N2 fixation of three legume species in an agroforestry system

Due to the strong soil erosion caused by agricultural activities and sloping landscape, agroforestry has been introduced as a potentially sustainable cropping practice in the mountainous NW Vietnam. Using legumes as understorey crops in these systems has a potential to limit erosion and smother weeds, as well as provide additional nitrogen (N) to the system. The approximate amounts of N added through dinitrogen fixation serves as an essential information for the comprehensive evaluation of agroforestry practices. Three factors, legume species (Arachis hypogaea, Arachis pintoi and Stylo guianensis), harvest time and trial, were evaluated regarding to the estimated N2 fixation from air (kg/ha). Results were obtained through field sampling and analysis of total N concentration and 15N abundance. Plants inoculated with rhizobia from the fields were grown in the lab in order to obtain data reflecting 15N discrimination during N2 fixation (B values). However, the data indicated N contamination, so two other sets of B value were used in the calculation of %Ndfa. Similar total N concentrations (2.56-3.09 %) were found in all legume species. Total N concentration (%) and %Ndfa were dependent on harvest time, with total N concentration higher at harvest 1 than harvest 2, while the %Ndfa of A. hypogaea on the contrary was higher at harvest 2. Even though all legumes showed N fixation from air (%Ndfa), A. hypogaea had the lowest proportion. Perennial legumes, Stylosanthes guianensis and A. pintoi, contained much higher amount of N than A. hypogaea. Between the two harvests in the same growing season, legume shoots had a significant decrease of total N concentration, but no significant difference in total N amount, which arose from the drastic difference of biomass. Estimated N amount in A. pintoi and S. guianensis were higher than previous studies. Using the two sets of B values in the calculations produced somewhat different values of %Ndfa, but both indicated the same trend. No difference was found between the three trials and could be considered as a random variable in further study. Some factors caused the uncertainty of the study, lack of local B values and the limited amount of reference for B value of A. pintoi. Besides, the different sampling time of N samples and biomass, as well as the limited amount of reference for B value of A. pintoi lead to the decrease of accuracy. Nevertheless, due to the consistent findings obtained from the two B values, this study offers an indication of the N supply to the agroforestry system by these different understorey legumes