MEMORANDUM
TO: 		Air and Radiation Docket EPA-HQ-OAR-2015-0293
FROM: 	EPA Office of Transportation and Air Quality
SUBJECT: 	Supporting Documentation for Jatropha Oil Production and Transport Greenhouse Gas Emissions

There are a number of spreadsheets on the docket containing data and calculations used in the analysis of greenhouse gases attributable to the production and transport of jatropha oil for biofuel. This memorandum provides additional explanation of the key assumptions used in our calculations and modeling.
Cultivation and Harvesting
Seed Yields and Oil Content 
For the purposes of our analysis, we project that in 2022, on average, one hectare of jatropha grown in southern Mexico (specifically Yucatan, Oaxaca and Chiapas states) will yield five tonnes of dry seeds per year, while one hectare in Brazil will yield four tonnes per year.  For Mexico, five tonnes per hectare reflects a middle to upper bound estimate of recorded yields in the literature, but is conservative compared to current yields reported in the GCEH petition.  Table 1 summarizes Mexican jatropha yields reported in literature reviewed by EPA, with comments on each study.
Table 1. Mexico Dry Seed Yield Studies (tonnes per hectare per year)
Study
Yield
Comments
Almeida et al. (2011)
4.3
Direct observations and literature collected covering Mexico, Brazil, India, and Tanzania. Observations are from "jatropha entrepreneurs" so production scale is uncertain.
Achten & Verchot (2011)/ 
Skutsch et al. (2011)
3.7 to 11
Achten & Verchot (2011) Mexico values are based on those reported by Skutsch (2011) which are in turn based Mexican government and farmer expectations and do not reflect observed data.
Achten et al. (2008)
5
Nicaragua from personal communication.

4
Paraguay from personal communication.
Trabucco et al. (2010)
1.5 to 5
Estimated productivity of naturally occurring jatropha as a function of environmental and climate factors, yield range for Southern Mexico only. Yields in other regions are lower.

There are fewer recorded observed yields in northeastern Brazil;  however, based on environmental and climate characteristics as modeled by Trabucco et al. (2010), we expect jatropha yield in this region will be somewhat lower than yields in southern Mexico.  Table 2 summarizes Brazil jatropha yields reported in literature reviewed by EPA, with comments for each study.
Table 2. Brazil Dry Seed Yield Studies
Study
Yield (t/ha/yr)
Comments
Bailis & Baka (2010)
2 to 6
Midpoint based on model runs by Lapola et al. (2009) and should not be considered separate data point. High and low estimates added arbitrarily for sensitivity analysis, not based on observed data.
Lapola et al. (2010a)
3.7
Models runs of smallholder farms in NE Brazil
Trabucco et al. (2010)
2.5 to 3.5
Estimated productivity of naturally occurring jatropha as a function of environmental and climate factors, yield range for NE Brazil only. Yields in other regions are lower.

Our projections for average jatropha seed yields in Mexico and Brazil in the year 2022 are conservative because intensive jatropha cultivation is relatively new, with lots of room for potential advances through genetics, breeding and improved agronomic practices.  
For jatropha grown in Mexico and Brazil in 2022 we project dry seed oil content of 35 percent by mass. This value is within the range of current values reported in the literature reviewed by EPA, and is also similar to the values reported in the GCEH petition.  (See Table 2-3 in the literature review conducted for this notice, which summarizes values reported in the literature.)  We view 35 percent seed oil content as a conservative projection for 2022 given the potential for increases through advances in breeding, genetics and crop management.
Preparation and Planting 
Chemical and energy inputs are required when jatropha is first planted.  There is limited information available on the inputs and energy requirements for preparation and planting of jatropha.  We used values from the GCEH petition for inputs of diesel and fertilizers.  Based on information in Bailis and Baka (2010), we also assumed that jatropha production in Brazil would require the addition of two tonnes per hectare of agricultural lime, as the soils in Brazil are highly acidic.  However, based on information provided by GCEH in support of their petition, soils in Mexico would not require lime addition.  We used a weighted average of the planting areas in Mexico and Brazil to calculate an average lime addition of 1.11 tonnes per hectare.
There is a range of values found in the literature on the lifetime of typical jatropha plantations.  For example, Achten et al. (2008) and Achten et al. (2010) suggest a jatropha lifetime of 30-50 years, whereas Bailis and Baka (2010) use lifetimes of 20 and 30 years for their lifecycle analysis.  Our analysis assumes that jatropha has a 20 year crop cycle, meaning that every 20 years the existing jatropha plants are removed and the crop is replanted.  Accordingly, our analysis assumes the GHG emissions associated with preparation and planting occur every 20 years.  A lifetime of 20 years is a conservative estimate, because it is at the low end of the range reported in the literature, and emissions from preparation and planting will occur more frequently than if we assumed a longer replanting cycle.  
Annual Inputs and Harvesting 
In addition to emissions associated with preparation, planting and replanting, growing and harvesting jatropha requires crop inputs and energy use on an annual basis. 
Fertilizer and Pesticide Inputs
We assumed that annual fertilizer inputs will have to at least make up for nutrients lost from harvesting jatropha fruits.  To determine the nutrients lost through removal of jatropha fruits, we used data from Bailis and Baka (2010), who also used this approach in their lifecycle analysis.  Bailis and Baka present data on the fertilizer requirements for a range of seed yields.  We interpolated these data to calculate the nutrients required for our assumed seed yields.  
To replace the nutrients lost through fruit removal we assumed that husks and seedcake will be applied to the fields as a source of nutrients, as this is an approach commonly reported in the literature and the GCEH petition.  To calculate the nutrients added from these sources, we used the nitrogen, phosphorus, and potassium contents of different parts of the jatropha plant presented by Reinhardt et al. (2008).  For the seedcake, we used data from the GCEH petition and the literature review to determine the percent shells, seedcake, and residual oil present in the seedcake.  We used a weighted average of the nutrient content of each of these components to calculate the nutrient content of the seedcake.  We found that an additional 9.3 kg/ha of inorganic nitrogen fertilizer would need to be added to make up for nutrients lost from fruit harvesting.  As a conservative estimate, we assumed that 9.3 kg/ha of potassium and phosphorus fertilizers would be added as well, even though this would replace more potassium and phosphorus than is lost through fruit harvesting.  Details about these calculations can be found in the spreadsheets posted on the docket.  
For pesticides, herbicides, and insecticides, we used values from Bailis and Baka (2010).  These value are based on surveys of jatropha growers in Brazil.  According to the surveys, two of six growers used herbicide, one used pesticide, and one used nematicides.  These growers typically applied these chemicals during the first 2-3 years of jatropha growth.  As a conservative estimate, we assumed that herbicides, pesticides, and nematicides would be used in the amounts reported by the growers that used these chemicals.  The emissions from pesticides, herbicides and insecticides are very small compared to the other GHG sources we evaluated.
Energy Use
We used inputs of diesel and electricity provided by the GCEH petition.  The electricity value used is a conservative/large estimate relative to other values in the literature, such as Fan et al. (2012), and the diesel value is within the range of values in the literature.  For example, Bailis and Baka (2010) assumed that jatropha fruit harvest would be manual, and that 5 L of diesel per hectare would be needed for mowing.  On the other hand, Reinhardt et al. (2008) assumed that 55-141 L of diesel per hectare would be needed for jatropha cultivation, depending on the yield.   
Nitrous Oxide Emissions
Equations for direct and indirect emissions from fertilizer and crop residues (including husks, seedcake, and above and below ground residue from trees that are removed at the end of a crop cycle) are from the Intergovernmental Panel on Climate Change.  Limited data are available on the nitrogen content of the husks, seedcake, and tree residue.  Therefore, we used the best data available to us.  The nitrogen content of the husks and seedcake are from Reinhardt et al. (2008).  According to Stratton et al. (2010), these data are based on a 30-hectare test plot in India.  The nitrogen content of the above and below ground residue is from Jongschaap et al. (2007), which compiled data from three sources.  
To calculate the amount of above ground residue, we used data from Bailis and Baka (2010) on the relationship between the above ground residue (excluding leaves) and the yield.  To this, we added the biomass from leaves, based on the ratio of leaf biomass to stem and branch biomass from Reinhardt et al. (2008).  The below ground residue was calculated based on the ratio of below ground residue to above ground biomass in Reinhardt et al. (2008).  For more information on these calculations, see the "Jatropha Oil Production and Transport GHG Calculations" spreadsheets posted on the docket for this notice.
Land Use Change and Agricultural Sector Emissions
Jatropha Biomass Carbon Stocks 
We estimated the average amount of biomass carbon sequestered by jatropha plantations in southern Mexico and northeastern Brazil, projected out to 2022.  Although there is relatively little data on jatropha plantations compared to other crops, jatropha biomass carbon stocks were estimated using available scientific information from the literature.  Reinhardt et al. measured basic data about jatropha plants, such as root to shoot ratios and biomass carbon content.  Bailis and Baka used the data from Reinhardt et al. to estimate biomass carbon stocks for different jatropha yield scenarios.  Using our projected jatropha yields of 5 and 4 tonnes per hectare per year for Mexico and Brazil respectively, we used the Bailis and Baka approach to estimate average biomass carbon stocks of 8.9 and 8.1 tonnes per hectare for ten year old jatropha plantations in Mexico and Brazil, respectively.  Per the methodology developed for the March 2010 RFS rule, we translated these estimates into average biomass carbon stocks over 30 years.  Assuming linear growth rates, a 20 year replanting cycle and pruning of any growth after 10 years to ensure fruit accessibility, we estimated average jatropha plantation biomass carbon stocks over 30 years to be 6.9 and 6.3 tonnes per hectare for Mexico and Brazil respectively.
EPA's estimates for jatropha biomass carbon stocks are within the range of estimates in the literature for jatropha plantations in these regions.  For comparison, Skutsch et al. measured 70 jatropha plants in Mexico and categorized them into two groups: fast growing and moderate growing, the difference between these relating to the soil conditions where the plants were found.  Assuming linear growth rates and a planting density of 1,600 trees per hectare they estimated carbon stocks of 9.1 and 3.6 tonnes per hectare at ten years for fast and moderate growing plants, respectively.  Skutsch et al. assumed a 20 year planting cycle, and that any growth after 10 years would be pruned out to ensure that fruits are accessible for harvesting, implying that the biomass carbon stocks would stay constant for stands from 10 to 20 years of age.  The GCEH petition estimated the average biomass carbon stocks for their jatropha plantations in Mexico to be approximately 7 tonnes per hectare.  Bailis and Baka used data on jatropha seed yields in combination with root to shoot ratios measured by Reinhardt et al. (2008) to estimate biomass carbon stocks for mature jatropha plantations in Brazil.  Assuming 1,250 jatropha plants per hectare, they estimated 5-9 tonnes of biomass carbon per hectare for high yield and low yield jatropha plants respectively.  Bailis and McCarthy found that jatropha plantations in Brazil and India store 7.3-9.1 tonnes of carbon per hectare in above ground biomass and litter when managed with regular pruning. 
There are other studies in the literature that report significantly lower biomass carbon stocks for jatropha plantations.  However, these studies looked at lower yielding jatropha production systems grown on less fertile land and/or using significantly lower crop inputs (e.g., fertilizer, irrigation).  Typically, these studies also looked at jatropha production in regions such as India, Africa and Southeast Asia that did not factor into our assessment.  We did not factor these studies into our assessment of jatropha biomass carbon stocks, because we focused our analysis on higher yielding, intensively managed plantations grown on traditional agricultural land in Mexico and Brazil.
Jatropha Grown on Unused Grassland Scenario 
As explained in the notice, the first main scenario that we evaluated considers jatropha grown on otherwise unused grasslands in southern Mexico and northeastern Brazil.  Mexico and Brazil grassland carbon stocks were estimated for the March 2010 RFS rule.  Per Harris et al. (2009), above- and belowground carbon stocks for grassland in Brazil were estimated to be 10.9 tonnes of carbon per hectare based on the average value reported in the literature for campo limpo (pure grassland).
To maintain a consistent global approach, for all countries except Brazil, carbon stocks in grasslands were estimated based on default biomass values given in Table 6.4 of the IPCC AFOLU Guidelines. These default values are presented by ecological zone. Therefore, grassland carbon stocks within each country reflect the area-weighted value based on the proportions of each ecological zone present within each country. Using this approach, biomass carbon stocks in Mexico were estimated to be 4.1 tonnes of carbon per hectare.
For grassland converted to jatropha, the only land use change emissions are associated with changes in biomass carbon stocks, meaning the difference between the average jatropha biomass carbon stocks and the average grassland biomass carbon stocks.  Per the methodology developed for the March 2010 RFS rule, the resulting emissions are annualized over thirty years.
Based on the methodology developed for the March 2010 RFS rule, we assume no change in soil carbon stocks averaged over thirty years when grasslands are converted to perennial crops such as jatropha.  For Brazil and Mexico we also assume that burning is not used as part of converting grasslands to jatropha.  For more information on the land use change calculations for this scenario, see the spreadsheets posted on the docket.
Jatropha Grown on Agricultural Land Scenario
In this scenario, we examined the possible impacts if jatropha directly displaced land that would have otherwise been used for agricultural production in southern Mexico and northeastern Brazil.  As described in the associated Federal Register (FR) notice, for this analysis we used the Food and Agricultural Policy and Research Institute international models as maintained by the Center for Agricultural and Rural Development at Iowa State University (the FAPRI-CARD model).  As explained in the FR notice, we conducted two simulations with the FAPRI-CARD model, one for jatropha cultivation in southern Mexico and one for cultivation in both southern Mexico and northeastern Brazil.
As described in the FR notice, we used a land displacement approach that allowed the model to shift or expand displaced crop and pasture areas to previously unused areas.  Using this approach, in both scenarios the FAPRI-CARD model projected that the area used for staple crop production in Mexico (e.g., corn, beans, wheat) would decline by more than the area displaced by jatropha, and in the scenario that included jatropha production in Brazil the model projected that pasture area would decrease by more than the area need for jatropha production and other crops displaced by jatropha.  These projections are the result of a number of factors including crop, livestock and land price changes projected by the model that incentivize intensification of crop and livestock production and other changes in crop and pasture land use.  As a conservative approach for GHG modeling, we adjusted the FAPRI-CARD results to eliminate the model's "extra" projected decreases in crop and pasture area in Mexico and Brazil beyond the area needed for jatropha and agricultural land displaced by jatropha.  In other words, as a conservative approach we assumed that growing jatropha in Mexico and Brazil would not result in less total crop and pasture area (including areas used for jatropha) compared to the control case that does not include jatropha.  The full results from these simulations are provided in spreadsheets available on the docket.
As explained in the notice, using the methodology developed for the March 2010 RFS rule we estimated the GHG emissions resulting from land use changes, changes in crop production (including for example emissions associated with changes in overall fertilizer use, energy use and nitrous oxide emissions), and changes in livestock production (including for example emissions associated with changes in energy use for livestock production, methane from enteric fermentation and nitrous oxide from manure).  The calculations for these emissions sources are included in the Jatropha Production and Transport GHG Calculations spreadsheets available on the docket.
Seed and Oil Transport 
Based on information in the GCEH petition, we assumed that jatropha seeds are transported by truck 20 miles to a facility (for a 40 mile round trip), and that pre-treated jatropha oil is transported 75 miles by truck and then 500 miles by barge.  These distances are converted to emissions using the emission factors and equations in the Jatropha Production and Transport GHG Calculations spreadsheets posted to the docket.
Oil Extraction and Pretreatment 
We assumed that oil is extracted by screw press.  The electricity and fuel needed for oil extraction is based on data in Bailis and Baka (2010).  As explained above in the "Seed Yields and Oil Content" section, we used a seed oil content of 35%.  We use an oil yield of 75% based on a review of the literature.  For mechanical screw press extraction, the reported extraction efficiency is 70-80%.
Inputs for oil pretreatment include chemicals (such as sodium hydroxide) and electricity.  For these inputs, we used values from the petition.  These values are within the range of values reported in the literature.  
Consideration of Other LUC Scenarios 
As explained in the FR notice and above, we evaluated two main scenarios: one where jatropha is grown on otherwise unused grasslands in southern Mexico and northeastern Brazil and one where it is grown on land in southern Mexico and northeastern Brazil that would otherwise be used for crops or pasture.  While it is possible that jatropha could be grown on other types of land such as shrubland or secondary forest that would result in GHG emissions that are higher than soybean oil, the RFS program's qualification requirements for renewable biomass would limit the amount of this type of land that can be used to grow jatropha for the RFS program.  Furthermore, since jatropha is not currently being produced on a commercial scale and biofuel production is by far the largest potential market for jatropha oil (e.g., it is not suitable for cooking oil), EPA's approval of jatropha biofuel pathways for the RFS is unlikely to result in a situation where growing jatropha on RFS qualifying land would increase the amount of jatropha being grown on non-RFS qualifying lands for purposes other than production of renewable fuels (a situation that could result in significant indirect impacts).  Below we discuss information we considered to reach these conclusions, including studies on historical jatropha land use in Mexico and Brazil, the RFS renewable biomass provisions, and land use policies in Mexico and Brazil.
Historical Jatropha Land Use in Mexico and Brazil
Skutsch et al. evaluated land use change impacts for three jatropha case studies in Mexico.  The Mexican cases, established in 2007, evaluated smallholder systems in Chiapas and Michoacan states and a commercial production operation in Yucatan. They found two jatropha plantations in Yucatan, with combined area of 2,350 hectares that had been planted on estates that had been partially reforested after a period of low-intensity grazing.  The observed jatropha area in Chiapas and Michoacan was described as small, with no specific area figure provided.  In Chiapas, the authors found that 14 of the 21 jatropha farmers interviewed (all smallholders) had planted on fields previously used for maize and peanuts, whereas 6 farmers used pasture land and one had cleared a forest plot.  In other words, in Chiapas the case study looked at jatropha established on land previously used for 66% annual cropping, 29% pasture and 5% secondary forest.  In Michoacan much of the jatropha was planted by growers practicing a shifting cultivation system, of these growers more than half had cleared a patch of secondary forest to plant jatropha and the rest had used fields that had been maize the previous year.  In Michoacan jatropha was also observed on land permanently used for agriculture, and in a few areas of prime, irrigated land.  Overall, the previous land uses in Michoacan were 25% secondary forest, 25% fallow land (fallow for 8-11 years) and 50% annual cropping.   More recently, Soto et al. studied jatropha adoption by smallholders in the Mexican state of Chiapas.  Their survey found that 83% of jatropha producers cultivated in land occupied by subsistence crops (mostly maize and beans) and 2.3% used land previously occupied by cash crops.  According to information provided by GCEH about their plantations in southern Mexico, the initial set of parcels purchased in 2008 were almost 90% used for cattle grazing, with the remainder being low-density brush.  Subsequently, GCEH acquired land from an adjacent pig farmer that was consolidating operations.  In Brazil, Bailis and Baka surveyed jatropha producers located in central and northeast Brazil.  The producers that were approached had used primarily pasture land, with smaller areas of food crops and natural vegetation also cleared.  In general, we believe these findings support our approach of evaluating the GHG impacts of jatropha production on otherwise unused grassland, and jatropha production on land otherwise used for crops or pasture.  As discussed below, we believe the RFS renewable biomass provisions will limit the amount of other direct or indirect land use change impacts that are not already accounted for in our modeling. 
Renewable Biomass Provisions
While it is possible that jatropha could be grown on other types of land such as shrubland or secondary forest that would result in GHG emissions that are higher than what we estimated through our modeling, the RFS program's qualification requirements for renewable biomass would limit the amount of this type of land that can be used to grow jatropha for the RFS program.  For example, oil extracted from jatropha seeds grown on land that was forested or not actively managed or fallow on December 19, 2007 would not meet the RFS program definition of renewable biomass at 40 CFR 80.1401.  Per the regulatory definition at 40 CFR 80.1401, renewable biomass includes, "Planted crops and crop residue harvested from existing agricultural land cleared or cultivated prior to December 19, 2007 and that was nonforested and either actively managed or fallow on December 19, 2007."  There are a number of different types of land that may be suitable to grow jatropha from an agronomic or economic perspective, but that would not align with the RFS regulatory definition for renewable biomass, for example this includes all of the following types of land (not intended to be an exhaustive list):
   * Previously undisturbed forest.
   * Land that was not previously cleared or cultivated prior to December 19, 2007.
   * Land that was forested on December 19, 2007.
   * Land that was not actively managed or fallow on December 19, 2007.
In other words, the RFS regulations exclude the use of many of the types of land that would result in higher land use change GHG emissions than the types of land used for jatropha in our modeling.  However, the following types of land if used for jatropha production could result in higher land use change emissions than the types of land considered in our modeling: land that was actively managed or fallow on December 19, 2007 that grew back as secondary forest before being planted with jatropha, and such land that would grow back as secondary forest if it is not planted with jatropha.
EPA also considered information on government policies that could influence the types of land used for future jatropha production. In Mexico, jatropha production is promoted by CONAFOR through the ProAborl Programme, which aims to provide financial and technical assistance to all types of landowners to protect, conserve, restore and sustainably manage their forest resources.  Information provided as part of the GCEH petition indicates that CONAFOR does not promote jatropha plantations that would replace forested areas.  While EPA does not have data to verify the impacts of this policy, it could discourage the use of secondary forest land for jatropha production.  Brazil also has a number of policies to reduce deforestation rates, but EPA has not conducted a detailed review of how these policies may impact the types of land used for jatropha production in Brazil.
Overall, our review of available information on jatropha land use in Mexico and Brazil suggests that the scenarios we modeled adequately capture the likely land use change impacts of future jatropha production for use in making biofuels for the RFS program.  Based on our modeling, if relatively small areas of high carbon stock land are cleared for jatropha plantations we believe the conclusions described in the notice preamble would still apply because the GHG impacts would be offset by jatropha grown on the types of land considered in our modeling.


References
Achten, Wouter M. J., and Louis V. Verchot. 2011. "Implications of Biodiesel-Induced Land-Use Changes for CO2 Emissions: Case Studies in Tropical America, Africa, and Southeast Asia." Ecology and Society 16 (4). doi:10.5751/ES-04403-160414.
Achten, W.M.J., L.R. Nielsen, R. Aerts, A.G. Lengkeek, E.D. Kjaer, A. Trabucco, J.K. Hansen, W.H. Maes, L. Graudal, F.K. Akinnifesi, and B. Muys. 2010. "Towards Domestication of Jatropha Curcas." Biofuels 1(1):91-107. doi:10.4155/bfs.09
Achten, W.M.J., L. Verchot, Y.J. Franken, E. Mathijs, V.P. Singh, R. Aerts, and B. Muys. 2008. "Jatropha Bio-diesel Production and Use." Biomass and Bioenergy 32(12):1063-1084. doi:10.1016/j.biombioe.2008.03.003
Almeida, J. W.M.J. Achten, M.P. Duarte, B. Mendes, and B. Muys. 2011. "Benchmarking the Environmental Performance of the Jatropha Biodiesel System through a Generic Life Cycle Assessment." Environ. Sci. Technol. 45(12): 5447-5453. doi:10.1021/es200257m.
Bailis, Robert E. and Jennifer E. Baka. 2010. "Greenhouse Gas Emissions and Land Use Change from Jatropha Curcas-Based Jet Fuel in Brazil." Environ. Sci. Technol. 44(22): 8684-8691. doi:10.1021/es1019178.
Fan, J., R. M. Handler, D.R. Shonnard, and T.N. Kalnes. 2012. "A Review of Life Cycle Greenhouse Gas Emissions of Hydroprocessed Jet Fuels from Renewable Oils and Fats." International Journal of Environmental Science and Engineering Research 3(3):114-138.
Harris, NL, S. Grimland and S. Brown. 2009. Land Use Change and Emission Factors: Updates since Proposed RFS Rule. Report submitted to EPA, EPA-HQ-OAR-2005-0161
Intergovernmental Panel on Climate Change (2006). "N2O emissions from managed soils, and CO2 emissions from lime and urea application."  2006 IPCC Guidelines for National Greenhouse Gas Inventories, Volume 4: Agriculture, Forestry, and Other Land Use. Eggleston, H.S., Buendia, L., Miwa, K., Ngara, T., and Tanube, K (eds). Published: IGES, Japan.
Jongschaap, R.E.E., W.J. Corré, P.S. Bindraban, and W.A. Brandenburg. 2007. Claims and Facts on Jatropha curcas L. Global Jatropha curcas Evaluation, Breeding and Propagation Programme. Wageningen University, Wageningen, Netherlands. Report 158. Plant Research International B.V. Available at: http://www.fact-fuels.org/media_en/Claims_and_Facts_on_Jatropha_-WUR. Accessed on August 10, 2008.
Lapola, David M., Ruediger Schaldach, Joseph Alcamo, Alberte Bondeau, Jennifer Koch, Christina Koelking, and Joerg A. Priess. 2010a. "Indirect Land-use Changes Can Overcome Carbon Savings from Biofuels in Brazil." Proceedings of the National Academy of Sciences (February 8). doi:10.1073/pnas.0907318107.
Lapola, David M., Ruediger Schaldach, Joseph Alcamo, Alberte Bondeau, Jennifer Koch, Christina Koelking, and Joerg A. Priess. 2010b. "Supporting Information for Indirect Land-use Changes Can Overcome Carbon Savings from Biofuels in Brazil." Proceedings of the National Academy of Sciences (February 8). 
Reinhardt, G., K. Becker, D.R. Chaudhary, J. Chikara, E. von Falkenstein, G. Francis, S.O. Gartner, MR.R. Gandhi, A. Ghosh, P.K. Ghosh, H.P.S. Makkar, J. Münch, J.S. Patolia, M.P. Reddy, N. Rettenmaier, S.C. Upadhyay. 2008. "Basic Data for Jatropha Production and Use." Heidelberg, Germany: Institute for Energy and Environmental Research Heidelberg GmbH.
Skutsch, Margaret, Emilio de los Rios, Silvia Solis, Enrique Riegelhaupt, Daniel Hinojosa, Sonya Gerfert, Yan Gao, and Omar Masera. 2011. "Jatropha in Mexico: Environmental and Social Impacts of an Incipient Biofuel Program." Ecology and Society. 16(4): 11. doi:10.5751/ES-04448-160411. 
Soto, I., Achten, W.M.J, Muys, B., Mathjis, E. 2015. Who benefits from energy policy incentives? The case of jatropha adoption by smallholders in Mexico. Energy Poicy, 79: 37-47
Stratton, R.W., H.M. Wong, and J.I. Hileman. 2010. "Life Cycle Greenhouse Gas Emissions from Alternative Jet Fuels." Partnership for Air Transportation Noise and Emission Reduction Project 28. Cambridge, MA: Massachusetts Institute of Technology.

Trabucco, Antonio, Wouter M. J. Achten, Colm Bowe, Raf Aerts, Jos Van Orshoven, Lindsey Norgrove, and Bart Muys. 2010. "Global Mapping of Jatropha Curcas Yield Based on Response of Fitness to Present and Future Climate." GCB Bioenergy 2 (3): 139 - 151. doi:10.1111/j.1757-1707.2010.01049.x.
