Page 186 - Breeding and regulatory opportunities, Renaud

Chapter 5

stable across trial environments; as was the case with neoglucobrassicin, δ- and

γ-tocopherol in our study. For traits where genotype played a more signiicant

role in contributing to variation, cultivars with a higher concentration level

tended to also be those that were most stable across environments as was seen

for lutein and glucoraphanin concentrations.

No signiicant diferences were found for cultivar performance in phytochemical

concentrations between genetic materials originating from two distinct

commercial sources, with the exception of lutein (data not shown). When the

full set of broccoli cultivars were analyzed for a correlation between date of

release and mean level of phytochemical content across trials, no signiicant

correlation was found with the exception of a negative trend for glucobrassicin

(Figure 5.3). Our data does not support the idea that modern breeding for high

yield performance and disease resistance necessarily leads to a trade-of in

level of phytochemicals. Previous reports examining the relationship between

year of release and performance had focussed on wheat vitamin and mineral

content (Murphy et al., 2008); Hussain et al. (2010); Jones et al. (2010), and mineral

content in broccoli (Farnham et al., 2011). However these authors did not study

phytochemical content and their results were equivocal on the question on

an innate biological trade-of between increased yield and nutritional content.

Not many studies have included two or more groups of phytochemicals. In

our study with three phytochemical groups we found that phytochemicals

demonstrating a negative correlation with one another (e.g. glucoraphanin with

the carotenoids), showed an inverse cultivar response: e.g. cultivars with highest

concentrations of glucoraphanin were the lowest in the carotenoids and vice

versa. When both horticultural traits and phytochemicals were analysed for their

phenotypic correlation, head weight was signiicantly and positively correlated

with glucoraphanin and negatively correlated with δ- and α-tocopherol and the

carotenoids. Farnham and Kopsell (2009) explained that negative correlations

may occur as a result of increased biomass accumulation in a certain genotype

that is not accompanied by increased carotenoid production, efectively

lowering the carotenoid concentration in the immature broccoli lorets when

pigments are expressed. Comparatively, head color was highly correlated to the

carotenoids and negatively correlated to the glucosinolates overall. The cultivar

168