Canopy dynamics of lucerne (Medicago sativa L.) genotypes of three fall dormancies grown under contrasting defoliation frequencies

This study evaluated the canopy dynamics of three lucerne (Medicago sativa L.) genotypes of different fall dormancy (FD) ratings when grown under three contrasting defoliation regimes over five regrowth seasons at Lincoln University, Canterbury, New Zealand. Crops were sown in October 2014 in a split-plot design with main plots as defoliation frequencies (28 days [DF 28], 42 days [DF 42] and 84 days [DF 84]) and genotypes as subplots. Genotypes varied in fall dormancy (FD) rating and were classified as dormant (cv. AgR Palatable, FD2), semi-dormant (cv. Grassland Kaituna, FD5), or non-dormant (cv. SARDI 10, FD10). All crops were grown through an establishment phase until January 2015, when defoliation treatments were imposed. The experiment terminated in May 2019. Defoliation treatments were imposed from early spring (August-September) to autumn (May) with a single clean-up graze in late June/July. Defoliation frequency x genotype interactions showed FD10 was the most productive in regrowth season 1 but least persistent and productive in regrowth season 5, particularly under the DF 28 and DF 42 treatments. Results suggest different partitioning responses amongst genotypes. In a decreasing photoperiod, the growth rate of FD2 decreased by 0.77 +/- 0.07 kg DM ha-1 degrees Cd- 1 compared with 0.67 +/- 0.09 kg DM ha-1 degrees Cd- 1 for FD5 and FD10. However, FD10 showed no response to an increasing photoperiod under the DF28 defoliation regime. The lower rate of partitioning to perennial biomass of FD10 was suggested by the rapid decline in shoot density (-0.000112 shoot m- 2 degrees Cd- 1) over time compared with - 0.000049 shoot m- 2 degrees Cd- 1 for FD2 and FD5. FD10 had heavier individual shoot mass over the last two years, but its canopy plasticity was insufficient to control the shoot size/density balance, so there was greater weed ingress. The DF 84 treatment enabled FD10 to produce > 15 t DM ha-1 year-1 over the five regrowth seasons, which was not different to FD2 and FD5. Under more frequent defoliations, FD2 and FD5 were more persistent and had higher yields than FD10.