WINTER CEREALS GROWN FOR GRAIN AND FOR THE DUAL-PURPOSE OF FORAGE PLUS GRAIN .2. WATER-USE AND WATER-USE EFFICIENCY

Water use of barley (Hordeum distichon L.) and triticale ( X Triticose cale Wittmack) grown for the dual purpose of forage plus grain as an a lternative to traditional grain monoculture (control) was assessed at Granada, southern Spain, in two experiments during three cropping year s. No significant differences in seasonal evapotranspiration (ET) and soil water depletion at maturity were found between production targets , irrespective of genotype, sowing date and year. A more efficient use of autumn and winter rainfall explained the higher water use observed in 1989/90 in early sown barley as compared to the normal sowing date . There were differences in water-use efficiency (WUE) between seasons in response to clipping: in 1987/88, clipping increased WUE in both e xperiments relative to the control, while it either reduced it (genoty pe experiment) or did not affect it (sowing system experiment) in the two following seasons. The response of WUE to clipping was a consequen ce of similar, differential responses between years observed in dry ma tter and grain production. Weather-crop interactions caused variable d rought and crop water-use patterns among years, which affected the dif ferent responses to clipping observed among years. In 1987/88, the cro ps experienced a prolonged drought period that started around clipping , and increased in severity until anthesis, when spring rains improved growth conditions for some weeks. Forage removal decreased water use and reduced plant water deficit, therefore allowing for enhanced growt h and increased water use around anthesis. Thus, clipped cereals had h igher grain numbers, dry matter and grain yield than the control at ha rvest. However, the water-use partitioning observed in the clipping-an thesis period of 1987/88 did not occur in following two seasons. This was possibly due to higher soil evaporation losses after forage remova l for clipped cereals in 1988/89, and higher water availability in the clipping-anthesis period in 1989/90. Thus, clipping, by reducing the photosynthetic surface after forage removal, yielded less biomass and grain, and resulted in lower WUE, relative to the control. Winter fora ge production was associated to ET in 1987/88 and 1989/90, and to WUE in 1988/89 and 1989/90, during the crop phases prior to clipping. Howe ver, the increases in ET did not imply lower water availability for su bsequent crop phases, as it was the result of more efficient use of au tumn and winter rainfall.