Analysis of inflorescence organogenesis in eastern gamagrass, Tripsacum dactyloides (Poaceae): the wild type and the gynomonoecious gsf1 mutant

Inflorescence organogenesis of a wild-type and a gynomonoecious (pistillate) mutant in Tripsacum dactyloides was studied using scanning electron microscopy. SEM (scanning electron microscope) analysis indicated that wild-type T. dactyloides (Eastern gamagrass) expressed a pattern of inflorescence organogenesis that is observed in other members of the subtribe Tripsacinae (Zea: maize and teosinte), family Poaceae. Branch primordia are initiated acropetally along the rachis of wild-type inflorescences in a distichous arrangement. Branch primordia at the base of some inflorescences develop into long branches, which themselves produce an acropetal series of distichous spikelet pair primordia. All other branch primordia function as spikelet pair primordia and bifurcate into pedicellate and sessile spikelet primordia. In all wild-type inflorescences development of the pedicellate spikelets is arrested in the proximal portion of the rachis, and these spikelets abort, leaving two rows of solitary sessile spikelets. Organogenesis of spikelets and florets in wild-type inflorescences is similar to that previously described in maize and the teosintes. Our analysis of gsf1 mutant inflorescences reveals a pattern of development similar to that of the wild type, but differs from the wild type in retaining (1) the pistillate condition in paired spikelets along the distal portion of the rachis and (2) the lower floret in sessile spikelets in the proximal region of the rachis. The gsf1 mutation blocks gynoecial tissue abortion in both the paired-spikelet and the unpaired-spikelet zone. This study supports the hypothesis that both femaleness and maleness in Zea and Tripsacum inflorescences are derived from a common developmental pathway. The pattern of inflorescence development is not inconsistent with the view that the maize ear was derived from a Tripsacum genomic background.