Agriculture Reference
In-Depth Information
8.2
Trial demonstrations of mixtures
Focusing on yield response, much experimental work has been carried out with cereals,
mainly wheat and barley. In spring barley, the level of disease control or yield increase
demonstrated has often been only moderate (Newton & Thomas, 1992; 1993; Newton
et al
., 2002; Mercer
et al
., 2006), but in winter barley yield increases of up to 17% have
been achieved on both small (Day, 1984; Newton
et al.
, 1998) and large plot scales
(Newton & Guy, 2008). The difference between the spring and winter crops of barley
may be attributable to lower disease levels and the shorter and less stressed season in
the former case. It may also refl ect the morphological similarities of the recent recom-
mended cultivars in the UK where many of the reported trials were grown and where
fewer complementary traits were available to interact benefi cially. More complementary
traits may have been present in elite germplasm some years ago as spring barley trials
carried out between 1978 and 1982 gave yield increases up to 16% and a mean of 5.69%
(Day, 1984). In contrast, Paynter & Hills (2007) generally found no overall yield increases
or disease control using current Australian winter barley germplasm. Using non-elite
germplasm such as doubled-haploid lines from a mapping population segregating for two
dwarfi ng genes grown in mixtures, yield enhancement of even two-component mixtures
of lines differing in dwarfi ng genes was 6.6% and 9.9% for three components (Newton
et al
., 2004).
The strategy of complementing cultivars with contrasting traits was demonstrated
using Canadian barley cultivars differing in height and maturity date, illustrating not only
the potential for yield enhancement with appropriate combinations, but also some which
can lead to negative interactions (Essah & Stoskopf, 2002), and similar contrasts have
been demonstrated in current wheat and barley mixture trials in Scotland (Newton AC &
Hoad S, unpublished data) and wheat mixtures in Canada (Knott & Mundt, 1990).
Combining varieties with contrasting traits does not result in the problems one might
expect due to trait convergence. Convergence of many characters has been reported
anecdotally, but there is also experimental data (Newton
et al
., 2002). Using mixtures
in which the components were physically distinct, it was possible to show some conver-
gence for malting quality attributes, as well as agronomic characters (Swanston
et al
.,
2005, 2006), but it has not been demonstrated whether a similar mechanism operates in
mixture components with similar morphology and genetic background.
Finckh & Mundt (1992) clearly illustrated why genetically heterogeneous populations
should be treated in a holistic way, as in their wheat experiments between 52% and 58%
of the yield variation was attributable to disease in monocultures whereas in mixtures
this dropped to between 10% and 31%, illustrating some of the potential different plant-
plant interaction effects. Day (1984) recorded powdery mildew reduction of around 35%
unrelated to growth stage or absolute level of infection, and Newton
et al
. (2002) found
that high inoculum pressure reduced mixtures effi cacy, all factors which confound the
causal relationships between yield and stresses.
The mechanism whereby mixtures achieve disease control can be attributed to dilution
of the density of susceptible plants, introduction of resistant plants as physical barriers
limiting pathogen spread, and induced resistance from enhanced frequency of avirulence
factors on adjacent susceptible plants (Chin & Wolfe, 1984). In the absence of expo-
sure to multiple hosts, races of pathogens tend to simple virulence patterns suggesting
