Nitrogen fixation and site nitrogen balances of fertilised dryland lucerne (Medicago Sativa L.) at Lincoln University, Canterbury, New Zealand

Abstract

Lucerne (Medicago sativa L.) is a high value perennial forage legume. It is deep-tap rooted, has high water use efficiency and is tolerant of drier environments. It produces high quantity and quality dry matter (DM). Yet, there was a lack of data on the impacts of using increasing amounts of N or other fertilisers or lime on nodulation, N fixation and N balances of dryland lucerne. A field experiment was conducted for more than one cropping year (Sep 2012-Nov 2013) on an established dryland lucerne stand at Lincoln University, Canterbury, NZ. The experiment consisted of six main treatment plots: control (0 lime/fertiliser), pelleted agricultural lime (L), S, and P (single superphosphate) + S (PS), L + S (LS) and L + S + P (LPS). The plots were split into subplots with one half receiving 0N and the other +N (urea, 46% N), resulting in 12 treatments, with four replicates in a split plot randomised complete block design. The rates used were 6000 kg L/ha, 128 kg S/ha (half as gypsum 8% S and the other half as elemental-90% S) and 50 kg P/ha. The main treatments were applied once at the start, while N fertiliser was applied after each harvest (cut-and-carry system) according to the N off-takes of the herbage. The cumulative N fertiliser applied was 813 kg N/ha/yr over six applications. The DM yields, N off-takes, N fixation and minerlaisable N were measured at each of the six sequential harvests. Two of these data sets were from spring 2012 and 2013 harvests. Annual parameters were obtained by omitting the 2012 spring data sets. Annual N balances and economic values of the extra DM yields from the +N plots were calculated. The main treatments did not affect the measured parameters, or the herbage regrowth at each harvest. Yet, they were affected strongly by the N fertiliser treatments in combination with the time of herbage regrowth. The cumulative mean yields were 27-30 and 23-25 t DM/ha/yr (P<0.001) for the +N and the 0N subplots, respectively. The annual DM yields, herbage N concentrations and N uptakes were higher on the +N plots compared with the 0N plots. N fixation and soil mineralisable N status were highest on the 0N plots. However, the %N concentrations and herbage palatability decreased with increasing DM yields. This was because these stands were harvested at 40-80 cm tall comapared with usual grazing heights of 25-40 cm. Cumulative mean fixed N were 467-501 and 390-440 kg N/ha/yr (P<0.001) for 0N and +N subplots, respectively. This was equivalent to 20-23 kg N/t DM on the 0N plots and 15-16 kg N/t DM for +N plots. Soil mineralisable N was depleted by 57% and 50% on the +N and 0N plots, respectively over the experimental period. Nitrogen balance calculations revealed that annual N fertiliser use efficiency of lucerne was 38%, resulting in 62% of the applied urea N being unaccounted for. This N was likely lost to either volatilisation or soil N leaching and immobilisation. N fertiliser application reduced N fixation by 81 kg N/ha/yr (15%). This indicates that lucerne strongly prefered fixed N rather than urea or soil N. Further research is required to confirm this result. Further studies quantifying N loss via the different N loss pathways are also required. The mean overall fertiliser N response ratio was low (5:1) and a gross margin deficit of $118 ha/yr was calculated. The use of increased amounts of P, S and lime on dryland lucerne was also unjustified for this lucerne stand.