Main content area

Seed Weight in Relation to Environmental Conditions in California

Baker, Herbert G.
Ecology 1972 v.53 no.6 pp. 997-1010
Ranunculus, altitude, dispersibility, drought, flora, growing season, habitats, herbaceous plants, herbs, humans, indigenous species, mesic conditions, mountains, nutrition, perennials, plant communities, reproductive performance, risk, root growth, root systems, roots, seedlings, seeds, shade, shoots, shrubs, trees, water stress, weeds, California, England
Analyses of the California flora involving nearly 2,500 taxa are presented which show that there are correlations between the weights of individual seeds and environmental conditions in which their producers normally grow. These differences in seed weight appear to be adaptive and result from compromises between increased nutrition of the seedling which would result from larger food reserves in heavier seeds and increased dispersibility and increased reproductive output which are provided when smaller seeds are produced in larger numbers. Literature and experiments show a general positive correlation between seed weights and rates of shoot and root growth, at least within species. Herbs (annual and perennial), shrubs and trees are necessarily treated separately in the calculations in this paper (for seed weights increase progressively in a series from annual herb to tree when to California flora or any particular community type within it are considered). Raw seed weights are reduced to a series of classes (on a logarithmic basis) before means for floras or community types are calculated. Generalizations arrived at from considerations of "native" species in "natural" plant communities are confirmed by their applicability to "introduced" species now forming various kinds of "weed" communities in California. Finally, species lists from actual stand analyses, including both "native" and "introduced" species, are utilized to provide the data for more precise analyses. For herbaceous plants, seed weights are higher, on the average, for taxa whose seedlings are exposed to the risk of drought soon after establishment (giving faster root—development). Such a relationship can be demonstrated for species of a single genus or, on a combination basis, for community types as a whole and can be put on a quantitative basis by subjectively ordinating community types (in relation to the likelihood of drought stress hitting the seedlings) and making rank correlations with mean seed weight for each community type. The relation holds even for desert communities (where large—seeded perennials produce large root—systems but small—seeded ephemerals complete at least their seedling development in temporarily mesic conditions). In coastal communities, the importance of wind—dispersal of the seeds of species whose seedlings become established in rock crevices outweighs any droughtiness of the habitat in favoring smaller seeds than typify the community types generally. Correlations of herb seed—weight with likelihood of seedlings becoming established in shade or in conditions of severe competition are less marked for California than Salisbury found them to be in England. For shrubs, shading and competitive stress appear to be more influential factors in promoting increased mean seed weight but for trees moisture availability again appears to be most important. Another kind of correlation is established; between mean seed weight and altitude at which the plants occur. With decreasing length of the growing season as Californian mountains are ascended the mean seed weight (whether measured on a subspecies, species or community type basis) also decreases. This appears to represent the selection of a strategy in which a reduction in the availability of photosynthate is reflected in smaller seeds rather than in reduced output as found in Ranunculus by Johnson and Cook. Although taxa introduced to California have fitted with the rules holding for native plants, they tend to have slightly heavier seeds than native species growing in climatically similar habitat types. This difference may be particularly related to human influence in making such habitats somewhat more xeric. Improved methods of analysis are suggested and further correlations which might be sought are discussed.