Genetic architecture of species level differences in Begonia section Gireoudia

Abstract

Begonia is one of the ten largest plant genera and is found throughout the tropics. I have used Begonia section Gireoudia to study the genetics underlying vegetative diversity in tropical herbaceous plants. Section Gireoudia is a large Central American group. The section is remarkably diverse in morphology and habitat preference. It ranges from wet rainforests to seasonally dry forests. I have investigated variation in morphological, anatomical and ecophysiological differences for 21 species in Begonia section Gireoudia. Based on the observed variation, species in Begonia section Gireoudia form a complex and unique group that stands out from currently analysed taxa in the global scale of variation on the basis of leaf function and resource use strategy traits as well as their peculiar leaf anatomy. Traits directly related to leaf function such as photosynthesis and stomatal conductance has very low values which overlap with those of CAM and aquatic plants. Values for traits indicative of resource use such as leaf mass area (LMA) and leaf dry matter content (LDMC) are also very low in Begonia when compared with the values observed globally. The trait- trait correlations across the species in section Gireoudia were also investigated and revealed patterns in micromorphology and ecophysiology.

Some of the traits measured are correlated with each other in apparently straightforward, well charaterised biological relationships e.g., the variation among Begonia species in stomatal conductance and net assimilation rate are positively correlated. On the other hand, the linkage of high Amass with high Nmass which is in large part the result of a direct causal relationship, has been observed at the global scale but this relationship is not significant in Begonia section Gireoudia.

I examined B. plebeja and B. conchifolia, two very closely related though ecologically divergent species from Meso-America, in more detail. I detected significant differences between the species for a number of phenotypic variables which may be related to their habitat preferences. This suggested that environmental conditions have driven divergent evolution of phenotypic traits for these two species.

Using a mapping population generated from hybrids between these two species I was able to examine the genetic basis of these differences. This revealed that although some traits (such as anthocyanin accumulation) appear to be under simple genetic control, most of the variation between species has complex genetic inheritance patterns. I used QTL analysis to identify significant QTLs for 20 physiological, anatomical and morphological traits which varied between these two species. Leaf shape traits appear to be largely influenced by a few loci of large effect, making these good potential targets for further analysis. The study also identified clusters of coincident QTLs for different correlated traits identifying pleiotropic genes or suites of linked loci.