The Phylogenetic Distribution and Frequency of Self-Incompatibility in Fabaceae

Premise of research. Surprisingly little is known regarding the phylogenetic distribution and frequency of self-incompatibility (SI) in some of the largest angiosperm families, including the legumes (Fabaceae). The ecological and economic significance of the family has motivated an enormous but scattered output of literature concerning its breeding systems and reproductive biology. Since the last synthesis of this literature, we have gained a clearer understanding of the various mechanisms responsible for SI, but data on their number and phylogenetic distribution remain limited.</p> Methodology. We compiled species-level information on legume breeding systems, pollination syndromes, longevity, stature, stem composition, geographic location, and ploidy. The resulting database contains entries for 1167 unique species across 184 genera, accounting for nearly 5% of the family. Using these data, we assess the phylogenetic distribution and frequency of SI in Fabaceae and evaluate the potential correlation between SI and three other traits: ploidy, stem composition, and longevity.</p> Pivotal results. We estimate that approximately one-fifth of legume species have SI, with the remainder being self-compatible. Appropriate controlled pollinations are available for only a small portion of species in our database, and genetic studies are limited to fewer than 0.5% of the family. Moreover, our survey reveals no convincing evidence that a single SI mechanism operates in Fabaceae. Despite uneven coverage and a sample of ca. 5% of Fabaceae, the results of our phylogenetic analyses strongly suggest that longevity is statistically significantly correlated with breeding systems, while ploidy is not.</p> Conclusions. Our collection reveals a great variety of SI expressions in the family. Compelling evidence indicates that one of the best-studied SI systems-RNase-based SI-is homologous across core eudicots and ancestral to the group. Previous studies surmise that this system may operate in Fabaceae, but this remains an open question. Whatever its genetic causes among legumes, the reduced seed and fruit set after selfing appears to evolve dynamically and exhibit uncommon variation. We highlight the need for empirical work evaluating SI and outline areas where future efforts may provide disproportionate rewards.</p>