Page 151 - Breeding and regulatory opportunities, Renaud

Broccoli crop improvement

selection directly in an organic environment could result in more rapid genetic

gain than indirect selection in a conventional environment, but because most

repeatibilities were close to 1.0, indirect selection in a conventional environment

would be nearly as efective as direct selection in an organic system. Also in our

trials we found larger genetic variances (broader minimum-maximum ranges)

compared to results under conventional management.

4.4.4 Breeding broccoli for organic systems

Determining whether broccoli cultivar development could better take place

under organic or conventional management systems to develop cultivars

optimized for organic agriculture is a complex proposition. Breeding in the

target environment is most efective for organic systems, where G × E interaction,

genetic diversity, and trait heritability are all taken into account (see e.g. Wolfe et

al., 2008). Driven by the need for eiciency, commercial broccoli breeders often

aim to reduce G × E interactions by selecting cultivars that are broadly adapted

to the range of their target environments. However, from our data location and

season and their interactions were the primary sources of variation identiied

for broccoli head weight and the other horticultural traits studied. This is

supported by our observations that the general location- and season-speciic

trend for head weight interacted with the cultivar’s maturity class designation,

where mid-to-late season cultivars were the highest ranking in Oregon in the

Fall, while in Maine early-to-mid season cultivars were the highest ranking. In

the Spring, best performing cultivars in both Maine and Oregon were in the

mid-to-late season maturing class. When comparing cultivar performance

between seasons and locations, we observed that the best performing early-

to-mid season cultivars in Spring trials and the mid-to-late season cultivars in

Fall trials for Oregon were a diferent group of cultivars than those in Maine of

the same maturity class.

Greater heterogeneity in organic management systems and G × M crossover

interaction observed on a local scale supports the idea that direct selection

(under organic management) of cultivars for organic agriculture would beneit

from evaluation in organic systems, particularly if the intent of the breeder is

to develop cultivars that support local adaptation. Annicchiarico et al. (2012)

found that the performance of lucerne (Medicago sativa) populations bred

in the location of intended use were better performers on organic farms in

133