Samples of onion plants were collected from 17 onion production sites in Burkina Faso from January to February 2017. The collection sites are located in four onion production zones (Figure 1): the Northern area including Yako, Ouahigouya, Titao, Korsimoro and Kongoussi sites; the Central zone comprising the sites of Mogtédo, Loumbila and Donsin; the Western zone including the sites of Kokologho, Koudougou, Réo, Gassan, Di, Tougan and the South-western area including the sites of Dano, Soumasso and Bama. At each site, samples were collected from three fields separated by at least 5 km, on the basis of ten plants collected at random per field. The isolation of the Fusarium isolates consisted of disinfecting each plant with 3% sodium hypochlorite and cutting it into small fragments. The fragments were incubated in a Petri dish on moistened blotting paper for five days. Based on the identification manual of fungi [13], plant fragments were examined under a stereo microscope and/or microscope to detect the developing Fusaria, regardless of the species. For each site, the number of Fusarium detected on the incubated onion plants was counted and the results were expressed as a percentage of plants attacked by the fungus. For each type of

Fusarium identified per plant, a small portion of mycelium was aseptically grown in a Petri dish containing PDA (Potato Dextrose Agar) culture medium. Successive transplantation of the isolates into the PDA medium has resulted in pure cultures of these fungi.

Monosporic cultures were produced from the pure cultures of Fusarium isolates. For each isolate, two drops of conidial suspension containing approximately 100 spores/ml of suspension prepared from a 10 day-old pure culture were spread on an Agar medium contained in a Petri dish. Two days after seeding, the germinating spores were transferred to Petri dishes containing the PDA medium using only one germinating spore per dish, and left to grow. Three monosporic isolates were produced per isolate. Based on the color of the mycelium and/or the shape of the conidia, one to four monosporic isolates were selected per site for molecular analysis.

For fungal DNA extraction, mycelium of the monosporic isolates was first cultured on Potato Dextrose Broth (PDB) liquid medium. For this purpose, a mycelial explant taken from each monosporic isolate in culture on the PDA medium was aseptically deposited in the PDB medium contained in an Erlenmeyer. To stimulate the development of the mycelium in the liquid medium, the bottles were incubated on an oscillator at the speed of 120 oscillations per minute, at room temperature in the laboratory, for two days. After removing the mycelial explant from the Erlenmeyer, the resulting mycelium was collected by filtration using a vacuum pump. For each monosporic isolate, 75 mg of mycelium was collected, transferred to an Eppendorf tube and stored at −80˚C. The fungal DNA was extracted using the method used by Sadfi-Zouaoui et al. [14] with some modifications. Six hundred microliters (600 µl) Cetyl Trimethyl Ammonium Bromide (CTAB) extraction buffer (1.4 M NaCl; 2% CTAB (w/v); 0.1 M Tris-Base pH8; 0.02 M EDTA pH8; 0.2 B-Mercaptoethanol (v/v)) were added to the 75 mg of mycelium and the mixture was stirred on a vortex for 10 min. The tubes were then placed in a water bath at 65˚C for 10 min, then 450 µl of phenol and 450 µl of chloroform alcohol-iso-amyl (composed of 49 ml of chloroform and 1 ml of alcohol-iso-amyl) were added to each tube and then mixed by inversion of the tube until a milky mixture was obtained. After centrifugation at 13,000 g for five minutes at 25˚C, the supernatant (approximately 500 μl) added to 400 μl of alcohol-isoamyl chloroform, was again centrifuged at 13,000 g for two minutes at 25˚C and then transferred to a new tube by adding 0.7 volume (300 µl) cold isopropanol to the solution to precipitate DNA. After a further centrifugation of the mixture, the pellet obtained was rinsed with 500 μl of 75% alcohol and centrifuged for 5 minutes at 13,000 g at 25˚C, this twice, before being dried in a hood and then dissolved in 50 µl of sterile distilled water and stored at −20˚C.

The DNA was then determined by spectrophotometer (Nanodrop 2000) and the initial concentration evaluated in nanograms per microliter (ng/µl). The different concentrations were diluted on a case-by-case basis so that the volume of DNA (3 µl) collected for PCR contains about 100 ng of DNA. The amplification reaction was performed in 25 µl containing 100 ng DNA (3 µl); 2.5 µl BD buffer, 10×; 0.5 µl dNTP (10 mM); 2 µl MgCl (25 mM); 0.2 µl Taq polymerase; 15.8 µl sterile distilled H₂O and 0.5 µl EF1 to 10 µM and EF2 to 10 µM primers. The primers used were:

EF1: ATGGGTAAGGARGARGACAAGAC [15];

EF2: GGARGTACCAGTSATCATCATGTT [16].

The amplification reaction was performed according to the following program: a denaturation cycle at −95˚C for 10 minutes followed by 35 consecutive cycles (denaturation at 94˚C for 30 seconds, hybridization at 52˚C for 30 seconds and elongation at 72˚C for 45 seconds). The electrophoresis was performed on a 1% agarose gel at 120 V for 20 minutes. The amplification products were visualized under UV light, using ethidium bromide previously incorporated into the agarose gel. Thirty microlitres (30 µl) of PCR product from each isolate were collected for sequencing.

The sequencing was carried out by Genewiz [17], with the P363 and CHOO9 primers. For the identification of Fusarium species, the sequences were processed, cleaned and aligned with the Chromas software. The consensus sequences obtained were compared with those in the NCBI (National Center for Biotechnologies Information) database on the site [18].

In order to determine morphological and cultural characteristics for a good differentiation of Fusarium species isolated from onion plants, three biometric characteristics related to the colony growth in culture, the conidia sizes and the number of cells (or divisions) per conidia were evaluated.

The experimental design used is a randomized complete block design with four replications. To assess the mycelial growth of the colonies, the diameter of the colonies was measured in each dish on the ninth (9th) day after the incubation of the explants in the PDA medium. The result, expressed in millimeters, was the average of two diameters perpendicular to each other.

The mycelial growth rate was calculated according to the formula:

V = [ ( D 1 − 4 ) + ( D 2 − 4 ) ] / N × 2

where V = Growth rate of the fungus in mm/day; D1 and D2 are the two perpendicular diameters of the colony, N the number of days after transplanting the explant and 4 the diameter of the explants.

For the evaluation of conidial dimensions, 25 conidia were randomly selected per monosporic isolate (one conidia representing a repetition), the length and width of each conidia were measured. The evaluation of the number of cells per conidia was coupled with the conidia size evaluation. The number of cells in each conidium was determined by counting the number of subdivisions within that conidium.

The biometric data collected were subjected to an analysis of variance using the Statistical Analysis System (SAS) software, version 8; 2001. The calculation and separation of the means of the measured parameters were performed according to the Duncan test (Duncan Range Multiple Test), at the 1% threshold.

A total of 510 onion plant samples were collected from the 17 production sites. Visual examination of these incubated plant samples allowed the detection and isolation of several types of Fusarium associated with the plants. The results of the analysis of variance showed that the prevalence of fungi of the genus Fusarium varied widely (p < 0.0001) depending on the collection sites (Table 1). The fungus was present in all onion growing sites where it infected 10% to 90% of the sampled plants with an average attack rate of 62.55%. The sites studied were classified into four distinct groups: Group 1, composed of nine (9) sites (Yako,

The values of the same column assigned the same alphabetical letter(s) are not significantly 5% threshold, according to the Duncan test.

Titao, Gassan, Di, Kongoussi, Korsimoro, Mogtédo, Dano and Réo), with the highest attack rates ranging from 66.67% to 90%. The second group includes the sites of Loumbila, Ouahigouya, Donsin, Koudougou and Kokologho, with 53.33% to 63.33% of contaminated plants. The third group includes a single site (Tougan) with a low prevalence rate of 30%. Finally, the fourth group includes the sites of Bama and Soumasso where the prevalence of Fusarium on plants was very low (10% - 13.33% of infected plants). Considering the onion production areas, the results also indicated that the prevalence of Fusarium was significantly higher (p = 0.0004) in the northern, western and central zones of the country than in the southwestern zone (Table 2).

Based on the color of the mycelium, the shape and/or the size of the conidia, thirty-six monosporic isolates were selected for molecular analysis. Chromatography results obtained on agarose gel using the primers EF1 and EF2 indicated that the molecular weights of the different products are approximately 700 bp for all isolates (Figure 2), indicating that these isolates are indeed fungi belonging to the genus Fusarium.

Analysis of the obtained sequences and their comparison with the sequences in the NCBI database revealed 99% of identity for 26 isolates, 98% for five isolates, 97% for four isolates and 93% for one isolate (Table 3). The analysis showed that the isolates belong to five species of Fusarium:

- Fusarium oxysporum which includes 16 isolates (44.44% of the studied isolates) with similarity rates of 93% - 99% with the reference accessions JF957830.1, KU985430.1 and JF957820.1 of the NCBI.

- Fusarium proliferatum which includes 15 isolates (41.66%) with 99% similarity rates to the NCBI reference accession KF222556.1.

- Fusarium solani which includes two isolates (5.55%) with 99% similarity rates to NCBI reference accessions KY486699.1 and KR816154.1.

- Fusarium fujikuroi which also includes two isolates (5.55%) with similarity rates of 97% to the NCBI reference accession LC055826.1.

- Fusarium thapsinum which comprises a single isolate (2.77%) with a 99% similarity rate to the NCBI reference accession KU508368.1.

F. o = Fusarium oxysporum; F. pro = Fusarium proliferatum; F. fujk = Fusarium fujikuroi; F. s = Fusarium solani; F. thap = Fusarium thapsinum. Yk = Yako; T = Titao; Ga = Gassan; Di; Kg = Kongoussi; K r = Korsimoro; Mg = Mogtedo; Da = Dano; R = Reo; Lm = Loumbila; Oua = Ouayigouya; Do = Donsin; Kd = Koudougou; Kk = Kokologho; Tg = Tougan; Sm = Soumasso; Ba = Bama.

The analysis of variance performed on biometric data including the mycelial growth speed of fungi, the length and diameter of conidia, the number of cells per conidium, revealed significant differences between the Fusarium species. In general, F. oxysporum, F. proliferatum, F. solani and F. fujikuroi had rapid mycelial development (7.72 - 8.27 mm/d) on the PDA culture medium compared to F. thapsinum which had slow growth (6.52 mm/d) (Table 4). The conidia of F. oxysporum, F. proliferatum and F. solani were relatively longer (4.74 - 5.96 µm) than those of F. fujikuroi (3.20 µm) and F. thapsinum (4.04 µm). Of the five species, F. solani had the largest conidia (1.88 µm wide) while F. oxysporum had the conidia with the highest number of cells per conidium (1.94). The results also showed variability between isolates within the same species. Indeed, 12 isolates belonging to the species F. oxysporum (i.e. 75% of the population) grew fast (7.88 - 9 mm/d) against four isolates (i.e. 25%) with slow growth (6.72 - 7.33 mm/d) (Table 5). Among the fast-growing isolates, F.o-8-Lm and F.o-60-Kd presented long (12.28 - 17.24 µm) and large (1.92 - 3.96 µm) conidia with more than three cells per conidium (3.40 - 5.80 cells per conidium). All other isolates of this species showed small conidia (2.52 - 10 µm × 1 - 2 µm) with 1 to 2.48 cells per conidium (Table 5).

In the F. proliferatum species, 12 isolates (i.e. 80%) grew rapidly (7.86 - 9 mm/d) against three isolates (i.e. 20%) with slow growth (Table 6). Among the fast-growing isolates, F.pro-4-Sm; F.pro-15-R; F.pro-3-Sm produced long conidia (9.44 - 11.24 µm) with 2.60 to 2.96 cells each (Table 6). In contrast, the other isolates had small conidia.

F. fujikuroi included two isolates, one fast-growing (8.98 mm/d) and the other slow-growing (6.81 mm/d). For the other characteristics, no significant differences were noted between these two isolates (Table 7). In F. solani, the isolate F. s-38-Tg was characterized by a rapid growth rate (9 mm/d) and large conidia (2 µm wide) and the isolate F. s-24-Kk by a slow growth of 6.45 mm/d and fine conidia (1.76 µm wide).

The values of the same column assigned the same alphabetical letter(s) are not significantly 5% threshold, according to the Duncan test.