Total average GSL concentration ranged from 3.1 mg g−1 DW (Buzz) to 11.6 mg g−1 DW (SR10). Both of these accessions are E.sativa, indicating the large degree of variability between accessions of this species, both commercial and germplasm. The lowest average accumulation for Diplotaxis was Wild Tirizia with 4.4 mg g−1 DW and the highest was 10.4 mg g−1 DW, (Wild Grazia). For glucosativin both the monomeric and the dimeric

forms were identified and quantified separately and concentrations of both forms varied significantly between check details accessions. On average 91.3% of the total GSL concentration was made up of glucosativin/DMB. This is much higher than the proportions presented in previous studies where values of around 60% have been generally given (Pasini, Verardo, Caboni, & D’Antuono, 2012). Other GSL compounds such as glucoraphanin and glucoerucin were not detected in all accessions. Again, previous studies have highlighted the prevalence of these compounds, but we found them to be relatively minor. Concentrations ranged from nil to 0.9 mg g−1 DW (Wild Grazia)

Wnt mutation for glucoraphanin and nil to 1.6 mg g−1 DW (SR16) for glucoerucin. Several other GSLs were quantified, and in some cases these were as high as the more generally accepted ‘major’ GSLs of rocket in concentration. The other compounds were: 4-hydroxyglucobrassicin, glucotropaeolin, glucolepiidin, glucoiberverin, glucoalyssin, glucoraphenin, diglucothiobeinin and glucoibarin. None of these GSLs discriminated between species. In general, the concentrations detected

were similar to those found in other studies. In some of these, plants were grown in field conditions and therefore subject to many different environmental stresses and inconsistencies. It is widely known that both GSLs and flavonols increase in concentration as Carnitine palmitoyltransferase II plants become stressed (Rohr, Ulrichs, Mucha-Pelzer, & Mewis, 2006). With this in mind it is somewhat unusual that the concentrations reported here were not lower, as stress was minimal in comparison to field conditions. Studies conducted in outdoor conditions are not directly comparable for this reason. Field conditions and climate vary greatly between growing regions and GSL proportions may change due to these variables. Our study represents GSL and flavonol accumulation in rocket varieties and species under conditions that can be easily replicated using controlled environment apparatus. This allows the basic genetic differences in GSL profile to be observed, rather than the differences between how accessions respond to their normal, field-based growing environment. A trial of five gene bank accessions used in this study have been grown under field conditions and will be analysed using identical LC/MS methods to determine the effects the outdoor environment has on GSL and flavonol profiles. Table 3 summarizes the range of concentrations of some GSLs previously reported in comparison with our own data.