Sources: Malaurie et al. (1998a)

Main edible species of yams are: 1) native of a continent, 2) and cultivated in the same continent, 3) or/and cultivated in an other. This observation implies very strong links to exchange problems. In Table 2, 36 countries have been observed by IBPGR in 1986 with Dioscorea germplasm. These countries are supposed to be concerned by an international exchange of yam germplasm. Some of them, have, in our knowledge, already developed in vitro germplasm collection.

Different genebank preservation levels exist. A first group concerns non aseptic germplasm conservation with in field genebank and seed genebank, where important disadvantages and heavy constraint of quarantine measures explain the choice of in vitro germplasm conservation (Hanson, 1986; Malaurie et al., 1998a) (Table 3).

Three levels of in vitro genebank preservation levels could be considered: 1) short term conservation: this conservation under normal growth conditions is suitable for temporary storage of germplasm collections, and for international distribution, 2) medium term conservation, which could be considered as an active in vitro genebank, 3) long term conservation, considered also as a base in vitro genebank.

These in vitro genebanks have been previously introduced in vitro from tuber or seed. These introductions have to be linked to an obligatory phytosanitary control from mother plants and from in vitro material after introduction. Medium term conservation, which correspond to in vitro culture under slow growth conditions, could be obtained by several ways: 1) physiological stage of the explant, 2) addition of osmotic agents and growth moderators, 3) low storage temperature, 4) low mineral or sucrose concentrations, 5) low oxygen pressure, 6) encapsulation in alginate (Charrier et al., 1991; Withers, 1991; Engelmann, 1991; Malaurie et al., 1998a).

At ORSTOM**, we choose to maintain the in vitro yam collection in a medium with low mineral nutrient and a low sucrose concentration. We succeeded in the introduction and maintenance of 14 species of yam (Malaurie et al., 1993). Since this time, this collection is continuously enriched by new genotypes and comprises 20 species (Table 4).

For yam, this simple solution of slow growth is used routinely and from these culture conditions a direct acclimatization is possible. This mode of conservation allows an international distribution of the material and corresponds to an active genebank (Malaurie et al. 1993,1998c, Malaurie and Trouslot 1995c).

This in vitro germplasm collection of yam is maintained in test tubes, at ORSTOM** (Montpellier, France), with a total of 6 test tubes by accession, with two different places of storage for the replicates ; the minimal growth conditions allow to maintain most of the accessions up to 2 years. Technical constraints in the collection management lead to subculture the accessions every 6-8 months (Malaurie et al., 1998c).

Different species maintained in the in vitro collection, such as D. cayenensis-D. rotundata complex are going to be enriched by cultivars from Burkina Faso for a sanitation program, and from Benin for a genetic program linkeds to virologic aspect. Others species supposed to be links to D. cayenensis-D. rotundata complex, such as D. mangenotiana, D. praehensilis, D. minutiflora, or D. abyssinica, D. praehensilis, are going to be enriched by further introduction.

Orstom** virologists are interested by D. trifida because of its strong sensibility to virus, which provoked in Guadeloupe, French West-Indies, its quite disappearance. Serological and molecular works are developed to explain this virus sensibility.

Long term conservation correspond to cryopreservation in liquid nitrogen, at -196 °C. Plant cryobiology, which begun in 1971 by Latta works on carrot cell suspension, benefited from results on animals cell by Polge et al in 1949. Since these dates, different techniques have been set up: 1) on one hand, the so-called conventional techniques, using two steps of slow freezing, with the addition of cryo-protector (Sakai, 1984), 2) and on the other hand, new techniques, characterized by a very rapid freezing, about 1000°C/ min, by direct immersion in liquid nitrogen (Table 5) (Dereuddre et al., 1990, 1991; Tannoury et al., 1991; Uragami, 1993). The aim of these techniques is to try to control water flow and ice formation, and tend to a vitrificated state, avoiding crystal formation during thawing, and to protect the cell from thermic shocks.

Most of the results about cryopreservation have been obtained from conventional techniques on suspension cells of medicinal yam, D. deltoidea being the most used ( Butenko et al., 1984; Popov and Fedorovskii, 1992; Popov and Volkova, 1994). More recent works have been done on rapid cryopreservation of callus (Chulafich et al., 1994), by direct immersion in liquid nitrogen, of two other medicinal yams (D. balcanica, D. caucasica).

Since 1996, new results have been obtained by two different research teams, using encapsulation/dehydration of shoot apices. On the one hand, Mandal et al. (1996) compared the survival capacities of apices after the osmotic and thermic stress of the technique of four species of yam - three edible (D. alata, D. bulbifera, D. wallichii), and one medicinal (D. floribunda). Four of them have survived after immersion in liquid nitrogen, with 26 to 71%, depending on the species. Meanwhile, only two of them (D. alata, D. wallichii) allowed the recovery into shoots after immersion in liquid nitrogen, with 21 and 37%, respectively.

On the other hand, ORSTOM** ability in different aspects of the long term conservation on tropical plants (Engelmann, 1991), and on encapsulation/dehydration technique applied on coffee, cassava, oil palm...etc, permitted to apply the process on apical shoot-tips of in vitro plantlets of yam (Malaurie and Trouslot 1996). Malaurie et al. (1998b) obtained survival rates over 50% for the two species (D. alata, D. bulbifera), and recovery to rooted leafy shoots after immersion in liquid nitrogen of at least 60% for D. bulbifera and 20% for D. alata, three months culture after thawing.

Comparatively to previous works on cryopreservation using encapsulation/dehydration technique, Malaurie et al. (1998b) have used higher sucrose concentration (0.9, 1.0 and 1.1M), a wider range of dehydration duration, up to 23h and a new and more accurate method for measuring of dry weight.

The new and more accurate method for measuring of dry weight used in our experiments consisted of desiccating alginate beads for 30d in airtight boxes containing dry silica gel, to avoid mass loss due to caramelization of sugar when drying at a temperature higher than 100°C. We obtained a strong linear correlation between dry mass (DW₃₀) and sucrose molarity for sucrose-pretreated alginate beads. During the whole experiment, we used DW₃₀ values estimated by linear regression (Table 6).

(1)Source: Malaurie et al. (1998b)
(2)From mean values over 13 to 15 replicates for each of the four sucrose concentrations
(y= 6.4319 + 29.872x; N= 4; r= 0.999). Similar results were obtained from replicate data
(y= 6.4177 + 29.883x; N= 55; r= 0.960). Data not shown.

For the best sucrose pretreatments depending species, Figure 1(source: Malaurie et al., 1998b) shows that D. bulbifera still has high survival with high sucrose concentration and after long duration dehydration (up to 23h). For the two species, the water content of encapsulated apices had to be decreased down to 0.15g H2O g-1 DW in order to obtain high survival after freezing. The percentage of water loss was of 67, 62, 58 and 55% FW (± 1%) for 0.75, 0.9, 1 and 1.1M sucrose pretreatments, respectively. Our results demonstrated that, in most cases, survival increased when dehydration was extended to a defined threshold, around 0.13-0.15g. H2O g-1 DW, which was obtained after desiccation periods from 10 to 18h. It seemed that, with this soft dessication process, we could rub out differences in residual water-free, which certainly exist between apices from a same plot.

In vitro germplasm conservation presents different advantages such as: 1) to be free from genetic erosion, 2) to have the possibility for the establishment of core collection with long term genebanks, 3) to be free from fungis and bacteria, 4) to be not expensive, when in vitro facilities are already present, 5) easy and convenient for international distribution. But International exchanges need more for safe international exchange. We need to know the plant material on genetic level, and over all on the phytosanitary level. On the phytosanitary level, various viruses have been described on edible and medicinal yams on their production area. Different works, depending virus and virus group, are reported (Table 7). Indexing techniques allow to highlight a certain number of viruses on yam: Poty, potex, badna and cucumo-viruses, where yam mosaic virus (YMV) provokes the most important loss.

¹ YMMV: is it a new potyvirus or a strain of the YMV ?
² D. trifida virus and DGBMV have been shown as YMV strains (Porth et al.,1987)
³ DaRMV should be a 'yam strain' of the beet mosaic potyvirus transmissible on N. benthamiana (Porth et al.,1987)
⁴ DaV is serologically links toYMV but differ by is non-transmissibility (Porth et al.,1987)

During the establishment of the yam in vitro germplasm collection, in the biotechnology laboratory of ORSTOM** (Malaurie et al., 1993), afterwards IIRSDA - Adiopodoumé research station, near Abidjan, Ivory Coast - clones were systematically indexed by ELISA directed to YMV, when introduced in vitro (Malaurie and Thouvenel, 1988; Malaurie et al., 1988a,b; Charrier and Hamon, 1991).

Later on, one indexation was carried out by a virologist team on the duplicate of the yam in vitro germplasm collection enriched by introduction of new genotypes. 92 samples, belonging to several yam species, were used for the indexation : D. alata, D. bulbifera, D. cayenensis- rotundata complex, D. dumetorum, D. esculenta, D. mangenotiana, D. praehensilis, D. shimperiana, D. togoensis, D. trifida. These samples were originating from various geographic areas: Africa, Caribbean, South America and Asia. Four viruses were fetched by ELISA technique: PVX (potato virus X, potexvirus) PVY (potato virus Y, potyvirus), CMV (cucumber mosaic cucumovirus), and YMV (Urbino et al. 1998). Results are presented in Table 8.

The use of in vitro techniques allows to be free from fungis, bacteria, and other pests. Only viruses could be present on the plant and have to be eradicated. Different techniques exist and are already applied on yam. There are meristem culture, thermotherapy and/ or chemotherapy (36°C during 1 to 2 weeks on in vivo or in vitro plants, use of chemicals such as vidarabine, ribavirin and 2-thiouracil). They could be used alone or associated (Table 9).

Success in meristem culture depends on the size and location of the explant excised, and on the growth regulator ratio. 'Meristem culture', on Table 9, concerns works using meristem-tips (0.2-0.5 mm long) as well as shoot-tips (0.6-2.5 mm long). Experiments on viability and in vitro morphological development of meristem-tips of two sizes, 'small' (0.3-0.5 mm) and 'large' (0.6-0.8 mm), have shown that it was better to use large meristem size to increase the shoot elongation percentage. The use of axillary or apical meristems did not induce difference and should allow an important yield in micropropagating such material from excised meristem-tips. Eleven months after meristem excision, production of plantlets was observed with a rate of 82% and 39% from the survivors, for a clone of D. cayenensis-D. rotundata complex and D. praehensilis genotype, respectively (Malaurie et al., 1995a,b).

Meristem cultures have been done on 8 clones of 5 Dioscorea species belonging to the in vitro germplasm collection. Morphological development has been observed and data were recorded 60 days after meristem inoculation. In our case, the production of rooted leafy shoots, 60 days after meristem inoculation, occurred in five clones out of eight, with percentage shoot leaf production of 5 to 26 %, depending on the clone. Six months later, the excised meristems of all clones developed into rooted leafy shoots, where D. bulbifera, and D. dumetorum was not, to our knowledge, mentioned in the literature (Table 10).