| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Departments of Ecology and Evolutionary Biology, Yale University, New Haven, Connecticut 06520-8106
2 Departments of Geology and Geophysics, Yale University, New Haven, Connecticut 06520-8106
* To whom correspondence should be addressed
Hart and Grosberg (1) claimed that demersal larvae remain in kin aggregations. A simple test of this proposition has recently become possible. In the athecate colonial hydroid Hydractinia symbiolongicarpus, allorecognition is controlled by a single chromosomal interval and fusion is expected only between closely related individuals (2). If Hydractinia larvae do aggregate in kin groups, then recruits from the same shell should fuse more frequently than those from different shells. We sampled a shallow subtidal population from Long Island Sound (Old Quarry Harbor, Guilford, CT) to identify hermit crab shells that bore two or more newly recruited Hydractinia colonies. The colonies were explanted from the shells, reared in the laboratory, and subsequently tested for fusibility. We observed fusion between 4.4% of co-occurring colonies and 1.8% of colonies from different shells—frequencies that do not differ significantly. These results provide no support for the claim that larvae remain aggregated in the field.
Hydractinia symbiolongicarpus (Buss and Yund, 1989) occurs most frequently as an epibiont on the shells of hermit crabs and is, therefore, a sessile organism on mobile substrata. Colonies are dioecious, with the gametes freely shed into the water column on a diurnal light cue. This natural history suggests that larval kin aggregations are unlikely; nonetheless, egg release is restricted to a period of less than an hour each day, the eggs are negatively buoyant, and the larvae crawl but do not swim. Thus, larvae might conceivably remain in kin aggregations if a hermit crab remained immobile; if the larvae did not crawl away from one another; and if tidal movement, currents, or bioturbation did not resuspend the mud and sand for a period of 48–72 h.
Should this concatenation of conditions occur, hermit crabs might be expected occasionally to pick up two larvae as they pass through a single aggregation. In contrast, different crabs would not be likely to traverse the same aggregation; hence larvae recruiting to different shells would be expected to be unrelated. A direct test of the claim that demersal larvae remain in kin aggregations is provided by comparing new recruits found on the same shell with those on different shells.
To make this test, we exploit new findings on the genetics of allorecognition in Hydractinia (2). When recruits encounter one another, one of two genetically alternative results are obtained. Colonies either fuse, establishing a chimera with a continuous gastrovascular system, or they fail to fuse (2–5). The failure to fuse is called rejection and triggers a nematocyst-based effector system, which in turn results in either the elimination of one colony or the development of the oft-observed "suture-lines" separating colonies on the same shell (6–8).
We have recently demonstrated that the decision to fuse or reject is controlled by a single chromosomal interval and confirmed this result by both classical genetic techniques and the development of molecular markers spanning the interval (2). Fusion occurs only if colonies share alleles at one or more loci in this interval (2). Thus, field-collected colonies are expected to fuse only if they are related or if both carry an allele common in the population. The latter is expected to occur infrequently, as allorecognition loci are said to be highly polymorphic (9). If larvae are aggregated in kin groups, then fusion among recruits on a single shell must exceed fusion among recruits from different shells.
To test Hart and Grosberg’s (1) larval kin aggregation hypothesis, we sought conditions most conducive to the formation of larval kin aggregations. Specifically, we avoided intertidal habitats and those with high currents, and we sampled from a calm, shallow subtidal embayment. Using scuba, we collected 1369 shells on four dates in 2002 (30 May, 4 June, 21 June, and 31 July) at a depth of 5 m on the seaward edge of One Tree Island, Old Quarry Harbor (Guilford, CT). Newly recruited colonies (<10 polyps) are easily identified by inspection of shells under a stereomicroscope (10). Forty-six shells clearly bearing two or more new recruits were identified. Colonies were removed from the shell with a scalpel blade, affixed to glass microscope slides with thread until they adhered, and maintained in 10-gallon glass aquaria filled with artificial seawater (Reef Crystals) where they were fed three times a week with 2- to 3-day-old Artemia nauplii. Explants were taken when the colonies had grown sufficiently, and the colonies were tested for their ability to fuse to one another, using colony-based or polyp-based fusibility assays. In the colony-based assay, small fragments cut from each colony were attached 2–3 mm apart with thread on a glass microscope slide and observed at daily intervals until they contacted; thereafter, daily observations continued until the assay was scored. In the polyp-based assay (11), one polyp cut from each colony was threaded onto a human hair and sandwiched between two agar blocks so that its cut end opposed that of the other polyp. After 2 h, the agar blocks and hair were removed, but the polyps continued to adhere. At 24 h, the assays were scored as fusions if the polyps fused to create a single gastric cavity and common ectoderm, or as rejections if they fell apart. Colony and polyp assays are interchangeable with respect to fusion and rejection (2); they differ only with respect to a third class of allorecognition phenotype in which colonies initially fuse but subsequently separate. This transitory fusion phenotype is detectable only in the colony assay; in the polyp assay it cannot be distinguished from fusion. Transitory fusion is also known to be controlled within the chromosomal interval (2), and it was not observed among our samples.
We performed 113 pairwise tests of fusibility between colonies recruited to a single shell. We compared these results to those obtained from an additional 110 pairwise assays performed with colonies derived from different shells. We observed fusion between 4.4% (5/113) of co-occurring colonies and 1.8% (2/110) of colonies from different shells. There is no significant difference between these two frequencies (G test for independence, G = 1.201, P > 0.25). The power of this analysis is low (0.23, 
= 0.05) because the frequency of fusion is so low. Indeed, if the frequency of fusion on the same shell actually exceeds that on different shells, a power of 0.8 would require a sample size of more than 1175 pairwise fusion assays to detect an effect as small that measured.
Our findings confirm the suggestion made by Hart and Grosberg (1) that fusion occurs at a detectable frequency in natural populations of Hydractinia. Moreover, our results are consistent with the extreme lower range of Hart and Grosberg’s (1) estimate that 2%–18% of co-occurring colonies are full siblings. Recall, however, that fusion can occur either because encounters are biased toward kin or as a consequence of allelic diversity in the population. Our observation that frequencies of fusion between recruits onto the same shell do not differ from frequencies of fusion observed on different shells suggests that the extraordinary claim of larval kin aggregates is not warranted on the basis of the data now available.
 |
Acknowledgments |
|---|
|
Literature Cited |
|---|
 TOP Literature Cited |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
