Computational identification of genes modulating stem height-diameter allometry

Abstract

The developmental variation in stem height with respect to stem diameter is related to a broad range of ecological and evolutionary phenomena in trees, but the underlying genetic basis of this variation remains elusive. We implement a dynamic statistical model, functional mapping, to formulate a general procedure for the computational identification of quantitative trait loci (QTLs) that control stem height-diameter allometry during development. Functional mapping integrates the biological principles underlying trait formation and development into the association analysis of DNA genotype and endpoint phenotype, thus providing an incentive for understanding the mechanistic interplay between genes and development. Built on the basic tenet of functional mapping, we explore two core ecological scenarios of how stem height and stem diameter covary in response to environmental stimuli: (i) trees pioneer sunlit space by allocating more growth to stem height than diameter and (ii) trees maintain their competitive advantage through an inverse pattern. The model is equipped to characterize 'pioneering' QTLs (piQTLs) and 'maintaining' QTLs (miQTLs) which modulate these two ecological scenarios, respectively. In a practical application to a mapping population of full-sib hybrids derived from two Populus species, the model has well proven its versatility by identifying several piQTLs that promote height growth at a cost of diameter growth and several miQTLs that benefit radial growth at a cost of height growth. Judicious application of functional mapping may lead to improved strategies for studying the genetic control of the formation mechanisms underlying trade-offs among quantities of assimilates allocated to different growth parts.

Keywords: functional mapping; height-diameter allometry; mathematical equation; quantitative trait loci.

Figure 1

Height–diameter developmental allometry for two parents, I‐69 (green line) and I‐45 (red line), and their progeny (grey line). (a) The allometric changes in stem height as a function of stem diameter. (b) The allometric change in stem diameter as a function of stem height.

Figure 2

Manhattan plots of −log(p) values over the Populus genome. (a) pi QTLs for the allometric scaling of height with diameter. (b) mi QTLs for the allometric scaling of diameter with height. The red horizontal line is the genomewide critical threshold at the 1% significance level determined through Bonferroni correction.

Figure 3

Temporal patterns of additive and dominant genetic effects by significant pi QTLs (a) and mi QTLs (b).

Figure 4

Developmental allometry of how stem height scales with stem diameter, modulated by pi QTL 5/4226091 with three different genotypes CC, CT and TT. Red lines show the differences of stem height growth among three genotypes at five representative ages.

Figure 5

Developmental allometry of how stem diameter scales with stem height, modulated by mi QTL 5/4220981 with three different genotypes TT, TC and CC. Genotype CC is much thicker than the other two (TT, TC) per the same amount of stem height growth.