The field study was conducted at the LaSBAD experimental site in Pissa, in the Guinean forest zone of the CAR. This climatic zone is characterized by two alternating seasons: a rainy season from May to October and a dry season from November to April. The average annual minimum temperature is 30˚65C, with a moderate difference of 11˚44C between maximum and minimum. Average annual rainfall is around 1600 mm/year. The soil is sandy loam. The village of Pissa 2 is located in the Mbaïki sub-prefecture of the Ombella M’Poko health region. It lies 75 km west of the capital Bangui, between latitude 04˚02'46.4'' north and longitude 18˚09'51.6'' east.

The plant material consisted of contaminated seeds collected from diseased plants and healthy seeds from asymptomatic plants of the CABMV-susceptible local accession kahkiri. Seed selection criteria are essentially based on their physiological age, the condition of their reserves, and their phytosanitary status. Seeds should not be too mummified, broken or too small, in order to achieve an excellent recovery rate. What’s more, large seeds have greater reserves than small ones; this factor helps the plant to settle in better.

The bain-marie was filled with a sufficient volume of water and once the equipment had been set to a precise working temperature, the temperature of the bain-marie was checked and recalibrated very regularly using a temperature probe. Once the temperature was stable, the cowpea grains were placed in a hermetically sealed 20 kg freezer bag and soaked in hot water at a precise temperature for 15 min. After treatment, the cowpea grains were removed from the bain-marie and dried at room temperature for 24 hours (see Table 1).

Table 1. Seed treatment conditions and parameters.

A three-repeat Fisher block design was used on the experimental site. The experimental plot covers an area of 2030 m². The individual plots have a surface area of 5 m × 5 m = 25 m² with a spacing of 1 m between rows and 1 m between bunches on the row in order to achieve a maximum density of 36 plants per individual plot, 756 plants on 21 individual plots. A spacing of 2 m between plots and 2 m between blocks, then 5 m between blocks and the edge. The study plot was installed in the center of the elementary plot, which has a surface area of 4 m × 4 m = 16 m² with a maximum density of 16 plants.

Cowpea was sown on 2 August 2022 at Pissa 2. 100 seeds were sown in monoculture and 2 to 3 seeds were planted per pot. Maintenance consisted of weeding and ridging. A total of 3 weeding operations were carried out from the 45th day after sowing (DAS), with an interval of 15 days between each operation. At 15 DAS, weeding was carried out with one plant per stake. The first three young leaves were collected in plastic bags. They were placed in a small cooler and sent to the laboratory for molecular biological analysis. The following agro-morphological parameters were measured: seed regeneration rate, length growth, disease incidence, symptom severity, seed yield and fresh biomass, and average weight of one hundred seeds. During the field observation periods, diseased plants were counted 10 days after sowing in all plots and then counted every 7 days from the tenth day after sowing until 90 days in all plots. Mosaic symptoms were observed, namely vein mosaic, mottled or interveinal mosaic, speckled mosaic and spotted mosaic. The severity of symptoms from 10 to 90 days after sowing was assessed in all plots using the Fargette (1987) scale [12]. This scale has five levels from 0 to 4. The following levels were defined: 0 = no symptoms; 1 = slight mosaics without deformations covering less than 20% of the leaf surface; 2 = mosaics and chlorosis covering about 50% of the leaf surface with occasional leaf deformations; 3 = mosaics covering most of the leaf accompanied by necrosis and leaf blade deformations; 4 = terminal stage characterised by plant death. The severity was calculated according to the following formula:

where:

S is symptom severity or symptom severity index (SSI);

ds is the degree of symptom severity on the leaf surface of diseased leaves;

Pt is the number of diseased plants at the control period.

The prevalence of virus infections was calculated using the following formula:

where:

P is the Disease incidence prevalence;

Pt is the number of diseased plants at the control period;

N is the total number of plants in the square.

The reverse transcription-polymerase chain reaction (RT-PCR) test was used to detect the viruses according to the following protocol.

2.6.1. RNA Extraction Using TRIzol

Mortars and pestles were first sterilised at 150˚C for 2 h and then placed at −80˚C at the same time as the fresh cowpea pods. The cold cowpea seeds were crushed and transferred to a cold 2 mL Eppendorf tube to which 1000 μL TRIzol (stored at 4˚C) and 200 μL cold chloroform were added before homogenisation. The tubes were kept on ice for 5 minutes. Centrifugation at 13,000 rpm for 15 min at 4˚C was performed to separate the aqueous phase. The supernatant was transferred to a new sterile 1.5 mL tube to which 550 μL of cold isopropanol was added and the tubes were placed at −20˚C for approximately 30 min. The mixture was then centrifuged at 13,000 rpm for 10 minutes at 4˚C. The resulting pellet was washed in 500 μL cold 75% ethanol by centrifugation at 12,000 rpm for 5 min, still at 4˚C. After drying, the pellet was resuspended in 30 μL sterile distilled water. RNA extraction with Trizol was followed by treatment with DNase.

2.6.2. Treatment of RNA with DNase

The purpose of DNase is to destroy any residual DNA present in the samples. For this purpose, 10 μg RNA was mixed with 2 μL 10× DNase buffer (Euromedex) and 10 μL DNase (1U/μL) Euromedex. This mixture was made up to 20 μL with sterile distilled water and incubated at 37˚C for 30 minutes. 1 μL of 25 mM EDTA was added to the mixture, followed by a second incubation at 65˚C for 10 minutes.

2.6.3. RT-PCR Amplification of CABMV

The RNA obtained had to be reverse transcribed into cDNA (complementary DNA) before amplification. To do this, the first step was to amplify the part coloured red in Figure 1. A reaction medium containing 5 µL of total RNA, 1 µL of Poty-GP1-5’CP-R primer (CAGCTGCGTCAGAGAAGTG) and 10 µL of sterile water was incubated at 70˚C for 10 minutes. After incubation, the reaction medium was immediately cooled on ice for 10 min and then a mixture containing 2 μL RT buffer (10×), 0.5 μL Ribogrip inhibitor, 1 μL FireScript reverse transcription at 200 U, 0.5 μL dNTPs at 20 mM was added to the previous reaction medium to adjust the final volume to 20 μL. Finally, the synthesis of the complementary RNA (cDNA) was performed after incubation of the reaction medium at 50˚C for 30 min and then at 85˚C for 5 min. For the partial amplification of the capsid protein gene, the PCR reaction itself was prepared by mixing (for a final volume of 50 μL) 10 μL of retro transcription product, 1 μL at 10 μM of each of the Poty-GP1-5’CP-R (CAGCTGCGTCAGAGAAGTG) and Poty-GP1-5’CP-F (CGARAAGGARTTRCAAAGG) primers, 0.5 μL at 20 mM dNTPs, 0.5 μL at 1 U Hot start TAQ DNA polymerase, 2.5 μL of MgCl2, 5 μL at 10× Hot start buffer and 29.5 μL of sterile water. Amplification was carried out in the following steps: an initial denaturation for 5 min at 94˚C, followed by 35 cycles (denaturation for 30 sec at 94˚C, hybridisation for 60 sec at 47˚C, elongation for 1 min at 72˚C) and finally a final elongation step for 7 min at 72˚C to complete the amplification reaction.

Figure 1. Gel electrophoresis (1%) of partial amplicons of the CABMV Faso capsid protein gene obtained with the Poty-GP1-5’CP-R/Poty-GP1-5’CP-F primers. Lines 1 to 5: CABMV-infected leaves from different samples. 59˚C, 65˚C and T 0˚C: positf; M: 1kb DNA size marker; T0: diseased and untreated seeds.

All RT-PCR reactions were performed on a Gene amp®, PCR system 2700 thermal cycler. At the end of amplification, each PCR product was electrophoresed on a 1% agarose gel prepared in 1× TAE buffer containing ethidium bromide (BET) at 0.5 μg/ml. Subsequently, 10 μL of PCR product (DNA) mixed with 1 μL of loading buffer (12.5% Ficoll in 5× TAE, bromophenol blue) was added to the gel wells for migration at 100 V for 35 min in the presence of a 1 kb molecular weight marker (Invitrogen™). Gels were visualised using a UV transilluminator.

The generalized linear model of binomial family was used to compare: 1) seed regeneration rates, 2) CABMV incidences between treatments and 3) proportions of symptom-free plants and those with different levels of CABMV symptom severity within untreated diseased seeds and between treatments. Data on agronomic performance (seed and biomass growth and yield) did not follow a normal distribution according to Shapiro’s test (P < 0.05). These data were then analyzed using a generalized model with a Poisson distribution. Tests were performed with R software (version 4.3.0) and the probability level for a significant difference was 0.05.

The RT-PCR assay is highly effective in detecting CABMV using the Poty-GP1-5’CP-R/Poty-GP1-5’CP-F primer pairs.

Healthy and untreated seeds were totally recovered, whereas diseased and untreated recovered at 79.2%. In treated seeds, the best recovery rate was obtained in samples for treatment at 65˚C (95%), following by the treatment at 61˚C (93%) and at 59˚C (91%). Grain recovery was significantly low for treatment at 67˚C (66.7%; P < 0.05) (see Table 2).

Table 2. Recovery rate by sample at the first month.

(*) = significantly lower rate according to a GLM with binomial distribution (P < 0.05).

Seven days after harvesting cowpea seeds, the overall phytosanitary status of the plants in the field was assessed on the basis of symptoms observed during the vegetative phase. The symptoms observed consisted of simple chlorosis, with a few plants showing symptoms of degrees 1 to 3. Different levels of intensity of CABMV symptoms were observed in the plot on plants resulting from diseased and untreated seeds (Control 1) and a few heat-treated plants. The same plant may present symptoms of very variable intensity (scale 1 to 2) on the surface of trifoliate leaves. The distinct symptoms on the leaves enabled us to classify the infection levels of the plants according to the Fargette scale. The results showed that almost all plants grown from heat-treated seeds showed a low rate of symptom progression (Figure 2). The prevalence of CABMV was low overall, but very high in Control 1, with high severity indices (IS = 1) reaching 45%. The rate of asymptomatic plants was high (80% - 85%) for all the treatments and for healthy and untreated seeds (Control 2), with no statistical difference (P > 0.05; Figure 2).

Figure 2. Percentage of plants following index of severity (IS) at month 1 in the experimental field. ns=no statistical difference; (*) = percentage significantly lower. Control 1 = plants emerged from the diseased and untreated seeds; Control 2 = plants emerged from the healthy and untreated seeds.

The disease prevalence increased from 15% - 20% at month 1 to 25% - 40% at the second month in all treatments except plants emerged from the diseased and untreated seeds, where the disease prevalence was significantly higher at 60% - 70% (P < 0.05; Figure 3). Three months after sowing, the disease prevalence reached 60% and 85% in the treatment at 59˚C and in Control 1, respectively, whereas prevalence remained stable in the other treatments. Control 2 (healthy and untreated seeds) showed a similar pattern of evolution in the disease prevalence as the heat-treated seeds (Figure 3).

Figure 3. Incidence of CABMV in yield squares over a three-month period. Evolution of the proportion of symptomatic plants as a function of time in relation to the treatments and Controls 1 and 2 in this study. The proportions of diseased plants were compared at the end of the crop cycle, taking Control 1 (diseased and untreated seeds) as the reference for these comparisons (Shapiro, *P < 0.05, **P < 0.005, ns = non-significant difference).

The severity of CABMV was assessed on the basis of observation of symptom development on leaf surfaces. Treatment inhibited symptom expression over the 3 months of experimentation, although severity was generally low. Indeed, the disease severity was 0.85 in the diseased and untreated seeds (Control 1) at 1 month, whereas this was significantly lower at around 0.25 for all other treatments (P < 0.05). Three months after sowing, the disease severity reached a significant proportion for Control 1, ranging from 1.5 to 2 (1.65 ± 0.3), whereas in the other treatments, the severity index remained inferior to 1. Comparative statistical analysis showed no significant differences in viral disease severity between thermo-treated seedlings and Control 2 (P > 0.05) (as shown in Figure 4).

3.6.1. Height Growth

Heights were compared at the end of the crop cycle (October), taking the untreated diseased plant (Control 1) as the reference for these comparisons. Better growths were recorded for higher temperatures of treatment (63˚C, 65˚C and 67˚C) with stem lengths around 120 ± 9 cm (as shown in Figure 5). The worst stem growth was recorded in Control 1 with an average stem length of 40 ± 4 cm three months after sowing.

3.6.2. Cowpea Seed Yields

The best yield was obtained by plants from infected seeds thermo-treated at 67˚C (114 kg/ha ± 35.96) followed, respectively, by 65˚C (95.77 kg/ha ± 34.71) and the healthy and untreated seeds Control 2 (95.07 kg/ha ± 27.08). However, there was a significant difference between the yield of Control 1 (43.18 kg/ha ± 27.53) and the others (P < 0.05), except for the treatment at the lowest temperature (59˚C). These results show the impact of the disease on the yield of Control 1, while the 67˚C treatment significantly improved yield (P < 0.005) (as shown in Figure 6).

3.6.3. Cowpea Biomass Yields

The highest biomass yield was observed at 65˚C (437.98 kg/ha ± 133.7), followed by Control 2 (441.98 kg/ha ± 132.29; Figure 7). The lowest yield was obtained by Control 1 (236.54 kg/ha ± 92.47). Biomass yield followed similar results as in grain yield presented above.

Figure 4. Evolution of CABMV symptom severity index over the time. (*) and (**) indicate P < 0.05 and P < 0.005, respectively.

Figure 5. Stem length (cm) of plants over the three months of field observation. Values with different letters are statistically different according to the generalized model with a Poisson family error distribution(P < 0.05). T1 (59˚C), T2 (61˚C), T3 (63˚C), T4 (65˚C) and T5 (67˚C).

Figure 6.Estimated seed yield in kilograms per hectare (kg/ha). Values with different letters are statistically different according to the generalized model with a Poisson family error distribution (P < 0.05).

Figure 7. Estimated above-ground biomass yields in the treatments. (kg/ha). Values assigned the same letter are significantly different according to the generalized model with a Binomial error distribution (P < 0.05).

The best yield was obtained with seeds from plants thermo-treated at 67˚C, 21.17 g ± 3.08, which is very significantly different from Control 1 (14.01 g ± 3.13; P < 0.05). There was no significant difference between treatment at 65˚C (20.75 g ± 3.12), 63˚C (20.53 g ± 3.03) and Control 2 (20.70 g ± 2.95; P ˃ 0.05). These results show the impact of the disease on seed enlargement status.

Seed Yield Losses

Yield loss in grain weight occurred in Control 1 (diseased and untreated seeds) and treatments at 59, 61 and 63˚C. Losses in seed yield were statistically more important in Control 1 and in treatment at 59˚C compared to those in treatments at 61 and 63˚C (Table 3). Increasing the temperature of treatment resulted in a decrease in the loss of seed weight. Treatment at 65˚C was the threshold temperature for the restoration of the grain yield. At 67˚C, a significantly high gain in seed (20.5%) weight was recorded (as shown in Figure 8).

Figure 8. Weight yield of 100 seeds in grams. Values assigned the same letter are not significantly different according to the generalized model with a Poisson error distribution (P < 0.05).

Table 3. Yield loss rates by seed weight of plants grown from thermo-treated seeds compared with Control 2 in experiments.