Reproductive success and compatibility among accessions of Jatropha curcas in Indonesia

Large scale exploitation of Jatropha curcas for biofuel production is restrained by low productivity hence economically not protable. One of the main constraints is due to limited number of female owers in an inorescence which eventually develop into fruits that bear seed. Investigation on oral biology and reproductive potential of the species had been carried out, but seems to vary among sites. The research was conducted to study reproductive success and the compatibility among Indonesian Jatropha accessions. The research was set up in two experiments. The rst experiment was conducted at the Jatropha seed orchard, Pakuwon, Sukabumi, West Java from April to July 2007 using four-year- old trees of accessions from Lampung, Banten, West Java, and Central Java. The second experiment was conducted at Leuwikopo Experimental Station of Bogor Agricultural University from April to July 2008 using one-year-old trees of accessions from Lampung, Bengkulu, Palembang, and Kediri. Flowers were obtained from controlled self and cross-pollinated as well as left for open-pollination. The result showed that reproductive success varied among accessions; the greatest success was 0.73 obtained from aWest Java accession and the least was 0.53 from Banten accession. The Jatropha accessions were categorized as partially self-incompatible as indicated by index of self-incompatibility (ISI) that ranged 0.93- 0.99. Banten accession produced highly viable and vigorous seeds regardless of pollination methods; accessions of Central Java produced higher viability seeds when cross-pollinated within accession, whereas those from West Java and Lampung produced higher viability seeds when self-pollinated. Keywords: accessions, Jatropha curcas , selfpollination, self-incompatibility, seed viability


Introduction
The need of renewable energy has encouraged many parties to nd sources of biofuel. is Jatropha curcas considered as one of the potential species not only for biofuel production but also biogas and compost (Openshaw, 2000, Tiwari et al., 2007, Achten et al., 2008).D species rought tolerant can survive in marginal land thus avoid competition with food crops which can only grow in fertile land.However, large scale exploitation is restraint to low productivity due to the low number of female owers, i.e. averaged 10 female owers per in orescence (Adikarsih and Hartono, 2007).Moreover a long-term systematic breeding program has not been set up which make the species is still considered as wild and exhibits variability in productivity among individuals (Achten et al., 2008).
The biodiesel development program in Indonesia has been set up to grow plantations consisting of 1.5 million ha of oil palm, 1.5 million ha of , 0.75 million ha Jatropha of sugar cane and 1.5 million ha of cassava.The development of 1.5 million ha of plantations Jatropha during 2007-2010 required about 900 ton of seeds (Hasnam, 2007) or about 0.6 kg.ha .Seed procurement -1 is also facing several limitations that contributed to low seed viability such as time of harvest in relation to fruit maturity, seed processing and drying (Sudjindro, 2008).
The project was started with the exploration of Jatropha germplasm in nine provinces which consisted of 54 districts and 200 000 stem cuttings and seeds had been collected from these areas.The collected germplasm were then grouped into 12 accessions, i.e.West Sumatera, Lampung, Banten, West Java, Central Java, East Java, West Nusa Tenggara, East Nusa Tenggara, South Sulawesi, Gorontalo and Maluku.Variations in plant morphology including leaf thickness, leaf shape and size, leaf color, pedicel length, fruit shape and size, seed weight and size, number of seed per capsule were found among the accessions (Hasnam, 2007).

Reproductive success and compatibility among accessions of Jatropha curcas in Indonesia.
Endah R. Palupi *, Ade S. Ahmad , Ro q Affandi , Abdul Qadir , and Eni Randriani planting.The second generation of the improvedpopulation (IP-2) had a higher productivity of about six to seven t.ha .yearat the fourth year after planting.The -1 -1 third generation (IP-3) was expected to have productivity of about eight to ten t.ha .yearafter the -1 -1 fourth year.After the third generation of improvedpopulation a long-term breeding plan will be required to increase productivity further, using either conventional or molecular techniques in order to become a Jatropha more economically attractive (Heliyanto et al., 2008).
Conventional technique for plant improvement through controlled pollination or hybridization requires information on owering phenology and pollination system as well as compatibility among accessions.Hartati (2008) studied the owering phenology of Jatropha in Pakuwon, Sukabumi during April-October and reported that pollination was best conducted in either late morning (when the stigma has completely splayed), or afternoon on the rst day of anthesis, although the pistillate (female) ower lasts for several days.This research was aimed at studying owering biology, reproductive success of several accessions and the compatibility among accessions of in Jatropha Indonesia.

Materials And Methods
The research was set up in two experiments.The rst experiment was conducted at the Seed Jatropha Orchard, Pakuwon, Sukabumi, West Java, about 450 m above sea level, with Latosol soil and type B climate (Oldeman) from April to July 2007.Four-year-old plants from four accessions were used i.e. accessions from Lampung, Banten, West Java, and Central Java with 20 plants from each accession.The second experiment was conducted at Leuwikopo Experimental Station, Bogor Agricultural University, about 250 m above sea level, from April to July 2008.One-year-old plants of accessions from Lampung, Bengkulu, Palembang, and Kediri (from stem cuttings) taken from the seed orchard in Pakuwon were used with 20 plants from each accession.
Type of in orescence, ratio of male to female owers, as well as time and duration of blooming were observed.Reproductive success, which was de ned as the number of ovules that develop into viable seeds (Wiens et al., 1987) was observed in open-pollinated population during the rst and second experiment which was started with the number of in orescence per plant and number of female or hermaphroditic ower per in orescence (Fl/In ).Fruit set (Fr/Fl) was calculated as the ratio of female/hermaphroditic owers that developed into mature fruits in an in orescence, whereas number of ovule per ower (O/Fl) was counted as averaged of 20 owers per accession, which resulted in three ovules per ower.Therefore seed set (Sd/O) was calculated as the proportion of the three ovules that developed into viable seeds in a fruit.Reproductive success was calculated from Fr/Fl ratio and Sd/O ratio.
The owers were tagged with color threads when once opened.The number and time of ower or fruits abortion were observed every alternate day.The experiment was arranged in a completely randomized design and analyzed using Duncan's multiple range tests (DMRT) to compare the reproductive success among accessions.
The compatibility among accessions in the rst experiment (four-year-old plants) was recorded from four types of hand controlled-pollinations, i.e.: 1. Intra-self-pollination (SP-intra): pistillate owers were pollinated with pollen from staminate owers of the same in orescence.
2. Inter-self-pollination (SP-inter): pistillate owers were pollinated with pollen from staminate owers of different in orescence of the same plant.
3. Intra-cross-pollination (CP-intra): pistillate owers were pollinated with pollen from staminate owers of other plants within the same accession.
4. Inter-cross-pollination (CP-inter): pistillate owers were pollinated with pollen from staminate owers of other accession.Accession from South Sulawesi was used as pollen donor for the four selected accessions because it had high staminate ower production.
All pistillate owers were bagged at 06.00-07.00hrs in the morning and pollinated during 10.00-12.00hrs.
Based on the results from the rst experiment self-and cross-pollination of the second experiment (on oneyear-old plants) were arranged as follows: 1. Self-pollination: pistillate owers were pollinated with pollen from staminate owers of the same or different in orescence of the same plant.
2. Cross-pollination: pistillate owers were pollinated with pollen from staminate owers of other accession.IP-1P was used as pollen donor for the four selected accessions.
The observation on each accession included fruit set from self-and cross-pollination, index of self-incompatibility. Seed quality was determined based on germination percentage, and germination speed was analyzed according to ISTA (2007).The data was analyzed using split-plot design, with accession as the main plot and pollination type as the sub-plots.

RESULTS AND DISCUSSION
The rst in orescence of the one-year-old plants was commonly unisexual.Staminate in orescences were observed in Lampung and Kediri accessions whereas pistillate in orescences in Bengkulu and Palembang accessions.The subsequent in orescences in all plants were bisexual.In most cases staminate and pistillate owers were born on the same in orescence (monoecious) and rarely staminate and hermaphroditic (andromonoecious) (Figure 1).
Most of the plants from Central Java produced staminate and hermaphroditic owers.The ower sex was indicated by the size of the ower bud; the hermaphroditic ower buds was the largest followed by pistillate and staminate ower buds.The staminate owers of the four-year-old plants usually bloomed around six days earlier than the pistillate owers and owering continued for several days.The pistillate owers bloomed only for two to ve days and the owering period coincided with that of staminate owers.On the other hand, the pistillate owers of the one-year-old plants bloomed 2 days earlier than the staminate owers and last for 2-6 days, followed by the staminate owers for several days.In this case only late blooming pistillate owers coincided with the blooming staminate owers.Raju and Ezradanam (2003) reported that is a cross pollinating species, Jatropha therefore different blooming time of male and female owers within an in orescence is bene cial to prevent sel ng.Battacharya et al. ( 2005) added that spp.Apis were the most frequent visitor of owers at the Jatropha same time they deposited pollen grains on the stigmatic surface, thus generating cross-pollination.1).The four-year-old plants were planted in the eld with 2 x 2 m spacing; while the one-year-old plants were planted in polybags.It is possible that the higher light intensity in the eld led to the formation of more male owers.Investigation on environmental factors affecting sex ratio will be valuable to increase production of female owers which could lead to an increase in fruit yield.
Accession from Banten produced the greatest number of in orescence whereas that of West Java had the fewest.The number of pistillate or hermaphroditic ower per in orescence as well as the number of fruits per in orescence was not signi cantly different among accessions, ranged 3.5-5.33 in the four year-old and ranged 6.3-9.1 in the one-year-old plants (Table 1).Again, the data showed that the performance of the oneyear-old was better than the four-year-old plants.However, the fruit to ower ratio (fruit set) from both plants was similar (Table 1), ranged 0.62-0.86indicating that the number of pistillate or hermaphroditic ower per in orescence was a major yield component and at the same time the limiting factor of kernel production.
The data implied that increasing the number of pistillate ower per in orescence would be the key to increase the productivity.Not all pistillate or hermaphroditic owers can develop into mature fruits.Bhattacharya et al (2005) reported that fruit set from pistillate owers was only 50%, due to fruit abortion during development.In this research, abortion at owering stage, i.e. owers that aborted within 4 days after anthesis, was around 5% in the four-year-old and 1% in the one-year-old plants.The abortion during fruit and seed development ranged 9-33%.
The average number of seeds per fruit in both experiments ranged 2.27-2.57and 2.66-2.80although on average three ovules were found in one ovary.Hermaphroditic owers in this experiment can produce up to four seeds per fruit.The seed to ovule ratio (seed set) was 0.76-0.85 in the four-year-old plants and 0.86-0.92 in the one-year-old plants, and did not vary among accessions.The data indicated that seed set was reasonably high although not all ovules developed into viable seeds.Insuf cient pollination could be the cause of the less than maximum seed set and could be overcome by increasing the population of the insect pollinators.Social bees, i.e. and Apis cerana Apis dorsata were reported as effective pollinators for Jatropha in Indonesia (Rianti et al., 2010), hence could be employed to increase seed production.In Nigeria, however, where the most effective insect Jatropha pollinators were not social bees, i.e.Chrysomya chloropyga Eristalis tenax (Diptera-Calliphoriade) and (Diptera-Syrphidae) (Alamu et al., 2013), introduction and maintenance of the insect population could be more dif cult.Data on reproductive success showed similar range from the four-year-old and the one-year-old plants, between 0.53-0.74,meaning that from 100 ovules produced, only 53-74 developed into viable seeds.This data indicated that seed production was lower than the yield potential; therefore cultural techniques need to be explored to increase fruit and seed set.
Abortion at the owering stage following controlled pollination was low, i.e. < 3% in cross-pollination and self-pollination during both experiments.However, abortion at the fruit development stage was as high as 32% in the four-year-old and 20% in the one-year-old plants.Fruit abortion in self-pollination was similar to that in cross-pollination.The data indicated that the abortion was not due to incompatibility as fruit set from self-pollination was not signi cantly different from that of Four-year-old plants Table 2. Fruit set and seed set from self and cross-pollination Note: Seed set was calculated from the number of sound seed divided by the number of total ovule assuming that 1 there were three ovules per ower ISI (index of self-incompatibility): ratio of fruit set from self-pollination to cross-pollination cross-pollination (Table 2).Fruit abortions occurred at different stages of development, but mostly were during the early stages when the size was less than half of the mature fruit.Observation on the causes of the abortion would be useful to nd measures to increase kernel production.
Fruit set and seed set of the four pollination types in the four-year-old plants (SP-intra, SP-inter, CP-intra and CP-inter) were not signi cantly different, so the data was presented as self-and cross-pollination (Table 2).Fruit set did not vary among accessions either in the rst or second experiments.It was surprising that cross pollination did not result in higher fruit and seed set than self-pollination, which indicated that was not Jatropha an obligate cross pollinated species.These results were in contrast to that reported by Raju and Ezradanam (2002) in India that fruit set from hand cross-pollination was higher (96%) than that of hand self-pollination (77%).Kaur et al (2011) also reported that in the central region of Punjab the fruit set from cross-pollination was higher (93.2%) than that of self-pollination (72.2%).The average index of self-incompatibility in this research was 0.99 and 0.93 in the four-year-old and one-year-old plants, respectively, so is categorized as a Jatropha partially self-incompatible species, supporting the research result of Wang and Ding (2012) that J. curcas has both self-pollination and cross-pollination systems.This is contradictory to Heller (1996) who stated that Jatropha was a cross-pollinating species.Given the data it is possible that the seeds produced in any Jatropha plants in Indonesia were the results of selfpollination with ants as the effective pollinator ( et Luo al. 2012).
The germination percentage of the self-pollination seeds was not signi cantly different from the crosspollination seeds, supporting the idea that was Jatropha not an obligate cross pollinating species.The four-yearold accession from Banten produced highly viable seeds from both self-and cross-pollination.Lampung and Central Java accessions produced highly viable seeds (> 80%) from self-pollination, whereas cross pollinated seeds from West Java accession had higher viability than self-pollinated (Table 3).
Slightly different results were obtained from the oneyear-old plants in which cross-pollination produced better quality seeds than self-pollination as indicated by germination rate of > 80% and germination speed of ≥ 9.46 %.day (Table 4).In general Banten, Bengkulu and -1 Kediri accessions produced highly viable seeds regardless of pollination systems.It would be interesting to learn the reproductive success of the self-and crosspollinated progenies.A study in Malaysia showed positive heterosis from crosses of particular superior parents (Islam et al, 2011) which could potentially lead to the development of hybrid varieties of Jatropha curcas.

15
Reproductive success and compatibility among accessions of Jatropha..........The germination speed, which is an indication of seed vigor, showed similar trends to germination percentage in which self-and cross-pollination resulted in nonsigni cantly different rates.The one-year-old plants tend to demonstrate higher germination speed than the four-year-old plants.Banten accession germinated at a faster rate (Table 4) regardless of pollination types.Selfpollinated seeds from West Java were slower to germinate compared to cross-pollinated ones.These data were in line with data on germination percentage that cross pollination between accessions could give negative effect on seed viability.In general, although Banten accession had the lowest reproductive success, hence seed production, this accession produced high quality seeds, which is an important character for a conventional breeding program.

Conclusion
Reproductive success varied among accessions, the highest was from West Java of 0.73 and the lowest from Banten of 0.53.Jatropha sccessions (Lampung, Banten, West Java, Central Java, Bengkulu, Palembang and Kediri) are partly self-incompatible with self-incompatibility index range of 0.93-0.99.Pollen source did not affect fruit set and seed set, but affected seed germination and speed of germination.

Figure 1 .
Figure 1. in orescence: (A) male (red arrow) Jatropha and female (yellow arrow) owers (B) male (red arrow) and hermaphroditic (yellow arrow) owersThe number of in orescence of the four-year-old plants varied among accessions and was fewer compared to that of the one-year-old plants.However, the

Table 1 .
Reproductive characters and reproductive success of each accession (from open pollination)