We examined seed-mass variation in 39 species (46 populations) of plants in eastern-central Illinois, USA. The coefficient of variation of seed mass commonly exceeded 20%. Significant variation in mean seed mass occurred among conspecific plants in most species sampled (by hierarchical ANOVA), averaging 38% of total variance. For most species, within-plant variation was the larger component of total variance, averaging 62% of total variance. Variation in seed mass among fruits within crops was significant in most species tested. We conclude that variation in seed mass among and within plants is widespread and common. There was little evidence of trade-offs between number of seeds and mean or variance of seed mass, and little correlational evidence of local competition for maternal resources. No consistent ecological (dispersal mode and growth form) correlates of variance of seed mass were evident.

This study examines the potential of small—scale environmental heterogeneity to influence population structure and fitness of individuals within a population. Two populations of Abutilon theophrasti were established at each of three densities in the greenhouse. At each density, one population was grown on substrate made heterogeneous before planting, with the addition of stone, localized nutrients, or sand, and soil compactions on a scale corresponding to individual seedlings. These were termed "patch types" and wee chosen to represent factors that would be unpredictable in both time and space to seedlings emerging in nature. Another population was established on homogeneous substrate at each density. Seedling height, final height, biomass, and seed number were measured. The variation in biomass of individuals within a population was significantly greater on heterogeneous than on homogeneous substrate at the lowest density. When the data where analyzed with ANOVA, there was a significant interaction of patch type with presence or absence of heterogeneity. Patch type explained 47—62% of the variance in 2—d seedling height and 27—33% or 30—37% of the variance in seed output or final biomass, respectively. A multiple regression model including patch type, seedling size, and random block together explained 58—76% of the total variance in final biomass, depending on density. Size of immediate neighbors showed no correlation with individual size. The data support our hypothesis that small—scale stochastic events can influence the position of an individual in a population hierarchy. This conclusion implies that genetic determinants of fitness may be overridden by commonly occurring stochastic events. The resulting lack of predictable selection against certain genotypes should contribute to the maintenance of genetic diversity in populations.

No abstract

No abstract

Net CO2 assimilation and water relations of Dryas integrifolia from Devon Island, N. W. T., were studied under controlled-environment conditions. Maximum net CO2 assimilation rate for single leaves was 18.7 mg g−1h−1. The optimum leaf temperature for net CO2 assimilation was between 9 and 14 °C. Positive net CO2 assimilation occurred at −5 °C. Respiration and net CO2 assimilation decreased with the onset of dormancy. Dark respiration was generally equal to or greater than that of other arctic and alpine species. The mean leaf water potential of actively growing plants was −11.4 bars. Values higher than −7 bars were found only infrequently. Turgor pressure was maintained at high values (+6 to +10 bars) over a wide range of water potentials, suggesting osmotic adjustment to soil moisture deficit.