Application of unfertilized ovary culture to identify elite mother palms of Cocos nucifera L . with regenerative potential †

Un-fertilized ovary is a promising explant for micropropagation of coconut. However, due to high heterozygocity in coconut, in vitro performance varies greatly among palms. In this study, callogenesis, somatic embryogenesis and plant regeneration potential of 35 Sri Lanka Tall (SLT) and 20 Dwarf x Tall hybrid (DxT) elite palms were assessed to select suitable mother palms for clonal propagation. Callusing was induced from ovaries excised from immature thidiazuron and activated charcoal at 0.1 %. The induced calli were proliferated in two cycles. Somatic embryos were induced 2-isopentyladenine. Germination of somatic embryos and the development of shoots occurred in Y 3 of gibberellic acid. Ovaries of SLT and DxT palms initiated callusing by 74.0 and 65.0 %, respectively. Percentage of explants that developed callus varied from 4 to 50 % in SLT and 9 to 57 % in DxT. Calli of DxT did not grow further to regenerate somatic embryos. However, SLT calli proliferated in two cycles and regenerated a mean of 12.5 well-developed embryogenic structures per ovary. The maximum welldeveloped embryogenic structures obtained were 70 in few ovaries. Twelve out of 35 SLT palms regenerated shoots with The results showed that not only individual palms but also individual explants respond differently to in vitro culture. The

The high potential of the un-fertilized ovary explants for cloning of coconut was reported recently (Perera et al., 2007).Plant regeneration from tissues in the un-fertilized ovary occurs through indirect somatic embryogenesis as reported for all other explants of coconut (Karunaratne & Periyapperuma, 1989;Fernando, 2001).This explant of Sri Lanka Tall palms gave a callusing frequency of 76 % and a plant regeneration frequency of 70 % (Perera et al., 2009).The study has reported that un-fertilized ovary is the most suitable explant for cloning coconut due to its somatic nature and high regeneration potential (Perera et al un-fertilized ovaries can be excised from the palm with minimal damage to the palm. Generally, the potential for in vitro regeneration shows strong genotype dependence in many plant species (Tomas & Smith, 1985;Cheliak & Klimazenska, 1995;Zhang & Yakahata, 2001) and past experience shows that coconut is not an exception (Fernando, 2001;Weerakoon, 2008).Due to the high heterozygocity, each coconut palm can be considered as an individual genotype.Thus, in vitro studying responses of ovary explants collected from a large number of elite mother palms.Therefore, this study was undertaken with the objective of applying the protocol developed by Perera et al. (2009), in identifying palms with high amenability to clonal propagation among elite mother palms of two commercially important cultivars [Sri Lanka Tall (SLT) and Dwarf x Tall hybrid (DxT)] (Vidhanaarachchi et al., 2010) .

Experimental material
old) high yielding mother palms showing a vigorous growth were selected for the study.The age difference between palms of the two cultivars could not be avoided due to limited availability of palms.The youngest open months was considered to be at maturity stage -4.Female -4 maturity stage (Perera et al., 2009).

Un-fertilized ovary culture
12 min and rinsed 5 times with sterile distilled water.The ovaries were dissected under stereo binocular microscope and the crushed explants were placed on callus induction medium, which consisted of basal CRI 72 (Karunaratne & Periyapperuma, 1989), supplemented et al., 2009).The number of ovaries that could be collected per palm varied from 6 to 71.Initial calli were further proliferated by dissection of embryogenic structures under stereo binocular microscope and subculturing into fresh callus induction medium (without TDZ).After 8 wks, the embryogenic structures that developed on callus same medium) for the second cycle.
After the 2 cycles, calli were transferred to regeneration medium as described by Perera et al. (2009), which (BAP).Well-developed embryogenic structures were individually transferred to CRI 72 medium supplemented were transferred to the same basal medium without plant growth regulators for 4 wks and to the basal medium somatic embryos were maintained in Y 3 medium acid (GA 3 ) by subculturing on to fresh media at 6 wk intervals until they were germinated and shoots were well-developed.
All culture media were supplemented with 0.1 % (w/v) activated charcoal (BDH 33033) and the cultures were maintained at 28 ± 1 °C in the dark until somatic embryos converted into shoots.The shoots were maintained at 28 °C under 16 h photoperiod (PAR: 25 -2 s -1 ).

Data collection and analysis
From all the initiated cultures, the number of ovary explants that initiated callus and the number of explants that produced embryogenic callus were recorded after 10 wks of incubation.Number of embryogenic structures proliferation, number of well-developed embryogenic structures that were transferred to regeneration medium after the second cycle of callus proliferation and number of shoots regenerated per palm were recorded.
The following calculations were done for each palm to identify the most responsive individuals.
callus/total number of ovary explants cultured x 100 Mean number of embryogenic callus per ovary (cycle 1) developed into callus/number of ovaries producing embryogenic callus Mean number of well-developed embryogenic structures well-developed embryogenic structures transferred to regeneration medium/ number of ovaries producing embryogenic callus

Statistical analysis
The variation in callusing frequency and the mean number of embryogenic callus per ovary (cycle 1) between the 2 cultivars was compared using the t-test.

Callogenesis
The crushed explants enlarged after one week in culture.
The signs of callusing were visible to the naked eye only after four weeks.After ten weeks, translucent calli appeared on some explants while the others became enlarged or turned necrotic.Ovaries from 74 % palms of SLT and 65 % palms of DxT produced calli.The callusing frequency per palm varied from 3.6 to 50.0 % and from 8.6 to 57.1 % with a mean of 13.8 and 19.0 % in ovary explants of SLT and DxT palms, respectively.The mean callusing frequency between the two cultivars was The embryogenic potential of an explant may be affected by several factors like the genotype, environmental conditions of mother palm, mother plant maturity, explant maturity etc., even under optimal culture conditions (Hornung, 1995b ;Fernando, 2001).In the from SLT and DxT palms varied from 40 to 55 cm.The (6 -71).Dissected ovaries were approximately 1 -3 mm in diameter and showed a gradient in development even These variations could also affect the uniformity of ovary explants resulting in differential responses to culture conditions applied in this study.Genotypic effects may also play a role in palm to palm variation in callusing frequency within a cultivar.These results are comparable callusing (Vidhanaarachchi & Weerakoon, 1997) and oil in explants collected from different mother palms.
Callus that developed from different ovaries also showed a variation in size and morphology.Ovaryderived calli of the cultivar DxT were smaller, developed fewer embryogenic structures and were more compact and less translucent when compared to that of SLT (Table 1).

Callus proliferation
Generally, coconut explants produce highly compact calli in small quantities.Previous attempts to increase the quantity of coconut calli were unsuccessful as cell suspensions could not be established due to the compact nature of the callus.Even though crushed callus developed into embryogenic callus, it did not show a fast growth (Fernando, 2001).Perez et al. (2006) introduced and somatic embryos induced from plumules could be repeatedly subcultured to achieve a large number of somatic embryos as high as 100,000 per explant.Perera et al. (2009) was able to increase the growth of coconut ovary-derived callus by the application of this method.However, the maximum number of embryogenic callus clumps possible was 25 per original ovary callus and this depended on the quality and the size of the original callus induced from these explants.In the present study, ovaries of 54 and 45 % SLT and DxT palms, respectively produced translucent nodular calli with embryogenic structures (Figure 1a), which were suitable for somatic embryo regeneration.In general, calli from DxT explants were smaller in size compared to those of SLT explants.Thus, the number of embryogenic embryogenic structures derived from DxT ovaries did not grow satisfactorily and developed a mean of 0.6 embryogenic calli per ovary, whereas those developed from SLT proliferated well and showed a mean of 2.8 embryogenic calli per ovary.At the end of the second cycle, a mean of 12.5 well-developed embryogenic structures per ovary per SLT palm regenerated and could be transferred to regeneration medium (Figure 1b).In a few individual ovaries, highly convoluted, translucent callus developed and they regenerated a large number of well-developed embryogenic structures.The highest number of well-developed embryogenic structures obtained per ovary explant (at the end of the second cycle) was 70.
The results clearly showed a difference in the ability among ovary explants to develop fast-growing callus that could regenerate somatic embryos, which may be due to genotypic effect or differences in stage of maturity of ovaries.

Somatic embryogenesis and plant regeneration
As in other plant species, the endogenous auxin concentration is one of the determining factors for the induction of somatic embryogenesis in coconut (Blake, 1990).Gradual reduction in auxin concentration with a corresponding increase in cytokinin level permits initiation and further development of somatic embryos.
Initially, ovary explants were subjected to a very transferring well-developed translucent embryogenic regenerated in the periphery.When these were transferred to hormone free medium, the nodular structures enlarged and turned opaque.Once they were they became globular and somatic embryos developed (Figure 1c).Embryo maturation culminated with the formation of elongated somatic embryos.These mature somatic embryos germinated when transferred to Y 3 medium (Eeuwens, 1976) with GA 3 .By maintaining them in the same medium, healthy shoots developed (Figures 1d & e).
Ovaries of 34 % of SLT palms under the study (12 palms did not develop callus from which shoots could be recovered.The number of shoots regenerated per the highly variable response of individual palms (Table 2).The shoot regeneration potential of several palms was higher than that was achieved by Perera et al. (2009).This again may be due to the genotypic differences of palms used in the two studies.Further, use of a higher cytokinin level and an increased callus production due to two cycles of callus proliferation before regeneration also may have affected higher somatic embryo regeneration.It is also worth reporting that a higher number of producing highly convoluted, fast-growing callus showing higher proliferation.Such callus proliferation might be a collective effect of genotype of the individual palm and optimum maturity stage of the ovary.
When compared to the performance of SLT palms, the poor performance of DxT may be due to the use of older mother palms or the use of culture conditions previously developed for SLT ovary culture.

CONCLUSION
good source of explants for clonal propagation of elite coconut palms.Initial callus derived from ovaries of SLT palms could be successfully proliferated in two cycles to increase the quantity.A variable response of genotypes on palm response at the stage of callus proliferation and plant regeneration could be observed.The SLT palms were more amenable to clonal propagation using the tested protocol.As a result of the present study, twelve high-yielding SLT palms with in vitro regeneration conditions, cultivar DxT did not perform well.Therefore,

Table 2 :
Variation in shoot regeneration potential of SLT palms