Pasture Intensifica.on and Double Cropping as Mechanisms to Mi.gate iLUC Andre Nassar
IEA Bioenergy ExCo 74 Workshop: Land use and Mitigating iLUC
Brussels, October 23rd, 2014 www.agroicone.com.br
iLUC Factors Calcula.on: Two Steps
Es5ma5ng iLUC in Hectares
• Global equilibrium models (general or par5al)
• Some aFempts to use alloca5on procedures based on historical data
• AIer years of using models, although have been strongly improved, models are s5ll incomplete
• The consequence is that they tend to overstate iLUC because they are very conserva6ve in yields improvement
Integrated or not integrated in the same tool/model
Transla5ng iLUC in GHG Emissions
• Different models available • Some are spa5ally explicit others not • They rely on emissions factors • Since global models not always inform the types of land converted (but only the amount of conversion on forests and pastures), emissions models also allocate “iLUC in ha” over types of “non produc5ve” land
CARB LCFS Analysis (September 29th, 2014)
Evolu.on of iLUC Factors Results Over Time (from models) Corn Sugar cane Sugar beet Palm oil Rape oil Soy oil Methodology Searchinger et al., 2008 104.0 111.0 n.a. n.a. n.a. n.a. FAPRI CARB, 2009 30.0 46.0 n.a. n.a. n.a. 62.0 GTAP EPA, 2010 26.3 4.1 n.a. n.a. n.a. 43.0 FAPRI with Brazilian model, FASOM Hertel et al., 2010 27.0 n.a. n.a. n.a. n.a. n.a. GTAP E4Tech, 2010 n.a. 8.0-‐27.0 n.a. 8.0-‐80.0 15.0-‐35.0 9.0-‐67.0 Causal-‐descrip5ve approach Tyner et al., 2010 15.2-‐19.7 n.a. n.a. n.a. n.a. n.a. GTAP Al-‐Riffai et al., 2010 n.a. 17.8-‐18.9 16.1-‐65.5 44.6-‐50.1 50.6-‐53.7 67.0-‐75.4 MIRAGE Laborde, 2011 10.0 13.0-‐17.0 4.0-‐7.0 54.0-‐55.0 54.0-‐55.0 56.0-‐57.0 MIRAGE Marelli et al., 2011 13.9-‐14.4 7.7-‐20.3 3.7-‐6.5 36.4-‐50.6 51.6-‐56.6 51.5-‐55.7 MIRAGE and JRC methodology for emissions calcula5ons Moreira et al., 2012 n.a. 7.6 n.a. n.a. n.a. n.a. Causal-‐descrip5ve approach GREET1_2013 9.2 n.a. n.a. n.a. n.a. n.a. GREET CARB, 2014 23.2 26.5 n.a. n.a. n.a. 30.2 GTAP Laborde, 2014 13.0 16.0 7.0 63.0 56.0 72.0 MIRAGE and JRC methodology for emissions calcula5ons EllioF et al. 2014 5.9 n.a. n.a. n.a. n.a. n.a. PEEL Harfuch et al., 2014 n.a. 13.9 n.a. n.a. n.a. n.a. BLUM
References "Al-‐Riffai, P, Dimaranan, B and Laborde, D (2010). Global Trade and Environmental Impact Study of the EU Biofuels Mandate. Specific Contract No SI2.537.787 implemen5ng Framework Contract No TRADE/07/A2, Final report March 2010. " CARB (2009). CARB Staff Report: Proposed Regula5on to Implement the Low Carbon Fuel Standard. California Air Resources Board: March,, 2009. CARB (2014). iLUC Analysis for the Low Carbon Fuel Standard (Update). California Air Resources Board: March, 2014. E4Tech (2010). A Causal Descrip5ve Approach to Modeling the GHG Emissions Associated with the Indirect Land Use Impacts of Biofuels. Final report. A study for the UK Department for Transport, October 2010. Elliot, J.; Sharma, B.; Best, N.; GloFer, M.; Dunn, J. B.; Foster, I.; Miguez, F.; Mueller, S.; Wang, M. A Spa5al Modeling Framework to Evaluate Domes5c Biofuel-‐Induced Poten5al Land Use Changes and Emissions. Environmental Science and Technology, 2014, 48 (4), pp 2488–2496 (DOI: 10.1021/es404546r). EPA (2010). Renewable Fuel Standard Program (RFS2) Regulatory Impact Analysis. Assessment and Standards Division, Office of Transporta5on and Air Quality, U.S. Environmental Protec5on Agency: February, 2010. GREET1_2013. GREET Model: The Greenhouse Gases, Regulated Emissions, and Energy Use in Transporta5on Model. Argonne Na5onal Laboratory. Harfuch, L.; Bachion, L.C.; Moreira M.M.R.; Nassar, A.M.; Carriquiry, M. 2014. Agricultural Expansion and Land Use Changes in Brazil: Using Empirical Evidence. In: Handbook of Bioenergy Economics and Policy (in press). Springer, 2014. "Hertel, T.; Golub, A. A.; Jones, A. D; O'Hare M.; Plevin, R. J.; Kammen, D. M. Globalland use and greenhouse gas emissions impacts of U.S. maize ethanol: es5ma5ng marketmediatedresponses. Biosci. 2010, 60 (3), 223–231; DOI: 10.1525/bio.2010.60.3.8." Laborde, D. (2011). Assessing the Land Use Change Consequences of European Biofuel Policies. Final Report, IFPRI, October 2011. Laborde, D.; Padella, M.; Edwards, R.; Marelli, L. Progress in Es5mates of iLUC with Mirage Model. Joint Research Center, Report EUR 26106 EN. Marelli, L.; Ramos, F.; Hiederer, R.; Koeble, R. (2011) Es5mate of GHG emissions from global land use change scenarios. JRC Technical Notes. EUR 24817 EN -‐ 2011 Moreira, M.; Nassar, A.; Antoniazzi, L.; Bachion, L. C.; Harfuch, L. (2012). Direct and indirect land use change assessment. In: Poppe, M. K.; Cortez, L. A. B. Sustainability of sugarcane bioenergy. Center for Strategic Studies and Management (CGEE), 2012. "SEARCHINGER, T.; HEIMLICH, R.; HOUGHTON R.A.; DONG, F.; ELOBEID, A.; FABIOSA J. Use of UScroplands for biofuels increases greenhouse gases through emissions from land-‐use change. Sciencev. 319, p. 1238–1240. 2008." "Tyner, W. E.; Taheripour, F.; Zhuang, Q.; Birur, D.; Baldos, U. Land use changes andconsequent CO2 emissions due to US corn ethanol produc5on: a comprehensive analysis; PurdueUniversity, West LafayeFe, Indiana, 2010;hFps://www.gtap.agecon.purdue.edu/resources/download/5200.pdf."
Op.ons for mi.ga.ng iLUC (the ones more relevant in my opinion)
• Reducing deforesta.on over .me through policies, monitoring, command and control sanc.ons, land use planning, zoning
• However, out of the scope of bioenergy systems
• Increasing the yields of individual crops, induced by technological improvement or price induced
• Making land more produc.ve • Reducing yield gaps on crops • Increasing produc.vity in grass-‐fed caWle systems • Integra.on systems: double-‐cropping and crop-‐livestock
• Developing crops suitable for marginal, degraded or low precipita.on lands
Op.ons for mi.ga.ng iLUC: a closer look Op6ons Opportuni6es/weaknesses
Reducing deforesta5on § Very important but long term § Requires government empowerment § Much broader agenda than biofuels § Models are shy in this issue
Increasing the yields of individual crops § Can bring posi5ve effects in the short term § But rate of yields increase in crops is decreasing (contribu5on is low) § GMO § Model capture that effect
Reducing yields gaps on crops § It can have huge effects § Require capacity building § Long term § Models capture but tend to be conserva5ve
Increasing produc6vity in grass-‐fed caFle systems § It can have even larger effects given that 2/3 of agricultural land is used for grazing
§ Models are conserva6ve and pasture intensifica6on is a consequence not a driving force (CETs and compe66on elas6ci6es are not calibrated to achieve real pasture intensifica6on
Integra6on systems: double-‐cropping and crop-‐livestock
§ It is a reality but it is not captured by models. § Short term
Developing crops suitable for marginal, degraded or low precipita5on lands
§ Long term § Probably will have lower effects than pasture intensifica5on
Double Cropping
Soy-‐Corn Double Copping System
• In the same raining season (Summer) soy and corn are cul.vated. Soy is planted firstly, in the beginning of Spring (September-‐October) and harvested in January. Corn is planted just a[er soy is harvested and it is harvested in mid Autumn.
• Short cycle soy varie.es were developed to allow corn being planted a[er soy • Requires 3 to 5 years to prepare the soil, recover fer.lity and organic maWer to op.mize produc.on
• 100% no .ll • 90% of corn planted area is rain fed • Very efficient in energy use and carbon footprint reduc.ons
• Potassium and phosphorus use is op.mized • Only addi.onal fer.lizer is nitrogen (because soy does not require nitrogen) • Requires herbicides for the no .ll cul.va.on
• The system has saved around 9 million ha in the last 10 years: reduc.on in the first crop area (around 3 million ha) and increase in the second crop (6 million)
Corn Produc.on in Brazil
• Currently, each addi.onal 1 ha of soy leads to: • 0.17 ha reduc.on in corn 1st crop • 0.50 ha increase in corn 2nd crop • Given that corn 2nd crop has higher yields than corn 1st crop, effects in produc.on are even higher
Variable Corn harvest 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Area (1000ha)
1st crop 9.690 9.381 8.581 9.280 9.686 9.422 8.511 6.864 7.508 6.895 6.749
2nd crop 3.276 3.030 2.968 3.333 4.082 5.022 4.381 5.043 5.711 7.304 8.967
Produc6on (1000 t)
1st crop 35.028 31.349 27.161 31.485 37.658 39.829 30.705 29.852 33.488 32.819 34.157
2nd crop 13.299 10.439 7.952 11.177 14.455 19.105 16.367 21.568 22.172 38.254 46.381
Source: IBGE.
Sources: CONAB; IBGE; ABRAF; ÚNICA; BLUM
Corn Produc5on (million tons) 1,000 ha Planted Area
0
2,000
4,000
6,000
8,000
10,000
12,000
0
10,000
20,000
30,000
40,000
50,000
60,000
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Soybean Corn 1st Crop CoFon
Eucalyptus and Pinus Sugarcane Corn 2nd Crop
34 31 27 31 36 40 34 34 36 34 36
13 11
8 11
15 19
17 22 21 39
46
0
10
20
30
40
50
60
70
80
90
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Corn 1st Crop Corn 2nd Crop
Corn 2nd crop 1,000 ha
Corn Double Cropping: Planted Area and Produc.on
Brazil: Annual Crops Expansion
11 Source: Agrosatelite
Soy-‐Corn System Environmental Indicators
20 30 40 50 60 70 80 90 100 110 120
Soybean Land Use Efficiency
Energy/tons Carbon Intensity/tons
Soybean-‐Corn (2004-‐2006=100)
2004-‐2006
2007-‐2009
2010-‐2012
Source: Agroicone
Pasture Intensifica.on
Pasture Intensifica.on
• Roughly 2/3 of agricultural area worldwide is occupied with pastures and meadows. We do not know how much is natural and how much is planted/managed. But we have a good idea of which regions can increase pasture produc.vity
• Situa.on in Brazil • Planted/managed pastures: 115 million ha (of which 10 million is considered degraded) • Natural (but used for caWle raising): 60 million • Produc.vity is growing but s.ll is below the poten.al
• Majority of pastures is located in areas non suitable for crops. However, in specific regions such as Brazil, central and eastern Africa, which represents 18% of total grasslands, there are strong poten.al for intensifica.on
• Pasture intensifica.on means • Grass-‐fed caWle raising systems with poten.al to produce more meat per ha, without increasing caWle herd, but reducing pasture area
• It is a func.on of adapted animals (gene.cally improved), managed pasture, rota.on grazing and some specializa.on (calf crop, yearling, finishing)
Pastureland Worldwide
Source: IIASA; FAO Hectares Arable land and Permanent crops 1,562,548 Permanent meadows and pastures 3,359,659
Factors Explaining Beef Produc.on Expansion in Brazil
Source: Martha Jr, G. B.; Alves, E.; Con5ni, E. Land-‐saving approaches and beef produc5on growth in Brazil. Agricultural Systems 110 (2012). hFp://dx.doi.org/10.1016/j.agsy.2012.03.001.
Produc.vity is growing
Source: Juliano Assunção, CPI.
Sources: IBGE, UFMG, INPE, BIGMA Consul5ng, Agroicone.
179,000
180,000
181,000
182,000
183,000
184,000
185,000
-
10.00
20.00
30.00
40.00
50.00
60.00
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012Livestock yield Pasture Area
(kg meat / ha) (1000 ha)
2002 2012 Varia6on CAGR (%)
Pasture area (1000 ha) 184,037 180,785 -‐3,252 -‐0.14% Herd (1000 Head) 185,349 213,239 27,890 0.98%
Meat produc6on (1000 MT) 7,139 9,748 2,609 2.64%
Livestock yield (kg of meat/ha) 39 54 15 2.78%
Milk produc6on (1000 liters) 24,172 33,996 9.824 3.6%
Milk produc6on per cow (liters/cow) 1,286 1,479 193 1.4%
Yield Improvement: livestock
Intensifica.on poten.al
Source: Bernardo B.N. Strassburg Agnieszka E. Latawie, Luis G. Barioni, Carlos A. Nobre, Vanderley P. da Silva, Judson F. Valen5m, Murilo Vianna, Eduardo D. Assad. When enough should be enough: Improving the use of current agricultural lands could meet produc5on demands and spare natural habitats in Brazil. Global Environmental Change 28, p. 84–97, 2014.
• Managed pastures: 115 million ha • Current carrying capacity: 94 million
animals (produc5vity of 0.81 AU/ha) • Poten5al carrying capacity: 274-‐293
million animals
159.4
316.3 386.6 391.8
254.3
105.6 85.2
236 290.3 275.9
121.4
-37.2 1 - 3 @/ha 3 - 6 6 - 12 12 - 18 18 - 26 26 - 38 @/
ha
77.2 130.7
342.3 540.3 574.9
817.5
5.2 12.8 185.8 427.9 460.7
702.9
1 - 3 @/ha 3 - 6 6 - 12 12 - 18 18 - 26 26 - 38 @/ha
73 175
317 364 484
712
00 76 158
233 351
579
1 - 3 @/ha 3 - 6 6 - 12 12 - 18 18 - 26 26 - 38 @/ha
R$/hectare/year
Gross profit Operating profit
Livestock technology: small improvement, large response
Calf breeding
Growth and termination
Investment, yields and returns
40% 31% 30% 30%
56% 64% 65% 65%
0%
20%
40%
60%
80%
100%
2012 Basline 2020 PNMC 2020 DZ 2020
Brazilian livestock technology profile
Low (0-3 @/ha) Average (3-6) High (>6)
• Technology profile (@/ha): low (0-3 @/ha); average (3-6); high (>6)
• 2012: 40% of production on low tech; 56% on average; 4% on high; average production per hectare 3.63 @/ha; 10 mm tons of beef; 181 mm of ha pasture
• 2020 Baseline: 31-64-5%; 4.65 @/ha; 3.5 mm ha less; 12 mm tons of beef
• 2020 ZD: 30-65-5%; 4.78 @/ha; 10 mm ha less; 12 mm tons of beef
• Calf breeding presents decreasing profitability when tech is higher than 12@/ha – limit to intensification;
• Termination with high tech: competitive profitability compared to grains, not considering investments costs.
Complete cycle