2.2 Evaluation of the external leaf morphology
Five-to-six fully expanded leaves were scanned (along with a scale) using a CANON CanoScan LiDE 400 scanner, and leaf areas calculated using Image J1.51d software (National Institutes of Health, Bethesda, MD, USA). These same leaves were then used to photograph the details of both adaxial and abaxial surfaces with a Zeiss Discovery v12 Dissecting microscope (objective 0.63x PanApo). Adaxially, the idioblast tips were easily visualized by their translucency, and, abaxially, the peltate hairs were counted using their central parts as references. Both idioblasts and peltate hairs were counted in 1 mm2areas at five positions on each leaf surface. To count the number of stomata, five leaves were fixed in FAA (formalin:acetic acid, Johansen 1940), washed three times in distilled water for one hour each, stained with Feulgen reaction (modified from Barell & Grossniklaus, 2005). Then, the leaves were cleared with a solution containing ethanol:benzoyl benzoate 3:1 (v/v) for six hours, ethanol:benzoyl benzoate 1:3 (v/v), and finally benzoyl benzoate: dibutyl pftalate 1:1 (v/v) for several days (Crane & Carman, 1987). Following removal of the peltate hairs, 1 mm2 samples were imaged with a Zeiss LSM700 Confocal Microscope connected to an AxioCam 512 camera and Zen Blue 2.3 software, using a 20x/0.8 M27 Plan-Apochromat objective. Note that the area occupied by the rounded base of the peltate hairs (3.45%) never contained stomata, and hence was subtracted from the total leaf area to get a more accurate estimate of stomatal density.