Agronomía Colombiana

Effect of the inducer on H. vastatrix urediniospore germination

Aqueous solutions of each inducer treatment were prepared at three different doses, considering the concentration of the active ingredient in the commercial product (Table 1). The pH of the water used ranged between 7.05 and 7.10. The source products were ASM (BION 50WG®, Syngenta Crop Protection LLC, Greensboro, NC, USA), SA (Re-Leaf®, Stoller International Inc., Houston, TX, USA), Potassium phosphite (PhytoGard Phosphyte®, Stoller International Inc., Houston, TX, USA), and Harpin protein (Messenger®, Plant Health Care, Holly Springs, NC, USA).

Table 1.: Products and doses used.

*According to the product manufacturer.

Coffee leaves (Coffea arabica var. Caturra) were collected from experimental plots at the Naranjal Central Station of Cenicafé (Chinchiná, Caldas) to obtain the rust inoculum. The leaves exhibited characteristic symptoms and signs of the disease. They were free of biological control agents such as Simplicillium sp. and Lecanicillium sp. mycoparasites of rust uredospores that are visually easy to identify due to the whitish color of the rust pustules. Subsequently, the leaves were stored in paper bags and taken to the Phytopathology Laboratory at the National Research Center of Coffee (Cenicafé), where they were placed for 48 h in moist chambers at 25°C in the dark to preserve the inoculum. Once the lesions showed abundant sporulation, the urediniospores were harvested by scraping the undersides of the leaves with a #22 scalpel blade and then storing them in gelatin capsules at room temperature for no longer than 24 h.

Inoculum solutions were prepared at a rate of 0.3 mg ml^-1 urediniospores in sterile distilled water, yielding a final concentration of approximately 4000 urediniospores ml^-1. These solutions were homogenized by ultrasound for 30 s at 50 Hz and maintained under constant stirring. Ten reading units were subsequently evaluated, each consisting of 4 µl of the inducing compound solution plus 4 µl of the suspension containing homogenized urediniospores. These tests were performed on slides covered with Parafilm® M. Once all the slide covers were prepared, they were placed in moist chambers at a temperature of 22°C in complete darkness. After 4 h, 2 µl of lactophenol cotton blue dye was added to each slide to stop germination and facilitate counting germinated and non-germinated spores using an Axio Lab. A1 stereomicroscope (Zeiss Company, Oberkochen, Germany) located in the Phytopathology Laboratory at Cenicafé. All processes of counting germinated and non-germinated spores were supported by Carl Zeiss® ZEN image acquisition software and processed using an algorithm designed for this purpose in ImageJ® (version 1.49).

Spores were considered germinated when they exhibited a germinative tube measuring more than one-third of the diameter of the urediniospore (Kushalappa & Eskes, 1989). The germination rate variable was calculated by randomly applying the treatment doses to the experimental units (slide mounts) and the doses of each inducer as the treatments. Since each molecule was evaluated separately, a daily evaluation of the absolute control (germination only in distilled water) was necessary to determine the quality of the inoculum. If the germination rate was lower than 50%, the treatment readings were not taken, and the trial was discarded. The percentage of inhibition in germination was obtained by percentage difference, considering the germination rates obtained for the absolute control of each treatment.

Effects of inducers on the control of coffee leaf rust

The coffee plants were obtained from the harvested seeds of self-fertilized C. arabica cv. Caturra populations to avoid genetic variability. These plants, located at the Naranjal Central Station of Cenicafé, are highly susceptible to all physiological races of H. vastatrix. The seedlings were transplanted into 17 x 23 cm seedbed plastic bags filled with soil supplemented with organic matter and placed in a mesh house under 40% shade, an average temperature of 25°C, and an average relative humidity of 70% until they reached six months of age or developed five pairs of fully expanded leaves. The application of fungicides or other products that could interfere with disease development was restricted. The seedbed was fertilized before treatment application with 2 g of diammonium phosphate (DAP) every two months, and irrigation was scheduled twice weekly.

When the plants reached the required number of leaves (four pairs of fully extended leaves), aqueous solutions of each inducer treatment were prepared at three different concentrations, as shown in Table 1. Each treatment was applied using a Gast® spray pump (Gast Manufacturing Inc., Benton Harbor, MI, USA) at a pressure of 5 psi, with a volume of 0.5 ml/leaf, maintaining a 30 cm distance between the applicator nozzle and the plant. For evaluations of the local and systemic response of the inducers, the second and fourth pairs of leaves were protected with a waterproof plastic cover to avoid direct contact during spraying (systemic response). The first and third pairs of leaves were applied on upper and lower surfaces to ensure uniform distribution (local response). Control treatments were applied using only distilled water, following the same spraying method.

Each treatment had an exposure time interval between product application and pathogen inoculation of 15, 30, and 45 d. Inoculation with H. vastatrix was performed on the same day for all intervals to avoid variations. As previously described, an inoculum solution of approximately 4 x 10⁴ spores ml^-1 was prepared. The inoculation was applied to the first through fourth pairs of leaves from the apex, including those previously protected using a Gast® spray pump at a constant pressure of 5 psi, with a volume of 0.5 ml/leaf, maintaining a 30 cm distance between the applicator nozzle and the plant. Once inoculated, the plants were placed in a dark room for 72 h at 23°C with a relative humidity of 90% or more, maintained by a Bahnson® humidifier (DnB Humidifier Manufacturing Inc., Winston-Salem, NC, USA), to favor pathogen germination and penetration. Afterward, the plants were returned to the mesh house under the same conditions (40% shade, average temperature 25°C, and average relative humidity 70%) until the first symptoms appeared.

The test was completely randomized with a 3x3x2 factorial arrangement (three doses x three-time intervals x two response types). The experimental unit was the seedling or nursery plant. The treatments were the interaction of the three doses (low, recommended, and high), the time intervals (15-, 30-, and 45-d post-inducer application), and the response signaling pathways (local and/or systemic). Each treatment, including the control, had six plants as replicates, each with eight leaves as reading units (n = 48). After 60 d post inoculation (the date on which the control reached maximum pustule development), the evaluation process of each treatment began. To do this, each leaf corresponding to the reading units was removed with the help of a scalpel, trying not to cause disturbance or loss of the generated pustules. The leaves were then placed in previously labeled trays and taken to the Phytopathology laboratory of Cenicafé.

Each treatment was evaluated by recording the percentage of the affected area as disease severity using a UMAX Powerlook® scanner (Model 2100XL). Each image was processed at 500 dpi on a black background with a quadratic area-preserving map. Images were individually processed with ImageJ® (version 1.49) software, calibrated to identify each lesion's minimum and maximum visible spectral ranges to determine the final severity on each reading unit. The percentage of control was obtained by percentage difference, considering the final severity obtained for the absolute power. The test duration was 6 months and was not repeated over time.

Statistical analysis

For both the in vitro germination rate of urediniospores and the disease severity in the leaves, we checked for normal distribution and homogeneity of variance with the Shapiro-Wilk and Levene test. Then, a descriptive analysis of variance (ANOVA) was performed using the statistical arithmetic mean, standard deviation, and coefficient of variation. Variables showing significant differences were subjected to Duncan's means comparison test (α = 0.05). Data analysis was performed with a Dunnett's means comparison test (α = 0.05) for the germination rate variable since each treatment was individually compared to the absolute control. All statistical analyses were conducted using R version 3.3.2 (R Core Team, 2023).

Effect of the inducer on H. vastatrix urediniospore germination

None of the treatments were statistically equal to the germination rates obtained for the absolute control (without product application) across the three evaluated doses (Fig. 1). On average, the germination rate for the absolute control was 67.9%, followed by potassium phosphite treatment (15.7%), ASM (15.1%), harpin (6.3%), and SA (0%). The treatment with the highest inhibition rate was SA (100%), followed by harpin (90.9%), potassium phosphite (76.4%), and ASM (74.5%).

According to the ANOVA, there were significant differences among treatments in terms of urediniospore germination after 4 h of exposure to the three different doses (F = 230.9, df = 4, P-value <0.0001). None of the evaluated molecules exhibited germination rates close to those of the absolute control (water alone without product application), which ranged from 59.2% to 76.2%. The germination inhibition displayed an inversely proportional relationship between the dose and the urediniospore germination rate. However, this effect did not occur with SA, which prevented spore germination even at the lowest dose, demonstrating the compound's marked direct inhibitory effect (Fig. 2).

Germination at the low dose (0.2 µl ml^-1) was only 27.9% germination, followed by the recommended dose (0.8 µl ml^-1) with 19.1% germination, and finally, the high dose (3.3 µl ml^-1) where germination did not occur. These findings are consistent with Costa et al. (2007), who evaluate the biocidal effect of 7.5 µl ml^-1 potassium phosphite on the in vitro germination of H. vastatrix urediniospores, reporting 100% inhibition compared to the absolute control, indicating a significant toxic effect of this molecule on pathogen germination. Similarly, Fernandes et al. (2013), studying the efficacy of copper and manganese phosphates at a concentration of 5 µl ml^-1 on the in vitro germination of H. vastatrix urediniospores, observes low germination rates of 22.8% and 20.8%.

Likewise, ASM showed adverse effects compared to those obtained in the absolute control. Germination at low doses (50 µg ml^-1) was only 29.9%. This was followed by the recommended dose (200 µg ml^-1) with 15.4% germination, and finally, the high dose (800 µg ml^-1) where germination did not occur. These results differ from those Guzzo et al. (2001) reported, who assessed ASM's biocidal effects on urediniospore germination and appressorium formation in H. vastatrix using fluorescence microscopy. Their study found no significant inhibition of germination or appressorium formation across all ASM concentrations tested (10 to 400 µg ml^-1). Similarly, Marchi et al. (2002) evaluated ASM across different concentrations (1 to 1000 µg ml^-1) on coffee leaves. They found no fungitoxic effects on the pathogen.

Figure 1.: The mean germination rate of urediniospores was exposed to three doses of potassium phosphite, Acibenzolar S methyl, harpin, and salicylic acid for 4 h, compared to the absolute control.

Figure 2.: The germination rate of urediniospores exposed to three different doses of A) potassium phosphite, B) ASM - Acibenzolar S methyl, C) harpin, and D) SA - Salicylic acid for 4 h.

Harpin also showed adverse effects compared to those obtained in the absolute control. Germination at the low dose (7.5 µg ml^-1) was only 17.7%, followed by the recommended dose (30 µg ml^-1) with 1.3% germination, and finally at the high dose (120 µg ml^-1) where germination did not occur. These findings contrast with those reported by Jesus (2009), who found no significant differences in the toxic effects of harpin on H. vastatrix urediniospore germination across a range of concentrations from 250 to 4000 µg ml^-1, results that were comparable to the absolute control. However, Galdeano et al. (2010) investigated harpin's effects on Cercospora coffeicola. They observed a stimulating effect on conidial germination and mycelial growth at concentrations ranging from 7.5 to 120 µg ml^-1. These discrepancies suggest that harpin may not exert a direct toxic effect on pathogen development but could instead stimulate growth, a phenomenon whose significance in disease control remains debated. Unlike our study, Jesus (2009) and Galdeano et al. (2010) applied harpin directly to leaf surfaces, allowing for absorption over approximately 6 h before pathogen inoculation (H. vastatrix and C. coffeicola), indicating that the molecule was not in direct simultaneous contact with pathogen spores.

Furthermore, it is essential to consider the pH of the aqueous dilutions used with Harpin treatment, which ranged between 5.40 and 5.57, representing the most acidic conditions in our experiment. Carré-Missio et al. (2012) investigated the effects of potassium silicate applications at different concentrations on H. vastatrix urediniospore germination across pH values of 5.5, 7.5, and 10.5. They report that at a dose of 8 g L^-1, germination inhibition rates are 13.9%, 15.6%, and 20.8%, respectively, indicating varying effects of the product with spores germinating more effectively under acidic conditions than neutral pH. Finally, SA showed inhibitory effects on germination for all evaluated doses, even at the lowest dose (0.3 µl ml^-1).

Although most authors did not report adverse effects on the germination of H. vastatrix, our results suggest a toxic effect on the germination processes of the spores, mainly depending on a high concentration of the applied product. Because of this, and in light of our results, it is not possible to ensure that the control of the disease corresponds to the direct activation of the plant's defense mechanisms but to the adverse effects that the molecules have on the germination of the spores, an essential aspect if we consider that this type of research seeks the selection of molecules without direct toxic effects (Steiner & Schönbeck, 1995), due to the implications that this has on the selection pressure of the pre-existing populations of the pathogen or the generation of a more significant number of resistant isolates.

Effects of inducers on the control of coffee leaf rust

According to the ANOVA, there were significant differences among treatments regarding disease control, regardless of the dose, application interval, and type of response signaling, compared to the absolute control (F = 26.08, df = 4, P-value <0.0001). The severity of the absolute control was 11.6%, followed by treatments with potassium phosphite (2.7%), harpin (2.5%), SA (1.6%), and ASM (1.3%). The treatment with the best control was ASM, achieving 88.6% control compared to the absolute control, followed by SA (86.1%), harpin (78.6%), and potassium phosphite (76.6%) (Fig. 3).

Figure 3.: Mean severity index at three doses, three different time intervals between product application and pathogen inoculation, and two signaling pathways in the responses to potassium phosphite, harpin, SA, and ASM compared to the absolute control.

Regarding the dose effects, only SA and ASM showed significant differences in disease control at high doses. Potassium phosphite, SA, and ASM exhibited substantial differences compared to harpin at recommended doses. In contrast, harpin, SA, and ASM showed significant differences compared to potassium phosphite at low doses. Regarding the effect of application intervals, harpin, SA, and ASM displayed substantial differences at a 15 d interval. Additionally, SA and acibenzolar-S-methyl (ASM) showed significant differences at 30 and 45 d application intervals. Concerning the effect of type of response signaling, SA and ASM showed substantial differences for local signaling. Furthermore, ASM was the only compound showing significant differences compared to the others for systemic signaling (Fig. 4).

Figure 4.: Mean severity index (left) and percent control (right) of variables. A and B) The molecule doses, C and D) The time intervals between product application and pathogen inoculation, E and F) The types of responses to potassium phosphite, harpin, SA, and ASM.

Potassium phosphite achieved the best results when applied at the high dose (3.3 µl ml^-1), followed by the recommended dose (0.8 µl ml^-1), and then the low dose (0.2 µl ml^-1), with disease control percentages of 83.4%, 82%, and 64%. Optimal disease control was observed when treatments were applied at a 15-d interval, followed by 30 and 45-d intervals, resulting in control percentages of 82.6%, 75.5%, and 71.2%. The predominant signaling pathway was more local (83.7%) than systemic (69.2%), indicating that disease control is more effective with higher doses applied at intervals not exceeding 15 d, achieving control rates ranging from 80% to 86% (Fig. 5A). These results contrast with Costa et al. (2007), who demonstrates effective disease control (57-65%) with 7.5 µl ml^-1 potassium phosphite applied at intervals close to 30 d. Similarly, Fernandes et al. (2013) found that manganese phosphite provided higher disease control (70%) than copper phosphite (56%) during longer application intervals, including 40 d against coffee rust.

Harpin exhibited the most effective results when applied at the highest dose (120 µg ml^-1), followed by the recommended (30 µg ml^-1), and then the lowest dose (7.5 µg ml^-1), achieving disease control of 81.9%, 76.9%, and 76.9% (Fig. 5B). Optimal disease control outcomes were observed with application intervals of 15 d, followed by 30 and 45 d, resulting in controls of 90.6%, 82.1%, and 63.2%, respectively, with a predominant signaling effect more local (82.3%) than systemic (74.9%).

Figure 5.: Severity index at three different doses (low, recommended, and high), at three different time intervals (15, 30, and 45 d) between the product application and pathogen inoculation, with the response signaling pathway being (L), a local effect, and (S), a systemic effect of A) potassium phosphite, B) harpin, C) SA, and D) ASM.

SA demonstrated optimal results at high doses (5 µl ml^-1), followed by the recommended (1.2 µl ml^-1), and then the lowest dose (0.3 µl ml^-1), achieving disease controls of 91.5%, 88.8%, and 78.6%, respectively (Fig. 5C). The most effective disease control occurred with application intervals of 15 d, followed by 30 and 45 d, achieving controls of 94.8%, 87.5%, and 76.6%, respectively, with a predominant local signaling effect (93.2%) over systemic (79.4%).

ASM showed optimal results at high doses (800 µg ml^-1), followed by the recommended dose (200 µg ml^-1). Then, the low dose (50 µg ml^-1) achieved 94.3%, 90.6%, and 82.9% control, respectively (Fig. 5D). The best disease control occurred at 15 d intervals (95.3%), followed by 30 and 45 d (92.3% and 80.2% control, respectively), predominantly through local (91.7%) rather than systemic (86.8%) signaling. Similar results were reported by Guzzo et al. (2001), who observes ASM's protective effects ranging from 66% to 97% locally and 83% to 94% systemically at higher concentrations (400 µg ml^-1). However, effectiveness significantly decreases when intervals between ASM application and pathogen inoculation exceeded 49 d, even at high doses. To study their effects on coffee rust, Sancho and Diaz (2006) conducted greenhouse experiments at Cenicafé using ASM and other resistance-inducing biological products at varying concentrations. Applying the BION product (ASM) at concentrations of 1, 10, 20, 50, and 100 µg ml^-1 effectively reduces the number of leaf lesions, with optimal application intervals observed at 8 d to induce robust plant resistance and minimize disease development time.

According to the results, treatments with these compounds consistently reduced disease severity and improved disease control compared to untreated controls. The efficacy of disease control was mainly influenced by the concentration of the product, which dictated the optimal frequency of application to avoid dilution of the treatment effects, with biweekly application intervals (every 15 d) being the best for disease control at a recommended dose. However, it is essential to note that in many of the compounds evaluated, the increase in concentration did not necessarily translate into longer application intervals (every 30 or 45 d), which led to adjusting the application intervals according to the product selected and the dose used.

In terms of signaling, local responses prevailed over systemic ones, and this behavior was primarily governed by the intrinsic properties of the molecule rather than other factors such as dose or application intervals. These types of local responses have economic and operational implications for preventive disease management plans since, as far as possible, the application process of the product should be carried out directly on the leaves to achieve better coverage, which increases application times and, therefore, the costs associated with management.