IS IT POSSIBLE TO DIAGNOSE AT HARVEST A PROBLEM OF POLLEN STERILITY IN WHEAT

We tested if various plant parameters, easy to measure at maturity in a wheat crop, could discriminate between situations with and without l imitations in grain set caused by pollen sterility. In a pot experimen t, plants were subjected to two treatments: shading during stem elonga tion, reducing growth; and climatic stress (low radiation and chilling ) at meiosis, inducing pollen sterility. The percentage of grain setti ng among the two oldest florets within each spikelet (LF) and the rati os of grain number to (i) biomass of the chaff and (ii) biomass of the straw plus chaff were markedly reduced by the stress at meiosis and e ither little or not affected by shading during stem elongation. Two in dices were tested in the field: LF and the ratio, RK, of the grain num ber per ear to an estimate, based on chaff biomass, of the potential n umber in the absence of sterility. These indices were measured in a ne twork of five field experiments with six cultivars. The fluctuations i n the indices were compared to those of a reference criterion that ide ntified sterility: the rate of grain setting among the two oldest flor ets of each spikelet, measured in bagged ears to prevent cross-pollina tion. Both indices were reduced when the reference criterion diagnosed a major sterility but they did not always decrease in case of minor s terility. Classification of the cultivars based on the reference crite rion was similar to that based on LF, but not on RK. LF diagnosis of s terility was in agreement with the occurrence of climatic conditions l ikely to induce sterility.