A comparison of methods of estimating lucerne population for monitoring persistence

Abstract

Two sowing rates, 4 and 12 kg/ha, and two sowing methods, broadcast and drilled into 18 cm rows, were used to give different population densities and spatial distributions of a semi-winter dormant lucerne, Condura 73 Brand, in mown dryland plots. Lucerne population in these plots was monitored for the first 12 months after sowing by either estimating the frequency of cells occupied by plants, using a 1 m2 steel mesh with a 10 cm x 10 cm or a 5 cm x 5 cm grid, or estimating plant population by taking crown counts at the soil surface. To estimate the true plant population for each sowing rate and method, plots were excavated regularly and plant tap roots counted. Crown counts consistently underestimated the true plant population over the range of 35-450 plants/m2. The underestimation was greater at the higher sowing rate. At populations greater than 150 plants/m2, frequency estimates were not sensitive to changes in plant numbers, but within the range of 20-80 plants/m2 they reliably reflected changes in plant populations. In a second experiment, six lucerne cultivars with a range of dormancy characteristics and crown structures were used to investigate the effect of crown size and type. Although crown counts underestimated (P<0.05) the actual plant population of a highly winter active, narrow-crown cultivar, there were no significant differences between cultivars in the number of plants or their frequency estimates.