Agriculture Reference
In-Depth Information
provide useful indications of specific and general mixing ability for predicting
performance of more complex mixtures of wheat (Knott and Mundt, 1990; Lopez
and Mundt, 2000; Mille
et al
., 2005), barley (Gacek
et al.,
1996a,b) and potato
(Phillips, 2004; Phillips and Wolfe, unpublished) in terms of both yield and disease
restriction.
A different approach was described by Suneson (1956) as evolutionary plant
breeding, focusing on composite cross populations of barley. Research on these
populations and others was reviewed by Phillips and Wolfe (2005) as a background
to the development of composite cross populations of wheat in the UK. One
objective in this programme is simply to develop a genetic resource for breeders.
A more speculative objective, however, is to allow population samples to undergo
natural selection under different environmental conditions. One question is whether
or not this will provide a rapid means of selecting segregating lines that 'nick' well
together to improve yield and yield stability under organic conditions.
10.3.6 Diversity and stability
Yield stability is considered to be one of the main advantages of cultivar mixtures
(Wolfe and Barrett, 1980). In practice, the definition of
yield stability
is
high
yield
over a range of environments (Eberhart and Russell, 1966) and this accounts for the
difficulty of finding an appropriate form of statistical analysis for stability (Crossa,
1988; Dubin and Wolfe, 1994). From the six analyses considered by Dubin and
Wolfe, the Westcott geometric analysis (Westcott, 1987) in combination with
regression analysis (Eberhart and Russell, 1966, modified by Mundt
et al.
, 1995)
gave a consistent and clear picture of the comparative performance of mixtures and
their components (Finckh
et al.,
2000).
Since different cultivars and species tolerate different ranges of environmental
variation, it is logical to expect that a mixture of cultivars or species will be more
stable in yield than any of the components. For example, the overall better and more
stable performance of soybean (Schutz and Brim, 1971) and wheat (Finckh and
Mundt, 1996) cultivar mixtures when compared with the mean of their pure stands
in different locations, was due to variable interactions over environments. Within
large data sets, the variance of the mixture is usually less than that of most if not all
of the components (e.g. Table 10.5; Allard, 1960; Dubin and Wolfe, 1994; Finckh
et al.
, 2000).
Such gains are not, of course, automatic. The environment plays a crucial role
together with the appropriate choice of system and variety. The choice of system
depends on being able to minimise competition and maximise facilitation. An
example of facilitation is the benefit often found from intercropping grass species
with legume species where the latter may increase the nitrogen available to the
former.
These considerations relating to yield are relevant also to stability of quality. For
example, Baumer (1983) in Bavaria found that the malting quality of barley variety
mixtures grown in different environments was more uniform than that of the
components grown in the same environments. Baking quality parameters (falling
