Coconut (Cocos nucifera L.) is native to the regions between 20oN and 200S of the Equator, where it plays a significant socioeconomic role in the local communities. There it is referred to as 'The Tree of Life', a eulogistic epithet describing its versatile use - more than 100 edible and non-edible products can be produced from it. Therefore the coconut palm is grown in about 90 tropical countries on more than 10 millions ha of land (Hamon et al., 1999).
Although coconut has a high local socioeconomic reputation, its production is experiencing many problems and consequently the area planted with this crop is declining. The conventional breeding approach using seed to replant land is very expensive due to the low production of seed for planting, and even when elite germplasm is available it takes decades to multiply up enough planting material for new areas (Adkins et al., 1999).
Hence over the past 40 years research has been directed towards developing a new technique for the micropropagation of coconut using somatic embryogenic approach. Throughout this time however one conclusion is repeatedly made -- coconut is very recalcitrant to somatic embryogenesis. And although the many obstacles to this are slowly being reduced, in order to successfully micropropagate coconut on a large scale bottlenecks in the protocol still exist, and those include inconsistency of the embryogenic response by explanted tissues, poor somatic embryo maturation and germination, low regeneration rate of the new plantlets and long time required to produce plants (1.5 years) (Samosir et al., 1998).
These bottlenecks and other problems were researched in the present study with the aim of trying to speed up the efficiency of coconut somatic embryogenesis process. Hence this thesis had the objectives to identify a starting protocol for coconut somatic embryogenesis; to select an appropriate for aim that explant; to optimize the production of embryogenic callus; to increase the rate of initiating coconut somatic embryos; to improve the maturation of somatic embryos and their germination efficiency; and to optimize the regeneration rate of the new plantlets.
In order to identify a starting protocol, preliminary work was conducted, where existing protocols for coconut somatic embryogenesis were compared in their efficiency to induce somatic embryos. The protocol that stood out as the best in producing most embryogenic callus and subsequently embryos, as well as having the least dead (in culture) explants, was that of Nikmatullah (2001). Therefore the latter was chosen to be used as a starting protocol for this study.
New sources of explants were investigated during the current work as well, using tissues from different parts of in vitro derived 8 months old coconut plantlets. Those however have shown to be unsuitable for somatic embryogenesis, since only non-embryogenic callus was developed by some of the inoculated tissues. The immature inflorescence explants were superior in producing embryogenic callus and somatic embryos; therefore they were selected as the preferred explant source to use in the next steps of the current study.
Optimizing the production of embryogenic callus was the first issue to address during the core work of this project. As a result of that the culture conditions were considerably improved by using vessels with larger headspace-medium ratio (3:1), as well as by selecting younger immature inflorescences and transversely segmenting the top half of the inflorescence spikes into smaller size (1 mm) sections. Further improvement was possible by studying the make up of the callus growth media. Amongst the administered for that purpose substances the applied together polyamines spermine (0.10 µ M) and putrescine (7.5 mM) have proven to play a notably positive role in the induction of callus from coconut immature inflorescence explants. Thidiazuron (TDZ, 10 µ M) too has shown a potential to improve the efficiency of the initial stage of coconut somatic embryogenesis, but only when applied in conjunction with other cytokinins (eg. BAP and 2iP). Smoke-saturated-water (SSW, 10 %) could only slightly diminish the amount of necrotising cultured explants, and high 2,4-D concentrations could not support the induction of callus from immature coconut inflorescences. Collectively taken, as a result of this current study the production of callus was improved by 300 %.
The rate of coconut somatic embryos formation was as well significantly increased (over 300 %), by the simultaneous application of suspension culture step, spermine (0.01 µ M), SSW (10 %) and high auxin concentration (500 µ M). Nevertheless the presence of TDZ and other cytokinins in the medium, as well as the absence of activated charcoal, were found to be unable to positively influence the somatic embryogenesis process.
Despite the considerable improvements made in the efficiency of inducing callus and initiating embryos, the poor maturation and germination (eg. 5 %, Verdeil et. al., 1999) of somatic embryos still remained a bottleneck to the whole somatic embryogenesis procedure. Therefore further work was conducted in that direction and discovered that embryo maturation and germination rate can be elevated to 55 % by administering ancymidol (30 µM) to the somatic embryo maturation medium. This plant retardant has exhibited here three potential modes of action towards the cultured coconut somatic embryos: a) as a promoter of somatic embryo maturation and germination; b) as a preventor of pre-germination death of the somatic embryos; and c) as a preserver of non-germinating somatic embryos, that still can possess the potential to germinate in the future.
The work during the next step of the process - regeneration of the new plantlets - has shown that the omission of plant growth regulators from the media was crucial for the development of germinated embryos into new plantlets, where otherwise no plant regeneration occurred at all. The achieved here plantlet regeneration rate in the PGR-free medim was 56 %, which is higher than the previously reported 20 % regeneration rate (Verdeil et al., 1994) for coconut plantlets produced from immature inflorescences explants.
As a result of this current work a new method was developed for somatic embryogenesis of coconut from immature inflorescences explants (Fig. 9.2). The overall efficiency of this protocol is over three times higher than that of the starting protocol (Nikmatullah, 2001) selected during the preliminary work. Furthermore, when using this new method the entire duration for regenerating clonal coconut plantlets (up to the stage of first root and shoot emerging) takes up to 8 months, which is the shortest reported time for producing coconut plantlets via somatic embryogenesis (eg. 36 months from inflorescences explants (Verdeil et. al., 1999) and 18 months from sliced zygotic explants (Samosir, 1999, Fig. 9.2), presenting an additional valuable advantage of this newly developed method, from the perspective of the potential to micropropagate coconut on a commercial scale.