Success Of Grass Danthonia Spicata Progeny From Different Flowers

Introduction

Darwin (1877) first contrasted vigor of progeny from CH and CL flowers with others scientists have subsequently published accounts of this phenomenon. Since CH flowers are self-compatible, it is not impossible for a plant with CH flowers to be 100% self-fertilized. Most plant species have reproductive systems other than complete self-fertilization. Comparisons of CH and CL progeny reflect differences in native environments, and unless seed size and other seed related factors are controlled for, genetic and non-genetic differences may be confounded.

The objective of this study was to compare genetically based differences in adult survival and fecundity of progeny from CH and CL flowers of the grass D. spicata in a natural field habitat.

Materials and methods

Ten progeny from CH flowers and ten from CL flowers from 15 different families were raised in the greenhouse from seed collected at random from a natural population of D. spicata located in a field on the Duke University campus, Durham, NC. The 300 plants were split into six ramets each, totaling 1,800 ramets. Three ramets of each plant were planted on a grid in each of two locations approximately 15 m apart in the same field.

Plants were at first divided into ramets of the same size. The ramets prepared for the far site continued to grow like that they averaged 5.8 tillers per ramet. In total, 74% of the ramets were between 2 and 6 tillers in size, and 95% were between 1 and 9 tillers in size. They attempted to control for differences in initial ramet size by tiller numbers at the time of planting and using this variable as a covariate in subsequent analyses. There were no differences in initial tiller number between CH and CL (4.54 and 4.51 tillers.)

The ramets were planted in November 1979. The following May, when D. spicata flowered, all flowering culm were collected, and the total number of CH and CL flowers produced by each culm were counted. The following November the ramets were again surveyed, noting ramets that had died since the last census. The next spring, fall, and following spring. Therefore, survival was determined at six-month period over a 30-month period, and fertility was determined during three flowering seasons over the same 30-month period.

Results

Survival

Approximately 80% of the ramets planted in the fall of 1979 were still alive in the spring of 1982. Mortality was higher in the 6-month period following flowering than in the 6-month period preceding flowering. Survival of CH ramets was greater than or equal to survival of CL ramets at each of the five census points, but at no time were differences statistically significant.

Flowering

Seventy-three percent of the surviving ramets flowered the first spring, 92% the second spring, and 77% the third spring. The percentage of CH ramets that flowered was a little higher than the percentage of CL ramets that flowered in the first and second springs but was a little lower in the third spring. There were no significant differences flowering between CH and CL ramets at any point.

Fecundity

Mean flwer production per reproductive ramet was approximately 70 the first spring, 320 the second, 155 the third, and 450 in total. The average, CL ramets produced fewer flowers than CH ramets each of the three-flowering season. The ratio of mean flower production per CL ramet/mean flower production per CH ramet was 0.948 the first spring, 0.978 the second, 0.910 the third, and 0.965 in total.

Discussion

The present study represents one of the first experimental attempts to measure genetically based fitness differences between seed derived from CH and CL flowers under natural field conditions. Many of previous studies have shown a fitness advantage for CH progeny (Darwin, 1877 with other scientists), but others have shown no difference or even an advantage for CL progeny (Wilken, 1982 with other scientists). Donnelly (1955) found plants of Lespedeza cuneata from CH flowers averaged 25% higher biomass and 40% more seeds produced than plants from CL flowers, although the differences were due in large part to differences in seed size (Cope, 1966).

Under what conditions might we expect there to be no fitness differences between CH and CL progeny?

Two such conditions include (1) a genetically uniform population and (2) complete self-fertilization of CH flowers. In a genetically uniform population, selfing and outcrossing are functionally equivalent. This possibility is clearly irrelevant for the study population of D. spicata; abundant genetic variation has been previously documented for the proportion of CL flowers and other quantitative characters (Clay, 1982 with other scientists). The second condition, complete self-fertilization of CH flowers, is more problematic.

This experiment was conducted in a natural habitat, whereas most other fitness comparisons of CH and CL progeny have been conducted in greenhouse or common garden environments. The greater spatial and temporal heterogeneity experienced to plants growing in a uniform greenhouse environment may obscure the impact of inbreeding due to the greater environmentally induced phenotypic variation among individual plants.

In this study, they focused on the adult stage of the life cycle and found relatively small and mostly insignificant differences between CH and CL progeny. A previous study of seed production and seedling establishment in natural populations of D. spicata found much larger fitness differentials at the juvenile found much of the life cycle (Clay, 1983). The results of this study, along with those of the previous one, suggest that the adaptive significance of cleistogamy in D. spicata is to be found at the seed and seedling stage rather than at the adult stage.