Variation in dormancy of the wetland sedge, Carex elongata L., was tested using seeds from three wild populations and the garden-grown progeny of one population. Germination experiments, comprising four combinations of temperature and light, were conducted either with fresh-matured or cold-stratified seeds, to assess the relative contribution of populations and mother plants to the total variation. Between-year variation was tested at the population level and, partly, at the individual level, using seeds collected in two consecutive years. Among-population variation accounted for 72%, and among-individual variation for 23%, of the total variance in the first experiment. Significant differences were apparent between one wild population and its garden-grown descendants. Differences in germinability among populations were maintained in the two consecutive years, but a proportion of the variance was due to the contribution of the maternal environment. Weak evidence for genetic control at the individual level was shown by a correlation across years in one population. However, the lack of a main effect at the individual level in the first experiment makes it difficult to assess the relative contribution of the mother plants to the total variation. Our results imply that germination patterns of C. elongata have a genetic basis, but are markedly modified by environmental conditions.

The flooding tolerance of Carex species was studied in relation to their field distribution and their capacity to form root aerenchyma under controlled conditions. In an alpine meadow, six Carex species were selected which were distributed in a clear zonation correlating with water content of the soil. Carex sempervirens and C. ferruginea were only found on nonflooded soil, the latter species preferring moister conditions. Carex davalliana and C. nigra were both associated with water-saturated soil, whereas C. limosa and C. rostrata preferred partially submerged conditions. Carex davalliana and C. limosa were bound to flooded soils with a relatively high redox potential and horizontally flowing groundwater. Carex rostrata and C. nigra grew in stagnant soil-flooded conditions with low soil redox potentials. The amount of aerenchyma in the roots of all species increased when grown in oxygen-deficient stagnant agar. This increase in root porosity, combined with increased root diameter, presumably improved internal aeration of the roots. Although all species survived experimental soil flooding, partial submergence was lethal to C. sempervirens and, surprisingly, also to the wetland species C. davalliana. Carex ferruginea showed a reduced growth rate during partial submergence. The three other species, all wetland plants, reached highest biomass production under soil-flooded and partially submerged conditions, with slower growth on free-draining soil. It is concluded that aerenchyma is not constitutive in the Carex species under study, and is best developed in Carex species from wetlands. Species with less aerenchyma perform poorly when soil-flooded, but conditions of partial submergence could even affect species with a considerable amount of root aerenchyma.