The authors grew the maize IBM (Mo17 xB73) mapping population in six different field sites in the US and South Africa over several years and then used kernel element concentrations and mathematical combinations (Principal Component Analysis) of them as phenotypes to identify QTLs controlling the kernel ionome.
With data coming from six different field sites the authors try different strategies to analyze QTL environment interactions (QEI) and find that the environment had a great effect on element kernel concentrations although high heritability values are found for some elements in plants growing in the same field.
The authors found that several single element QTLs overlapped with PCA QTLs highlighting the idea that element concentrations behave on a concerted manner.
Overall, this is a great piece of work that provides a very complete view of the genetic architecture of the maize kernel ionome and should help in other efforts to understand how we can manipulate nutrient and toxic elements in maize.
Element kernel concentrations are the result of several processes including root acquisition, xylem loading, phloem transport and kernel loading. Identifying which steps are more important to increase the concentration of nutritionally valuable elements or to reduce it from toxic ones in edible parts of the plants is crucial to develop more nutritious and safer to eat foods.
Given the current information available on gene expression in different tissues at different developmental stages (e.g Stelpflug et al 2015) I wonder if the authors, based on the their QTL analysis, can identify key tissues or developmental stages determining kernel element concentrations in maize.
some minor comments:
- fig 1, 3, 8. Figures indicating QTLs found in different environments across chromosomes. Dashes crossing vertical lines separating different chromosomes are confusing. Is difficult to distinguish if the QTL is at the end of one chromosome or at beginning of the next.
- l.98 Ukulima?
- l.250: reference needed.
- l.531: Specify what lines were grown in each site. Is there a set of lines that was grown in all the sites? What do the authors consider a replicate? Pooled ears from a row from which a kernel was randomly sampled? Several kernels from pooled ears? Several kernels from individual ears per row?