Coarse- versus fine-grained water stress in Arabidopsis thaliana (Brassicaceae).

The effects of coarse vs. fine-grained variation in water regime on 13 accessions of Arabidopsis thaliana were compared and it was found that the scale of environmental variation can lead to differences in which traits are affected and in the magnitude and direction of effect. Coarse-grained (between individual) variation was studied by comparing plants that received high or low amounts of water on a daily basis. Fine-grained (within individual) variation was distributed across the coarse-grained treatments; plants in the fine-grained treatment received the same amount of water per week as the corresponding high and low coarse-grained treatments; however, in the fine-grained treatments, plants received their entire weekly water allotment over 2 days. The number of leaves produced, days to bolting, length of bolt, number of basal and lateral branches, and survival to 30 days after bolting were measured and analyzed by ANOVA. Plants grown in high water generally produced more leaves, taller main axes, more branches and had increased survival, relative to those in low water. Bolting date showed a significant genotype specific effect of water level. Plants grown under fine-grained variation in watering had taller main axes relative to constant water. There were significant interactions between both types of variation for bolting date, main axis height, and survival. Principal components for all traits were loaded such that the components themselves were also independently affected by genotype, water level, variability of water level (grain), and level-by-grain interactions in ways that were largely predicted by their component traits. Overall, the effects of fine-grained variation were subtler than those of coarse-grained variation: fewer traits responded, and the magnitude of the responses was smaller. Nevertheless, the responses to fine-grained variation are distinct from those of coarse-gained variation, differing with respect to which traits respond and in the direction of response, and must therefore be treated as independent sources of environmental variation.