*2.9. Colletotrichum higginsianum Infection*

*C. higginsianum* Sacc culture, isolate IMI349061, originating from *Brassica rapa* (CABI Bioscience), was prepared as outlined in Liu et al. (2010) and Liu et al. (2007) [51,52]. Spores were suspended in ddH2O, with a final concentration of 106 spores/mL. Plants were inoculated by applying multiple single 1 μL droplets of inoculant to the oldest leaf, avoiding the mid-vein.

Throughout the course of infection, leaves were maintained on moist filter paper in Petri dishes kept under a 16–8 h/light–dark period (150 ± <sup>10</sup> <sup>μ</sup>mol m−2s−1) during the first 2 and last 2 days of the period of infection, at the third day the plates were transferred to complete darkness. Complete darkness was used to promote germination of the appressorium, while the leaves were transferred back to a day/night light cycle on days 4–5 to prevent chlorosis at approximately 20 ◦C throughout.

Infection was then evaluated every 24 h based on the percent total leaf area infected (Equation (5)). The lesion area was measured using ImageJ (Version 1.53a).

$$\text{Percent leaf area infected} = \left(\frac{\text{Lesion area (cm}^2\text{) at }T(\text{x})}{\text{Total leaf area (cm}^2) \text{ at }T(\text{x})}\right) \times 100\tag{5}$$

Equation (5). Percent leaf area infected by *Botrytis cinerea* and *Colletotrichum higginsianum*, where *T*(*x*) was the time point of interest.

#### *2.10. Botrytis cinerea Infection*

*B. cinerea* grown on potato dextrose agar (PDA) for 7 days at 22 ◦C was used to inoculate four-week-old *A. thaliana* leaves [53]. The leaves were inoculated using a solution containing *B. cinerea* mycelium suspended in ddH2O, with a final concentration of 10<sup>6</sup> spores/mL. Three 1 μL droplets of the mycelium containing solution were applied to one half of the leaf, while three 1 μL droplets of water were applied to the other half. Caution was taken to avoid the leaf mid-vein. Throughout the course of infection (120 h), the leaves were maintained on moist filter paper within sealed Petri dishes and kept the dark at 23 ◦C throughout the course of infection.

The infection was evaluated based on the percent total leaf area infected (Equation (5)).

#### *2.11. Statistical Analysis*

Analyses related to shear force, water loss, protoplasmic streaming, electrolyte leakage, and fungal pathogens were performed using ANOVA tests, with Tukey's tests as a method of post hoc analysis (*p* < 0.05). Two-tailed two-sample *t*-tests (*p* < 0.05) analyzed calcium localization according to differences in RGB pigmentation and differences in the intensity of green pixels following FDA staining.

Rheological results were analyzed using generalized additive models (GAMs). Models were initially fit based on each fixed explanatory variable separately, prior to fitting models at all fixed explanatory variables together. Data were log-transformed. GAMs were generated using the gam function from the "mgcv" package [54,55]. The Akaike information criterion (AIC) score was used when building models to select for the optimum model. Function gam.check was used for model checking. GAMs were analyzed using ANOVA, with F-tests. Figures of each GAM were individually created for each of the pectin solutions using function "plot.gam", in addition to the packages "mgcViz" and "rgl" [56,57]. The package "devtools" was used to download the color palette, and "inauguration\_2021" was used for coloration figures [58].

All statistical analyses were performed with the RStudio statistical software (Version 1.2.5033). In addition to the previous packages mentioned, "ggplot" and "ggplot2" were also used [58].

### **3. Results**

Pectin viscosity significantly (*p* < 0.05) increased under the application of boron alone in both HM pectin and GENU BETA sugar beet (GB) pectin (Figure S1, Table S2). Boron had a more pronounced impact on 8% HM pectin viscosity at 5 ◦C, increasing the viscosity of HM pectin to just over 6000 Pa.s 5 ◦C compared to 8% GB pectin with boron, which had an average viscosity of less than 4000 Pa.s at 5 ◦C (Figure 1). Boron also had a more notable impact on HM pectin at 12 ◦C, with the trend only changing at 20 ◦C, when boron increased GB pectin viscosity to levels greater than HM pectin viscosity (Figure 1). Interestingly, calcium application alone did not have a significant effect on pectin viscosity (Table S2). As outlined in Section 2.2, calcium and boron were added directly to the pectin solution. However, calcium did have a significant effect on pectin viscosity when its influence was considered in combination with increasing pectin concentration and temperature (Table S2). The effect of boron on pectin viscosity remained significant when its effect was considered in combination with increasing pectin concentration and temperature (Table S2).

**Figure 1.** Change in viscosity (Pa.s) as temperature (◦C) increased. Viscosity was analyzed across two concentrations of pectin (4% or 8%), two types of pectin (high methylated citrus pectin or GENU BETA sugar beet pectin) and with either calcium (0.05 M CaCl2), boron (0.05 M H3BO3) or no additional element.

The analysis of the force required to shear *A. fistulosum* sheaths revealed the application of calcium (applied as a soil drench) resulted in a significantly (*p* < 0.05) higher Allo–Kramer shear force (265.0 N g−1), compared to *A. fistulosum* plants that had not received calcium (210.6 N g−1) (Figure 2). This increase in shear force is representative of an increased toughness within the sheaths. Toughness is a mechanical property that describes the ability for a material to resist fracture [59,60]. The toughness of a material is influenced by both its strength and ductility [59,60].

Calcium was spatially localized within *A. fistulosum* epidermal cell walls (Figure 3). Xray microscopy imaging showed that the exogenous application of calcium to *A. fistulosum* increased the concentration of calcium within the cell layer, with the additional calcium primarily localizing to the apoplast (Figure 3). The spatial localization of calcium to the apoplast is evident in Figure 3B, where epidermal layer cell walls can be observed because of increased calcium concentrations in the apoplast (more intense lighter blue color). This calcium appears to be primarily distributed to the radial cell walls. These areas of interest are identified using white arrows. Small flecks of white, light blue and green are also visible across both Figure 3A,B. These, in addition to red flecks, which are more pronounced on Figure 3B, represent calcium contamination and are not related to structures within the cell layer.

**Figure 3.** 2D Calcium map obtained using the 20-ID beamline at the Advanced Photon Source, Lemont, IL, and the X-ray microscopy technique. Calcium concentration is represented by color changes within the map, with darker blue indicating a lower concentration of calcium to red indicating a higher concentration of calcium. White arrows point to sample areas where calcium appears to localize to the apoplast. (**A**) Single *Allium fistulosum* epidermal cell layer obtained from a control plant without calcium treatment. (**B**) Single *Allium fistulosum* epidermal cell layer obtained from a plant treated with a soil drench of 100 mL of a 0.05 M CaCl2 every second day for four weeks.

In addition to increasing the toughness of *A. fistulosum* sheaths, calcium application significantly (*p* < 0.05) reduced percentage water loss, as did increasing pectin concentrations (Figure 4B). An 8 percent high methylated citrus pectin with calcium (proxy for *Allium*) lost the least amount of water after 6 h (48.5%), while 8% GENU BETA pectin with calcium (proxy for *Arabidopsis*) had a water loss of 58.3% after 6 h (Figure 4A). However, the effect of calcium in mitigating dehydration stress in *Allium* species tissue was not significant (*p* > 0.05) (Table S4).

**Figure 4.** (**A**) Percent water loss over 6 h over a 6 h period for the 12 pectin solutions of interest. Amongst the 12 solutions, there are two concentrations (4% or 8%), two pectin types (high methylated (HM) citrus pectin or GENU BETA (GB) (sugar beet) pectin), and the addition of calcium (Ca) (0.05 M CaCl2), boron (B) (0.05 M H3BO3) or no additional element. (**B**) Boxplot of average percent water loss overa6h period for the 12 pectin solutions of interest. Amongst the 12 solutions, there are two concentrations (4% or 8%), two pectin types (high methylated (HM) citrus pectin or GENUA BETA (GB) (sugar beet) pectin), and the addition of calcium (Ca) (0.05 M CaCl2), boron (B) (0.05 M H3BO3) or no additional element. See Tables S5 and S6 for statistical analysis.

The freezing-tolerant *A. fistulosum* was also significantly (*p* < 0.05) more resistant to dehydration stress compared to the freezing-sensitive *A. cepa* (Figure 5B). After 17 h, *A. fistulosum* sheaths had a percentage water loss of 27.1%, while sheaths obtained from *A. cepa* lost 33.1% water (Figure 5A). *A. fistulosum* continued to lose less water as the period of dehydration was extended from 16 h to 18 h, retaining 5.2% more water compared to *A. cepa* (28.1% compared to 33.3%) (Figure 5A).

**Figure 5.** (**A**) Percent water loss over 16–18 h in *Allium fistulosum* and *Allium cepa*. Plants treated with calcium (CA) received 100 mL of a 0.05 M CaCl2 every second day for four weeks. Control plants (NCA) did not receive calcium. (**B**) Boxplot showing average percent water loss over 16–18 h in *Allium fistulosum* and *Allium cepa* sheaths. Each bar represents both control and calcium treated plants combined. See Table S4 for statistical analysis.

In addition, epidermal cell layers from dehydrated and subsequently rehydrated *A. fistulosum* sheaths had a significantly (*p* < 0.05) higher percentage viability based on protoplasmic streaming compared to *A. cepa* (Figure 6B). After 16 h of dehydration, cell layers obtained from *A. fistulosum* had a 57.6% viability compared to 7.1% in *A. cepa* (Figure 6A). The trend of increased viability in *A. fistulosum* continued as the length of dehydration was extended. After 18 h, *A. fistulosum* epidermal cell layers had 39.2% viability while those obtained from *A. cepa* had a percent viability of 2.5% (Figure 6A).

In addition, there was a higher level of FDA-based "greenness" within calcium treated cell layers, indicating an increased level of viability (Figures S2 and S3). Generally, cell layers obtained from *A. fistulosum* also had a significantly (*p* < 0.05) increased level of greenness compared to those obtained from *A. cepa*, indicating greater viability following dehydration stress (Table S8). In addition, differences amongst other relationships were also found to be statistically significant (*p* < 0.05) (Table S8).

Boron also significantly (*p* < 0.05) reduced water loss in pure pectin solutions (Figure 4B). After 6 h, 8% GENU BETA pectin with boron (proxy for *Arabidopsis*) had a 59.7% water loss compared to 63.7% for the GENU BETA pectin control (Figure 4A). While similar dehydration stress tolerance trends were observed in *Arabidopsis* boron transporter mutants, the differences were non-significant (*p* > 0.05) (Figure 7B). Moreover, the over-expression of *PMEI5* did not significantly influence dehydration stress resistance (*p* > 0.05) (Figure 7B).

**Figure 6.** (**A**) Percent protoplasmic streaming in *Allium fistulosum* and *Allium cepa* epidermal cell layers following dehydration over 16–18 h. Plants treated with calcium (CA) received 100 mL of a 0.05 M CaCl2 every second day for four weeks. Control plants (NCA) did not receive calcium. (**B**) Boxplot showing percent protoplasmic streaming for *Allium fistulosum* and *Allium cepa*, following a 16–18 h dehydration period. Each bar represents both control and calcium treated plants combined. See Table S7 for statistical analysis.

However, mutations in boron transporters and the over-expression of *PMEI5* were found to influence the rate of infection for both fungal pathogens (Figure 7). The analysis of the rate of *B. cinerea* infection amongst the genotypes of interest found *bor1* to be the most susceptible to the pathogen, consistently having the greatest percent area of infection from 1 dpi (days post-inoculation) (Figure 7). In general, the percentage of leaf area infected in the *bor1* leaves was significantly (*p* < 0.05) greater compared to all the other genotypes of interest (Figure 7B). At 5 dpi, 100% of the *bor1* leaf tissue was covered in a *B. cinerea* lesion (Figure 7A). By comparison, *nip6;1* had 60.7% of its leaf area infected, leaves obtained from *nip5;1* were 61.4% covered in a *B. cinerea* lesion, while *p35S::PMEI5* leaves were 68.4% infected (Figure 7A). Col-0 leaves were 71.4% covered by the lesion (Figure 7A).

In general, over the entire course of the infection, *bor1* was also significantly (*p* < 0.05) more susceptible to *C. higginsianum* compared to Col-0 (Figure 8B). At 5 dpi, over 60% of *bor1* leaf tissue was covered in a lesion caused by *C. higginsianum*, while all other genotypes of interest had less than 40% of their leaf tissue covered by a lesion (Figure 8A). At 1 dpi, the percentage leaf area covered by lesions from *C. higginsianum* was over 30% larger compared to the other genotypes analyzed (Figure 8A).

**Figure 7.** (**A**) Average lesion size on *Arabidopsis thaliana* leaves caused by *Botrytis cinerea*. Five genotypes were examined (*bor1* (boron transporter mutant), *nip5;1* (boron transporter mutant), *nip6;1* (boron transporter mutant), *p35S::PMEI5* (*PMEI* overexpression), and Col-0 (wild type)) over a 0–120 h period post-inoculation. (**B**) Boxplot showing average lesion size on *Arabidopsis thaliana* leaves caused by *Botrytis cinerea*. Five genotypes were examined (*bor1* (boron transporter mutant), *nip5;1* (boron transporter mutant), *bor1* (boron transporter mutant), *nip6;1* (boron transporter mutant), *p35S::PMEI5* (PMEI mutant), and Col-0 (wild type)) over a 0–120 h period post-inoculation. Error bars represent standard error. See Tables S9 and S10 for statistical analyses.

**Figure 8.** (**A**) Average lesion size on *Arabidopsis thaliana* leaves caused by *Colletotrichum higginsianum*. Five genotypes were examined (*bor1* (boron transporter mutant), *nip5;1* (boron transporter mutant), *bor1* (boron transporter mutant), *nip6;1* (boron transporter mutant), *p35S::PMEI5* (PMEI mutant), and Col-0 (wild type)) over a 0–120 h period post-inoculation. (**B**) Boxplot showing average lesion size on *Arabidopsis thaliana* leaves caused by *Botrytis cinerea*. Five genotypes were examined (*bor1* (boron transporter mutant), *nip5;1* (boron transporter mutant), *bor1* (boron transporter mutant), *nip6;1* (boron transporter mutant), *p35S::PMEI5* (PMEI mutant), and Col-0 (wild type)) over a 0–120 h period post-inoculation. Error bars represent standard error. See Tables S11 and S12 for statistical analyses.

#### **4. Discussion**
