Rc. Izaurralde et al., CROP AND NITROGEN YIELD IN LEGUME-BASED ROTATIONS PRACTICED WITH ZERO-TILLAGE AND LOW-INPUT METHODS, Agronomy journal, 87(5), 1995, pp. 958-964
Though legumes are beneficial in crop rotations, there is limited info
rmation on how tillage system-crop sequence interactions influence cro
p yield and N production, To see if biomass and N yields in shortterm
legume-based rotations under zero tillage (ZT) and low-input (LI) prod
uction methods can equal those in cereal monocultures under ZT and con
ventional tillage (CT), field experiments were conducted in Alberta, C
anada, from 1989 to 1992 at Ellerslie (Typic Cryoboroll soil) and Bret
on (Typic Cryoboralf). Treatments at each site consisted of (i) two 4-
yr rotations, each with the same crop sequence but different tillage m
ethods, and (ii) four continuous barley treatments in 2 x 2 factorial
combination of tillage and fertilizer N, The crop sequence was barley
(Hordeum vulgare L.)-barley and field pea (Pisum sativum L.) intercrop
-barley-fababean (Vicia faba L.). At Ellerslie, tillage for weed contr
ol and seedbed preparation was either CT-LI or ZT. At Breton, one rota
tion used the LI approach; the second used deep tillage (DT). Weeds on
CT and ZT were controlled with either pre- or postemergence herbicide
s. No herbicides were applied to LI treatments at either site or to th
e DT treatment at Breton, Yields of barley following legumes under ZT
were similar to those of fertilized continuous barley. With nonchemica
l weed-control methods, weed competition reduced yields of barley foll
owing legumes by 24% compared with fertilized continuous barley, The i
ncreased fababean yield measured under DT was associated with improved
rooting conditions and water extraction. The equivalent N-fertilizer
value of legume residues with similar weed-control levels averaged 19
kg ha(-1), Except for the LI system, legume-based rotations produced,
over the 4-yr cycle, amounts of N equivalent to continuous cereal syst
ems. Resource use efficiency of legume-based rotations, as measured by
net-N yields, was equivalent to continuous systems at Breton, but som
ewhat reduced at Ellerslie.