Fruits of T. tetraptera were harvested from five approximately 10-year-old trees in Kekem, West Region of Cameroon. This tree was chosen according to their approximate age (more than 10 years old), their health status and the quality of fruits.

After air-drying at room temperature, the fruits were broken with a hammer to extract the seeds, which were stored in sterilized, airtight glass containers. Leaves of Tephrosia vogelii, Callistemon viminalis, and Cymbopogon citratus, used for extracts, were collected between 6 and 7 a.m. at the research farm of the Faculty of Agronomy and Agricultural Sciences (FASA), University of Dschang, West Region. The selection of the 3 plant extract species was based on several criteria such as:

• Their antifungal activity: The chosen species have shown proven efficacy against various pathogenic fungi in previous studies.

• Their accessibility and sustainability: These species are often locally available or easily grown, making them easy to use in practical applications.

• The presence of bioactive compounds: Chemical analyses have revealed the presence of compounds such as flavonoids, alkaloids and terpenes, known for their antifungal properties.

Seeds underwent thermal scarification (immersion in water at 80˚C for 2 (T1) or 4 (T2) hours) or mechanical scarification (abrasion with sand paper on the proximal part near the hilum, at 1 (T3) or 2 (T4) mm depth). The choice of these scarification depths is motivated by technical and biological considerations:

• Scarification efficiency: It is known that these scarification depths promote better moisture and nutrient penetration, increasing germination [17] .

• Embryo protection: Scarification that is too deep could damage the embryo, while these depths balance efficiency and safety.

• Variability of the species: This choice also takes into account the specific characteristics of the seeds of the species studied, where some require more or less intense scarification. The control batches received no treatment (T0). Post-scarification, seeds were disinfected in a 3% hypochlorite solution for 3 minutes, then rinsed three times with distilled water (at 5, 10, and 15 minutes). They were placed in 15 cm-diameter Petri dishes lined with three layers of moistened blotting paper and incubated at 22˚C ± 1˚C under a 12-hour light/12-hour dark photoperiod, with watering every two days. Each treatment consisted of 100 seeds, replicated three times. Germination (G) was calculated as G = (g × 100)/N, where g is the number of germinated seeds and N is the total number of seeds sown [18] . The infection rate (IR) was determined by IR = (i × 100)/N, where i is the number of infected seeds [18] . The germination rate (GR) was calculated as GR = Σ n/Σ (n × DAS), where n is the number of seeds germinated per day and DAS is the number of days after sowing [18] .

Fungi were cultured on potato dextrose agar (PDA) supplemented with 1 g/L chloramphenicol, sterilized at 121˚C for 15 minutes. Seeds, symptomatic or not, were disinfected in a 3% hypochlorite solution for 2 minutes, then rinsed with distilled water [19] . Ten seeds were aseptically placed in 90 mm Petri dishes containing 20 mL of PDA and incubated at 22˚C ± 1˚C. Fungal colonies visible after 5 days were purified on PDA [20] . Isolation frequency (IF) was calculated as IF = (NF/NT) × 100, where NF is the number of samples with a given fungus and NT is the total number of samples [21] . Identification relied on morphological characteristics (mycelium and fruiting bodies) observed under a microscope, following mycological keys [22] .

For each 10-day-old isolated fungus, a spore suspension was prepared by adding 10 mL of sterilized distilled water to PDA Petri dishes, gently brushing with a fine paintbrush. A drop of Tween 80 was added to homogenize the spores, and the suspension was filtered (mesh < 1 mm) to remove mycelial fragments [23] . Spore concentration was determined using a hemocytometer (Thoma cell) and adjusted to 10⁶ spores/mL per Mathur and Kongsda’s formula [24] . Disinfected seeds [19] were placed in Petri dishes with three layers of moistened, sterilized blotting paper (100 seeds per dish). A 10 mL suspension of each fungus was sprayed onto the seeds, the dishes sealed with parafilm, and incubated at 22˚C ± 1˚C. Each treatment was replicated three times, and the seed infection rate was assessed.

Freshly harvested leaves were washed with running water, shade-dried for two weeks, and ground into a fine powder. For aqueous extracts, 100 g of powder was macerated in 500 mL of distilled water for 24 hours with two stirrings, protected from light. The mixture was filtered through muslin and Whatman No. 4 paper, and the filtrate constituted the crude extract [25] . Extracts were dried separately in an oven at 50˚C.

In vitro antifungal activity was tested using the PDA diffusion method at concentrations of 2.5, 5.0, and 7.5 mg/mL (The choice of the different concentrations was made on the basis of a test previously carried out and efficient concentrations used in the literature with others plant extracts) against three identified pathogenic fungi: Alternaria alternata, Aspergillus fumigatus, and Cercospora sp. The negative control for this experiment was water and the positive control was the chemical fungicides Momtaz at the recommended concentration of 5 mg/kg, this fungicide is homologated in Cameroon for seed treatment. Radial growth (RG) was measured as RG = (d1 + d2 − 2d0)/2, where d0 is the explant diameter, and d1 and d2 are orthogonal culture diameters. Inhibition percentage (%I) was calculated as %I = 100 × (Dc − Df)/Dc, where Dc is the control growth diameter and Df is the diameter on extract-amended medium [26] .

Scarified and disinfected seeds (200 per treatment) were soaked in 25 mL of aqueous extracts at concentrations of 2.5, 5.0 and 7.5 mg/mL. Sterilized distilled water (25 mL) and the fungicide Momtaz (5 mg/kg) were used as negative (T−) and positive (T+) controls, respectively. Following soaking, the seeds were oven-dried at 40˚C for 15 minutes [27] and then placed in Petri dishes lined with moistened paper, maintained at 22˚C ± 1˚C under a 12-hour photoperiod. Each treatment was replicated three times. Daily observations were conducted to calculate germination and infection rates. The seedling vigor index was determined using the formula: Vigor Index = Germination Rate × (root length + shoot length) [28] .

The analyses were carried out using the statistical analysis software R version 3.5.1. at the 5% probability threshold. Since the data did not follow the normal distribution and the homogeneity of the variance was not respected, the Kruskal-Wallis test was used for the separation of the means.

Thermal and mechanical scarification significantly increased the germination rates of T. tetraptera seeds compared to the control, where the germination rate was zero ( Figure 1 ). The germination rate was highest with mechanical scarification at a 2 mm depth (92.33%), followed by mechanical scarification at 1 mm (81.52%), while thermal scarification treatments for 2 hours and 4 hours showed very low germination rates of 1.27% and 0.67%, respectively. There was no significant difference between the infection rates obtained from the different treatments, but the seed infection rate was higher in the heat scarification treatment boxes for 2 hours (42.00%). The control boxes had no germinated seeds but exhibited a fairly high infection rate (28.67%). The different treatments showed significant differences in the percentage of infection. The highest seed infection rate was observed in seeds subjected to thermal scarification (T1) for 2 hours (42.00%). In contrast, seeds scarified with sandpaper (T4) and heat for 4 hours (T2) had lower infection rates of 20.67% and 27.67%, respectively. Seeds that had not undergone any treatment did not germinate but showed fairly high infection rates. Figure 2 shows the effect of the various treatments on the infection rates of T. tetraptera seeds at 14 days after sowing (DAS). Figure 3 shows the germinated and infected seeds at 3 DAS, where visible mycelium of the pathogens is observed on seeds. The different results between thermal and mechanical scarification could be explained by the fact that at 2 mm, the embryo is already visible and develops without hindrance due to the integument [28] . It has been shown that scarification using a nail clipper could increase germination capacity by 80% to 90% after three days [28] . The dormancy-breaking methods used have been evidenced by other studies, namely mechanical scarification using sandpaper [29] [30] and thermal scarification with hot water (80˚C), which made it possible to remove the waxy cuticle from the seeds [31] . Some authors highlighted the interest in involving local populations in the domestication and in situ and ex-situ conservation strategies of plant species of interest [32] .

Mechanical scarification also resulted in a high germination rate (92%) and a reduction in average germination times. Similar results were obtained on the seeds of Acacia origena, Acacia pilispina, and Pterotobium stellatum [33] , and in Parkinsonia aculeata [34] . Mechanical scarification leads to the rapid imbibition of the seed coat and the entry of water into the reserves, which allows the rapid emergence of the radicle and the triggering of metabolic reactions of the embryo and cotyledons [35] . This could therefore enable rapid and high germination of seeds with integumentary dormancy (because the integument is hard and impermeable to water and oxygen).

Scarification treatments had significant effects on the duration of first germination. The first germination was observed on seeds scarified with heat for 4 hours (T2) and sandpaper at 2 mm (T4) at 2.33 and 2.67 days, respectively. On the other hand, seeds that had undergone mechanical scarification with sandpaper at 1 mm (T3) and thermal scarification for 2 hours (T1) showed slightly delayed first germination times of 4.67 and 4.91 days, respectively. Seeds that had not undergone any treatment (T0) did not germinate, and therefore, the duration of first germination was zero ( Figure 4 ).

Seven species of seed-borne fungi were identified: Rhizoctonia sp., Aspergillus niger, Aspergillus flavus, Aspergillus fumigatus, Trichoderma sp., Alternaria alternata, Cercospora sp., and Pestalotiopsis microspora. The most frequently isolated species were A. flavus (21.78%), A. niger (18.48%), and A. fumigatus (16.31%); Pestalotiopsis microspora had the lowest isolation frequency ( Table 1 ). The pathogenicity test was positive for A. alternata, A. fumigatus, and Cercospora sp., with infection rates of 45.75%, 32.25%, and 22.12%, respectively, 5 days after inoculation ( Table 2 ). The seeds’ response to artificial inoculation showed rapid seed death by these three seed-borne pathogens within 5 days ( Figure 5 ). Criteria for evaluating and classifying a seed as infected include:

*Data are means ± standard deviations. Means with different superscript letters in the same column are significantly different as per the Kruskal-Wallis test at p = 0.05. Legend: +++ = highly pathogenic; ++ = moderately pathogenic; + = pathogenic; − = not pathogenic.

• Visual symptoms: Presence of spots, discoloration, or alteration of the seed surface, damping-off.

• Seed health: Observation of fungal filaments or spores on the seeds.

• Germination tests: Seeds are considered infected if the germination rate is significantly reduced compared to a healthy control group.

• Biochemical tests: Analysis of the production of secondary metabolites indicative of fungal infection.

The isolated fungal species are well-known as damaging species of ENTFP seeds. Most of these species have already been identified on Ricinodendron heudelotii and Garcinia kola, two ENTFPs of high socio-economic value in Cameroon, as being responsible for seed losses [36] . Additionally, species like A. flavus and Cercospora sp. have been reported as fungi responsible for post-harvest losses of Monodora myristica [37] . Most of these fungi are responsible for high post-harvest losses of some fruits [38] . Cercospora sp. was also isolated from seeds of Persea americana with high occurrence frequencies [39] .

Similar results were obtained on the seeds of other woody species. This is the case for Aspergillus niger and Trichoderma sp. in Oroxylum indicum [40] , Azadirachta indica (neem), Leucaena leucocephala, and Gmelina arborea [41] , where they cause various diseases such as anthracnose and damping-off. Highly polyphagous species such as A. flavus and A. niger have been inventoried on the seeds of several woody species: Terminalia microcarpa, Cassia fistula, Sterculia foetida, Acacia mangium, and Eucalyptus grandis [41] [42] . The high presence of extremely polyphagous fungi such as Aspergillus could be explained by the fact that humidity is high under the seed canopies, which favors their development. Pods for production should be collected directly from seed trees when they have reached maturity to reduce fungal infections of seeds. As these fungi are extremely polyphagous, it would be wise to exercise caution when handling them and during storage to avoid possible contamination of the seeds.

Table 3 shows that aqueous extracts of plants at different concentrations significantly reduce the growth of pathogens. The extract of C. viminalis at a concentration of 7.5 mg/mL significantly reduced the radial growth of A. alternata and Cercospora sp. by 80.89% and 69.08%, respectively. The extract of C. citratus at the concentration of 7.5 mg/mL significantly reduced the radial growth of A. alternata, Cercospora sp., and A. fumigatus by 60.25%, 41.77%, and 19.49%, respectively. The extract of Tephrosia vogelii totally inhibited the development of the three fungi at all concentrations tested. The effectiveness of aqueous extracts of T. vogelii and C. viminalis on the development of these microorganisms varies according to concentrations and has been demonstrated by studies on C. viminalis [43] and T. vogelii [44] . Other studies showed that C. citratus and T. vogelii extracts exhibited antifungal activity against the potato late blight pathogen Phytophthora infestans [45] . The reduction of infection of the different fungal species by extracts of T. vogelii, C. viminalis, and C. citratus shows that these plants contain antifungal substances. The antifungal activity of these plant extracts is thought to be due to the action of oxygenated monoterpenes. The aqueous extract of T. vogelii completely inhibits the development of all three fungi. The biological actions of T. vogelii are mainly due to the presence of secondary metabolites of the alkaloid family. These include rutenones and their enantiomers, and then rutilines [46] , the in vitro and in vivo efficacy of plant extracts including C. viminalis against late blight [47] , the inhibition of Aspergillus sp. by C. citratus extract [48] , and the inhibitory activity of T. vogelii on the radial growth of seed fungi of cowpea in stock [49] . This suggests that extracts from these plants are indeed effective in combating fungi.

*Data are means ± standard deviations. Means with different superscript letters in the same column are significantly different as per the Kruskal-Wallis test at p = 0.05.

The effects of plant extracts on seed germination were significantly similar. However, it is noted that the germination rate was highest with T. vogelii and C. viminalis extracts at 2.5 mg/mL (97.33%) and 5.0 mg/mL (96.67%), respectively. There was no significant difference between the different concentrations of the extracts and the two controls on the germination of T. tetraptera seeds ( Table 4 ). Additionally, the lowest seed infection was obtained with the same extracts, and the values obtained were significantly (p < 0.05) comparable to the positive control (synthetic fungicide). Specifically, seed infection with T. vogelii was 4.54% with at the concentration of 7.5 mg/mL, while in the positive control, it was 3.33%. Significant differences were observed between the different concentrations of the aqueous extracts and the two controls on the percentage of infection of T. tetraptera seeds ( Table 5 ). Table 6 shows that the maximum vigour index of plants (674) was obtained with C. viminalis at 7.5 mg/mL, while the positive control obtained the lowest vigour index (506). However, significant differences were observed between the different doses of aqueous extracts and the two controls on the vigour index of T. tetraptera seeds.

*Data are means ± standard deviations. Means with different superscript letters in the same column are significantly different as per the Kruskal-Wallis test at p = 0.05.

^*Data are means ± standard deviations. Means with different superscript letters in the same column are significantly different as per the Turkey test at p= 0.05.

*Data are means ± standard deviations. Means with different superscript letters in the same column are significantly different as per the Kruskal-Wallis test at p= 0.05.

The T. vogelii extracts protected the seeds against fungal attack more effectively than the others, followed by C. viminalis, E. saligna, and C. citratus extracts. This activity of the aqueous extracts could be due to their chemical composition of antifungal substances. These results corroborate those of studies showing the antifungal potential of E. saligna extracts against fungi responsible for the deterioration of green beans post-harvest [46] , the in vitro and in vivo efficacy of plant extracts including C. viminalis against black nightshade downy mildew [47] , the inhibition of Aspergillus sp. by C. citratus extract [48] , and the inhibitory activity of T. vogelii on the radial growth of fungi in stored cowpea seeds [49] . This suggests that the extracts of these plants would be effective in combating fungi associated with ENTFP seeds. The antifungal activity of these plant extracts could be due to the action of oxygenated monoterpenes [50] and phenolic compounds, including sterols, flavonoids, condensed tannins, coumarins, and alkaloids [45] .