REDUCED SEED SET IN ELYTRIGIA-REPENS CAUSED BY ALLELOPATHIC POLLEN FROM PHLEUM-PRATENSE

Earlier studies have shown that extracts from pollen of Phleum pratens e reduce pollen germination and seed set in Elytrigia repens (L.) Nevs ki (Poaceae), but the effect of in situ pollen from P, pratense on see d set in E. repens in the field has not been previously demonstrated. By clipping the inflorescences of P. pratense just prior to flowering, we reduced pollen dispersal in fields at three old-field sites. In ad jacent fields at all three sites, P. pratense was allowed to flower un impeded. In the clipped fields, the mean number of pollen grains of P. pratense per stigma of E. repens was reduced to less than 1 versus 9- 10 grains in the unmanipulated fields. Mean percent seed set of E. rep ens in the clipped fields was approximately 65-70%, whereas in the unm anipulated fields it was approximately 15-20%. In the following year, when no treatments were applied (i.e., P. pratense was allowed to disp erse naturally in all fields), mean percent seed set in the same plant s of E. repens was uniformly reduced to less than 15% in all fields. S everal lines of evidence suggest that this reduction in seed set in E. repens was caused by allelopathic pollen of P. pratense, rather than by physical occlusion: (i) previous studies using extracts of P. prate nse pollen elicited similar decreases in seed set in E, repens, and th e seed set decreases were nonlinear (i.e., not indicative of displacem ent); (ii) the number of P. pratense pollen grains on E. repens stigma s needed to cause pollen allelopathy (based on previous pollen extract studies) existed in the unmanipulated fields but not in the clipped f ields; and (iii) E. repens pollen outnumbered P. pratense pollen on E. repens stigmas by at least 5:1 (even when P. pratense inflorescences were not clipped).