FLOWERING AND FRUITING SEASONAL CHANGES OF SIX ACCESSIONS OF JATROPHA CURCAS L. IN A SEMI-ARID REGION OF SENEGAL

* Ibrahima Diedhiou 1 , Roger Bayala 1 , Moustapha Diere Sagna 1 and Papa Madiallacke Diedhiou 2 . 1. Ecole Nationale Supérieure d’Agriculture (ENSA), Université de Thiès (UT), BP A 296, Route de Khombole, Thiès, Sénégal. 2. UFR des Sciences Agronomiques d’Aquaculture et de Technologies Alimentaires (UFR S2ATA)/Université Gaston Berger de Saint-Louis (UGB), Route de Ngallèle BP 234 Saint-Louis, Sénégal. ...................................................................................................................... Manuscript Info Abstract ......................... ........................................................................ Manuscript History


2139
Under these conditions improvement of J. curcas yield becomes imperative to enhance profitability of the plantations. To realize this perspective, it is necessary to have reliable data on flowering, fruiting and relationship between them. Indeed, these data are essential to develop and plan strategies selection and breeding (Kaur et al., 2011).
In wet regions of Asia such as India, Malaysia and Indonesia, many studies focused on J. curcas phenology characteristics and especially its floral biology. These aspects have so far received little attention in North America, where it is noted exceptionally, the study of Nietsche et al. (2014) carried out in south Florida. It emerged from these investigations that J. curcas is a monoecious and diclinous (with inflorescences of unisexual flowers) species. However, sometimes J. curcas might also produce hermaphrodite flowers (Fresnedo-Ramirez, 2013). Flowering begins in first year if environmental conditions are favorable (Silip et al., 2010). However, flowering is more important in the second and third years after plantation. The flowering cycle of J. curcas can happen the whole year in wet areas (Vidal, 1962). It is noticed that plants from cuttings get flower earlier that those from seeds.
Fruiting of J. curcas can starts in the first year. According to Silip et al. (2010), flowering to fruit set occurs within one to eight days. Fruits develop physiologically or reach mature green stage within 21 to 35 days. The species can have until four cyclical fruiting peaks per year (Alam et al., 2011). Fruit to female flower ratio (fruit set) varies from 0.37 to 0.79 (Gosh and Singh, 2008;Kaur et al., 2011;Nietsche et al., 2014). The yield of J. curcas has wide variation (from 32 to 745.72 g/tree) according to origin, age, soil, climate and farming practices (Heller, 1996;Rao et al., 2008;Wijaya et al., 2009;Wani et al., 2012;Diédhiou et al., 2012;Shabanimofrad et al., 2013;Singh et al., 2013). Although J. curcas is considered tolerant to drought, several studies (Kaur et al., 2011;Ma et al., 2016) showed that water has a significant effect on its phenology and particularly fruiting.
If J. curcas phenology seems to be well documented in Asia, while in West Africa and especially in the Sudano-Sahelian areas where climate and soil conditions are different, there is little information on this question (Ahoton and Quénum, 2012).
In these conditions, the objective of this study was to analyze seasonal changes in flowering and fruiting of six J. curcas accessions in a semi-arid region of Senegal.
The climate is tropical semi-arid with short rainy season (June to October) and long dry season (November to May). The total annual precipitation is 480 mm (Kizito et al., 2006). During the experiment, the rain was collected with Decagon weather station (Decagon Devices, Inc., Pullman, WA, USA). The rainfall started in the third 10 days of June and become more and more important from July. August and September were wetted months with respectively, 233.6 and 226.8 mm cumulative rainfall. The quantity of precipitation of these months represents 81% of the total annual rainfall (568.2 mm). Overall, there was a good distribution of precipitation during that year. The mean annual minimum ambient temperature is 19°C and a mean annual maximum ambient temperature of 33°C. The minimum and maximum of relative humidity range from 49 to 91% in rainy season and from 27 to 75% in dry season.

2140
The study site lies on a leached and disturbed ferruginous sand soil classified as a ferruginous Oxisol (FAO, 1998). The top horizon (0 -0.7 m), is sandy with a friable continuous structure and no distinct horizonation with low clay content of about 5%. The organic matter and nitrogen content is about 0.5 and 0.3% respectively and water-retention capacity ranging from 60 to 100 mm (Zanté, 1983).

Experimental Design:-
A set of six accessions of J. curcas, where four were from Senegal, one from Tanzania and one from Mozambic was used for the experiment. Accessions codes in ENSA's collection and the rainfall of their ecological areas are showed in table 1. Seedlings of the six accessions were produced from vigorous seeds in a nursery using polybag containing a mixture of soil and compost in the ratio of 1:1. Four months old seedlings were established in the field during August 2008. Seedlings were irrigated during 90 days after planting and at 15 days intervals. No fertilization and pruning have been done. Weeding was performed as needed, usually in rainy season. The experimental design is a randomized block design (RBD) with six treatments (accessions) and four replicates. The sub-plot was a square of 6 m side and had 16 trees. Planting spacing was 2 m in-row x 2 m between-rows, having 2500 plants per hectare.
At the time of the study, trees were 42 months old. All accessions had started flowering at the second year after plantation.  The total number of inflorescences per tree was counted every 14 and 7 days respectively in dry season and rainy season.
The flowers sex parameters were recorded by simple count method once in May (dry season) and September (wet season). In the wet season, this survey was carried out on the six accessions; but in the dry season it was conducted on only four accessions (CE6, CE95, CE6 and CE97) which produced recognizable flowers in this season. Fift een inflorescences per each of the three selected trees per accession were sampled, so a total of 45 inflorescences were considered per accession.

Fruiting phenology:-
The fruiting survey was performed during 10 months (February-November 2012) on four trees per accession, randomly selected in four blocks of the experimental plot. Observations were made every 14 and seven days respectively in dry season and rainy season.
The yield (seed weight per hectare for the density of 2500 plants/ha) and its components (number of fruit per tree, fruit weight per tree, fruit weight and seed weight per tree) were calculated for each season and also for the whole duration of the survey.
The number of fruit was recorded by simple count method. The fruit weight and the seed weight were measured by electronic balance in the laboratory.

Data Analysis
The data were submitted to a two ways analysis of variance (ANOVA) using Statistix. Means separation among treatments was performed by Tukey's test at the 5% probability level based on the F-test of ANOVA.

Results:-
Flowering phenology:-The flowering rhythms of the six J. curcas accessions are summarized in Fig 1. All accessions produced inflorescences continuously throughout the observation period. However the full flowering occurred between April (dry warm season) and July (early in rainy season) for all accessions. Accessions had two peaks of flowering. The first peak occurred in dry season (April, 24 th -June 6 th ) and the second, with the highest number of inflorescences, arrived in wet season (8 -22 July).  Table 2. There were significant differences between seasons (p < 0.05) for the number of female flowers per inflorescence but not for the number of male flowers per inflorescence. The number of female flowers per inflorescence was significantly higher in the wet season than in the dry season. Consequently, the male to female flowers ratio was lower in the wet season compared to the dry season. The influence of the accession on number of male flowers per inflorescence, number of female flowers per inflorescence and male to female flower ratio is presented in table 3. There were significant differences between accessions (p < 0.05) for all flowering parameters. Number of male flowers per inflorescence ranged from 127.10 to 149.88 and the number of female flowers per inflorescence ranged between 3.68 and 6.88. These parameters were significantly higher for the accessions CE6 and CE97; consequently the ratio values of these accessions were the lowest.  The influence of the interaction between accession and season on flowering parameters is summarized in Table 4. The data showed a significant effect (p < 0.05) on all flowering parameters. The number of male flowers per inflorescence varied from 103.77 to 187.90. Maximum value of this parameter was observed for CE6 during wet season, followed by CE50 and CE97 during dry season.  Fruiting phenology:- Figure 2 summarizes the fruits production of accessions from February to November. All accessions showed continuous fruit production during the experiment survey and average fruits production patterns were fairly similar for all accessions. Fruiting patterns is highest concentration between September 15 th and October 3 rd . During dry (end of April) CE50, CE95 and CE6 had low fruit production while CE97, CE98 and CE14 had nearly no production. In full production period, CE95 and CE97 produced 66 % and 97 % of total production respectively. For EC50, EC6, CE98 and CE14, average was 73 %, 87 %, 94 % and 95 % of total fruit production.
The effect of the season on yield parameters is shown in table 6. The two way analysis of variance revealed that the season had a significant effect (p<0.05) on all the yield parameters except one fruit weight (Table 6). Values recorded for the wet season were significantly higher than those observed in the dry season. Seed yield, noted for the wet season (1282.10 kg ha -1 ) was 13 times higher than that of the dry season (99.40 kg ha -1 ). The influence of the accession on yield components is summarized in Table 7. There were significant differences between accessions for yield parameters, out of fruit weight. Fruit weight per tree ranged from 243.20 g (CE98) to 826.35 g (CE6) while seed weight per tree varied from154.38 g (CE98) to 547.90 g (CE6). Seed yield per hectare were between 385.90 (CE98) and 1369.80 (CE6). For all these parameters, the accession CE6 had the highest values, followed by CE97 and CE95.  The effect of the interaction between accession and season on yield parameters is shown in Table 8. Results indicated that except fruit weight, all yield parameters (fruit weight per tree, seed weight per tree, seed yield) was significantly affected by the interaction between the season and the accession (p <0.05).
The highest fruit weight per tree was recorded in CE6 (1552.70 g) and CE97 (893.50 g) during the wet season. The lowest value of this parameter was observed in CE97 (16.70 g) during the dry season. But there was no significant difference between that value and those of CE14, CE98, CE50, CE6 and CE95 during the dry season. Seed weight per tree was significantly higher in CE6 (1030.10 g) and CE97 (569.10 g) during the wet season. It was significantly lower in CE97, CE14, CE98, CE50, CE6 and CE95 in the dry season. The trend is similar for seed yield.
Annual yield:-Cumulative values (from February to November 2012) of yield parameters are shown in table 9. Whatever the parameter considered, highest values were recorded in CE6, CE97 and CE95, while lowest value was observed in CE98.

Discussion:-
This study has shown significant effect of the studied factors (season, accession) and interaction between them on floral traits of J. curcas. The observations revealed that flowering of the six accessions studied was continuous with two peaks. The first peak occurred in hot dry season while the second, with the highest number of inflorescences, appeared early in the wet season. These results were reported in Asia (Sukarin et al., 1987;Kaur et al., 2011) and Florida (Nietzsche et al., 2014). Full bloom of J. curcas occurred between the beginning of April and the end July. Flowering parameters including the number of female flowers per inflorescence and male to female flower ratio change significantly with the season. Overall, wet season has significantly fostered female flowers development and induced low male to female ratio. Thus, the number of female flowers ranged of 2.47-11.87 in the wet season with an average of 7.13. In the dry season, this parameter varied from 1.67 to 5.73 with an average of 4.56. This is consistent with observations reported for Florida (Nietzsche et al., 2014). The period of full bloom of the accessions and installation of the second flowering peak coincides with the warmer months with temperatures ranging between 27.5 (June) and 27.7 °C (July). Also, the wet season begins in the study region, between June and July. Therefore, the best performance of flowering in the starting of the wet season (higher peak and highest production of female flowers), established in this study, may be related to climatic conditions particularly the high temperatures and increase of soil moisture with installing of rainfall. These observations suggest that floral initiation and development of female flowers of J. curcas requires high temperatures and moderate humidity. In a study conducted by Nietzsche et al. (2014) in Florida, it has been observed that J. curcas flowering begins with rising temperatures and precipitation. Wu et al. (2011) reported that female sites on inflorescences of J. curcas occur mostly in seasons with high temperature and rainfalls. They also observed that the right amount of rainfall may possibly enhance the number of female flowers for each inflorescence, while too much or too less would not favor the occurrence of female flowers.
However, we found in this study, that effect of season on the number of male flowers was not significant. This result suggests that the production of female flowers is more environmentally sensitive than the male flowers that would depend more on genetic factors. It is consistent with the observations of Anandalakshmi et al. (2015) that showed that female flower production is more sensitive than male flowers. In a study conducted in China to understand organization of the flower sexes in J. curcas, Wu et al. (2011) documented that, male and female flowers have similar tissues and organs before female primordia emergence. So, they proposed that there are factors that regulate sex differentiation in this plant, which may selectively affect the action of homeotic genes in one whorl, such as only restraining the initiation of a gynoecium meristem in male J. curcas flowers.
This study also showed that the flowering characteristics vary significantly among accessions. In wet season, CE6 had more male (191.60) and female (11.87) flowers per inflorescence with lowest male to female flower ratio (16.40 The present study has shown significant effect of the studied factors (season, accession) and interaction between them on fruiting traits of J. curcas. It showed that the fruiting rhythm of J. curcas in semi-arid area in Senegal is continuous with a short period of full production (September -October). Considering the average length of fruit set to physiological maturity is 3 to 5 weeks (Silip et al., 2010), that indicates that the majority of fruits are from fecundation that took place during the first ten days of August. Some accessions (CE14, CE97 and CE98) had no fruit set in dry season but their full production ranged from 94 to 97 % of total production. In contrast CE6, CE50 and CE95 had fruit set in dry season and their full production varied from 66 to 87 % of total fruit production. These results indicate that J. curcas harvest should be done in wet and dry season according to accessions. This would reduce the high labor costs of harvesting J. curcas fruit (Everson et al., 2013). This study also noted that the season has a significant influence on yield parameters. Overall, production of fruits and seeds was better in the wet season than in the dry season. Thus, seed weight per tree ranged from 11.40 to 85.40 g in dry season with an average of 39.78 g; in wet season it was between 295.90 and 1030.10 g with an average of 512.83 g. High fruit and seed production of accessions in the wet season are due to better high flowering with a greater number of female flowers (Ferry, 2006). In addition, the values of male to female flowers ratio observed show that there are enough male flowers to ensure good pollination. However, the low number of female flowers is a major constraint of production. Our results can be explained by the fact that in the wet season, the efficiency of flowering (number of fruits / number of female flowers per inflorescence) is higher and humidity conditions are more favorable to fruit growth. Ma et al. (2016) reported that J. curcas have high demand for water during flowering and fruiting period in the dry-hot valley of Chin-sha River China. Anandalakshmi et al. (2015) mentioned that the annual increase in fruiting success of a three year old plantation in Tamil Nadu, India, was due to sufficient rainfall accompanied by availability of nutrients. In South Florida, where summer season is characterized by high average precipitation (1473 mm/yr) and high average temperatures (29 °C), Nietzsche et al. (2014) reported that fruit set average is highest in that season (75.5%) compared to fall, indicating a positive effect of climate on this characteristic. Also, our results showed that there are significant differences among accessions for the cumulative yield (dry and wet seasons). In terms of yield, CE6 is the better accession with a cumulative seed yield of 1095.50 g per tree or 2739.3 kg ha -1 . CE98 and CE50 with lowest number of flowers in wet season produced 308.50 and 350.80 g per tree as cumulative seed yield, respectively. Previous works (Rao et  In this study, the highly significant variability of floral and fruiting traits found among accessions, could be associated with the genetic differences between these accessions and the climate conditions, particularly temperature, since plants grown in relatively homogeneous soil conditions. Therefore, it is important to select high yielding accessions to improve the profitability of plantations. In this perspective, as suggested by Singh et al. (2010), the number of female flowers borne by the plant may be given more weightage rather than selecting plants with better male to female ratio and number of inflorescences borne by the plant. However, our results suggest that we can also improve the performance of plantations through water supply at the beginning of flowering and dry season.

Conclusion:-
This work demonstrated that flowering and fruiting of J. curcas were continuous during ten months, with periods of full bloom and full fruit production. The flowering cycles of different accessions showed one to two peaks, one in the dry season and another one early in the rainy season.
The characteristics of flowering and fruiting change significantly according to the accessions and the season. The wet season has been more favorable for the production of flowers. Also during this period, J. curcas had between 66 and 97% of total fruit production. Four accessions had highest seed yield (≥1 t ha -1 ); and could be promising for the production of biofuel.
Based on our results, to better understand the reproductive biology and phenology of J. curcas, it is important to determine the role of climate and soil moisture in the flowering and fruiting of this species.