STYLE MORPH RATIOS IN TRISTYLOUS DECODON VERTICILLATUS (LYTHRACEAE) - SELECTION VS HISTORICAL CONTINGENCY

Tristylous plant populations should exhibit equal frequencies of the three style morphs at equilibrium. In contrast, New England and central Ontario populations of Decodon verticillatus (Lythraceae) show a marked deficiency of the mid-styled (M) morph. This pattern was also evident in an independent sample of 30 populations from eastern Ontario; all nine cases of morph loss involved the M morph, and this morph also occurred at low frequencies, especially in large trimorphic populations. The hypothesis that a fitness disadvantage to the M morph accounts for its low frequency was not supported by data from a series of held and glasshouse experiments. Measures of reproductive success from two populations showed no disadvantage to the M morph, but revealed substantially higher seed set in the mid- (M) and short-styled (S) morphs than in the long-styled (L) morph. A pollen addition experiment showed that low seed set in the L morph can only partly be explained by reduced pollen receipt. Comparison of morph frequencies between adult plants and their open-pollinated progeny in three populations failed to reveal any morph-specific fitness differences. Analysis of growth and flowering of open-pollinated progeny from three populations in a 2-yr glasshouse experiment revealed differences among morphs for some parameters in some populations, but no consistent patterns were evident. The deficiency of the M morph may have resulted from an historical accident during post-glacial colonization of parts of the northern range. Computer calculations examining the return of morph frequencies to equilibrium after major perturbations indicated that populations of long-lived clonal species like D. verticillatus may preserve skewed morph ratios for > 10 000 yr. Moreover, the distribution of fossil seeds indicates that this species has occurred in its post-glacial range in southern Ontario for only approximate to 5000 yr. It is important to distinguish between ecological and evolutionary time scales when testing selective interpretations of microevolutionary patterns involving clonal taxa in glaciated regions.