Forage and seed yield response of lucerne cultivars to chemically weeded and non-weeded managements and implications for germplasm choice in organic farming

摘要

Organic livestock production is steadily increasing in Europe. Lucerne (Medico go saliva L) can play a key role in organic crop-livestock systems of southern Europe, owing to its positive attributes of adaptation, forage quality and nitrogen provision. This study tackled the unprecedented question on whether specific variety recommendation and, possibly, specific breeding are necessary for lucerne in organic systems. In lowland of northern Italy, where competition against weeds is expected to be the main determinant of success in organic farming, eight cultivars were evaluated in one environment under a chemically weeded management and an 'organic' management mimicked by no weed control, with the objectives of: (i) assessing the extent of cultivar x management interaction for dry matter (DM) and seed yield traits; and (ii) comparing different germplasm types for competitive ability and suitability to organic farming. The experiment was designed as a split-plot holding managements on main plots and cultivars on subplots. The test germplasm encompassed varieties bred under conventional or organic management and landraces from northern and central Italy, as well as germplasm with semi-erect or semi-prostrate habit. The recorded traits were lucerne DM yield over 3 years (13 harvests), DM yield over the first harvest of both the fourth and fifth year (as a measure of persistence), seed yield and its components assessed in the third year, and competitive ability against weeds as expressed by lucerne proportion under 'organic' management. Weed competition caused a severe reduction of DM yield and lucerne proportion in the 'organic' management but just a slight, non-significant reduction of seed yield, for which the greater stand thinning in this management relative to the chemically weeded one was largely compensated by increases in racemes per stem and seed weight. No cultivar x management interaction was observed for 3-year DM yield, seed yield and persistence (P>0.05), suggesting that specific variety testing under organic management is not necessary. Greater relative growth rate, implying both better DM yield in the absence of severe weed competition and greater competitive ability under organic management, was inferred as the main reason for the consistency between managements for response of best-performing germplasm. The low competitive ability of semi-prostrate germplasm reflected its low vigour. The results did not suggest any specific adaptation to organic conditions of landraces or material bred under organic management. Specific adaptation to the region of evaluation was the main driving factor for cultivar adaptation to both managements.