FIELD AND GENETIC-STUDIES TESTING OPTIMAL OUTCROSSING IN AGAVE SCHOTTII, A LONG-LIVED CLONAL PLANT

In this study we combine field experiments, designed to test the predi ctions of optimal outcrossing theory in Agave schottii, with molecular genetic studies, using RAPD (random amplified polymorphic DNA), polym erase chain reaction to assess the underlying genetic hypothesis of op timal outcrossing theory. Initially, 48 ''females'' of A. schottii wer e hand-pollinated with pollen collected from 1 m, 10 m, 100 m, and 250 0 m distances. Each female received all four distance treatments. Addi tionally, a subset of the focal females and their pollen donors were u sed in an analysis of genetic similarity across the four distances. Re sults of hand-pollinations showed that crosses of 1 m had significantl y lower seed set than 10 m and 100 m crosses. Crosses of 2500 m had in termediate seed set. Combined relative fitness was significantly lower for 1 m crosses compared to 10 m crosses, while 100 m and 2500 m cros ses were intermediate. Thus, A. schottii experiences inbreeding depres sion and a trend toward outbreeding depression. Genetic analyses showe d a similar pattern: individuals 1 m apart had on average higher genet ic similarity (proportion of bands shared) than individuals separated by greater distances, with a trend toward lower genetic similarity for plants located 2500 m distant. The observed spatial genetic patterns are likely maintained by the combined effects of clonal reproduction, clone longevity, limited seed dispersal and the substantial number of inbred progeny produced, counteracting distant allele transfer which t ends to reduce population genetic structure. The correspondence betwee n our ecological and genetic results indicates that RAPD markers are u seful tools for assessing ecological phenomena.