Is allorecognition specificity in Hydractinia symbiolongicarpus controlled by a single gene?

I N many benthic, colonial invertebrates, the capacity to distinguish conspecific self from nonself, or allo-recognition, plays a crucial role in determining the nature and outcome of somatic, and perhaps gametic, interactions (BUSS 1982, 1990; GROSBERG 1988). One of the most striking features of invertebrate allorecogni-tion systems is the extraordinary precision with which self can be distinguished from nonself, even in very large populations. To the extent that allorecognition specificity is heritable, this precision suggests that populations of colonial invertebrates carry unusually high levels of genetic variation, rivalling levels found at MHC Class I and I1 loci (e.g., P o n s and WAKELAND 1990; PARHAM and OHTA 1996) and self-incompatibility loci in angiosperms (reviewed in CHARLESWORTH 1995). However, unlike the vertebrate major histocompatibil-ity complex, or many plant self-incompatibility systems, the question of whether one or a few loci (with high levels of allelic diversity) or multiple loci (with more modest amounts of variation per locus) control invertebrate allorecognition remains largely unexplored. Some of the first attempts to decipher the formal genetics of invertebrate allorecognition met with considerable success. For example, among ascidians in the genera Botryllus and Botrylloides, a single locus with multiple codominant alleles appears to control the acute fusion/ rejection response; juxtaposed individuals sharing one or both alleles at this locus somatically fuse, whereas those not sharing an allele reject each other (OKA and WATA-and FELDGARDEN 1992). However, comparable studies of cnidarians, especially on members of the genus Hydrac-tinia, yielded inheritance patterns that did not conform to the predictions of the simple Botryllus genetic model used an inbred strain of H. symbiolongzcarpus to reexamine the transmission genetics of allorecognition specificity in this species and surpris ingly concluded that, " The results are in complete agreement with a model of Hydrmtinia allorecognition as a one-locus trait, with codominant expression of alleles, such that one shared allele yields a fusible phenotype. " In this paper, we first show that a simple and powerful prediction of the one-locus model proposed by MOKADY and Buss (1996) does not withstand an empirical test. This raises the question of how MOKADY and BUSS (1996) reached the conclusion that a single locus controls allorecognition specificity in H. syrnbiolongicarpus. To address this question, we evaluate the limitations of the mating scheme described in their study, which could potentially bias the results in favor of the preferred hypothesis. Second, we show that their sample sizes are …