To Thine Own Self be True? Yes! Thou Canst Not Then be False to Any Other. A Reply to Grosberg.

Abstract

Our original paper (1) was motivated by the philosophy that every healthy scientific discipline should have at least two alternative hypotheses (2); our goal was simply to incorporate a consideration of selection favoring self fusion in future work on the selective forces operating on allorecognition systems. We did not intend to resolve this issue, and we do not feel that a resolution is possible given the limitations of current information. Consequently, we welcome Grosberg’s commentary (3) on our previous efforts and are thankful for this opportunity to extend the discussion. Grosberg raises a number of interesting points about our argument, three of which merit additional commentary. First, Grosberg feels that we have misinterpreted a series of models developed with J. F. Quinn (4). We acknowledge that our discussion of their results was too brief to fully justify our interpretation. Nevertheless, we do feel that our interpretation is valid. Grosberg and Quinn did consider the costs and benefits of fusion and aggression at the individual level. However, the costs and benefits were explicitly those associated with fusion with kin. Neither fusion with self, nor any other selective force, was considered as an alternative. Our contribution was an explicit consideration of the selective forces of self fusion, which can be invoked only at the individual level. We did not intend to imply that past work had ignored all forms of selection at the individual level, just the selective pressures of fusion with self An additional aspect of our interpretation of Grosberg and Quinn’s results involves their predictions from a series of models that consider the consequences of variation in costs and benefits for three cases (4): aggression alone, fusion alone, and fusion as an alternative to aggression. The aggression only model predicts a monomorphic population under all conditions. In both of the remaining cases, polymorphism occurs only when the costs exceed the benefits [as cited in Grosberg’s quote (3)]. When the benefits are greater than the cost, the initially most frequent allele increases to fixation, and the population becomes monomorphic. We interpret this result as preliminary evidence against fusion with kin generating and maintaining high levels of polymorphism. The scenario of costs exceeding benefits is compatible with neither the basic assumptions behind the evolution of kin fusion, nor the continued existence of allorecognition systems. While Grosberg may now feel that these models show that “individual selection could easily maintain allotypic polymorphism” (3, p. 454) he and Quinn originally reached rather a different conclusion that is in accord with our own interpretation of their work: “allotypic polymorphism can be maintained directly by the individual costs and benefits of fusion provided fusion carries a net fitness cost. This raises the question of how fusion conditioned on relatedness can be evolutionarily stable. Our results suggest that selection acting at the level of clonal or kin-aggregations, rather than at the level of the individual, may provide an explanation for the evolution of allotypic specificity through aggression or fusion.” (4, p. 157) “Because the individual costs and benefits of fusion and aggression cannot readily account for why these behaviors are conditioned on allotypic identity, other explanations must be sought.” (4, p. 165)