Crops that require less water but produce the same yield would aid agriculture. We show that engineering lower stomatal density in wheat leaves can improve water-use efficiency, yet maintain yield.

(K.S.); 0000-0002-6438-9188 (R.S.); 0000-0002-9703-0745 (A.J.F.); 0000-0001-9972-5156 (J.E.G.).Stomata are formed by a pair of guard cells which have thickened, elastic cell walls to withstand the large increases in turgor pressure that have to be generated to open the pore that they surround. We have characterized FOCL1, a guard cell-expressed, secreted protein with homology to Hyp-rich cell wall proteins. FOCL1-GFP localizes to the guard cell outer cuticular ledge and plants lacking FOCL1 produce stomata without a cuticular ledge. Instead the majority of stomatal pores are entirely covered over by a continuous fusion of the cuticle, and consequently plants have decreased levels of transpiration and display drought tolerance. The focl1 guard cells are larger and less able to reduce the aperture of their stomatal pore in response to closure signals suggesting that the flexibility of guard cell walls is impaired. FOCL1 is also expressed in lateral root initials where it aids lateral root emergence. We propose that FOCL1 acts in these highly specialized cells of the stomata and root to impart cell wall strength at high turgor and/or to facilitate interactions between the cell wall and the cuticle.

Genetically identical plants growing in the same conditions can display heterogeneous phenotypes. Here we use Arabidopsis seed germination time as a model system to examine phenotypic variability and its underlying mechanisms. We show extensive variation in seed germination time variability between Arabidopsis accessions and use a multiparent recombinant inbred population to identify two genetic loci involved in this trait. Both loci include genes implicated in modulating abscisic acid (ABA) sensitivity. Mutually antagonistic regulation between ABA, which represses germination, and gibberellic acid (GA), which promotes germination, underlies the decision to germinate and can act as a bistable switch. A simple stochastic model of the ABA-GA network shows that modulating ABA sensitivity can generate the range of germination time distributions we observe experimentally. We validate the model by testing its predictions on the effects of exogenous hormone addition. Our work provides a foundation for understanding the mechanism and functional role of phenotypic variability in germination time.

No abstract