Regeneration ecology of Jatropha curcas L. in Africa: implications for its biofuel production and invasiveness

Jatropha curcas L. is a tropical tree belonging to the Euphorbiaceae family. The high oil content and suitable fatty acid composition of the seeds are two arguments to promote J. curcas as a promising biodiesel crop. However, little information is available on the regeneration ecology and productivity of the species. In addition to this, the higher number of male flowers compared to female flowers results in a low yield. A better knowledge of the reproductive ecology of the species is imperative (e.g. pollination process, flowering, fruiting, seed yield, seed dispersal and establishment) for yield optimization and invasiveness risk assessment and management. Therefore this research first focused on the current discourse and it questions how to increase the seed and oil yield of J. curcas through agronomic approaches such as pollen manipulation, pruning, spacing and fertilization. The second question was the invasiveness issue of J. curcas. To answer both questions we formulated the following research objectives: (i) identify the possible potential arthropod pollinators of J. curcas, (ii) find out the best pollination method to increase seed yield, oil quantity and quality, (iii) identify the best agronomic management practices and timing to modify flower sex ratio and to increase seed and oil yield of J. curcas, and (iv) assess the invasiveness risk of J. curcas and seeding potential to adjacent land ecosystems. For seed yielding crops such as J. curcas it can be expected that successful pollination is a key process that affects seed yield, oil quantity and oil quality. Insight into the diversity and movement of potential pollinators forms an important aspect of the needed knowledge. In the first part of this dissertation we report on J. curcas flower visitors observed in two Southern African countries. Observations at two sites, each on 10 trees, during 10 days showed that 41 insects and 2 Arachnida in Zambia and 29 insect species in Malawi visited J. curcas flowers. Diptera and Hymenoptera were the largest group. The most abundant insect visitors were Apis mellifera and Chrysomya chloropyga. A. mellifera visits more inflorescences within shorter period than C. chloropyga and shows a frequent appearance in both sites.Floral phenology studies and pollination experiments (natural and artificial) in two sites in Zambia (2 year and 5 year old plantations) and one site in Malawi (5 year plantation) showed J. curcas is not only of protandrous nature, as previously reported. It can be both protandrous and protogynous and able to produce seeds through both self-and cross-pollination. The male to female flower sex ratios were 22:1, 17:1 and 10:1 respectively for 2 year and 5 year old plantation in Zambia and 5 years old plantation in Malawi. The flower longevity periods (mean± SE) were 1.80±0.07 days for male flowers and 4.5±0.18 days for female flowers. Fewer fruit set and lower seed yield were recorded from autogamous pollination in Zambia sites. In the case of Malawi, more matured fruits resulted from autogamous pollination but the fruits contained fewer seeds. High fruit and seed yield were recorded for open pollination similar to pollen-supplemented pollination at Zambia sites, which indicates there was no pollen limitation in the study sites. In the Malawi site, there was no seed yield difference between pollination treatments. The natural fruit set and seed yield in Zambia indicates that conserving the natural pollination will improve J. curcas fruit set and seed yield, as a base for better oil production.The seed morphology, seed oil content (g), seed oil concentration (%) and oil fatty acid composition study after different pollination treatments in Zambia showed the effect of pollination more reflected in total oil yield per inflorescence and tree than in unit seed oil content and concentration. Up to 70% and 29% oil yield reduction was observed for autonomous autogamy and self-pollination treatments, respectively, compared to open pollination. Cross- and self-pollination resulted in longer seeds than open pollination but not in oil content and concentration differences. The oil has high unsaturated fatty acid content (80%) and is composed of 9 fatty acids. Pollination treatments had an effect on fatty acid composition of individual seed oil from mature trees (5 year old), but did not for 2 year old trees. The oleic acid content, a determinant fatty acid component for quality biodiesel production, was lower for artificial self-pollination (9% reduction) compared to open pollination. This research demonstrates overall oil yield and quality are the highest under natural pollination. This shows the importance of safeguarding the quality and effectiveness of open natural pollination. This can be done by enhancing the presence of insects identified as J. curcas pollinators, particularly those pollinators enhancing cross-pollination. Proper application of fertilizer and canopy management such as pruning and planting density are believed to increase J. curcas seed and oil yield. The 2009 to 2011 field experiments in Balaka, Malawi showed J. curcas yield can be best improved without pruning. Spacing and pruning did not show flower sex ratio differences among the treatments. We found up to 55% seed and final oil yield reduction after pruning, particularly for trees pruned in the dry cold season, and beginning of rainy season compared to non-pruned trees. It is claimed that J. curcas can be grown on soils with low nutrient content but this study revealed that yield was low for non-fertilized trees, even on good soils. Fertilizer application altered the flower sex ratio of J. curcas. Seed yield and oil yield increased at higher application rates of Phosphorus (at 200 % P level, relative to the recommended rate) and Nitrogen (at 143 % N level, relative to the recommended rate) with an increase of seed yield and oil yield (mean ± SE) to 217±41% and 217±38% and 203±42% and 204±45% respectively compared to non-fertilized trees. Based onthe integrated invasiveness research experiments (spontaneous occurrence of seedlings, seed dispersal mechanisms, seed predation by animals, and germination success of dispersed seeds) in Zambia, J. curcas is not an invasive species. No spontaneous regeneration was observed in land use systems adjacent to J. curcas plantations. Primary seed dispersal was limited, and predominantly under the canopy of the mother plant. Rodents and shrews dispersed and predated J. curcas seeds and fruits. They transported the seeds up to 23 m from the sources and repositioned them in their burrows up to 0.7 m deep, but none of these seeds could establish. Germination experiments in adjacent land use systems revealed 4% germination success at the soil surface, and 65% if buried artificially at 1-2 cm depth, yet the latter is unlikely to occur under natural conditions. These findings show that J. curcas seeds may be dispersed by animals to adjacent land use systems, but no natural recruitment was observed given low germination on the surface and none in burrows. All together these results suggest that the plant currently does not show an elevated risk of invasion to adjacent land use systems, at least not in the investigated case study. The overall study elucidates the importance of understanding the critical steps and processes of J. curcas regeneration ecology to optimize seed yield, oil yield and to avoid the invasiveness risk of the species in Africa. The research methodology can be also applied for other similar current tropical biofuel crops. As such, this study adds to the scientific knowledge of the regeneration ecology of tropical biofuel crops in tropical and sub tropical regions of the world.status: publishe