae and there were fewer samples taken here. There were no significant differences between densities over crustose coralline algae-covered rock surfaces vs. invertebrate-covered rock surfaces during any of the replicate sampling days (Table 2).

Mussel pediveligers. - There was a significant difference (Wilcoxon's signed-ranks test, P [less than] 0.05) between densities of pediveligers 1-5 cm over invertebrate-covered rock surfaces vs. 1-5 cm over crustose coralline algae-covered rock surfaces on wall 1 ([ILLUSTRATION FOR FIGURE 7 OMITTED], spring/summer, fall); more pediveligers were found over invertebrates than over crusts (Tables 1, 2). There were no significant differences in pediveliger densities between crust-covered surfaces and invertebrate-covered surfaces on walls 2 or 3 [ILLUSTRATION FOR FIGURE 7 OMITTED]. In the 14 August 1990 replicate sample, there was a significant difference (Kruskall-Wallis test, P [less than] 0.05) between densities of pediveligers over crustose coralline algae-covered rock surfaces vs. invertebrate-covered rock surfaces; more pediveligers were found over the crust-covered surfaces (Table 3). There were no significant differences found on the other three replicate sampling days.

Mussel veligers. - Although there were no differences in most sampling periods, there was a significant difference (Kruskall-Wallis test, P [less than] 0.05) in density of veliconcha-stage mussel veligers over crustose algae-covered surfaces vs. invertebrate-covered surfaces on 29 August 1990. Much higher densities of veliconcha-stage veligers were found over invertebrate-covered surfaces (Table 2).

Hydroid actinula larvae. - There was a significant difference (Wilcoxon's signed-ranks test, P [less than] 0.05) between samples collected 1-5 cm over invertebrate-covered surfaces vs. samples collected 1-5 cm over crustose coralline algae-covered surfaces on wall 1 and wall 2 [ILLUSTRATION FOR FIGURE 8 OMITTED]. More actinula larvae were found over invertebrate-covered surfaces (Table 2). The invertebrates covering these surfaces included large numbers of hydroids, especially Tubularia, in summer. In the 14 August 1990 sample, there was a significant difference between densities of actinula larvae found 15 cm over invertebrate-covered surfaces vs. 1-5 cm over crust-covered surfaces. No significant differences were found on other sampling days (Table 3).

Anomia veligers. - There were no significant differences between densities of Anomia larvae found over invertebrate-covered surfaces vs. crust-covered surfaces on any of the three walls (Table 2). The replicate samples also showed no difference in density of veligers 1-5 cm over crustose coralline-covered rock surfaces vs. 1-5 cm over invertebrate-covered surfaces (Table 3).

Other larvae. - There were no significant differences in density of barnacle nauplii and cyprids between crust-covered surfaces and invertebrate-covered surfaces (Table 2). Polychaete larvae, spionid setigers, and various postsetigers also showed no significant differences.

Flow regime and larval distribution

Ascidian tadpole larvae and mussel pediveligers had very significant density differences among 1-5 cm, 1 m, and [greater than]200 m from the rock walls during days of both high and low flow. However, the P values for the Wilcoxon's signed-ranks test were higher (less significant difference) on days of high flow (mean flow speeds each 0.5 s for 5-10 min were 6.7-15.4 cm/s, Interocean S4 current meter data) than on days of low flow (mean flow speeds each 0.5 s for 5-10 min were 1.0-6.6 cm/s) in all cases for ascidian larvae (see Table 6). The ascidian tadpole larvae were more abundant 1 m from the wall on days of higher flow than they were on days of lower flow, compared to the layer 1-5 cm from the walls. For pediveligers, the significance levels were of similar magnitude in both flow conditions. However, differences between 1-5 cm and [greater than]200 m were much greater on low-flow days for both ascidians and pediveligers.

Actinula hydroid larvae had significantly higher densities 1-5 cm from wall compared to 1 m from the wall on days of low-flow speed but showed no significant differences on days of high flow. Actinula were therefore more thoroughly mixed on days of higher flow. Other larvae did not show clear and consistent differences between samples overall and thus were not tested for flow-related differences.

DISCUSSION

There are at least four types of dispersal among encrusting invertebrates. The first, exhibited by mussels, barnacles, and other common intertidal organisms, is long-distance dispersal of planktotrophic larvae (Thorson 1950, Scheltema 1986, Levin and Bridges 1995). These larvae stay in the water column for weeks or months and can be carried many hundreds of kilometres from their point of origin. These are referred to as teleplanic ([greater than]2 mo in plankton) and actaeplanic (coastal, [less than]2 mo) larvae (Scheltema 1989). The second type includes lecithotrophic larvae that are in the water column for hours to days and are primarily dispersed within a few kilometres (anchiplanic, Scheltema, in Levin and Bridges 1995). The third pattern, common for subtidal organisms, is short-range dispersal on the order of seconds to minutes (Potswald 1978, Olson 1983, Sebens 1985). Some of these larvae (also termed anchiplanic) are ready to settle almost immediately but can delay settlement until an appropriate cue is encountered (ascidians, Millar 1971; bryozoans, Keough 1984). The fourth dispersal strategy (aplanic, Levin and Bridges 1995) is very short-range dispersal, such as that of the octocoral Alcyonium siderium, which has a crawling benthic planula larva (Sebens 1983). These differences in dispersal ability should produce characteristic patterns of abundance in the water column. Animals with crawling larvae would almost never be found in the near-substratum plankton. Animals with short to intermediate range dispersal should be found more frequently in the water very close to the rock surface. Finally, animals with long-distance larvae would be found most frequently far from rock surfaces, until they are ready to settle, or could have a completely random pattern.

Recent studies on subtidal invertebrates indicate that even swimming larvae of some common sessile species may not disperse very far from their point of origin (e.g., bryozoans, Keough and Downes 1982). However, there is a clear adaptive significance to pelagic larval dispersal, whether short or long term (Hedgecock 1986). Long-distance dispersal allows larvae to encounter isolated habitats, and areas recently cleared of competitors by disturbance events. Also, long-distance transport of larvae and the resultant gene flow can keep populations genetically mixed; speciation events through isolation may be rare for such species (Scheltema 1971, Strathmann 1974, Roughgarden et al. 1988).

Water flow affects larval dispersal on all levels. Larvae released on subtidal rock walls, for example, are often subjected to bidirectional (oscillatory) flow and are likely to be wafted back and forth, with their actual horizontal displacement being small at slack tides. Also, when larvae are released close to the substratum they tend to remain in slower moving water layers and thus have a high probability of contacting the substratum by random swimming, eddy entrapment, or turbulence (Hannan 1984). Sebens and Koehl (1984) found that ascidian larvae were abundant 1-3 cm from a rock wall but were virtually absent 80 cm away. This could occur because the ascidians are stronger swimmers than other invertebrate larvae, which employ ciliary motion, and could maintain their position near the wall, or the larvae could be trapped in the slow-moving water layers near the substratum (see also Svane and Young 1989). The latter pattern could be a passive result of larvae having been released very close to the rock surface. Rock surfaces are irregular and, because of acceleration and deceleration during oscillatory flow over the rough surface, a stable boundary layer cannot form. The resulting turbulence may benefit organisms relying on flow for larval dispersal away from parental microhabitats (Eckman 1990). Moderate to strong flow, and lack of a stable boundary layer, allow larvae to be picked up from points of release and moved rapidly above the substratum or transferred to mainstream flow by eddy diffusion (Denny 1988). Therefore, the type of development and the swimming capabilities of larvae could be very important in determining whether larvae are carried away from the point of release and if so, how far (Butman 1987).

Density patterns in relation to dispersal ability

The actual dispersal capability of benthic organisms, and the ecological implications of dispersal, are complex; data from field sites are badly needed. The purpose of this study was to provide an assessment of larval densities adjacent to subtidal rock walls, the sites of origin and recruitment for a particular group of species. The potential dispersal capabilities of these organisms were considered when forming hypotheses. The results of this study show a clear horizontal pattern of distribution away from vertical rock substrata for ascidian tadpole larvae, mussel pediveligers, veliconcha-stage mussel larvae, Anomia veligers, hydroid actinula larvae, and others.

The first hypothesis tested was that differences in dispersal ability or larval development time will be reflected in planktonic distributions next to vertical rock walls. Benthic (nonplanktonic) larvae (crawl-away) are unlikely to be found in any of the water-column samples. Larvae that spend a very short time in the plankton, and thus have low dispersal ability, should be rare at greater distances from the walls of origin. Such larvae may not be getting away from the walls, into dispersive currents, and are thus primarily repopulating the parental habitat. Since the dispersal abilities of most of these larvae are known in general terms, certain patterns of density could be predicted (Table 3). Larvae with very short dispersal should be found at high density 1-5 cm from rock walls, rarely at 1 m from the wall, and should be absent [greater than]200 m from the wall. Larvae with moderate dispersal should be common 1-5 cm from wall, sometimes 1 m from wall (especially during days of high flow) and not [greater than]200 m from vertical walls.

Larvae with long-distance dispersal (weeks or more), which are fully developed and ready to settle, should be found at high density 1-5 cm from the wall, at low density 1 m, and should be rare or absent [greater than]200 m from walls. These larvae are not expected to have originated from these particular walls but are in the process of settling on them. If they settle rapidly, accumulation will not occur. If, however, they have a prolonged search phase (testing the surface, reentering the water column), such accumulation would be expected. Finally, larvae with long-distance dispersal, but which are early in their larval development, would be expected to have greater densities [greater than]200 m from the walls and low density 1 m or closer to the rock surfaces. Alternatively, they might be expected to show no significant pattern and to be found everywhere at similar densities. If such larvae can actively avoid the substrate, as do some holoplankton (Sebens and Koehl 1984), the 1-5 cm sample would have the lowest larval densities.

Actual results for larvae collected adjacent to these walls are depicted in Table 3. Ascidian tadpole larvae, which have short-to-moderate dispersal showed the expected pattern; they were found at higher densities 15 cm from rock surfaces, were usually rare at 1 m, and were almost never found farther away. The most common ascidians on these rock walls were a colonial ascidian, Aplidium glabrum, and a solitary ascidian, Molgula citrina (Durante and Sebens 1994). A. glabrum and M. citrina tadpole larvae are competent to metamorphose immediately or within a few minutes of release, and such larvae can search for a settlement site for at least 1-2 h (Grave 1926, Durante 1989, Svane and Young 1989). Ascidian tadpole larvae are vigorous swimmers, but are not very directed, which suggests that they may use their swimming ability to stay close to the substrata of origin and to search for appropriate settlement cues. Alternatively, some ascidian larvae exhibit positive phototaxis upon release, then reverse direction away from light. This behavior could assist their entry into mainstream flow away from the substratum and increase their potential dispersal distance. Hydroid actinula larvae, also short-distance dispersers, showed a similar strong pattern; they were found primarily close to invertebrate-covered substrata with abundant Tubularia colonies. These larvae are released from adult Tubularia and are ready to settle immediately.

Mussel veligers have a long planktotrophic pelagic larval phase (Thorson 1950). The veliconcha stage did not show a pattern of greater abundance near the vertical rock surfaces, often the reverse, but the pediveligers did so strongly (Table 1). Pediveligers are in the "search" phase of their larval life (Bayne 1965). They are on and off the substratum, crawling and swimming, while searching for an appropriate settlement spot. It is not surprising, therefore, that this late larval stage showed a pattern of greater abundance near the rock surfaces than farther away.

The distribution of barnacle cyprids in the water column in relation to recruitment patterns on adjacent shores has been studied a number of times (Grosberg 1982, Gaines et al. 1985). Cyprids can demonstrate a vertical stratification pattern in the water column that reflects their pattern on shore. Barnacle cyprids and nauplii were common only in the spring and early summer, but a clear pattern of cyprids being sampled only close to the walls was not demonstrated in this analysis; this may be due to spatial scale. Patches or "clouds" of barnacle larvae may be large and thus may not be evident on this small scale. Barnacle nauplii were significantly more abundant [greater than]200 m away from the wall than close to the wall during spring on wall 2, which also agrees with the expected pattern of long-distance dispersers early in their development not having a pattern of greater abundance close to the rock walls.

There are other obvious patterns based on hypotheses derived from the behavior and ecology of each larval type. For example, larvae which develop in the plankton for weeks or more would be expected to avoid the turbulent regions near the bottom until they are ready to settle so they could avoid being damaged or preyed on. Veliconcha larvae showed a pattern of being more abundant offshore ([greater than]200 m) than at 1-5 cm over invertebrate areas on walls 2 and 3. Barnacle nauplii larvae were also found to be more abundant [greater than]200 m from wall 1 (significant) and wall 2. Once these nauplii become cyprids and veliconcha larvae become pediveligers, and are thus ready to settle, they have the potential to accumulate near the rock substrata. They appear to accumulate better in areas covered with erect invertebrates (acting as roughness elements). The presence of roughness elements on the substratum lowers the critical velocity by an amount that depends on the size of the roughness elements, which results in less shear stress on the substratum between them (Denny 1988). On smooth surfaces, large energetic eddies called "sweeps" periodically dip down and touch the substratum, resulting in a burst of turbulence and higher velocities. The time between these sweeps can be [less than]1 s (Denny 1988). With tall invertebrates present, the high velocity "sweeps" would not have a chance to develop and larvae can thus "hide" near the substratum. Hiding, or passively accumulating, near the substratum could be a problem if the larvae were prey for the adult organisms living on the walls. Pediveligers were not found to be common prey items for anemones in a previous study, possibly because of the time of sampling (Sebens and Koehl 1984). However ascidian tadpole larvae were common prey for both anemones and octocorals. Risk of predation could have important ramifications on how long the larvae spend searching on or near the substratum.
TABLE 4. (A) Expected density pattern 1-5 cm over different
substrata, for larvae ready to settle. "Attractor" designates
whether larvae are hypothesized to be attracted to (or retained
over) these substrata preferentially, by any mechanism (passive,
active). (B) Actual density distribution patterns 1-5 cm over
different substrata ([ILLUSTRATION FOR FIGURES 2-9 OMITTED],
analysis in Table 1.)


A) Expected density distribution


                                             Crustose      Inverte-
Dispersal type                Attractor      algae(*)      brate(*)


I. Nonplanktonic, not         none           Abs           Abs
reaching crustose
algae


II. Short distance            a. crust       Mod           High
                              b. invert.     Low           High
                              c. none        Low           High


III. Moderate distance        a. crust       High          Mod
                              b. invert.     Mod           High
                              c. none        Mod           Mod


IV. Long distance             a. crust       High/mod      Mod/low
                              b. invert.     Low           High/mod
                              c. none        Mod/low       Mod/low


B) Actual density distribution


                                           Crus-
                                           tose
Larvae                      Type          algae(*)
Invertebrate(*)


Alcyonium(**)               I               Abs           Abs
Spirorbis(**)               I               Abs           Abs
Anomia                      IV c.           Mod           Mod
Barnacle cyprid             IV c.           Low           Low
Mussel pediveliger          IV b.           Low           Mod/high
Ascidian larva              III b, c.       Low           High
Polychaete postsetiger      IV c.           Low           Low
Hydroid actinula            III b, c.       Low           High


* High ([greater than]50 larvae/1000 L), Mod (10-50 larvae/1000 L),
and Low (0-10 larvae/1000 L) refer to absolute densities of larvae
relative to each other. Abs = absent or very rare.


** Not enough larvae for statistical comparison in these samples,
but common on rock surfaces (Sebens 1983, 1984).


Three common inhabitants of these rock walls were the polychaetes Spirorbis spp., various bryozoans, and the octocoral Alcyonium. These (except Alcyonium) were found on settling plates put out at the same time plankton samples were collected (K. P. Sebens, unpublished data). These organisms are important because they are common on the wall and are also interesting because much work has been done with Spirorbis setigers (Knight-Jones 1951), bryozoan cyphonautes larvae (Keough 1984), and crawling Alcyonium planula larvae (Sebens 1983). However, such larvae were extremely rare (Spirorbis, cyphonautes) or nonexistent (Alcyonium planula) in the plankton samples. Spirorbis larvae recruited to settling plates on the rock walls very abundantly, yet these larvae were almost nonexistent in the plankton samples. Settling plates [greater than]1 m from the rock walls received very few Spirorbis recruits, compared to those on walls (K. P. Sebens, unpublished data). Spirorbis larvae are competent to settle immediately and are short lived in the water column (minutes to hours). It is very likely they are staying on, or only millimetres from, the rock surfaces and thus sampling 1-5 cm from the wall without scraping or disruption was not close enough to collect them.

Density patterns over invertebrate-covered surfaces vs. crustose coralline algae-covered surfaces

The second hypothesis addressed was whether larvae are found at similar densities over crustose coralline algal-covered surfaces compared to nearby invertebrate-covered areas. Breitburg (1984) found that 11 of the 14 most abundant invertebrate taxa she studied were less likely to settle on crustose coralline algae than on rock surfaces. However, in our study, it may be that there were fewer larvae present in the water and available to settle over algal areas because of their distance from the parental source. Larvae released close to vertical substrata in invertebrate-dominated areas, and which are ready to settle, may accumulate there and not accumulate over crusts because of the lack of "roughness elements" to trap them. The relatively smoother patches of crustose coralline algae may thus receive equal numbers of larvae but not retain them.

Larvae with short-distance dispersal, and which are attracted to crustose algae, could be found over algal-covered surfaces (end of planktonic development period) and also over invertebrate-covered surfaces (beginning [TABULAR DATA FOR TABLE 5 OMITTED] of planktonic period). Larvae with long-distance dispersal, and which are attracted to crustose algae, would be found over algal crusts (at the end of the planktonic period) but at lower densities over invertebrate-covered surfaces. Larvae with short-distance dispersal, but which are not attracted to crusts, would not be found at high densities over crust-covered surfaces but would be abundant over invertebrate-covered surfaces where they were released. Finally, larvae with long-distance dispersal, but which are not attracted to crusts, would be equally or less abundant over crustose algae than over invertebrates (depending on stage of development).

The actual results for larvae in this study are varied (Table 4). Aplidium glabrum and Molgula citrina are short-distance dispersers, which were probably not attracted to crustose algae; other ascidian tadpole larvae [[TABULAR DATA FOR TABLE 6 OMITTED] avoided settling on crust-covered surfaces (Breitburg 1984). These larvae were found, as expected, in significantly higher densities over invertebrate-covered surfaces in our study. This agrees with Breitburg's (1984) findings that crustose algae can inhibit settlement of ascidians, but our results indicate that such larvae may not even be venturing over crust-covered patches in large numbers on these walls. Wall 3 did not show this pattern; however, this wall had much smaller patches of crustose algae than did the other two walls. Oscillatory flow on this small spatial scale could eliminate the pattern. Also, the samples taken on wall number 3 were all taken in the fall and there were fewer larvae present at that time. Actinula larvae showed a pattern of greater abundance over invertebrate-covered than over algal crust-covered areas. This is probably because these larvae can settle very quickly after release from the parent colonies, and because they are being dispersed rapidly enough that they become "diluted" a metre or more from their release point.

Barnacle cyprids were expected to be attracted to crustose algae; they were found to settle abundantly on crust-covered surfaces by Breitburg (1984). Cyprids showed no clear pattern of abundance over crust-covered surfaces or over invertebrate-covered surfaces. They settle abundantly on Lithothamnion as well as on bare rock, and they search for areas of small-scale roughness and cracks (Crisp 1974); barnacles do not settle well (or cannot) on smooth plates. Data from the vertical walls at the Shags Rocks show that barnacle cyprids settle on Lithothamnion spp. and on bare rock and barnacles were found to settle preferentially on Lithothamnion, which is very rugose (K. Paull, unpublished data). Overall, these cyprids were not very abundant in the plankton samples, which may account for the lack of statistically significant differences. No larvae except Anomia (on wall 2, during spring and summer) were found to be more abundant over crust-covered surfaces. K. Paull (unpublished data) found Anomia settling abundantly on Lithothamnion. Since these larvae are in the plankton for weeks, they may be collecting over crust-covered areas because of attraction to this substratum type. Alternatively, they could be consumed by anemones, hydroids, and other predators in dense invertebrate stands.

Significantly more mussel pediveligers were found over invertebrates than over crustose algae on wall 1 during spring and summer. Since these larvae are not originating from that wall, but are arriving from a great distance, they may be responding to cues from other invertebrates or from adult mussels. It is also quite possible they are found in higher densities over invertebrate-covered walls due to the physical structure (roughness elements) resulting in slower or more turbulent flows in that area. Pediveligers were found at lower densities over crust-covered surfaces on wall 1 only during spring and summer. As stated earlier, wall 1 had wide, spatially separate ([greater than]3 m apart), crust- and invertebrate-covered areas. No significant patterns were found during the other three sampling seasons. These larvae may also be inhibited by the crustose algae, as Breitburg (1984) suggested, and may not even be entering a search phase there.

Changes in density patterns in relation to flow regime

Bidirectional (oscillatory) flow induced by waves and unidirectional tidal currents are the important components of flow at the walls studied (Patterson 1984). Days with higher flow would be expected to have better mixing of larvae near the rock surfaces. Larval distribution patterns would thus be less distinct, and larvae normally found only close to the wall would be found at greater distances than during times of low flow. The third hypothesis was that larvae with short-distance dispersal, which are found in higher densities 1-5 cm from the rock surfaces during times of low flow, will also occur at high densities at greater distances away during times of high flow. This expected pattern is illustrated in Table 5. Larvae with short and moderate dispersal should be found in much lower relative densities one metre away during days of low flow than during days of high flow.
TABLE 7. Results of Wilcoxon's signed-ranks test on all data (all
walls, seasons) separated by S4 regime. Table entries are
significance levels (P values). Boldface numbers are significant at
P [less than or equal to] 0.05.


                                                 S4 low     S4 high


                                                 (N = 13)   (N = 8)
Ascidian       1-5 cm vs. 1 m                     0.152      0.42
               1-5 cm vs. [greater than]200 m     0.002      0.005
               invert vs. crust                   0.38       0.42


                                                 (N = 8)    (N = 3)
Actinula       1-5 cm vs. 1 m                     0.66       1.0
               1-5 cm vs. [greater than]200 m     0.009      0.102
               invert vs. crust                   0.67       0.285


                                                 (N = 11)   (N = 8)
Pediveligers   1-5 cm vs. 1 m                     0.016      0.66
               1-5 cm vs. [greater than]200 m     0.28       0.47
               invert vs. crust                   0.016      0.67


Ascidian tadpole larvae, actinula hydroid larvae, and mussel pediveligers all showed a pattern of greater abundance 1-5 cm from the vertical rock surfaces (Table 6). Therefore, these three were separated by flow intensity (measured on that sampling day, within 2 h of the sample time) and were tested separately to see if there was an increase in density 1 m from the wall on days of high flow vs. days of low flow. Ascidian tadpole larvae [ILLUSTRATION FOR FIGURE 9 OMITTED] and mussel pediveligers still were significantly more dense 1-5 cm from walls compared to 1 m from walls on days of high flow, but this difference was less significant. We interpret this to mean that larvae were better mixed, less stratified, on days of higher flow. This same pattern was demonstrated for actinula hydroid larvae; there were no significant differences in density between the two distances on days of high flow, whereas there were significant differences on days of low flow.

It is clear from the results of this study that predictions of dispersion in relation to rock substrata, based on larval development time, locomotion, and competence, are generally reliable. Larvae that emerge ready to settle are generally not found in the water column at even moderate distances from suitable habitat (e.g., vertical rock walls). Larvae that have completed a long planktonic development can accumulate near rock surfaces as they prepare to settle and metamorphose. Nonplanktonic larvae, released on or almost on the substratum, essentially never appear in water samples even as close as 1-5 cm from the rock. Water motion and turbulence affect the observed stratification of larvae; periods of high flow can thereby effect better transfer of larvae from one rock wall to others nearby. This may also result in lower recruitment to sites on the same wall, or part of wall, as the parent inhabits. These processes are very important for models of community structure, which depend on knowing the relationship between adult abundance and recruitment rate on a range of scales. For communities subdivided into semi-isolated subunits (e.g., rock walls), the probability that a larva released on one wall will reach another, even a few metres distant, is critical. Further descriptive and experimental studies are needed to determine the factors that produce the patterns we have illustrated and to provide the link between studies of pattern in the field, and magnitudes of larval advection among areas of appropriate habitat.

ACKNOWLEDGMENTS

We wish to thank: Randy Olson, Jon Witman, Charlie Ellis, Trish Morse, Jan Witting, Sean Grace, Melody Chen, and Karen Vandersall for assistance with project design and with the manuscript and Gene Cusack and Joe Bradley for assistance with tools and equipment. This work could not have been done without the significant diving assistance of: Kathy Paull, Ted Maney, Chuck Arnold, Jan Witting, Doug Updike, and Marco Calavetta. This research was funded by N.S.F. grants OCE 8900144 and OCE 9296093 to K. Sebens. This is M.S.C. contribution number 211.

LITERATURE CITED

Bayne, B. L. 1965. Growth and the delay of metamorphosis of the larvae of Mytilus edulis (L.). Ophelia 2:1-47.

Breitburg, D. L. 1984. Residual effects of grazing: inhibition of competitor recruitment by encrusting coralline algae. Ecology 65:1136-1143.

Butman, C. A. 1987. Larval settlement of soft-sediment invertebrates: the spatial scales of pattern explained by active habitat selection and the emerging role of hydrodynamic processes. Oceanography and Marine Biology Annual Review 25:113-165.

Butman, C. A., J. P. Grassle, and C. M. Webb. 1988. Substrate choices made by marine larvae settling in still water and in a flume flow. Nature 333: 771-773.

Chia, F.-S., J. Buckland-Nicks, and C. M. Young. 1984. Locomotion of marine invertebrate larvae: a review. Canadian Journal of Zoology 62: 1205-1222.

Connell, J. H. 1985. The consequences of variation in initial settlement vs. post-settlement mortality in rocky intertidal communities. Journal of Experimental Marine Biology and Ecology 93:11-45.

Crisp, D. J. 1974. Factors influencing the settlement of marine invertebrate larvae. Pages 177-265 in P. T. Grant and A. N. Mackie, editors. Chemoreception in marine organisms. Academic Press, London, UK.

Day, R., and L. McEdward. 1984. Aspects of the physiology and ecology of pelagic larvae of marine benthic invertebrates. Pages 93-120 in K. A. Steidinger and L. M. Walker, editors. Marine plankton life cycle strategies. CRC Press, Boca Raton, Florida, USA.

Denny, M. W. 1988. Biology and the mechanics of the wave-swept environment. Princeton University Press, Princeton, New Jersey, USA.

Durante, K. M. 1989. Reproductive ecology and larval behavior of three north temperate ascidians. Master's thesis. Northeastern University, Nahant, Massachusetts, USA.

Durante, K. M., and K. P. Sebens. 1994. Reproductive ecology of the ascidians Molgula citrina Alder & Hancock, 1848 and Aplidium glabrum (Verrill, 1871) from the Gulf of Maine, USA. Ophelia 39:1-21.

Gaines, S., S. Brown, and J. Roughgarden. 1985. Spatial variation in larval concentrations as a cause of spatial variation in settlement for the barnacle, Balanus glandula. Oecologia 67:267-272.

Gotelli, N. J. 1987. Spatial and temporal patterns of reproduction, larval settlement and recruitment of the compound ascidian Aplidium stellatum. Marine Biology 94:45-51.

Grave, C. A. 1926. Molgula citrina (Alder and Hancock), activities and structure of the free-swimming larva. Journal of Morphology 42:453-471.

Grosberg, R. K. 1982. Intertidal zonation of barnacles: the influence of planktonic zonation of larvae on vertical distribution of adults. Ecology 63:894-899.

Grosberg, R. K., and J. F. Quinn. 1986. The genetic control and consequences of kin recognition by the larvae of a colonial marine invertebrate. Nature 322:456-459.

Hannan, C. A. 1984. Planktonic larvae may act like passive particles in turbulent near-bottom flows. Limnology and Oceanography 29:1108-1116.

Hedgecock, D. 1986. Is gene flow from pelagic larval dispersal important in the adaption and evolution of marine invertebrates? Bulletin of Marine Science 39:550-564.

Jablonski, D. 1986. Larval ecology and macroevolution in marine invertebrates. Bulletin of Marine Science 39:565-587.

Jablonski, D., and R. A. Lutz. 1983. Larval ecology of marine benthic invertebrates: Paleobiological implications. Biological Reviews 58:21-89.

Keough, M. J. 1983. Patterns of recruitment of sessile invertebrates in two subtidal habitats. Journal of Experimental Marine Biology and Ecology 66:213-245.

-----. 1984. Kin-recognition and the spatial distribution of larvae of the bryozoan Bugula neritina (L.). Evolution 38:142.

Keough, M. J., and Downes, B. J. 1982. Recruitment of marine invertebrates: the role of active larval choices and early mortality. Oecologia 54:348-352.

Knight-Jones, E. W. 1951. Gregariousness and some other aspects of the setting behaviour of Spirorbis. Journal of the Marine Biological Association of the United Kingdom 30: 201-222.

Levin, L. A. 1990. A review of tracking methods for labelling and tracking marine invertebrate larvae. Ophelia 32:115-144.

Levin, L. A., and T. S. Bridges. 1995. Pattern and diversity in reproduction and development. Pages 1-48 in L. McEdward, editor. Ecology of marine invertebrate larvae. CRC Press, Boca Raton, Florida, USA.

Millar, R. H. 1961. Scottish oyster investigations 1946-1958. Journal of Marine Research Scotland 3:1-76.

-----. 1971. The biology of ascidians. Advances in Marine Biology 9:1-100.

Olson, R. R. 1983. Ascidian-Prochloron symbiosis: the role of larval photoadaptations in midday larval release and settlement. Biological Bulletin 165:221-240.

-----. 1985. The consequences of short-distance larval dispersal in a sessile marine invertebrate. Ecology 66:30-39.

Olson, R. R., and R. McPherson. 1987. Potential vs. realized larval dispersal: fish predation on larvae of the ascidian Lissoclinum patella (Gottschaldt). Journal of Experimental Marine Biology and Ecology 110:245-256.

Patterson, M. 1984. Patterns of whole colony prey capture in the octocoral Alcyonium siderium. Biological Bulletin 167:613-629.

Pechenik, J. A. 1990. Delayed metamorphosis by larvae of benthic marine invertebrates: does it occur? is there a price to pay? Ophelia 32:63-94.

Potswald, H. E. 1978. Metamorphosis in Spirorbis (Polychaeta). Pages 127-143 in F. S. Chia and M. E. Rice, editors. Settlement and metamorphosis of marine invertebrate larvae. Elsevier, North Holland, New York, USA.

Roughgarden, J., S. Gaines, and H. Possingham. 1988. Recruitment dynamics in complex life cycles. Science 241: 1460-1466.

Scheltema, R. S. 1971. Larval dispersal as a means of genetic exchange between geographically separated populations of shallow-water benthic marine gastropods. Biological Bulletin 140:284-322.

-----. 1977. Dispersal of marine invertebrate organisms: Paleobiogeographic and biostratigraphic implications. Pages 73-198 in E. G. Karffman and J. E. Hazel, editors. Concepts and methods of biostratigraphy. Dowden, Hutchinson, and Ross, Stroudsburg, Pennsylvania, USA.

-----. 1986. On dispersal and planktonic larvae of benthic invertebrates: an eclectic overview and summary of problems. Bulletin of Marine Science 39:290-322.

Sebens, K. P. 1983. The larval and juvenile ecology of the temperate octocoral Alcyonium siderium Verril. I. Substratum selection by benthic larvae. Journal of Experimental Marine Biology and Ecology 70:1-17.

-----. 1984. Water flow and coral colony size: interhabitat comparisons of the octocoral Alcyonium siderium. Proceedings of the National Academy of Sciences (USA) 81: 5473-5477.

-----. 1985. Community ecology of vertical rock walls in the Gulf of Maine, U.S.A.: small-scale processes and alternative community states. Pages 346-371 in P. G. Moore and R. Seed, editors. The ecology of rocky coasts. Hodder and Stoughton, Toronto, Canada.

-----. 1986. Spatial relationships among encrusting marine organisms in the New England subtidal zone. Ecological Monographs 56:73-96.

Sebens, K. P., and M. A. R. Koehl. 1984. Predation on zooplankton by the benthic anthozoans Alcyonium siderium (Alcyonacea) and Metridium senile (Actinaria) in the New England subtidal. Marine Biology 81:255-271.

Sebens, K. P., and E. J. Maney. 1992. A portable, diver-operated plankton sampler for near substratum use. Pages 167-172 in L. B. Cahoon, editor. Proceedings of the American Academy of Underwater Sciences. American Academy of Underwater Sciences, Costa Mesa City, California, USA.

Strathmann, R. R. 1974. The spread of sibling larvae of sedentary marine invertebrates. American Naturalist 108: 29-44.

Svane, I., and J. N. Havenhand. 1993. Spawning and dispersal in Ciona intestinalis. Marine Ecology 14:53-66.

Svane, I., and C. M. Young. 1989. The ecology and behaviour of ascidian larvae. Oceanography and Marine Biology Annual Reviews 27:45-90.

Thorson, G. 1950. Reproductive and larval ecology of marine bottom invertebrates. Biological Reviews 25:1-45.

Young, C. M. 1986. Direct observation of field swimming behavior in larvae of the colonial ascidian Ecteinascidia turbinata. Bulletin of Marine Science 39:279-289.
COPYRIGHT 1996 Ecological Society of America
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1996 Gale, Cengage Learning. All rights reserved.

 
Article Details
Printer friendly Cite/link Email Feedback
Author:Graham, Krista R.; Sebens, Kenneth P.
Publication:Ecology
Date:Apr 1, 1996
Words:9715
Previous Article:The seasonal decline in tree swallow clutch size: physiological constraint or strategic adjustment?
Next Article:Reproductive strategies of modular organisms: comparative studies of reef-building corals.
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters