For reprint orders, please contact                                                            ...
Review  Achten, Nielsen, Aerts et al.Key terms                                                                    Species ...
Towards domestication of Jatropha curcas  Reviewout outside the natural range of the species [23–25] . The      Table 2. F...
Review  Achten, Nielsen, Aerts et al.Table 3. List of insects observed to visit Jatropha flowers and corresponding forage ...
Towards domestication of Jatropha curcas  Reviewgenetic differentiation in Jatropha populations deserves          breeding...
Review  Achten, Nielsen, Aerts et al.Key term                                   ƒƒ Quantitative variation within        pr...
Towards domestication of Jatropha curcas  Reviewthat plants from some provenances in Mexico con-                  ten of t...
Review  Achten, Nielsen, Aerts et al.     Conventional breeding                                          mechanism behind ...
Table 5. 24 microsatellite primer pairs in nontoxic and toxic varieties of Jatropha with indication of polymorphism.      ...
Towards Domestication of Jatropha Curcas
Towards Domestication of Jatropha Curcas
Towards Domestication of Jatropha Curcas
Towards Domestication of Jatropha Curcas
Towards Domestication of Jatropha Curcas
Towards Domestication of Jatropha Curcas
Towards Domestication of Jatropha Curcas
Towards Domestication of Jatropha Curcas
Upcoming SlideShare
Loading in …5

Towards Domestication of Jatropha Curcas


Published on

Towards Domestication of Jatropha Curcas

Published in: Education, Technology, Business
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Towards Domestication of Jatropha Curcas

  1. 1. For reprint orders, please contact ReviewTowards domestication of Jatropha curcasBiofuels (2010) 1(1), 91–107 Wouter MJ Achten1, Lene R Nielsen2, Raf Aerts1, Ard G Lengkeek3, Erik D Kjær2, Antonio Trabucco1, Jon K Hansen2, Wouter H Maes1, Lars Graudal2, Festus K Akinnifesi4 & Bart Muys1† Jatropha curcas L. attracts a lot of interest as a biofuel crop, triggering large investments and rapid expansion of cultivation areas, and yet, it should still be considered as a (semi-)wild, undomesticated plant. To use the full potential of Jatropha and to support further expansion and systematic selection, breeding and domestication are a prerequisite. This review reveals and identifies gaps in knowledge that still impede domestication of Jatropha. Prebreeding knowledge is limited. In particular, the regeneration ecology and the degree of genetic diversity among and within natural populations in and outside the center of origin are poorly studied. There is only a limited understanding of the Jatropha breeding system and the effect of inbreeding and outbreeding. This review presents all currently available and relevant information on the species distribution, site requirements, regeneration ecology, genetic diversity, advances in selection, development of varieties and hybridization. It also describes possible routes to a better Jatropha germplasm, gives recommendations for tackling current problems and provides guidance for future research. We also discuss the participatory domestication strategy of Jatropha integration in agroforestry.Jatropha curcas L. (further referred to as Jatropha) is a fuel production, poverty reduction and large returns onrapidly emerging biofuel crop currently attracting a lot investments [11] . However, many of these claims are yet toof interest and investments [1] . This stem-succulent [2] be supported by scientific evidence [4,12] . Major knowledgedeciduous tree or shrub produces seeds rich in toxic oil gaps concerning basic ecological and agronomic properties(27–40% [3]). The oil can be extracted easily with tech- (growth conditions, input responsiveness of biomass produc-niques of different sophistication levels [4,5] . The crude tion, seed yield and the species’ genetics), make seed yieldJatropha oil meets the fuel quality standards of rape- poorly predictable [4,10] . Considering the current expan-seed [6] and can be easily converted into biodiesel, meet- sion [1] , this situation might hold considerable sustainabil-ing US and European standards [7,8] . However, although ity risks (economic, social and environmental) [3] . Amongthe downstream processing is well known, the species’ other issues, the water requirement [2] and water footprintagronomy and, as such, the potential seed production, is (amount of water needed per GJ biodiesel) [13,14] of Jatrophastill shrouded in uncertainty. are still poorly understood. A better knowledge of these Popular claims on drought tolerance, low nutri- agronomic properties is vital for the further application ofent requirement, pest and disease resistance and high the species. Jatropha can still be considered a (semi-)wild,yields  [9] have triggered a Jatropha hype [4,10] with sky- undomesticated plant showing considerable performancehigh expectations on simultaneous wasteland reclamation, variability [4,10,11] .† Author for correspondence1 Division Forest Nature and Landscape, K.U.Leuven, Celestijnenlaan 200E 2411, BE-3001 Leuven, Belgium; Tel.: +32 1632 9721; Fax: +32 1632 9760;E-mail:; bart.muys@ees.kuleuven.be2 Division Forest Genetic Resources, Forest & Landscape Denmark, Hørsholm Kongevej 11, DK-2970 Hørsholm, Denmark; Tel.: + 45 3533 1500;Fax: +45 3533 15173 The Tree Domestication Team, Agrobusiness park 76, 6708 PW Wageningen, The Netherlands; Tel.: +31 31 742 12414 World Agroforestry Centre (ICRAF), Southern Africa Regional office, Chitedze Agricultural Research Station, PO Box 30798, Lilongwe, Malawi;Tel.: +26 51 707 328; Fax: +26 51 707 323 future science group 10.4155/BFS.09.4 © 2010 Future Science Ltd ISSN 1759-7269 91
  2. 2. Review  Achten, Nielsen, Aerts et al.Key terms Species distribution & site requirements In order to reduce the risk of futureJatropha curcas L.: Perennial stem unsustainable practices and to improve The delineation of the original area of species distribu-succulent shrub of the spurge family, future crop performance, furthertion of Jatropha has been the subject of a long debate [17]the seeds of which contain an inedible selection, breeding and domestication but, currently, there is a growing agreement that theoil that can be converted into biodiesel of Jatropha is primordial. However, Jatropha center of origin is Mexico and continentalBiodiesel: Vegetable oil or animal fats substantial prebreeding knowledge is Central America [17–19] .converted to (m)ethylesters as analternative fuel for diesel engines important to facilitate and guide an The Portuguese learned about Jatropha’s medicinal effective and robust route towards its properties in the 16th Century, and later establishedGrowth conditions: Conditions underwhich a species can naturally occur domestication [15] . Knowledge about commercial plantations for soap and lamp oil produc- the degree of genetic diversity among tion on the Cape Verdian Islands and Guinea Bissau [17] .Breeding: Changing the genetics ofplants or animals for the benefit of and within natural populations in Later, Jatropha genotypes adopted in Western Africahumankind (e.g., through selection or and outside the center of origin is were spread across other Portuguese colonies in Africamolecular techniques) required to gain the first ideas about (Mozambique, Angola) and into Asia (India, China andGenetic diversity: Total number of where to find potentially valuable Indonesia) [17] . Jatropha now grows pantropically, fromgenetic characteristics in the genetic genetic m­ terial. Further knowledge a Brazil to the tropical islands of Fiji [201] .makeup of a species is needed on the reproductive biol-The climatic conditions of locations where theSpecies distribution: The spatial ogy of Jatropha, including its phenol- species is found under natural conditions within the(uniform, random or clumped) andtemporal arrangement of a species center of origin are given in Table 1. Jatropha distribu- ogy, mating patterns in populations,throughout its range tion occurs naturally under an annual precipitation possible pollinators and the breedingRegeneration ecology: Collective ranging between 944 and 3121 mm, and a length of system, in order to design suitablename of flowering, fruiting and other growing season (the number of months in which the breeding and deployment strategies.reproduction characteristics of a species mean precipitation is higher than half of the potential Finally, a screening of origins across the distribution area for key agro- evapotranspiration) between 5 and 11 months [20] . nomic traits will help in choosing the basis for furtherIn its natural area of distribution, the species is most work and support the initial selection efforts. abundant in tropical savanna and monsoon climates At present, no major systematic exploration and evalua- (A m and Aw climate types according to Köppen clas- tion of genetic resources has been published and only little sification [21]) and in temperate climates without a dry and scattered knowledge is available on the basic repro- season and with a hot summer (Cfa [21] ), while it is ductive biology of the species. Also, limited information uncommon in semiarid climates (Bs [21]) and totally has been published on quantitative genetic variation, such absent in arid climates (Bw [21]) [20] . The Cape Verdian as heritability, and genetic variance components, geno- Islands, where Jatropha was first successfully intro- type by environment interaction, germplasm pathways duced and grown for commercial use, have a typically and juvenile mature correlation. Such genetic variables are Sahelian climate, with a strong oceanic influence [22] . of key importance for conducting breeding programs [16] . Mean annual vertical rainfall varies between 100 mm The lack of knowledge will limit planting programs from on the coastal flats to 900 mm at the highest eleva- using the full potential of Jatropha. Establishment and tions, with strong support of oceanic humidity (unre- public sharing of prebreeding knowledge is therefore corded by standard pluviometers) and mild annual important for effective domestication of Jatropha. and daily temperature changes. In its distribution of The objective of this review is to synthesize relevant herbarium specimen locations recorded in Eastern information on the ecological attributes and breed- Africa, where it was probably introduced using gen- ing potential required for germplasm improvement, otypes from Cape Verde, Jatropha is established in u­ ilization and domestication of Jatropha. t locations with annual precipitation ranging between 650 and 2500 mm, and length of grow-Table 1. Mean, optimal range (25–75% percentiles) and total range (5–95% percentiles) ing season between 4 and 12 months.of climate variables for the locations of Jatropha specimens found in the area of natural These latter results were derived bydistribution of Jatropha (n = 241). replicating the published methodol- ogy for Jatropha s­ ecimen locations in pStatistic Tmean (°C) Tmin (°C) Tmax (°C) Pa (mm/year) LGS (# months) AI (/) Eastern Africa [20] .Mean 24.4 16.5 32.5 1689 7.3 1.04Optimal range 23.4–26.2 14.4–19.4 31.5–34.0 1207–2001 6–9 0.73–1.19 Regeneration ecologyTotal range 19.3–27.2 10.5–21.2 27.4–35.7 994–3121 5–11 0.55–1.99 Jatropha is monoecious – with male andAI: Aridity index: the ration of mean annual precipitation to total potential evapotranspiration; LGS: Length of growing female flowers on the same plant and inseason; Pa: Annual precipitation; Tmax: Maximum daily temperature of the warmest month; Tmean: Mean annualtemperature; Tmin: Minimum daily temperature of the coldest month. the same inflorescence [23,24] . Studies onData from [20]. pollination biology have been carried92 Biofuels (2010) 1(1) future science group
  3. 3. Towards domestication of Jatropha curcas  Reviewout outside the natural range of the species [23–25] . The Table 2. Flower and pollen characteristics of Jatropha curcas.size of the raceme inflorescences, with dichasial cymepattern, can vary considerably (5–9.5 cm in length and Male per Female per Male:female Total P P:O Ref. inflorescence inflorescence per male4.5–12.5 cm in diameter) and with this variable size,the number of flowers also varies (Table 2) . Note that, 25–93 1–5 29:1 655 6332:1 [23]although the number of male and female flowers per 17–105 2–19 1617 ± 100 539:1 [25]inflorescence varies a lot among the observations [23] 49–238 0–17 24.5:1 1597–5763 13,015–46,968:1 [24] O: Ovule; P: Pollen.(male: 25–238 and female: 1–19), the male-to-femaleflower ratio of the two studies reporting it is similar observations, Chang-wei et al. conclude that this mass(24.5:1 [24] and 29:1 [23]). opening mechanism of female flowers can promote out- Male flowers are small and plate shaped [23] . The crossing and minimize self-pollination [24] . Based onfive sepals (5  mm long) and five petals (7  mm) are their estimation of the outcrossing index (OCI = 4) [26] ,free [24] . The latter are connivent at the flower base, Chang-wei et al. categorize Jatropha as an out-crosserforming a short tube [23] . Ten diadelphous stamens are that is self-compatible and needs pollinators [24] .arranged in two tiers of five. The lower tier is free, while The adhesiveness of the pollen, the smoothness ofthe upper tier is united [23] . The anthers are yellow, the stigma of 1.62 mm in diameter and a pollen flowapproximately 2 mm, dithecous and dorsifixed. Five by wind of 2.8 grains cm-2 make wind pollinationoval-shaped glands are present at the villous flower base. almost impossible [24] . The bright yellow anther color,The anthers dehisce 1 h after flower opening by longitu- male flowers opening evenly spread over the inflores-dinal slits. The pollen produced are globular and their cence life span, the fragrance and nectar availabilitysize ranges from 52 to 89 µm (Table 2) [23,24] . The floral of both male and female flowers and the large pollenbase c­ ntains 0.3 µl nectar. o (52–89 µm), with many verrucae on their exine with Female flowers show a similar shape as the male flow- adhesion, suggest insect pollination to be the major pol-ers  [23,24] . The flowers contain three styles and bifid lination method [24] . Observed insect visitors mainlystigmas. The ovary has three carpels, each with a single belong to the order of the Hymenoptera (Table 3) . Rajulocule producing one ovule. The floral base is villous and Ezradanam observed that of the total foraging vis-and contains five yellow elliptical glands under the its made by insects on male flowers, bees contributedovary  [23,24] . The stigmas are receptive during 3 days 34%, ants 61% and flies 5% [23] . On female flowers,after the opening of the flower. The flower base secretes bees made up 28%, ants 70% and flies 2% of the total.nectar [23] . Although Raju & Ezradanam report that Bhattacharya et al. observed the major abundance inthe female flowers secrete the same quantities of nectar the Apis genus (honey bee; 71%) [25] .as the male flowers  [23] , Bhattacharya et al. observeda higher nectar production in female flowers (4.54 ± Genetic resources, diversity & characterization of0.82 µl in 1200 h) than male flowers (1.92 ± 0.44 µl in breeding systems in Jatropha curcas1200 h) [25] . The reported pollen ovule ratios are given ƒƒ Breeding, inbreeding & the potential of releasingin Table 2. vigor & productivity through smart outcrossing Research on the existence of a system to promote Inbreeding depression in tree species is the process byout-crossing and minimize self-pollination in Jatropha which self- or related matings lead to homozygosityis scarce. Temporal dio­ cism is often seen in monoe- e and the accumulation of deleterious mutations [27–29] .cious plants with unisexual flowers, but observations of Inbreeding depression reduces individual fitness, sur-opening periods of male and female flowers that over- vival and growth variables [30] , and raises the possibilitieslap rejects this temporal dioecism hypothesis  [23,24] . of population and/or species extinction [31–33] . The nega-However, Chang-wei et al. described the opening tive effects of inbreeding in trees are well documentedsequence of the flowers  [24] . They observed that male and include embryo abortion, limited fruit set, reducedflowers start opening from the first or second day of overall seed yield and lower germination rate [27] .the inflorescence life (13–19 days). Although the male Furthermore, selfed or inbred progeny can sufferflowers open evenly spread over this period, a small from lower seedling vigor and poor growth form, andpeak could be observed in the open-to-total male-flow- end up being less productive when they reach matu-ers ratio during the ninth and 13th day. The pollen rity  [34–38] . Inbreeding depression is worsened by theviability reduces considerably from the second or third large variations in fecundity often observed in tree spe-day after dispersal. The opening of female flowers is cies  [39,40] . This phenomenon, in which a small num-more concentrated between the third and fifth day. ber of trees contribute disproportionately to the seedIn this period, 60% of the total female flowers open. crop, can result in the effective population size of a treeFemale flowers remain open for 3 days. Based on these (Ne) – the size of an ‘idealized’ population that would future science group 93
  4. 4. Review  Achten, Nielsen, Aerts et al.Table 3. List of insects observed to visit Jatropha flowers and corresponding forage type Jatropha is not only able to reproducethey collect . via selfing, but also through apomixis (without sexual reproduction)  [24] .Order Family Genus Species Forage type Ref. However, the issue with contrast-Coleoptera [24,25] ing traits apparently present in theDiptera Calliphoridae Chrysomya megacephala Nectar [23,24] breeding system seems not to beHemiptera [24] fully clarified (Table 4) . PreliminaryHymenoptera Apidae Apis florae, indica, dorsata, Nectar and [23–25] studies indicate very low variation in mellifera, cerana pollen microsatellite simple sequence repeat Anthophoridae Ceratina simillima Nectar and pollen [23] (SSR) markers within populations of Eumenidae Eumenes conica Nectar [25] Jatropha even in its natural distribu- Formicidae Camponotus compressus spp. Nectar [23] tion (Mexico) [47] . Pamidimarri et al. Crematogaster spp. Nectar [23] applied SSR, amplified fragment- Solenopsis geminata Nectar [23] length polymorphism (AFLP) and Pheidole spathifer Nectar [23] random amplification of polymor- Halictidae [23] phic DNA markers to discriminate Vespidae Vespa spp. Nectar [25] between two Mexican accessions ofLepidoptera Pieridae Catopsilia pomona Nectar [24] Jatropha (one toxic and one non-Thysanoptera Thripidae Scirothrips dorsalis Nectar and pollen [23] toxic) [48] . Although they could dis- Trips hawaiiensis Nectar and pollen [23] criminate between the accessions,Adapted from [23]. they found no variation between individuals within each accession. have the same genetic properties as that observed for This could be an indication of a population structure a real population – being much lower than the census with a high level of homozygosity as well. size, and lower than that required to maintain heterozy- It is important to confirm or deconfirm the hypoth- gosity and productivity [41,42] . Ne is also lowered if the esis of nonpanmictic breeding, because understanding reproductive connectivity between trees in a landscape the breeding pattern is central for design of domesti- is weak. Connectivity depends on the density and even- cation strategies. Breeding, large-scale mass propaga- ness of distribution of sexually mature individuals in tion and distribution across landscapes will obviously the landscape and, if a species relies on animal pollin­ be much easier if the species is reproducing by natural ators and/or seed dispersers, on the presence of these selfing without inbreeding depression or, especially, if agents to facilitate gene flow [43] . it reproduces by apomixis [49] . As explained above, Jatropha can set seed after both insect and self-pollination. However, self-pollinated fruits ƒƒ Genetic diversity between populations are lighter in general [44] and aborted before maturation The genome size of Jatropha is fairly small (2C DNA in 25% of cases [23] . This could be due to early acting content of 0.850  ±  0.006  pg or C DNA content of inbreeding depression [45,46] and, thus, may reflect a high 0.416 × 109 base pairs) compared with other members of natural outcrossing rate. Chang-wei et al. suggested that Euphorbiaceae [50–52] . The level of genetic diversity and Table 4. Overview of life history traits with ecological and genetic importance. Trait and alternate state Remarks Breeding system Sexual Apomixis? Apomixis unusual Outcrossing Self-fertilization Pollination by insects Self-incompatibility? Self-compatibility Monoecy/dioecy Hermaphroditism Age at maturity Precocity Delayed First fruits at the age of (2–)4–5 years Seed crop variation Masting Nonmasting? Seeds: ƒƒ Size Large Small Seeds ellipsoid, 1–2 cm long ƒƒ Dispersal Far Near? ƒƒ Dormancy Dormant Nondormant Orthodox storage Senescence Senescent Immortal Lifespan of 30–50 years Traits of Jatropha curcas discussed in the literature are printed bold. Data from [23,24,44,50,51].94 Biofuels (2010) 1(1) future science group
  5. 5. Towards domestication of Jatropha curcas  Reviewgenetic differentiation in Jatropha populations deserves breeding system (see flower mor- Key termspecial attention due to its introduction history as an phology above). Both the small Agroforestry: Collective name for landexotic species in many countries. In such a situation, population sizes during introduction use systems and technologies, whereplant populations may result in a complex genetic history, history and the losses of introduced woody perennials are deliberately used on the same land unit as agriculturalincluding several potential genetic bottlenecks  [53,54] . genotypes due to imbalanced clonal crops and/or animals, either in someGiven the successive introductions of Jatropha and its propagation by farmers may have led form of spatial arrangement orability of clonal mass propagation within a short time, to purging of recessive, deleterious temporal sequenceit is possible that all African and/or Asian populations alleles. This could have counteractedresult from a narrow germplasm origin [51,55] . Genetic inbreeding depression  [65] . Owing to the lack of know­bottlenecks resulting from such founder effects are also ledge, it is indeed possible that the Jatropha landraces haveknown in other important  (a­groforestry) crops, most reduced growth, due to imbedded inbreeding depression.notably coffee and banana [56–58] . ‘One-off’ introduc- Given the low genomic diversity in Jatropha landra-tions are of particular concern if they are already of ces, we believe that ‘smart’ outcrossing between superiornarrow genetic base, and/or represent low quality or Asian individuals with new introductions from Americaspoorly adapted material. Many tropical trees in farm should be performed. Such crosses should release anylandscapes also demonstrate both extremely low densi- inbreeding depression and thereby increase vigor andties and highly aggregated distributions, which – even fruit production if genetic diversity of American landra-if long-distance pollen transfer is possible – will reduce ces is effectively larger. The introduction of genetic vari-effective population sizes and promote inbreeding [27] . ability can be performed by intra­ pecific and/or interspe- sData on other tree species that propagate vegetatively cific crossing. Actually, interspecific crossing experimentsshow comparable concerns of genetic erosion [59] . Species have successfully been carried out between Jatropha cur-that are reproduced vegetatively are vulnerable for clone cas L. and Jatropha gossypifolia L., Jatropha glanduliferalosses, unless new clones are introduced or old clones Roxb., Jatropha integerrima Jacq., Jatropha multifida L.,redistributed. Generally, a certain number of individual Jatropha villosa (Forssk.) Müll. Arg. and Jatropha mahesh-clones respond more successfully to propagation and warii Subram. & M.P.Nayar [66] . F1, F2 and F3 hybrids,simple mathematical simulation models show that, after as well as various back-crossings, are available and shouldsome generations, only a few clones may d­ minate an o be evaluated with respect to seed production, oil yieldarea [57] . and performance when grown under different climatic Recent studies based on genetic markers uncov- conditions [67] . The results from such studies will provideered surprisingly low levels of genetic diversity in important insight in the next few years. Random amplifi-Jatropha landraces from China [47] and only modest cation of polymorphic DNA markers and AFLP markerslevels of diversity in India [60,61] , indicating that the are available for interspecific hybrids identification [68] .gene pool applied at a large scale may rest on a fairlyfragile genetic foundation. Tatikonda et al. studied the Experience with testing of seed sources &diversity of 48 accessions from India based on AFLP development of varietiesmarkers and found 680 polymorphic fragments, which Until now, little information has been published on theprovided discriminative power for the classification production variability between provenances of Jatropha.of germplasm accessions into five major clusters [62] . In a study with 13 provenances and landraces tested inThere are limited published data characterizing the Senegal (two sites) and Cape Verde (two sites), significantgenetic variation in the African gene pool. Basha et al. differences of vegetative growth were found at all sites.included accessions from Egypt and Uganda, which At one of the test sites at Cape Verde, provenances werein general clustered closer to the Asian landraces com- significantly different concerning the number and weightpared with the Mexican accessions [63] , although the of capsules and the number and weight of seeds per shrubemerging pattern was not completely clear. Regarding 25.3 months after planting [17] . Genotype by environmentthe level of genetic variation within African popula- interactions between sites were significant in Senegal, buttions, preliminary results based on SSR markers and not in Cape Verde [17] . Ginwal et al. compared plantsAFLP markers indicate surprisingly low levels of genetic from ten Indian landraces after 6–24 months, and founddiversity in landraces from Mali, Kenya and Tanzania large significant variations, attributing more than 80%[Nielsen LR et al., Unpublished Data] . Correlations of growth of the total phenotypic variance to seed sources [69] . Thisand oil production with DNA polymorphism have not is a relatively high level of genetic variability comparedyet been investigated. with what is usually found in tree species of tropical dry Species with naturally high levels of inbreeding (‘self- zones  [70,71] , and is especially surprising given the lim-ers’) are expected to show less inbreeding depression [64] ited level of variation observable at the DNA level, asbut, at present, very little is known about Jatropha’s discussed previously. future science group 95
  6. 6. Review  Achten, Nielsen, Aerts et al.Key term ƒƒ Quantitative variation within products [27] . However, published information on thisSeed sources: Provenance of provenances & landraces aspect of Jatropha is very scarce. To our knowledge,planting material Studies of the quantitative genetic there are no studies on genetic material over a range variation within populations are of environments published for Jatropha, other than the sparse in Jatropha. Phenotypic studies are several, but Senegal and Cape Verde tests reported by Heller [17] . they are often made for seed properties of genotypes This lack stresses the importance of establishing genetic standing at different sites, making it impossible to sepa- tests over a range of environments, in order to reveal rate the effect of the genotype from the effect of the possible genotype by environment interactions (prov- site [69,72] . Oil content assessments have shown a pheno- enance trials), and to determine suitable seed collection typic range from 28 to 39% and 100 seed weight from zones for superior germplasms. Actually, genotype by 49.2 to 64.9 g in accessions from Indian landraces [72] . environment interaction in Jatropha is expected from Similar results were found in another study with Indian investigations on other crop species of Euphorbiaceae landraces, where the oil content ranged from 29.9 to (e.g., Ricinus communis L. [74,75]). 37.1% and the 100 seed weight from 57 to 79 g in can- didate plus trees (plants yielding more than 2 kg of dry ƒƒ Toxic & nontoxic Jatropha: a potential key trait in seed per plant and with ages above 5 years) [18] . However, breeding programs in both cases, these estimates were influenced by the An important aspect of domestication of Jatropha is the site of the candidate trees. In an experiment with cut- toxicity of the plant [48] . Consumption of Jatropha seeds tings from 29 candidate trees evaluated after 34 months, may result in various symptoms, including vomiting the broad-sense heritability was high (ranging between and diarrhea, and has proven lethal in animal experi- 0.63 and 0.88) for height, number of branches, number ments [76–78] . The most problematic toxic components of flowers, ratio between female and male flowers, days in Jatropha are probably a number of phorbol esters that, from initiation of flowering to fruiting, days from fruit- in general, are found present in high concentrations in ing to maturity and seed yield per plant [18] . The genetic the seeds [79–81] . Phorbol esters are compounds known correlation of plant height with seed yield per plant and to cause severe biological effects, including inflam- female-to-male flower ratio was 0.36 and -0.23, respec- mation and tumor promotion [82,83] . Removal of the tively, and the genetic correlation between the female- phorbol esters during processing is possible [84] , but this to-male flower ratio and seed yield per plant was 0.48. is not an easily deployed process and the presence of The number of branches was moderately correlated with possibly toxic phorbol ester degradation products after seed yield per plant (genetic correlation of 0.61). The treatment cannot be ruled out [81] . genetic correlation between number of flowers and seed From the user’s perspective, the toxic phorbol esters yield was 0.29 and the yield per plant was moderately provide a potential health risk for workers and limit the correlated (0.32) with days from fruiting to maturity [18] . use of the protein-rich press cake that otherwise would The low genetic correlation of seed yield with the height be suitable for animal diets [77,83] . Still, phorbol esters and number of flowers suggests that the possibility of may protect the plants against pests, and will, there- increasing seed yield through selection for these two fore, be important when testing if phorbol ester-free traits is poor. Slightly better opportunities to increase Jatropha plants are more susceptible to damage from seed yield are present in the case of (indirect) selections pests. To our knowledge, no such experiments on the aiming at an increase in female-to-male flower ratio or insect-resistance of nontoxic versus toxic plants exist, number of branches. and it is, therefore, advisable to include such studies in High broad-sense heritability and a small phenotypic any breeding program that aims at removing the toxic variation was found for plant height and root collar phorbol esters from the phenotype. The fate of degra- diameter in ten accessions [73] . dation of phorbol esters in soils is also an important aspect, because residuals from Jatropha oil productions ƒƒ Genotype by environment interactions (e.g., the press cake) are often applied as fertilizers [4] . Jatropha is cultivated under variable conditions, and Although it is claimed that the phorbol esters of the one aspect of domestication is the degree to which dif- press cake decompose completely within 6 days after ferent genotypes perform best under different growth application to the soil [85] , it is still uncertain whether conditions [16] . Another aspect of domestication is the crops fertilized by the Jatropha press cake (e.g., horti- performance of each provenance under different con- culture and cereals) can take up phorbol esters during ditions. In general, a wide genetic base is essential to that period [4] . prevent inbreeding depression and allow for adapta- For these reasons (e.g., human health, soil and seed tion to changing environmental conditions (e.g., cli- cake use), breeding for nontoxic Jatropha provides inter- mate change) and to altering markets for tree species’ esting prospects. In this context, it is highly interesting96 Biofuels (2010) 1(1) future science group
  7. 7. Towards domestication of Jatropha curcas  Reviewthat plants from some provenances in Mexico con- ten of these are directly involved in breeding [90] . Thistain very low or nondetectable levels of phorbol list is by no means exhaustive and an extensive webesters  [63,79,86–89] . The presence of naturally occurring search reveals thousands of web pages that refer to com-plants with low levels of phorbol esters is very interesting panies and organizations involved in improvement ofin a domestication context, because it makes it likely Jatropha cultivars, some even for subtropical areas [202] .that plant material without phorbol esters can easily Still, surprisingly limited peer-reviewed documenta-be developed without the use of advanced molecular tion of superiority of specific planting material basedbreeding or transgenic modification. However, intro- on progeny/clonal testing is available at present.ducing nontoxic material may raise new complications, Both decentralized, low-input breeding approachesas nontoxic and toxic Jatropha are morphologically based on farmland seed sources as more centralizedalike. Additionally, close proximity of toxic and non- germplasm procurement systems can be considered [91] .toxic Jatropha can t­ igger unexpected traits through r We think that farmers’ involvement in testing, procure-cross-pollination. ment and deployment of improved planting stock will Jatropha seeds also contain a trypsin inhibitor, lectins be important, in order to ensure large-scale access toand phytate that are antinutritional factors in relation to improved planting material. Still, given the limitationsa potential use of the press cake for animal feed [79] . The in genetic knowledge on Jatropha as reviewed earlier,trypsin inhibitor and lectins can be removed by heat a number of steps seem to be required on the route totreatment and phytate is neutralized by the addition of better germplasm domestication programs. Below, wemicrobial phytase [88] . list a set of relevant steps that will, in our opinion, be of key relevance for low-input local breeding, as well asRoute to better germplasm for more intensive breeding initiatives:The few published results from field testing with Jatropha ƒƒ Exploration of the genetic resources of global l­ ndraces asuggest quantitative genetic control of a number of traits compared with the Central American gene pool;of importance for the cultivation of the species [18,73] .Experiences from a number of dry-zone woody species ƒƒ Analysis of the breeding system of Jatropha by use ofhave, in general, revealed very large genetic differentia- molecular markers;tion, in terms of production and suitability for growth ƒƒ Determination of the effect of inbreeding andunder different climatic conditions [70] . The a priori outbreeding;expectation is, therefore, that development of improvedplanting material, through testing and deployment of ƒƒ Exploration of the genetic mechanism behind thegenetically improved sources of reproductive material, expression of toxicity by quantitative trait locus map-will be an effective tool for enhanced production. ping and to evaluate the genetics of toxicity and oil Fortunately, Jatropha is suitable for quick and effi- production (i.e., type and degree of inheritance;cient domestication compared with other woody species. dominance/codominance of toxicity, epigenetics);This is due to a number of important features regarding ƒƒ Estimation of the potential to improve biomass, espe-production of oil, its propagation, establishment and cially oil production, traits of importance for mecha-adaptation of the plants: nized harvesting, disease resistance and drought tol-ƒƒ Short generation turnover. Recurrent selection can be erance of toxic and nontoxic Jatropha when grown performed in short intervals; under different climatic conditions;ƒƒ Easy to propagate clonally; ƒƒ Development of breeding and seed-transfer guidelinesƒƒ Native gene pools still exist; based on climate and soil variables, taking different farming systems and farmer preferences into account;ƒƒ Techniques and experience with controlled crosses are available; ƒƒ Association studies based on cosegregation betweenƒƒ Precocity (short juvenile phase), indicating that produc- genotypes and molecular markers with high resolu- tivity can be assessed early, once the trees start to fruit; tion (e.g., SSR, ALFP and single nucleotide polymor- phism markers) and exploring their use for effective Publically available information suggests that a marker-aided selection, either for direct deploymentnumber of different commercial and public entities are of identified and selected clones/seed or for inclusioninvolved in the development of improved germplasm. in decentralized multiple breeding populations;Carels lists 22 organizations from eight countriesas major international organizations involved in the ƒƒ Application in farmers’ fields and governmental orgenetic improvement of Jatropha [90] . Carels notes that private intense plantations. future science group 97
  8. 8. Review  Achten, Nielsen, Aerts et al. Conventional breeding mechanism behind toxicity. By localizing the locus (or The most important traits for the selection of candi- loci) responsible for (non)toxicity in the respective par- date plus phenotypes of Jatropha are seed yield, seed size ent with molecular markers, it will be possible to use and oil yields [92] . In the case of mechanical harvesting, the markers for future use in breeding. Additionally, dwarfing of stocks, branching suppression and uniform established clonal field trials can serve as material for ripening will also become important [93] . By setting up association studies based on cosegregation between a seed-based breeding approach – based on breeding genotypes (e.g., clones with high oil content and Indian of offspring from provenances and open-pollinated Andhra Pradesh accessions [100]) and molecular markers families – several other breeding objectives could be as, for instance, microsatellites (SSRs), AFLPs and/or achieved. Apart from obvious yield criteria, it is impor- inter-SSR markers. Still, more segregation studies are tant that such a breeding approach develops planting needed to resolve the mode of heritance of toxicity and material adapted to local environmental conditions. its molecular background, as such understanding will Clonal propagation is highly efficient to obtain high guide the design of multiple breeding programs. genetic gains [94,95] , since all genetic effects (additive The application of genetic markers (e.g., polymor- as well as nonadditive effects) are transferred by the phic microsatellites [Table 5]) also makes it possible to propagation and since selections based on clonal tests carry out earlier selections and, thus, reduce the time are much more precise, due to the fact that all the genetic between the recurrent selections and increase the genetic background of a genotype is tested. J. curcas is easily gains per year. Finally, the use of the markers will also cloned from cuttings. Pretreatment of spring branch help increasing breeding efficiency. The ability to reveal cuttings with auxins (IAA, IBA and NAA) and thia- genetic markers associated with certain traits depends mine stimulates rooting and sprouting [96,97] . However, on the size of the material, sufficient polymorphic rooted cuttings do not develop a taproot (in contrast m­ rkers and precise estimates of genetic values. a to seedlings and are probably more prone to drought Transgenic approaches may also prove important and wind  [98] . This might limit the use of cuttings to in the future [19,93] . Jatropha can be transformed based areas with irrigation. In this context, the indication on Agrobacterium tumefaciens infection [101,102] . The of apomixis in Jatropha [24] is encouraging, as it may approach can target removal of toxicity and anti­ utrient n enable breeders to grow genetically identical clones as components, but work has also been carried out on the in cuttings, but with taproots. biosynthesis of fatty oils [103] . If the improved value High yielding varieties developed in a clone-based of byproducts as livestock feed is a primary target for breeding approach can be deployed as clones directly, breeding, reduction of antinutrient factors can poten- or subsequently as components in clonal seed orchards tially be an interesting goal for molecular breeding, as (in cases where deployment of seedlings are preferred). suggested in other crops [104] . This step could also include the propagation of high- producing varieties that are genetically transformed to Future perspective remove toxicity, for example [93] . The current public and private interest in Jatropha has triggered large-scale investments and expansion of its ƒƒ Molecular breeding plantations [1] . As such, genetic improvement, selection As toxicity in terms of phorbol esters seems to be and domestication initiatives currently focus on improv- expressed qualitatively [99] , it may be regulated by only ing Jatropha for its performance in large-scale mono­ one or a few genes. The inheritance of toxicity is not culture plantations. However, monoculture expansion settled, but Sujathha et al. suggest that the phenotype holds risks of unsustainable practice that will not be of the mother tree is passed on to the seed (i.e., non- solved by domestication (e.g., land use conflicts, land toxic mothers give nontoxic seed and toxic mothers right conflicts and loss of ecosystem services) [105] . give toxic seeds, independent of the phenotype of the Furthermore, there are specific problems with Jatropha father) [99] . Maternal inheritance (of chloroplast-specific monocultures, ranging from unknown best practices, markers) was also observed in natural and artificial to pest and disease management and to unsure eco- intraspecific hybrids  [66] . This pattern of inheritance nomic viability. Several of these general and specific could be explained in different ways, including hypoth- problems could be overcome by following a pathway of eses of apomixes or involvement of suppressor genes. community-based small-scale Jatropha initiatives (e.g., Potentially, the nontoxic phenotype may be driven by agroforestry systems) in specific areas [106] . As such, we a single suppressor gene that inhibits the production of believe that integrating Jatropha into existing small- all phorbol esters when present. Established F1 hybrids holder agroforestry systems, such as hedge and bound- between toxic and nontoxic Jatropha and first-genera- ary planting, can form a robust and sustainable base tion backcrossing could be used to detect the genetic for Jatropha domestication and expansion [105] , and can98 Biofuels (2010) 1(1) future science group
  9. 9. Table 5. 24 microsatellite primer pairs in nontoxic and toxic varieties of Jatropha with indication of polymorphism. SSR primer and Primer sequence (5´–3´) Ta (°C) Repeat motif Expected allele size (bp) Polymorphic NCBI GenBank ID Nontoxic Toxic/not specified Nontoxic Toxic/not specified jcds10 (EU586340) F:CATCAAATGCTAATGAAAGTACA; 46.5 (TG)6CACGCA(TG)4 108/122 108/122 +future science group R:CACACCTAGCAAACTACTTGCA jcds24 (EU586341) F:GGATATGAAGTTTCATGGGACAAG; 51.0 (CA)5(TA)8(CA)4… 204/210 204/216 X X R:TTCATTGAATGGATGGTTGTAAGG (TA)3GA(TA)4 jcds41 (EU586342) F: AACACACCATGGGCCACAGGT; 56.5 (CA)6(TA)2 102/114 102/114 + R:TGCATGTGTGCGGGTTTGATTAC jcds58 (EU586343) F:TCCATGAAGTTTGCTGGCAAT; 54.0 (GT)4(GA)5 104/112 104/112 + R:AGGTCATCTGGTAAAGCCATACC jcds66 (EU586344) F:CCTACGAGTGATTGGATAGTTTCTCA; 54.0 (CT)2(GT)3ATTGCA(AT)4 216/228 216/228 + R:TCTTCCATCAAGAGTCGTTGGGCA jcps1 (EU586345) F:GAGGATATTACAGCATGAATGTG; 47.5 (TG)4…(GT)3…(GT)4 132/162 132/162 + R:AATCAATCAATCTTTGGCAAA jcps6 (EU586346) F:CCAGAAGTAGAATTATAAATTAAA; 44.0 (AT)3G(TA)3…(CT)3… 288/305 288/380 X X R:AGCGGCTCTGACATTATGTAC (GT)5CT(GT)3 jcps9 (EU586347) F:GTACTTAGATCTCTTGTAACTAACAG; 48.0 (GT)3GC(TG)2A(GT)3 140/132 140/132 + R:TATCTCTTGTTCAGAAATGGAT jcps20 (EU586348) F:ACAGCAAGTGCACAACAATCTCA; 55.0 (TG)12(GA)22 271/260 260/278 X X R:TACTGCAGATGGATGGCATGA jcps21 (EU586349) F:CCTGCTGACAGGCCATGATT; 54.8 (CA)2…(CA)4 189/200 189/208 X + R:TTTCACTGCAGAGGTAGCTTGTATA jcms21 (EU586350) F:TAACCTCTTCCTGACA; 43.0 (CA)7 81/89 75 X + R:ATAGGAAATAAGAGTTCAAA jcms30 (EU586351) F:GGGAAAGAGGCTCTTTGC; 48.5 (GT)5T(TG)2 135/144 144/148 X X R:ATGAGTTCACATAAAATCATGCA JcSSR-18 (EU099518) F:GGCGACAGGAAGAGCATG; 62 (TA)3(GT)18 394 + R:GCAATCTTGGACAGGAAACG JcSSR-19 (EU099519) F:CTTGAAAGTTTTTGTAATTTC; 50 (AC)21 214 + R:CGCCAATCATAGATC JcSSR-20 (EU099520) F:GGCTGAACTTGCGCC; 60 (AC)10 260 + R:GCCCTGATTTCTGGTC JcSSR-21 (EU099521) F:CTGAAATGGAGAAATTGG; 50 (C)7(A)5(CA)9 249 + R:ACATATCGAAGATAGGG JcSSR-22 (EU099522) F:GAATCTCAACAGTGCCC; 52 (TC)16 152 + Towards domestication of Jatropha curcas  Review JcSSR-24 (EU099524) F:CACACACAACCAAACTGG; 56 (C)6G(C)6(AC)5 287 R: GGTTCTCTGAGATCCTC JcSSR-26 (EU099526) F:CATACAAAGCCTTGTCC; 55 (CA)18 211 + R:AACAGCATAATACGACTC T a: Annealing temperature; bp: Basepairs; NCBI: National Center for Biotechnology Information SSR: Simple sequence repeat; X: Polymorphism shown on gel; +: Number of alleles reported > 1. Data from [48,63,102,203].99