Biological invasions provide opportunities to study evolutionary processes occurring during contemporary time scales. Here, I combine a literature review of common garden studies of invasive plant species, with field and glasshouse experiments on populations of the outcrossing, perennial, wetland invader Lythrum salicaria (purple loosestrife - Lythraceae), to investigate the evolutionary genetics of local adaptation in reproductive and life-history traits. A review of 32 common garden studies of 28 introduced species identified previously unrecognized latitudinal clines in phenotypic traits in both native and introduced populations. To obtain direct evidence for clinal variation and local adaptation, I investigated populations of L. salicaria sampled along a latitudinal gradient of growing season length in eastern North America. Controlled pollinations of plants from 12 populations provided no evidence for the breakdown of self-incompatibility to self-compatibility towards the northern range limit. However, a quantitative genetic experiment involving 20 populations revealed latitudinal clines in population mean, variance and skew for days to flower and vegetative size. Broad-sense estimates of genetic variance were significant for most traits; however, strong inter-correlations among traits suggested that fitness trade-offs have constrained population divergence. The observed clines supported a model of selection for early flowering in northern populations constrained by a trade-off between age and size at flowering. A comparison of variance-covariance matrices of family and population means (G and D, respectively) of life-history traits demonstrated that populations have evolved in response to selection under genetic constraints, rather than through neutral processes. A reciprocal transplant experiment involving six populations and three common gardens spanning the latitudinal range provided direct evidence for local adaptation in flowering phenology. Populations maintained the same rank-order for time to flowering and vegetative size at each site, and southern populations had the highest fecundity at the southern site but the lowest at the northern site. Finally, a phenotypic selection analysis in each common garden involving 61 F2 families of crosses between a northern × southern populations confirmed that selection favours earlier flowering in northern populations. These results demonstrate that natural selection on reproductive phenology has accompanied the invasive spread of L. salicaria in eastern N. America.

Biological invasions provide opportunities to study evolutionary processes occurring during contemporary time scales. Here, I combine a literature review of common garden studies of invasive plant species, with field and glasshouse experiments on populations of the outcrossing, perennial, wetland invader Lythrum salicaria (purple loosestrife - Lythraceae), to investigate the evolutionary genetics of local adaptation in reproductive and life-history traits. A review of 32 common garden studies of 28 introduced species identified previously unrecognized latitudinal clines in phenotypic traits in both native and introduced populations. To obtain direct evidence for clinal variation and local adaptation, I investigated populations of L. salicaria sampled along a latitudinal gradient of growing season length in eastern North America. Controlled pollinations of plants from 12 populations provided no evidence for the breakdown of self-incompatibility to self-compatibility towards the northern range limit. However, a quantitative genetic experiment involving 20 populations revealed latitudinal clines in population mean, variance and skew for days to flower and vegetative size. Broad-sense estimates of genetic variance were significant for most traits; however, strong inter-correlations among traits suggested that fitness trade-offs have constrained population divergence. The observed clines supported a model of selection for early flowering in northern populations constrained by a trade-off between age and size at flowering. A comparison of variance-covariance matrices of family and population means (G and D, respectively) of life-history traits demonstrated that populations have evolved in response to selection under genetic constraints, rather than through neutral processes. A reciprocal transplant experiment involving six populations and three common gardens spanning the latitudinal range provided direct evidence for local adaptation in flowering phenology. Populations maintained the same rank-order for time to flowering and vegetative size at each site, and southern populations had the highest fecundity at the southern site but the lowest at the northern site. Finally, a phenotypic selection analysis in each common garden involving 61 F2 families of crosses between a northern × southern populations confirmed that selection favours earlier flowering in northern populations. These results demonstrate that natural selection on reproductive phenology has accompanied the invasive spread of L. salicaria in eastern N. America.