DINITROGEN FIXATION IN PEA - CONTROLS AT THE LANDSCAPE-SCALE AND MICROSCALE

Topography influences the distribution of water and N, thereby exertin g an indirect control on N-2-fixing activity of legumes. A study was c onducted to assess the (i) variability of N-2 fixation in pea (Pisum s ativum L.) in a rolling field, and (ii) degree of landscape-scale cont rol on N-2 fixation. A 90 by 100 m sampling grid, with 10-m spacings, was established in the field. Each sampling point was classified as ei ther footslope or shoulder. The percentage of N derived from the atmos phere (% Ndfa) was estimated using natural N-15 abundance and the A-va lue approaches. Spring soil water content and inorganic N were most co ncentrated in the footslopes, whereas mean estimates of % Ndfa showed inverted spatial distribution patterns. At flowering, natural N-15 abu ndance estimates of % Ndfa did not differ between shoulders and footsl opes. In contrast, A-value mean estimates of % Ndfa were 84% for shoul ders and 92% for footslopes. The two approaches gave similar mean esti mates of % Ndfa at maturity, with values of approximately 72 and 84% f or footslopes and shoulders. A random spatial pattern existed for tota l aboveground N accumulation, indicating that pea was able to adjust i ts N-2-fixing activity according to the available soil N. Despite simi lar spatial patterns for % Ndfa estimated using the two approaches, th e correlation between the A value and natural N-15 abundance approache s was poor (r = 0.213 at flowering and r = 0.377 at maturity). The poo r correlation suggests that N-2 fixation by pea was partially controll ed at the Landscape scale, whereas strong micro-scale controls may hav e existed that ultimately regulated N-2 fixation.