A predictive model for weed biomass in annual intercropping

Chunfeng Gu*, Wopke van der Werf, Lammert Bastiaans

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Intercropping has frequently been reported to provide good weed suppression. Here, we tested the hypothesis that the weed biomass in annual intercropping systems can be adequately forecasted based on the weed biomass obtained in pure stands of the component crop species. The aim of this analysis was to enlarge the understanding on the weed suppressive ability of intercrops, and specifically to shed light on the relevance of the mechanisms at the basis of weed suppression in intercrops and the factors governing this weed suppression. Based on a literature survey, a dataset covering 76 experiments, a total of 35 different crop species combinations and a total of 339 records was composed, with each record containing weed biomass of the intercrop as well as that of pure stands of the component species. Three models were evaluated and compared using the dataset. The first two models, the arithmetic mean (model 1) and the weighted arithmetic mean (model 2a and 2b) of the weed biomasses obtained in pure stand, resulted in a systematic overestimation of weed biomass in intercrops. This result confirms that the ability of intercrops to suppress weeds is in general well developed. A third model was constructed based on an extended version of the hyperbolic yield-density equation. Mathematical elaboration of this equation suggests the weed biomass in intercrops to be equal to the weighted harmonic mean of weed biomasses in pure stands (model 3), whereby weighting is based on the relative densities of the component species in intercrops. Comparison between observed and predicted data showed that the model accurately predicted weed biomass of simultaneous intercrops in mixed and row design. The harmonic mean indicates that weed biomass in intercrops is the outcome of the joint competitive effect of the component species, whereby the more strongly weed suppressive species contributes a more than proportional share. Such dominance of one of the species is generally referred to as selection. For intercrops with a less intimate entanglement of the two component species, either due to temporal (relay intercrops) or spatial (strip intercrops) separation, the harmonic mean tended to underestimate observed weed biomass. A contribution of complementarity, following from niche differentiation or facilitation between component species, to the weed suppression of intercrops is not accounted for in the harmonic mean model. The high accuracy of the predictions thus suggests that the density and the selection effect are the main mechanisms responsible for weed suppression in intercrops. Choice of component species, their mixing ratio and the total plant density of the intercrop were all shown to clearly influence the weed suppressive ability of the intercrop. With a prominent role of the selection effect, care should be taken that an overemphasis on weed suppressive ability in the design of intercropping systems should not result in the poorer weed suppressive species being outcompeted. Another important implication of the finding that weed biomass in intercrops can be accurately predicted based on weed biomass in pure stands of the component species is that variety selection for intercrops with improved weed suppression can simply be confined to variety selection in pure stands.

Original languageEnglish
Article number108388
JournalField Crops Research
Volume277
DOIs
Publication statusPublished - 1 Mar 2022

Keywords

  • Complementarity
  • Hyperbolic equation
  • Intercropping
  • Plant competition
  • Selection
  • Weed

Fingerprint

Dive into the research topics of 'A predictive model for weed biomass in annual intercropping'. Together they form a unique fingerprint.

Cite this