*2.5. Bulk Wax Determination*

Leaf lamina and petiole wax were extracted and quantified according to the methods used on pea [19]. Details of the method used for bulk wax extraction and quantification is exactly as presented by Tafesse et al. [29]. The method can be summarized as follows: first, bulk was extracted from the leaf surfaces by dipping the leaf tissue (lamina or petiole) samples into 10 mL chloroform for 15 s in 100 mL glass tubes. To evaporate the chloroform, the tubes were then placed in a water bath at 70 ◦C for 30 min. Then 5 mL reagent (acidic potassium dichromate) was added to each tube containing the wax, and boiled at 100 ◦C for half an hour. After cooling, 5 mL distilled water was added to each tube, vortexed, and spectral absorbance was measured at 590 nm with a spectrophotometer. Finally, wax concentrations were calculated from the spectral data using a standard curve equation that was developed from a linear (R<sup>2</sup> > 0.98) relationship of known concentrations of beeswax [29].

#### *2.6. Spectral Reflectance and Vegetation Indices*

Spectral reflectance measurements on stipules were taken on three to five occasions per plot for each of the six environments during the crop reproductive phase using a portable spectroradiometer (Model PSR-1100F, Spectral Evolution Inc., Lawrence, MA, USA). This instrument enabled hyperspectral readings with a range of 320–1126 nm and

1.6 nm sampling interval, and a total of 512 discrete narrow bands. A 1-m fiber-optic cable with industry-standard interface with the instrument, controlled by a PSR-1100 Pistol Grip, enabled us to specifically capture reflectance from stipules for spectral measurements. A stipule of a fully expanded leaf at the second or third node counting from the tip of the pea main stem, fully exposed to the sun, was measured on sunny and usually hot days around solar noon (between 11:00 and 14:00 h) from the same direction, avoiding shadows, cloud, and any other interference we could control. Before measurements, reflectance was taken on a white plate that provided maximum reflection, and leaf reflectance was measured by holding the fiber sensor within 3 cm from the stipule surface approximately within a viewing angle of 80–90◦. The reference reflectance was repeatedly taken every 15 min (equivalent to once every 12 plots) to adjust for the changing irradiance from the sun, and more frequently if clouds stopped measurements.

Vegetation and pigment indices, including normalized difference vegetation index (NDVI), green normalized difference vegetation index (GNDVI), photochemical reflectance index (PRI), normalized pigment chlorophyll ratio index (NPCI), and water band index (WBI), were each calculated according to Rouse et al. [30], Gitelson et al. [31], Gamon et al. [32], Peñuelas et al. [33], and Peñuelas et al. [34], respectively, as follows:

$$\text{NIDVI} = (\text{R}\_{\text{rirr}} - \text{R}\_{\text{r}}) \div (\text{R}\_{\text{rirr}} + \text{R}\_{\text{r}}) \tag{6}$$

$$\text{GNDVI} = (\text{R}\_{\text{rir}} - \text{R}\_{\text{g}}) \div (\text{R}\_{\text{rir}} + \text{R}\_{\text{g}}) \tag{7}$$

$$\text{PRI} = (\text{R}\_{531} - \text{R}\_{570}) \div (\text{R}\_{531} + \text{R}\_{570}) \tag{8}$$

$$\text{NPCI} = (\text{R}\_{531} - \text{R}\_{570}) \div (\text{R}\_{531} + \text{R}\_{570}) \tag{9}$$

$$\text{WBI} = \text{R}\_{900} \div \text{R}\_{970} \tag{10}$$

where; R, reflectance; nir, near infrared band (bandwidth 760–860, center band 820 nm), r, red band (bandwidth 650–700 nm, center band 675 nm); g, green (bandwidth 530–580, center band 555 nm). The center bands were rounded to the nearest whole number (for example 530.5 nm was 531 nm). For vegetation indices calculated from two or more single bands such as WBI, the nearest whole number band was used as the center band.
