RESEARCH WORK
Morphological and productive characterization of Jatropha curcas L. provenances
R. Machado
Estación Experimental de Pastos y Forrajes «Indio Hatuey»
Central España Republicana CP 44280, Matanzas, Cuba
E-mail: rmachado@indio.atenas.inf.cu
ABSTRACT
The objective of this work was to characterize the morphological and productive variability of Jatropha curcas provenances. For that purpose the following indicators were measured or estimated: height; diameter of primary branches and stem; height of the first branch; diseases; number of primary, secondary and tertiary branches; fruit number and weight; seed yield; weight of 100 seeds; seed size and oil content. In order to interpret the results PCA and cluster analysis were used. Height was the most variable indicator, followed by diameter of primary branches, number of secondary and primary branches and stem diameter, which accounted for 57,7% of the variability. Based on these indicators six groups were formed. The highest values, in all morphological indicators, were found in group I (SSCS-3), and the lowest ones in group IV (SSCS-7 and SSCS-6). Only six from the ten seeded provenances produced fruits; while the planted ones did it in all cases. Among the seeded ones the following stood out: San Miguel, D-1 and SSCS-3, with estimated yields of 2 325; 243 and 169 kg of seed/ha; and in the planted ones: Cabo Verde, SSCE-10 and SSCE-7 with 814; 327 and 93,2 kg of seed/ha, respectively. In SSCE-10 and Cabo Verde the oil percentage reached 32,8 and 35,0%, respectively. The results allowed identifying the outstanding provenances, in the morphological (SSCS-3) as well as the productive indicators (Cabo Verde, SSCE-10 and San Miguel).
Key words: Characterization, Jatropha curcas L., provenance.
INTRODUCTION
The germplasm characterization of any crop is a procedure that is normally used to describe the morphological, phonological and productive traits that identify the species or accessions; as well as to verify the variation degree of the collections of useful, but potentially different, materials represented by a higher or lower quantity of individuals.
Olivera et al. (2009) stated that the morphological and agronomic characterizations are complementary activities which consist in describing the attributes of the accessions and, with that, determining their usefulness; but in turn it allows identifying the promising types for the selection, breeding processes or other purposes (Bonilla et al., 2008).
In order to carry out the characterization process known descriptors are used. When the genetic diversity among species and within the species is easily observable, the morphological descriptors provide information with which the duplication of the same material can be prevented and the overestimations of the existing diversity can be minimized (Becerra and Paredes, 2000).
In the particular case of Jatropha curcas, qualitative descriptors such as leaf color, petiole pigmentation and flowering onset, among others have been used; however, those of quantitative origin stand out for their number and practical importance, for example: days to emergence, plant height, number of branches, number of fruits per plant, number of racemes per branch, number of fruits per raceme and per tree, seed size and oil percentage, seed yield per plant and per area unit, and incidence of pest insects and diseases (Mendoza, 2008), which are conveniently analyzed based on descriptive statistical analyses (Flores et al., 2009) or by means of multivariate techniques (Wencomo et al., 2003; Machado and Olivera, 2008); which facilitates differentiating the accessions and the knowledge of their agronomic value (Campusano,2009).
The objective of this work was to characterize the morphological and productive variability of a J. curcas collection, and identify the most outstanding individuals based on quantitative indicators of interest.
MATERIALS AND METHODS
Procedure. The morphological and productive characterization of the J. curcas provenances was conducted in the germplasm bank of oil plants established at the Experimental Station of Pastures and Forages «Indio Hatuey» on June 18th, 2009, according to the indications made by Machado (2011).
The provenances planted from propagules (cuttings) or directly seeded, were arranged in 8 x 2 m plots, spaced at 2 m between rows and 2 m between plants, separated by 3 m-alleys in both senses.
During the development period of the plants the corresponding agricultural works were maintained, consisting in manual weeding of the plots and alleys, so that the plants became fully developed and without the damage weeds can cause them.
The morphological and productive characterization was made in all the plants that formed the plots, when they reached one year of age. For that purpose the following indicators were taken into consideration:
Diseases. The disease percentage in the trees (leaf curl and infestation produced by fungi). For that purpose a scale of six values was used, proposed by the National Plant Health Department (2005).
Weight of 100 seeds. With the use of the analytical scale. Seed length and width were measured with a caliper.
In addition, the number of primary, secondary and tertiary branches and the number and weight of harvested fruits were counted.
Some of these indicators served to estimate the productivity of fruits per tree; as well as the total seed yield; seed productivity per tree and seed yield per area unit.
Indicators were also added related to the physical characteristics of the fruits and the oil content, which were determined at the ICINAZ (Cuban Institute of Sugarcane Research).
The measurements and estimations were made in all the trees that composed each plot.
Statistical analysis. The morphological characterization data were processed through a principal component analysis, and the indicators which preponderance value was equal to or higher than 0,70 were selected as influential. From the pattern of principal components a classification was made, in group, of the characterized materials, by means of a cluster analysis, for which only the variables that fulfilled the above-mentioned requisite were taken into consideration.
RESULTS AND DISCUSSION
Table 1 shows that the accumulated variability, related to the measured morphological indicators, was high. Plant height was the most variable indicator, followed by diameter of primary branches, number of primary branches, number of primary branches and diameter of stem base, which accounted for most of the variability extracted in CP1 (57,7%). These morphological traits were positively related, which showed the existing high complementarity degree in these structural features as plants grow.
Normally, fruit trees and other ligneous plants, as in the case of J. curcas, usually have a high variability degree because they are essentially heterozygous and of crossed pollination, which leads to genetic segregation in the offspring (Caso, 1992).
The existing variability regarding the height of the first branch was conveniently explained in CP2, which indicates that this structural characteristic was not related to the other morphological variables, in correspondence with the principles for this type of analysis (Philippeau, 1986).
In the case of the number of tertiary branches the variability was also explained, partly, by this component, and it turned out to be the least variable indicator (together with the above-mentioned one), with only 16,9% of the variance extracted in this component. This indicates that in the first year of life the number of tertiary branches did not decisively influence as variable morphological factor among these provenances, perhaps motivated by the narrow variation range existing for this indicator, related to the development of plants at that age. J. curcas is known to be a shrubby species capable of reaching between 3 and 5 m (Manurung, 2007) and up to 6 m or more in full development (five years or more), moment in which the tertiary branches and those of other types can reach their highest degree of quantitative expression, together with the rest of the morphological and productive structural traits.
Table 2 shows the width of variation ranges and the average values of morphological indicators for the seeded provenances as well as for the planted ones; the outstanding types in these indicators are also shown. In general, the seeded provenances showed higher width, but particularly higher extreme values in the outstanding types, when compared to the ones planted through cuttings, except in the number of tertiary branches. This indicates that the former behaved as more precocious types in their structural development for vegetative traits, in the first year. Such performance is in contrast with the one reported by Montes Osorio et al. (2008) regarding the fact that the plants propagated by cuttings do not produce main root, but grow faster than the seeded ones. However, it does not imply that such performance pattern for these traits will remain inalterable in future years, aspects which study must be continued.
It is important to emphasize that in the number of secondary and tertiary branches, null lowest values were found, in some provenances, which was detected in the case of SSCS-7 and SSCS-6, which did not develop secondary branches; while D-1, D-2, SSCS-4 did not develop tertiary branches, indicating a more delayed development.
Table 3 shows the six groups, formed from the means of the morphological indicators represented in CP1. The extreme values were found in group I (SSCS-3) and in group IV (SSCS-7 and SSCS-6), both propagated by seed; the former with the highest values for all morphological traits and the latter with the lowest values. The other provenances reached intermediate values, which occupied groups where seeded or planted materials were indistinctly found, except in group III, which individuals were propagated by seed.
It is important to emphasize that the values found in height, diameter of primary branches and number of primary and secondary branches, in the provenances of group III, showed that the two provenances of this group had very low growth rate in the analyzed year, and they could not practically ramify, aspect which was manifested in a poor development of these materials from the point of view of their particular structure. This indicates that in the populations of this species genotypes can be found which development occurs rapidly, while others are remarkably delayed, independently from the fact that they are seeded, method that presupposes better plant establishment.
Nevertheless, independently from the existing differences in these morphological indicators, the evident variability within the J. curcas population should be stressed. Similar results were reported by Rao et al. (2008), who found significant variation in height (with a maximum of 134 cm) and branch number (with average values of 6,3 branches/plant and maximum values of 12,2 branches/plant), and by Sunil et al. (2009), who found high variation coefficients for several morphological characteristics, among them number of primary branches and plant height.
The values of some traits related to fruit production, productivity per tree, number of harvests and percentage of leaves with curling and infestation by fungi, are shown in table 4. Only six of the ten J. curcas provenances propagated by seed produced fruits; while all those planted by cuttings reached, with higher or lower intensity, the mature seed phenophase (harvest).
Among the provenances directly seeded in the field, the most outstanding were: San Miguel, D-1 and SSCS-3, but the first one showed better performance in the percentages of trees that produced harvest with regards to the population; as well as acceptable productivity per tree; while D-1, in spite of producing a lower fruit quantity, was slightly higher in productivity per tree. It is noticeable that provenance D-2, similar to D-1 from the morphological point of view (group III, table 3), did not produce fruits, which could be associated to other characteristics, such as physiological or environmental response ones, not to the possible influence of diseases, because the values in both were pretty similar.
From the planted provenances, the most outstanding were SSCE-10, Cabo Verde and SSCE-7. The first one reached the highest number of harvested fruits and the highest percentage of trees which produced fruits with regards to their population. Yet, the Cabo Verde provenance was better in terms of productivity per tree. Similar results as this one were reported by Machado and Suárez (2009).
Although in the germplasm bank there was presence of Lepidoptera, trips, leafhoppers (Empoasca sp.) and an unknown insect which decapitates the flower by the peduncle, which presence is worrisome because it limits seed production, the most outstanding fact was infestation by fungi, presumably from the genera Colletotrichum and Cercospora; as well as a mosaic symptom with leaf curl possibly caused by a viral agent. Freire and Parente (2006) reported in Brazil the presence of leafhoppers; trips; anthracnose (Colletotrichum gloesporoides and C. capsici), Oidum, Phytophthora and fusariosis, among the most frequent pest insects and diseases in that country.
In general, the affectation degree by these diseases was moderate to high and it was indistinctly manifested, in the seeded as well as the planted materials.
It is not possible to state whether there is or not relation between the intensity degree of these diseases and fruit production, because provenances were detected that reached a high number of fruits and in which the values of these diseases were high, as in D-1, or on the contrary, very low fruit production with an intensity similar to the above-described, as in the so-called Población Silvestre (Wild population). Even such provenances as SSCE-10, Cabo Verde, SSCE-7 and SSCE-8 produced large fruit quantities in plants which infestation values were higher, as compared to the Las Tunas provenance, which with the lowest value produced only six fruits.
It is not possible either to state whether there is tolerance of the more productive provenances to these diseases, because the amount of fruits they could produce if the diseases were not present is unknown. It is not known whether the provenances that produced little fruit or none would have reached higher or lower productions in the above-mentioned context.
It was clear that some provenances were less affected, such as San Miguel, SSCS-6, Cabo Verde and particularly the one called Las Tunas.
Table 5 shows the productive and morphological characteristics related to the seed, in the provenances which produced higher fruit quantities.
Among the seeded provenances, SSCS-3 reached the highest weight of 100 seeds and the lowest number of seeds in one kilogram (as they were larger), but it was surpassed by San Miguel and D-1 in total seed yield and productivity per tree.
In the planted provenances Cabo Verde stood out in the weight of 100 seeds, size, total yield and productivity per tree. The lowest weight of 100 seeds was reached by SSCE-10, which has smaller seeds (larger number in one kilogram); but it was higher than SSCE-7 in terms of yield and productivity per tree.
It is important to state that the seed yield estimate (kg/ha) was low for the most outstanding accessions in this indicator. However, it can be considered normal because it was reached in the first year, moment in which low productions can be expected (Jongschaaps et al., 2007); but it is in the range indicated by these authors for such countries as Brazil during the first year (335 and 190 kg/ha), when sowing frameworks of 4 x 3 m and 8 x 2 m, respectively were used. This was in correspondence with the yield per tree, which differed remarkably from the one expected for the reference species (higher than 2 kg of seed/ha/year), according to the review made by Cerels (2009).
It is valid to emphasize that these seed yields were higher, in some cases, than the ones reached on saline soils of the Paraguay farm in Guantánamo (340 kg/ha), with irrigation, and the ones obtained on friable black soils of Bayamo, without irrigation, where the yield was 168,2 kg/ha (BIOMAS-CUBA, 2010); and also similar to the ones found by González (2007) on soils which had previously been dedicated to sugarcane cultivation (Saccharum sp.).
Table 6 shows some properties related to the fruit characteristics, oil content in the seeds and estimated oil production per area unit in two of the outstanding provenances (the only ones that could be analyzed), independently from the propagation form. In these traits Cabo Verde turned out to be higher, although the differences were slight regarding fruit composition and oil content, encouraging aspect for SSCE-10 collected on national territory. This allows stating that in the Cuban populations of this species individuals can be found with acceptable potential of fruit production and oil content, similar to that of the outstanding introduced provenances, such as Cabo Verde, which percentage is comprised in the reported range for the species (Olivera et al., 2008).
The results allow concluding that there was an acceptable variability for vegetative traits, although a relatively small number of provenances were characterized, which allowed grouping them through the indicators with higher preponderance value and identifying the most outstanding individuals.
The degree of vegetative development reached through seeding was higher, in general terms; nevertheless, the indicators that identify fruit and seed production, in the outstanding types, corresponded to materials which were planted.
Among the accessions collected in the country, contrasting individuals were found in terms of morphological and productive traits, independently from the modality used for their establishment, which allowed identifying high-potential materials. This was evident in the provenances San Miguel, SSCE-10 and SSCS-3.
The infestation by fungi and the presence of leaf curl did not allow reaching a clear understanding of the role they could have played in plant development and fruit production.
Outstanding provenances were identified, in morphological indicators (SSCE-10) as well as in productive and reproductive indicators (Cabo Verde, SSCE-10 and San Miguel).
To continue the collection, introduction and characterization of J. curcas provenances is recommended, in order to increase the existing genebank and document the accessed material; as well as to pay especial attention to the stress which could be caused by disease infestation and the presence of potentially-pest insects.
