Modelling nitrogen fixation of pea ( Pisum sativum L.)

Abstract

Nitrogen fixation was simulated for a leafless variety (Delta) of pea (Pisum sativum L.) in central Sweden. It is assumed that N2 fixation is basically proportional to root biomass, but limited by high root N or low substrate carbon concentrations. Input data on root carbon and nitrogen were estimated from observations of above-ground biomass and nitrogen. The simulated N2 fixation was compared with estimated values from observations using the 15N labelling technique. Test data were taken from pea monocultures and pea-oat mixtures with varying pea biomass levels during 1999. Simulated within-season accumulated N2 fixation correlated to the estimated N2 fixation with a correlation coefficient (R 2) of 0.74. For seasonal simulations, the predictability was higher (R 2=0.93). Two alternative non-dynamic models, estimating seasonal N2 fixation as proportional to above-ground biomass and above-ground N, respectively, gave lower predictability (R 2=0.83 and 0.80, respectively). The models were also applied to a second year (1998) and two other sites by comparison with accumulated N2 fixation estimated by the Difference method. A halved specific N2 fixation rate (expressed per unit of root biomass) in 1999, compared with 1998, corresponded to essentially dryer and warmer soil conditions during 1999. It was indicated that the variations in soil moisture were more important than soil temperature. It was concluded that the abiotic responses might be of great importance for modelling N2 fixation rate under different soil conditions.