Lucerne yield, water productivity and persistence under variable and restricted irrigation strategies

Lucerne (Medicago sativa L.) has the potential to be grown widely under water-limiting conditions in the dairy region of northern Victoria and southern New South Wales, Australia, possibly because of its greater water productivity and because irrigation management of lucerne can be more flexible compared with other forage species. A large-scale field experiment was conducted at Tatura in northern Victoria, over 5 years to determine the effects of limiting (deficit) and non-limiting irrigation management on the dry matter (DM) production, water productivity (irrigation and total water productivity) and stand density (or persistence) of lucerne. Nine irrigation treatments were imposed that included full irrigation, partial irrigation and no irrigation in either a single, or over consecutive, irrigation seasons. In the fifth year of the experiment, all plots received the full irrigation treatment to examine plant recovery from the previous irrigation treatments. In any one year, there was a linear relationship between DM production and total water supply (irrigation plus rainfall plus changes in soil water) such that DM production decreased as the total water supply - due to deficit irrigation - decreased. Over the 5 years, annual DM production ranged from 1.4 to 17.7tDMha(-1) with the highest production occurring in plots that received full irrigation. Irrigation water productivity was inversely related to the amount of water used and was higher in the treatments that had only been partially irrigated for that year compared with the treatments that had been fully watered for that year. Total water productivity values were significantly lower only in the treatments that had not been irrigated for that year, and there was little difference between the treatments that were only partially watered during the year and the fully watered treatments (range 9.1-12.2kgDMha(-1)mm(-1) for Year 4). There was no significant reduction in plant density or plant persistence in those plots where deficit irrigation had been imposed. However, the high irrigation regime and poor drainage in the fully irrigated border-check plots significantly reduced plant density and allowed weed infestation in the fifth year of the experiment. These results suggest that, although lucerne DM production is directly related to total water use and may be significantly reduced in the irrigation regions of south-eastern Australia in seasons when water is restricted, the lucerne stand is able to fully recover once a full irrigation regime is resumed. This makes lucerne an ideal forage species for situations when water is limiting.