Style morph frequencies in Minnesota populations of Lythrum (Lythraceae)

Abstract

Style morph frequencies (short∶mid∶long) were determined for a total of n = 11 918 plants in 16 Minnesota populations of Lythrum salicaria L. Nine populations were in the establishment phase, with population sizes ranging from n = 56 to n = 2 192. Most of these populations exceeded previously reported population sizes in the native European habitat. A nonparametric statistical test, the chi-square (χ²), can be used to determine if populations are at isoplethic equilibrium (1∶1∶1, short∶mid∶long); a χ² value >5.99 is significant at the 5% level. Only one established population (White Bear Lake, n = 1991, χ² = 3.0) fitted the null hypothesis for isoplethy, although all established populations contained all three style morphs. Pooled values for these populations indicated an excess of mids and longs, with shorts being deficient. Colonizing populations had a higher percentage of mids (54%) when compared to established populations (33.7%). Short styles were almost nonexistent (8%) in colonizing populations. Five out of the seven populations lacked at least one style morph. A review of the literature reporting style morph frequencies in tristylous L. salicaria revealed that no statistical analysis for isoplethy has been performed. Darwin originally assumed that all populations would be isoplethic, possessing equal numbers of all three style morphs, but concluded, without statistical analysis, that, instead, populations were anisoplethic. Since tests for statistical deviations from the expected frequencies (1∶1∶1) have not been used, χ² analysis was performed. Several of these populations were at isoplethic equilibrium (Nadder χsu2 = 1.7, Blelham χ² = 1.69, Potsdam χ² = 1.5, Vestfold χ² = 0.4, Buskerud χ² = 5.62, Kilchberg χ² = 0.35, Lausanne χ² = 3.32, Canberra χ² = 5.29, Massachusetts χ² = 3.13), suggesting that the general conclusion of anisoplethy in tristylous L. salicaria is inappropriate.