The plant material was constituted of seven cassava varieties plantlets RB 89509, Atiwè, ATI, OKI, Dossi, Somity (local varieties from Benin) and BF92/0267 variety from Burkina-Faso. These varieties were included in the WAVE (Central and West African Virus Epidemiology) germplasm and theirs plantlets were produced at the Central Laboratory of Plant Biotechnology and Plant Breeding of the University of Abomey-Calavi in Benin (LCBVAP/UAC).

Cassava plantlets with good root system, two nodes at least, and a minimum of three leaves were gently removed from the growing media. They were then calibrated according to their length (cm) in five modalities: [1 - 5], ]5 - 10], ]10 - 15], and ]15 - 20], and after sown in the pots containing various substrates.

Substrates were obtained by mixing compost, potting soil, sawdust and palm kernel fibers with distilled water (Table 1) and then sterilized in the autoclave at 121˚C for 20 min [18]. This was followed by cooling for one hour and then potting in pots (8.6 cm in diameter and 5.3 cm in height), pierced with 4 holes, washed and disinfected in bleach for 24 h. A batch of seeded pots was placed directly in a tunnel where average internal temperature varies between 27.32˚C and 28.91˚C and the relative humidity varies between 88.38% and 99.98%. The second batch was in an air-conditioned room with an internal temperature of 25˚C and a relative humidity of 95% for seven days before being transferred to tunnel for five weeks. This humidity was kept saturated by regular sprinkling of water.

As concern the watering, the shive solution [12] control solution (SN1), and the NPK fertilizer solutions composed of 20-10-10 (SN2); 13-17-17 (SN3) constituting a modified combination derive from the recommended combination (10-52-10) for cassava acclimatization [17]. The nutrient watering solutions were obtained by diluting 2 g of fertilizer in 1 L of water and applied at 80 ml per pot [19] immediately after transplanting into the substrates and once a week for the six weeks of the experiment.

The experimental set-up was a random block made up of three rows of plastic pots. Each row representing a watering solution consists of three different types of substrate, the seven varieties with four repetitions; making 84 pots per row for a total of 252 pots for the three rows. Thus, each acclimatization condition (tunnel and air-conditioned chamber + tunnel) was made up of 252 pots, for a total of 504 pots for the conduct of the entire experiment. The data providing information on the survival or not of tissue culture plants under pre-acclimatization conditions were subjected to a binary logistic regression. A generalized linear regression model was applied in order to determine the relationship between the

S = Substrate.

number of leaves and the substrates on the one hand and the nutrient watering solutions on the other. The height parameter of the plantlets was subjected to an analysis of variance (ANOVA) to reveal the differences between the treatments. Duncan’s test was applied for treatments significant at the 5% level. Data obtained were analyzed using XLSTAT version 14 software.

The chi-square test showed a highly significant difference (p < 0.0001) in the acclimatization processes. Also, varieties and the Pregnant × Variety interaction had a highly significant influence (p < 0.0001) on plantlet survival. Under tunnel, the rate of survival varied from 24.44% (ATI) to 37.78% (Atiwè) against 44.44% (ATI) to 77.78% (OKI) under acclimatization conditions (Table 2, Figure 1).

*** = very highly significant difference (p < 0.0001); DDL: degree of freedom, Pr: probability.

The Chi-square test revealed a highly significant influence (p = 0.023) within a height of different survival plantlet. However, the interaction between varieties and the variety × Height class interaction had a very high influence (p < 0.0001) on survival rate acclimatization (Table 3). All varieties with a stem length between 1 and 5 cm exhibited a survival rate from 40% to 70%; but those with a stem length between 5 and 10 cm exhibited a survival rate ranging from 20% to 53%. Plantlets with stem length classed between ]10 - 15] cm, had a survival rate close to zero. However, with the class ]15 - 20], the varieties Dossi and BF exhibited respectively 50% and 100% of survival rates (Figure 2).

The results indicated that the survival rate of acclimatized plantlets was not influenced by the substrates (p > 0.05) (Table 4). In addition, genotype and the interaction genotypes × substrates had a significant influence on the survival rate (p = 0.003). This survival rate varied from 33.33% (ATI) to 56.67% (OKI) on substrate S1; from 32.67% (BF92/0267) to 60% (Atiwè) on the S2 substrate and from 30% (RB 89509) to 63.33% (Atiwè) on the S3 substrate (Figure 3).

The number of leaves emitted is not influenced by either substrate or the substrate × variety interaction. On the other hand, it is significantly influenced by

* = significant difference (p < 0.05), ** = highly significant difference (p < 0.01); *** = very highly significant difference (p < 0.001) DDL: degree of freedom, Pr: probability; Class. Ht: height class.

NS = non-significant difference; ** = highly significant difference (p < 0.01); DDL: degree of freedom, Pr: probability.

the variety (p < 0.004). The best phylogenesis was observed in RB89509 variety with 1.87 leaves on substrate S2. Furthermore, the coefficient of determination is low R² (23.84%) < 50% (Table 5).

Likewise, growth in height was very significantly influenced by cassava genotypes (p < 0.0001), substrate and substrate × variety interaction had no influence on height growth. Thus, the BF92/0267 variety has a greater height whatever the substrate with values 0.90 cm, 1.2 cm and 1.13 cm respectively on the substrates S1, S2 and S3. The lowest height gain was obtained in local varieties OKI (0.06 cm) on substrate S1 and Dossi (0.05 cm) on substrate S3. In addition, it should be noted that the coefficient of determination is high R² (%) > 50 (Table 5).

The Chi-square Test showed no significant difference (p > 0.05) between the watering nutrient solutions on the survival rate on cassava plantlets (Table 6). However, the varieties and the variety × nutrient solution interaction had a highly significant effect on the survival rate of plantlets (p < 0.05). The survival rate per watering nutrient solution varied from 26.67% (BF92/0267) to 53.33%

NbreM. F: Average number of leaves; HT gain (cm): Height gain; The values followed by the same letter are not significantly different; NS = non-significant difference; *** = very highly significant difference (p < 0.001).

S = not significant difference; * = significant difference (p < 0.05), ** = highly significant difference (p < 0.01); DDL: *** = very highly significant difference (p < 0.001); degree of freedom, Pr: probability; Nut. Nutrient solutions.

(Atiwè) for SN1 solution; from 36.67% (Somity) to 63.33% (Atiwè) for SN2 solution, and from 30% (RB 89509) to 53.33% (OKI) for the SN3 solution (Figure 4).

The nutrient solution and the nutrient solutions × varieties interaction showed no significant influence on the number of leaves emitted. The varieties presented a significant effect (p < 0.05) on the emission of new leaves with a better phylogenesis observed on RB 89509 variety (2.13 leaves with SN2) and the lowest observed on OKI variety (0.33) using SN3 nutrient solution. Furthermore, the coefficient of determination is low R² (30.57%) < 50 (Table 7).

Height growth was very significantly influenced by genotypes (p < 0.0001) with coefficient of determination is high R² (58.6%) > 50) (Table 7). The nutrient watering solution and the nutrient watering solution × variety interaction had no influence on height growth (Table 6). Plantlets of the variety BF92/0267 had a higher height regardless of the nutrient watering solution. Thus, the gain in height was 0.8 cm, 1.18 cm and 1.12 cm respectively for SN1, SN2 and SN3. The lowest height gains were obtained in local varieties OKI (0 cm with SN3), ATI (0.05 cm with SN3) and Sommity (0.025 cm with SN1).