This paper examines indicators and methods currently used in intercropping
research. A review of papers in Experimental Agriculture and Journal of Agr
icultural Science (Cambridge) in the years 1990-1999 gave 50 papers on inte
rcropping from which 72 experiments were examined. The objectives of experi
ments, apart from virtually all being concerned with crop yield under inter
cropping, were dominated by interest in management methods affecting interc
ropping benefit, with concerns about the economic analysis of intercropping
benefits and of the sustainability of tested systems and in the mechanisms
underlying intercropping effects. Concerns about stability and sustainabil
ity were not as central as might have been expected. Few experiments were m
ainly concerned with exploring the way in which species interact in mixture
as an explanation of intercropping effects (e.g. through manipulation of t
he timing of planting or examination of root and shoot competition). Close
to half the studies used the additive series design with pairwise, replacem
ent series and response model designs each accounting for about 17% of case
s. Only two studies measured at the level of the individual plant and only
two studies included initial indicators of plant size or other biological s
tarting point. Analysis of the combination of experimental design used, the
measurements taken and the statistical analysis performed showed that whil
e most experiments could provide some valid indicators and thus inferences
on crop yields few of them could be used to address questions on the inter-
and intraspecific interactions leading to the intercropping outcomes. The
duration of experiments was 1 year for about half of experiments involving
annual species only and about one-third of experiments involving at least o
ne perennial species. The overall conclusion is that considerable care need
s to taken to ensure that the indicators and experimental methods used are
appropriate to the development and testing of sustainable systems based on
intercropping. (C) 2001 Elsevier Science B.V. All rights reserved.