Comparative assessment of einkorn and emmer wheat phenomes: II—phenotypic integration

Phenotypic integration, as the outcome of the number and strength of (co)variation among phenotypic traits in the genetically-related diploid einkorn (Triticum monococcum L. subsp. monococcum; genome AA) and tetraploid emmer wheat [Triticum turgidum subsp. dicoccon (Schrank) Thell.; genome BBAA] was estimated for seven modules (i.e., area, density, dimensions, dry weight, ecophysiology, structure, and yield components) based on 110 traits measured or estimated on plant, tiller, leaf, spike, spikelet and kernel samples at three growth stages during four growing seasons. Classical methods of phenotypic integration assessment using algorithms for dimensionality reduction, matrix correlation, ordination, discrimination, multidimensional scaling and functional relationships, generated insightful but inconsistence estimates of phenotypic integration; thus, making it challenging to compare the strength of integration across species and modules using a single and reliable phenotypic integration index. Divergence between einkorn and emmer due to polyploidy, although evident at the mean phenotypic index at whole plant phenotypic level (0.37 ± 0.08 and 0.59 ± 0.09, respectively), was manifested more at the reproductive than at the vegetative level; while at a multidimensional scale, emmer exhibited larger correct classification (86.0%) than einkorn (69.5%). A standardized z-score, which was based on partial least squares analysis of trait variation, when adjusted for plant size, provided unbiased estimates of phenotypic integration indices comparable across modules and species (minimum z-score of 16.9 ± 1.1 for density module in einkorn to 32.1 ± 2.0, for yield components module in emmer). Despite the polyploidy diversity bottleneck in emmer, both species overlap in the range of their genetic variation for many traits and seemingly share some phenotypic intermediate forms. Phenotypic variances are unequal among einkorn and emmer; their differences could be attributed to emmer’s polyploidy diversity bottleneck and manifested at the different numbers of significant common principal components shared between the species, with a range from three (area module) to nine (dimensions module) shared common components. Future research needs to explore how inter- and intraspecific phenotypic variation affect population dynamics and performance under field conditions. As potential alternative crops in future cropping systems, considerable opportunities and benefits, in influencing the structure and function of agroecosystems, are expected if intraspecific trait databases are developed for these early domesticates.