The plants produced in vitro, well rooted, were transferred under a shade for acclimatization. For this, they were removed from the tubes and freed from the agar by washing their base with sterile distilled water and then transplanted into a substrate made up of a mixture of sand and potting soil (v/v), previously sterilized. The substrate was potted in black polyethylene bags. Two weaning conditions were tested:

- A batch (1) of 12 young plants was placed in the shade for 2 days then transferred over the bench to the light;

- A second batch (2) of 12 young plants was placed in a mini-greenhouse with a transparent plastic cover with adjustable opening. Thus, the plants were maintained in conditions of relatively high humidity. They were watered with the [20] nutritive solution every two days. After the 2^nd week, the plastic cover was opened gradually.

The number of plants that survived after 15 and 30 days of acclimatization was used to determine the survival or recovery rate.

For a 5-day induction with IBA at 50 mg·L⁻¹, no roots appeared at the base of the vitroplants in the expression medium after 30 days of culture. In addition, significant callogenesis was noticed (100%) as well as rooting (33.33%). The best rooting rate (58.33%) was obtained for an induction period of 3 days with IBA (50 mg·L⁻¹) whereas for the control, this rate was 8.33% (Table 4; Figure 3). For

In the same column, the values followed by the same letter are not significantly different at the 5% threshold of the Newman-Keuls’s test. IAA: Indole-3-acetic acid; IBA: Indole-3-butyric acid.

the 3-day induction period with IBA, the average number of roots is 2.2, a value significantly higher (p = 0.016) than those obtained with the other induction times.

For the length of the roots, the analysis of variances reveals a very significant medium effect (F = 30.20; p ≤ 0.001). For an induction period of 1 day with IAA, the average length of the roots is equal to 3.75 cm, a value not significantly different from the results which are higher than the elongations obtained with the other induction periods, particularly, that with IBA (1 day-induction period) with which, an average root length of 2.38 cm was noticed (p = 0.36). The lowest root elongation at 1.8 cm was observed with the control plants. For an IAA induction period of 3 days, the mean length of the vitroplants is 3.56 cm (p = 0.004). a value significantly different from those noticed with IAA (5 days) and IBA (1 day), i.e. root lengths of 2.4 cm (p ≤ 0.001) and 2.5 cm (p = 0.009), respectively.

The results recorded in Table 5 revealed that only 2 explants (25%) installed directly under the bench in the shade room survived after 15 days. Indeed, 3 days after the start of the ex vitro weaning, the leaves of most young plants begin to necrose. After 15 days of weaning, the survival rate for the young plants of batch 2 is equal to 50%. After partial opening of the plastic sheeting, the young plants behave well. After 30 days (Figure 4), the survival rate remained unchanged.

Lot 1: young plants kept directly in the shade in the greenhouse, and then on the bench. Lot 2: young plants kept in a mini-greenhouse, then under natural conditions.

Double disinfection with mercuric chloride and then with bleach was very beneficial for the disinfection of nodal explants, with a low infection rate i.e. 12.5%). These results are consistent with those obtained by [21] for Atriplexhalimus. According to [22], depending on the origin of the explants, two methods are conventionally used: disinfection with mercuric chloride, generally used for potentially infected material, and disinfection with calcium hypochlorite. Several researchers [23] [24] [25] [26] ) used the first method to disinfect explants of Jatrophacurcas originating from greenhouses, nurseries or fields or nodal explants of Tectonagrandis, a woody tree [27].

In addition, a high level of mercuric chloride at 1% on Atriplexhalimus micro-cuttings, taken from woody mother-plants aged 30 years, allowed to obtain a high survival rate of 85% [28] [29]. Ultimately, it emerges that the effectiveness of the disinfectant product, HgCl₂, would be linked to the inherent factors of the plant, namely the age, the species considered, the diameter and size of the explants and other parameters such as the optimal dose of the disinfectant and the immersion time. The optimum dose of mercuric chloride to be applied is also decisive to avoid the toxicity and the lethality of the disinfectant on the reactivity of the explants.

The base medium of [16] provided the best reactivity rates. In terms of the highest average number of shoots and the best elongation, the best results were obtained with hormone-free environments. These results show that these media allow better multiplication in this species. The medium of [16] is one of the most widely used media for the in vitro vegetative propagation of plant species, especially in numerous studies of the Annona genus. This is how it was used, e.g. by [30] [31] in A. Cherimola [32] in A. muricata [33], in the hybrid species, A. cherimoya cv (A. cherimola × A. squamosa) or [34] in A. glabra. [19] used this medium to in vitro multiply A. squamosa in accordance with the results obtained by [35]. On the other hand, Woody Plant Medium (WPM) [18] is an increasingly used medium for the in vitro vegetative propagation of fruit essences [22].

Overall, the nodal explants reacted better in media with hormones than in the control one. The MS medium supplemented with 2 mg·L⁻¹ of BAP led to the best reactivity rate (91.66%), the greatest average number of newly formed shoots (2.3) and the greatest number of nodes (5.3). The M14 medium [MS + BAP 2 mg·L⁻¹ + ANA 0.1 mg·L⁻¹] made it possible to have better elongation of the shoots (8.25 cm). Thus, BAP used alone had a better effect on explant bursting but rather had a detrimental effect on the elongation and the average number of nodes when the concentration is high (4 and 5 mg·L⁻¹). A concentration of 0.5 mg·L⁻¹ of NAA decreases considerably the multiplication rate, the length of shoots as well as the average number of nodes. Adding low concentrations of NAA to the culture medium increased significantly the number of newly formed shoots in A. cherimola [36], Atriplexhalimus L. [21] and Pisumsativum L. [37]. In Annonaglabra, the best average lengths were obtained by using the concentration of 0.5 mg·L^–1 of BAP while higher concentrations (1 and 1.5 mg·L^–1) of this phytohormone induced a reduction in shoot length [34]. According to [21], the addition of BAP improves the rate of bud break and the mean number of axillary shoots increases with decreasing BAP concentration. [38] reported that cytokinins begin to have an adverse effect on organogenesis when their concentration is greater than 2 mg·L^–1. Working on Atriplexhalimus, a sylvo-pastoral shrub species of the Chenopodiaceae family, [28] and [21] showed that BAP increases the propensity to bud in this species, as well as the presence of several buds in a leaf axilla. This phenomenon has also been observed in vitro in Acacia cyanophylla [39] and Annonasquamosa [40]. However, according to [41], high concentrations of cytokinins generally favor the production of large numbers of cuttings. Thus, the use of BAP concentrations at 2 and 4 mg·L^–1 has been shown to significantly increase the number of cuttings produced from explants of A. muricata [42]. Thus, in Annonaglabra, the low concentrations of BAP, less than 0.5 mg·L^–1, are insufficient to induce a significant increase in multiplication rates [34].

In some cases, the hormonal combination of BAP-NAA has been shown to be harmful to the in vitro organogenesis of plant species. This is the case with Solanumsessiliflorum in which the combination of BAP (0.25 and 0.5 mg·L⁻¹), and NAA (0.1 and 0.3 mg·L⁻¹) did not have a beneficial effect on in vitro multiplication according to [43]. Thus, some authors recommend the use, instead of NAA, of other auxins such as IAA or IBA [44] [45]. With Macluratinctoria [46] and Nothofagusalpina [47], adding NAA to the medium that already contained GA₃ combined with BAP did not improve the multiplication rate.

The traditional methods of vegetative propagation were inefficient, due to the low morphogenetic potential of A. senegalensis, and the low rooting rate of its conventional cuttings. The micropropagation technique can be applied successfully to this species and other Annona sp. to overcome these problems [48] [49] [50].

During the in vitro rooting stage, a rooting rate of 8.33% is obtained in the MS/2(0) expression medium without prior induction with auxins. This could be explained by a high endogenous auxin level in this species. [51] made the same observation with Parkiabiglobosa. It could also be explained by a drop in the pH in the medium due to a decrease in the concentration of sucrose. Indeed, according to [52], the level of sucrose in the culture medium 4 weeks after cultivation decreases from 20 mg·L^–1 to 5 mg·L^–1 while the medium becomes more acid with a pH decrease from 5.7 to 5.0. In many Annonaspecies such as A. glabra, this drop in pH allows a better response to root formation by increasing the percentage of rooting and the average number of roots produced by vitroplants [34]. However, a 5.0 pH of the medium has a negative effect on root length when the auxin concentration or the induction time is high, which would explain the decrease in root length as the induction time increases. IBA had a better effect on the rooting rate and on the average number of roots. Many authors have reported that an induction period of 3 to 5 days in the dark and then 7 to 10 days in full light in a medium containing a concentration of IBA of 50 to 100 mg·L⁻¹ with 2 - 2.5 g·L⁻¹ of activated charcoal and 20 mg·L⁻¹ of sucrose, followed by transfer to the MS/2(0) expression medium led to good rooting of shoots from mature material in many plants as well as for Annonaceaespecies. This is the case with Annonacherimola [48] [49] to attempt the plant breeding of this species and A. atemoya [30] [31] [32] [33]. This relative efficacy of IBA compared to other auxins has been observed in other fruit tree species such as Guava sp. [53]. Adversely, for A. muricata vitroplants, rooting was stimulated by galactose and NAA, after explants were preconditioned for two weeks, in a plain medium containing 10 g·L⁻¹ of activated charcoal [42].

The weaning of young plants by direct transfer to the atmosphere of the shade house resulted in high mortality. This result is explained by the very high relative humidity inside the culture tubes during in vitro culture. Indeed, under these conditions, the leaves of young plants have a high number of stomata compared to those of adult plants adapted to in vivo conditions [54] [55] [56] [57]. Adding to this are the combined effects of the presence of sucrose in the medium and the accumulation of ethylene, which negatively impact the development and closing capacity of the stomata once the young plants are subjected to natural conditions [1] [58] [59]. According to other authors, this high frequency of stomata and their poor closing capacity are the cause of the desiccation of young plants after their transfer to natural weaning conditions [56] [59] when relative humidity drops rapidly. [60] noticed this phenomenon in A. glabra.

On the other hand, gradual weaning results in fairly high survival rates. In fact, as has been observed for the juvenile material, keeping the young plants in a mini-greenhouse before their transfer under natural conditions makes it possible to maintain the explants under conditions close to those in vitro,i.e. very high relative humidity conditions, which lead to saturation of the atmosphere, hence to the protection of leaf surfaces of plants not yet fully adapted to natural conditions.