Page 148 - Breeding and regulatory opportunities, Renaud
P. 148
Chapter 4
that not only yield as such but also yield stability across years and seasons are
important breeding objectives for organic conditions. ‘Batavia’, ‘Belstar’ and
‘Green Magic’ had the highest combined head weight and head weight stability
in both management systems, while ‘Arcadia’ was one of the top performing
cultivars in organic, but not in conventional trials. Not all cultivars that performed
well in head weight were stable, such as ‘Maximo’. These examples demon-
strate that some cultivars may be more tolerant to abiotic and biotic stress than
others, and therefore more suitable for organic management systems. A strong
positive correlation of top performing cultivars between management systems
was also found by Burger et al. (2008) for maize who recommended as a result of
these indings that cultivar performance under conventional conditions could
provide a good prediction for the average cultivar performance under organic
conditions in a breeding program. They also recommended that the use of
organic test sites would increase the chances of identifying broadly adapted
genotypes when aiming at cultivars for both systems.
To further examine the question of whether diferences in ranking at
the individual paired conventional and organic sites were signiicant, we
performed Spearman’s rank correlation on cultivar performance between
paired conventional and organic environments. Correlation coeicients were
large and statistically signiicant as would be expected when mean genotype
ranking was similar between management systems (data not shown). However,
when correlation was performed on Fs only (leaving out the inherently lower
1
yielding OPs and inbreds), signiicant correlation was observed in the trial
combinations for Maine Spring 2008, and Oregon Spring 2007 and 2008, but
not the other ive trial combinations (Table 4.6). It is apparent that the signiicant
correlations observed on the full set of cultivars was a function of hybrids always
being higher yielding than OPs and inbreds, but a much weaker association
was revealed within the hybrid sub-group. The weak correlation among
hybrids is in agreement with the crossover interaction that was observed at a
local level between management systems described above (Table 4.7a and b).
Przystalski et al. (2008) analyzed performance of cereals grown under organic
and conventional systems in multiple locations, and determined that despite
high overall genetic correlation for yield and associated traits, there were
exceptions on the individual cultivar ranking level that could be relevant to
the selection process. For example, a cultivar that produced an average yield
130
