Contributions of Fixed N by Legumes

The ability of legumes to progressively improve the N status of soils has been utilized for thousands of years in crop rotations and traditional farming systems. The 163 million ha of legume crops grown each year, legume components of the 200 million ha under temporary pastures or forage crops, the 10-12 million ha of perennial legume cover-crops in rubber and oil-palm plantations, and the legume trees and shrubs in agroforestry systems all contribute fixed N to agriculture. Collectively experimental and on-farm data both suggest potential inputs of several hundreds of kg of fixed N/ha per year (Table 1), often with maximum rates of N2 fixation of 3-4 kg shoot N/ha per day (Unkovich, Pate 2000). The amounts of N2 fixed with most legumes are regulated by environmental or management constraints to plant growth associated with soil nutrients, water supply, diseases and pests. Legumes commonly fix around 20-25 kg of shoot N for every tonne of shoot dry matter accumulated across a range of environments (Figure 1) unless their capacity to fix N is restricted by local practices which either limit the presence of effective rhizobia (no inoculation, poor inoculant quality), or directly affect soil N fertility (excessive tillage, extended fallows, fertilizer N, rotations; Peoples, Herridge 2000). In addition to fixed N in shoots, legume roots may contribute up to 0.7 kg of fixed N for every kg of fixed N accumulated in shoots (Kelner et al. 1997; see also Peoples, Herridge 2000).

Table 1. Estimates of the annual amounts of N2 fixed in different agricultural systems21.

N2-fixing organism System

Range measured (kg shoot N/ha)

Commonly observed (kg shoot N/ha)

Free-living Associative



Tropical grasses



Actinorrhizal trees Green manure legumes Forage legumes Crop legumes Tree legumes










0-15 10-20 25-65 10-50 30-75 50-150 50-250 30-200 50-300

a Adapted from Ledgard and Giller (1995) with additional information from Peoples and Craswell (1992); Peoples et al. (1996); Unkovich, Pate (2000); Maskey et al. (2001); and Peoples et al. (2001).

0 0

Post a comment