Male and female fertility, seed germination, and progeny fertility were used to determine cultivar fertility in species of Lythrum. One short-, 11 mid-, and six long-styled cultivars were included in this study. Duplicates of several cultivars from different nurseries and three unknown cultivars from Minnesota gardens were also collected. Plants from 17 Minnesota and one Wisconsin population of L. salicaria served as fertile male and/or female testers. Pollen stainability (usually 100%) showed low levels of male gamete abortion. Pollen size within and among anther type varied widely; possible 2n gametes were present in primarily the short- and mid-anther morphs. Seed production per capsule from legitimate cross-pollinations, using cultivars as male parents with Minnesota or Wisconsin female testers, averaged 48 ± 36 across style morphs. Cultivars differed as males, as did anther morphs. With female fertility tests, seed set per capsule ranged from zero to 152 and averaged 54 ± 40 in legitimate pollinations (i.e., pollinations between stamen and styles of the same length). Seed set for other crosses showed similar trends. Only `Morden Gleam' produced no seed with all legitimate pollinations, although illegitimate selfs or interspecific crosses produced seed. Seed from legitimate crosses of L. salicaria × cultivars had 30% to 100% germination. Common male and female parents within each legitimate crossing group were not significantly different. This study showed that the cultivars are highly fertile when used as male or female parents with wild purple loosestrife, native species (L. alatum Pursh.), or other cultivars. Thus, cultivars grown in gardens could serve as pollen or seed sources for the continued spread of purple loosestrife. The implications of cultivar fertility, especially interspecific F1 hybrids, is discussed in relation to the spread of noxious weeds in wetlands.
Neil O. Anderson and Peter D. Ascher
Hai-fei Yan, Xue-jun Ge, Chi-ming Hu, and Gang Hao
, 1997 ; Pemberton et al., 1995 ). Given has P. obconica with a heteromorphic self-incompatibility system ( Stevens and Murray, 1981 , 1982 ), we consider inbreeding not the main reason for the heterozygote deficiency. However, Wahlund effect is at