Stomatal bioassay experiments were performed as described (Pei et al., 1997). Abaxial epidermis strips were loaded onto glass slides with 100 μL of buffer A (50 mM KCl and 10 mM Mes, pH 6.1). After staining with 0.2% toluidine blue O for 20 s, the sample was rinsed twice with distilled water and imaged using a compound microscope (Leica, Wetzlar, Germany). For each sample, the percentage of stomata that were open (defined as having a width >1 μm) was determined from at least 400 stomata. The width and length of the opening (i.e., aperture) of only those stomata that were open were measured and used to calculate the average stomatal aperture (width/length). The width and length of at least 30 stomatal pores were measured. The width and length of open stomatal apertures also were used to calculate the stomatal aperture area [π × (width/2) × (length/2)], which together with the percentage of stomata that remained open, were used to calculate the total open stomatal area per unit leaf area containing 100 stomata [i.e., (average stomatal aperture area) × (percentage of stomata that remained open) × 100].

Induction of stomatal closure by ABA and H₂O₂ was investigated using epidermal strips that were first incubated in CO₂-free buffer A for 2 h at 22 to 25°C under a photon flux density of 200 μmol m⁻² s⁻) to promote stomatal opening. ABA (50 μM final concentration, dissolved in 95% ethanol, with equal volume of ethanol used as a control) or H₂O₂ (1 mM final concentration) was added to the buffer. After the indicated time, samples were stained with toluidine blue O for image recording or loaded with fluorescence dye to determine the production of H₂O₂.