Modelling soil carbon changes after converting pasture to maize cropping

Finding the optimal balance between maximising productivity and minimising soil carbon (C) losses requires a good understanding of the C dynamics of agricultural systems. In New Zealand’s dairy farms, maize silage is often used as supplemental feed when there is insufficient pasture growth. A previous study based on eddy-covariance and C balance measurements over two successive years found that conversion of a paddock grazed by dairy cattle to growing maize for silage resulted in substantial C losses. Here, we compared these previously observed C balances with simulations using the ecophysiological model CenW. The simulations indicated that over the cropping period, the maize site lost 5.0 tC ha −1 in year 1 and 3.7 tC ha −1 in year 2, respectively. These modelled losses were slightly smaller than the C losses inferred from eddy-covariance and harvest measurements alone (-8.5 tC ha −1 and −4.4 tC ha −1 in years 1 and 2, respectively). Even though the maize-silage system fixed a large amount of C during the growing season, it nonetheless lost C overall because nearly all fixed C was exported in silage, with only minor returns through root growth and litter fall. We further conducted a scenario analysis to assess the extent of C losses when different fractions of maize biomass were harvested. In particular, we tried to understand how C dynamics would be affected if only cobs were harvested instead of all plant material. The scenario analysis suggested that retention of greater biomass residuals after harvesting could minimise or even prevent soil C losses from cropping. These results provide valuable insights for the sustainable use of supplemental feeds in agricultural systems. Modelled carbon (C) balance of a maize crop with different percentages of maize above-ground biomass harvested and removed off-site. Maize silage with 92% of all biomass removed can lead to a large C loss, but production systems with less biomass removal, such as systems where only cobs are harvested (50% C removed), would retain more C on-site. The less C is removed in harvests, the more favourable are the resultant outcomes for site C balances. • We modelled the carbon (C) balance of a maize silage crop grown on a pasture site. • Under maize, we modelled a loss of ≈ 4 tC ha −1 in each of two successive seasons. • While maize fixed much carbon, nearly all was exported in silage. • Maize feeding causes more C loss on producing than expected gain on receiving land. • Integrated production systems must consider both C gains and C losses.