Search

All photosynthetic organisms balance CO2 assimilation with growth and carbon storage. Stored carbon is used for growth at night and when demand exceeds assimilation. Gaining a mechanistic understanding of carbon partitioning between storage and growth in trees is important for biological studies and for estimating the potential of terrestrial photosynthesis to sequester anthropogenic CO2 emissions.(1,2) Starch represents the main carbon storage in plants.(3,4) To examine the carbon storage mechanism and role of starch during tree growth, we generated and characterized low-starch hybrid aspen (Populus tremula x tremuloides) trees using CRISPR-Cas9-mediated gene editing of two PHOSPHOGLUCOMUTASE (PGM) genes coding for plastidial PGM isoforms essential for starch biosynthesis. We demonstrate that starch deficiency does not reduce tree growth even in short days, showing that starch is not a critical carbon reserve during diel growth of aspen. The low-starch trees assimilated up to similar to 30% less CO2 compared to the wild type under a range of irradiance levels, but this did not reduce growth or wood density. This implies that aspen growth is not limited by carbon assimilation under benign growth conditions. Moreover, the timing of bud set and bud flush in the low-starch trees was not altered, implying that starch reserves are not critical for the seasonal growth-dormancy cycle. The findings are consistent with a passive starch storage mechanism that contrasts with the annual Arabidopsis and indicate that the capacity of the aspen to absorb CO2 is limited by the rate of sink tissue growth.

Norway spruce (Picea abies) is an important boreal forest tree species of significant ecological and economic importance. Hence there is a strong imperative to dissect the genetics controlling important wood quality traits in Norway spruce. We performed a functional genome-wide association mapping of 17 wood quality traits in Norway spruce using 178101 single-nucleotide polymorphisms (SNPs) generated from exome genotyping of 517 mother trees. The wood quality traits were defined using functional modelling of wood properties across annual growth rings. Association mapping was performed using a multilocus LASSO penalized regression method and we detected a total of 51 significant SNPs from 39 candidate genes that are involved in wood formation. Our study represents the first functional multi-locus genome-wide association mapping (AM) in Norway spruce. The results advance our understanding of the genetics influencing wood traits, identify novel candidate genes for further functional studies and support current Norway spruce breeding efforts.