1 Title: Soy Expansion in Brazil’s Cerrado
2 Authors: Lisa L. Rausch1*, Holly K. Gibbs2, Ian Schelly3, Amintas Brandão Jr. 4, Douglas C. Morton5, 3 Arnaldo Carneiro Filho6, Bernardo Strassburg7, Nathalie Walker8, Praveen Noojipady9, Paulo Barreto9, 4 Daniel Meyer10
5 Affiliations:
6 1. Corresponding author: Center for Sustainability and the Global Environment (SAGE), 7 Nelson Institute for Environmental Studies, University of Wisconsin-Madison, 1710 8 University Ave, Madison, WI 53711, USA, [email protected], telephone: +1.608.890.0037, 9 fax: +1.608.265.4113 10 2. Department of Geography, Center for Sustainability and the Global Environment (SAGE), 11 Nelson Institute for Environmental Studies, University of Wisconsin-Madison, USA, 12 [email protected] 13 3. Center for Sustainability and the Global Environment (SAGE), Nelson Institute for 14 Environmental Studies, University of Wisconsin-Madison, USA, [email protected] 15 4. Center for Sustainability and the Global Environment (SAGE), Nelson Institute for 16 Environmental Studies, University of Wisconsin-Madison, USA, [email protected] 17 5. NASA Goddard Space Flight Center, Greenbelt, MD, USA, [email protected] 18 6. Global Canopy, Oxford, United Kingdom, [email protected] 19 7. Department of Geography and the Environment Pontifical Catholic University of Rio de 20 Janeiro (PUC-Rio) and International Institute for Sustainability, Rio de Janeiro, Brazil, 21 [email protected] 22 8. National Wildlife Federation, Washington, DC, USA, [email protected] 23 9. NASA Goddard Space Flight Center, Greenbelt, MD, USA, [email protected] 24 10. Global Canopy, Oxford, United Kingdom, [email protected] 25 26 27 Running title: Soy Expansion in Brazil’s Cerrado 28 29 Keywords: Brazil, commodity agriculture, deforestation, land use change, policy analysis, property- 30 level analysis, soy, supply-chain policies 31 32 Type of article: Letter 33 34 Number of words in abstract: 140 35 36 Number of words in manuscript (main text): 3,000 37 38 Number of references (in main text): 40 39 40 Number of Tables and figures (main text): 6 41
1 42 Abstract:
43 The Cerrado Biome is Brazil’s breadbasket and a major provider of ecosystem services, though these
44 dual roles are increasingly at odds. We conducted a comprehensive assessment of Cerrado conversion
45 for the rapidly expanding, high value crop of soybeans using 580,000 property boundaries and time
46 series of satellite data. Soy was a major driver of Cerrado conversion during 2003-2014 (22%) and only
47 15% of all Cerrado conversion exceeded Forest Code restrictions for clearing on private properties.
48 However, soy farms were five times more likely to be non-compliant than other farm types. As a leading
49 cause of both Cerrado conversion and non-compliance, the soy sector has environmental and economic
50 incentives to shift production to existing cleared lands. Combined, property boundaries and suitability
51 maps highlight sustainable solutions that mitigate the environmental impacts of further soy expansion
52 across both old and new cropland frontiers.
53 Introduction
54 The Brazilian Cerrado is a vast neotropical savanna and Brazil’s most productive agricultural
55 region (Lambin et al 2013; Rada et al 2013; Cardoso da Silva & Bates 2002), with 61% of the country’s
56 soy area and half the cattle herd (Noojipady et al 2017; Lapola et al 2014). By 2015, 42% of natural
57 Cerrado vegetation (86 Mha) had been cleared for agricultural use. One-fifth (17 Mha) of this cleared
58 land was planted with soy and another 44% (38 Mha) was used as pasture. Much of this conversion
59 occurred decades ago (Strassburg et al 2017, Sano et al 2010; Klink & Machado 2005), leading to the
60 Cerrado’s current mosaic formations including secondary vegetation from abandonment or rotational
61 management (Jepson, 2005).
62 Extensive clearing is possible because the Cerrado is afforded relatively little legal protection,
63 especially compared with the Amazon biome. Brazil’s Forest Code (“FC”) permits agricultural use on 65-
64 80% of a property with Cerrado vegetation, compared to only 20% of properties with Amazon forest
2 65 vegetation. Protected areas cover 13% of the Cerrado, compared to 46% of the Amazon (Strassburg et
66 al 2017; Carranza et al 2014). Additionally, private-sector zero-deforestation commitments for soy and
67 cattle have been in place in the Amazon since 2006 and 2009, respectively, but do not currently include
68 the Cerrado (Gibbs et al 2015; Gibbs et al 2014; Macedo et al 2012).
69 Continued Cerrado conversion will have major impacts on ecosystem services, which may
70 accelerate the projected transition to hotter temperatures and lower rainfall, thereby threatening
71 agricultural productivity in both the immediate and the longer terms (Coe et al 2017; Spera et al 2016;
72 Arante et al 2016). Ongoing clearing also threatens the region’s endemic flora and fauna (Strassburg et
73 al 2017; Vieira et al 2017). Loss of remaining carbon stocks in the Cerrado vegetation could significantly
74 reduce Brazil’s ability to meet Nationally Determined Commitments for greenhouse gas emissions (Coe
75 et al 2017; Aguiar et al 2016). Ongoing conversion also threatens water supplies to more than 29 million
76 Cerrado residents (Strassburg et al 2017). In recognition of the need for a balanced development
77 strategy, a growing number of large companies with substantial economic interests in the Cerrado have
78 made public commitments to eliminate deforestation from their supply chains (United Nations 2014;
79 Consumer Goods Forum 2010), which raises the stakes for ongoing clearing.
80 Here, we provide a detailed study of the role of soy expansion in recent land use trends and a
81 farm-level assessment of the potential conservation and sector impacts of efforts to limit conversion in
82 the Cerrado. We characterized land use at the farm level, including FC compliance on properties
83 observed to produce more than 25ha of soy (“soy farms”) and those farms that produced cattle or other
84 types of crops (“non-soy farms”), to demonstrate the role that soy plays in Cerrado clearing and to
85 assess pathways for future soy expansion. We also considered potential impacts of zero-conversion
86 policies in the Cerrado, based on updated maps of soy suitability, evidence from field visits, and a survey
87 of over 500 soy buyers.
3 88 Our study builds on previous work that considered expansion of all types of croplands, and that
89 estimated legal reserve surpluses and deficits for rural properties overall (Noojipady et al 2017; Freitas
90 et al 2017; Gibbs et al 2014; Soares Filho et al 2014; Sparovek et al 2012). We go further and provide
91 analyses of soy properties, along with regional details, that can support evidence-based decision-making
92 by stakeholders in Brazil’s soy sector, for whom details about soy properties compared to other
93 properties are fundamental. Previous studies have lacked the combination of property boundaries, soy
94 expansion maps, and soy suitability maps used here to isolate the impacts of soy from other land uses
95 and to assess how current and potential soy properties would be affected by new policies.
96 METHODS
97 Assessing causes of annual clearing
98 To assess clearing for soy in the Cerrado, we created a map of annual soy expansion between
99 2001 and 2014, based on maps of annual soy extent from Agrosatélite and NASA. To determine how
100 much clearing was caused by soy expansion each year in the Cerrado, we intersected the annual LAPIG
101 conversion maps with our dataset of annual soy expansion, and then tracked land use for the three
102 years following deforestation (using an approach similar to Gibbs et al 2015; see Supplement sections 2
103 and 3.2).
104 Property-level assessments
105 We created a map of private properties by combining 573,638 boundaries from Brazil’s Rural
106 Environmental Registry (Portuguese acronym: CAR), combined with an additional 140, 311 private
107 property boundaries from the National Institute of Colonization and Agrarian Reform (Incra) (see
108 supplement section 5). Our final property dataset covers 62% of the areas under or eligible for private
109 ownership and 77% of the 2014 soy area.
4 110 We assessed the legality of annual clearing and the potential for continued legal clearing using
111 our map of private properties. The required area of the Legal Reserve set-asides on each property was
112 determined using a map by (Soares Filho et al 2014) that spatially allocates legal reserve requirements
113 according to the FC (see supplement section 5.1). We used LAPIG to estimate the annual cleared area on
114 each property and determined whether this clearing went beyond these thresholds, at which point it
115 was considered illegal. We conducted a robustness analysis using 12th order watersheds as proxies for
116 private properties following (Soares Filho et al 2014).
117 Properties with the potential for legal clearing were identified by overlaying a natural vegetation
118 map created by combining non-cleared areas in Lapig with native vegetation areas identified by Inpe’s
119 TerraClass 2013 map (see Supplement section 5.2). Properties where natural vegetation exceeded
120 required amounts were considered to have the potential for additional legal clearing.
121 Land available for soy expansion
122 We estimated the area and soy suitability of different types of land in the Cerrado: 1) areas
123 under native vegetation, according to the map described above; 2) areas of native vegetation that can
124 be legally cleared under the FC; and 3) areas that have already been cleared (those areas outside of
125 native vegetation areas and excluding 2014 soy production areas, including areas planted to corn and
126 cotton). We used 12th order watersheds so that we would have complete coverage.
127 To determine the amount of vegetation that could be legally cleared under the FC, we summed
128 the native vegetation identified by the map described above in each watershed. We considered any area
129 of remaining native vegetation that exceeded the legal reserve requirement for a given watershed, as
130 determined by the map described in section 5.1 of the supplement, to be available for legal clearing.
5 131 We then overlaid these maps with our suitability maps. For watersheds with a surplus of native
132 vegetation beyond the Legal Reserve requirement, we assumed that any suitable areas would be the
133 most likely to be cleared and gave those areas preference when summing the vulnerable suitable area.
134 For example, in a watershed with 100ha of native vegetation that may be cleared, and 150ha of area
135 that is suitable for soy, we assumed that all 100ha of native vegetation that could be cleared would
136 include suitable area.
137 RESULTS
138 Soy expansion is a major driver of Cerrado clearing
139 Soy expansion played an important role in recent clearing in the Cerrado. Direct conversion for
140 soy accounted for 16-32% of annual clearing between 2003 and 2014 [fig 1 and table 1]. In total, soy
141 replaced at least 1.3 Mha of native vegetation during this period, an average of just over 108,000 ha per
142 year. Soy-driven conversion was highest in Mato Grosso and in Matopiba, a region that encompasses
143 the portions of Maranhão, Tocantins, Piauí, and Bahia states that fall in the Cerrado biome. Between 6
144 and 33% of annual clearing in the Cerrado biome portion of Mato Grosso and between 24 and 46% of
145 annual clearing in Matopiba was planted directly to soy. Soy expansion was responsible for only 2 to
146 11% of the conversion each year in the Southern Cerrado.
147 Illegal clearing is concentrated on soy properties
148 Most Cerrado conversion between 2003 and 2014 did not exceed the legal limits established by
149 the FC. Just 15% of clearing that took place on our mapped properties in 2003 or after converted the
150 “legal reserve” set-aside areas of native vegetation required by the FC, though this amounts to a non-
151 trivial amount of nearly 700,000 ha [table 2]. At 35%, the Cerrado portion of Mato Grosso had the
6 152 highest annual rates of clearing in these legal reserves, with the lowest rates in older frontier areas of
153 the Southern Cerrado (9%).
154 Soy farms were disproportionately likely to have cleared in excess of FC allowances. They
155 represent only 7% of the properties we mapped (42,256 out of 584,161 total properties), but 44% of the
156 7,080 properties that cleared beyond their legal limits between 2003 and 2015. The region with the
157 most excess clearing on soy properties has shifted over time. Soy farms with excess clearing between
158 2003 and 2008 were concentrated in Mato Grosso (59%, 1247 out of 2,120 total farms) while those that
159 exceeded clearing allowances after 2008 were concentrated in Matopiba (52%, 725 out of 1,389 total
160 farms). Including clearing that took place prior to 2003, at least 51% of soy farms across the Cerrado lack
161 the on-property legal reserves required to clear more land without violating the FC, a figure nearly five
162 times the rate for non-soy properties (11%).
163 Millions of hectares of native Cerrado are vulnerable under the FC
164 More than one third of the remaining Cerrado vegetation outside of protected areas could be
165 legally cleared under the FC [38 Mha; table 3]. Of the land that could be legally cleared, one quarter (8
166 Mha) is highly suitable for soy, and another 7 Mha is potentially suitable according to more general
167 criteria. Nearly half of the highly suitable land that could be legally cleared is in the agricultural hotspot
168 of Matopiba (3.8 Mha, 48%) where recent, rapid soy expansion puts vegetation at particular and
169 immediate risk [fig. 2, table 3]. Nearly the same amount falls within areas identified by the government
170 as particularly important for conservation (3.3 Mha, 42%). Highly suitable land that can be cleared
171 legally covers nearly 5 times the area cleared for soy since 2003, which suggests that conversion could
172 continue legally for decades at current rates, although this conversion could generate feedbacks that
173 ultimately reduces the suitability of many regions (Spera et al 2016; Pires et al 2016).
7 174 Most of the highly suitable land that could be legally cleared is located on non-soy properties
175 (83%; 4.7Mha). Only 12% (4,865) of soy farms had ≥25 ha of these lands, and only 4% (1,829) had ≥100
176 ha [Table 4]. The highest concentration was in Matopiba, where nearly 19% of soy farms can still
177 expand into native vegetation.
178 Soy area can expand without additional clearing
179 The Cerrado’s soy sector can grow without additional conversion of natural vegetation. In 2015,
180 there were approximately 23 Mha of cleared land that were highly suitable for soy and another 15 Mha
181 that were potentially suitable (38 Mha total, outside of the current soy extent). The highly suitable
182 cleared areas are enough to more than double the current production area. Soy areas could as much as
183 triple if potentially suitable lands were also utilized.
184 Most cleared lands suitable for soy are accessible to markets. The majority of cleared land
185 deemed highly suitable for soy was within 10 km of a major federal or state highway (19 Mha, 82%) and
186 within 100km of a silo or crushing facility owned by a major soy trading company (18 Mha, 79%). A
187 survey of agricultural storage facilities (“silos”) indicated that there are hundreds of additional
188 commercial soy silos across the Cerrado that receive soy on behalf of major trading companies or that
189 sell soy to these major traders; accounting for these silos would at least double the number of units
190 receiving soy (see Supplement section 7).
191 However, other activities such as ranching and production of sugarcane will limit how much of
192 this area is ultimately available for soy (Strassburg et al 2014; Cohn et al 2014). As little as 2.8 Mha (12%)
193 of the total highly suitable and cleared area was found on current soy producing properties. Over 80%
194 of the highly suitable and cleared lands (18 Mha) are located in the Southern Cerrado [fig. 2], where
195 recent soy expansion has been slow compared to Matopiba and Mato Grosso. Sugarcane occupies
8 196 around 12% (2.8 Mha) of the cleared and suitable land, nearly exclusively in the Southern Cerrado, and
197 most of the rest is under pasture (70% of the cleared suitable land; 16 Mha).
198 DISCUSSION
199 Our results show that soy has played a critical role in recent conversion in the Cerrado. Indeed,
200 conversion rates in the Cerrado were higher than deforestation rates in the Amazon prior to the “Soy
201 Moratorium” zero-deforestation commitment by major soy companies there (Supplement section 4.1).
202 This loss of native vegetation may put the future of the industry at risk by reducing precipitation and
203 increasing temperatures in important soy producing regions (Arante et al 2016). Reductions in
204 biodiversity and carbon stocks could generate impacts felt not just locally, but well beyond the extent of
205 the Cerrado (Coe et al 2017).
206 Recent moves by the government to target companies that bought soy from land embargoed for
207 illegal deforestation and to launch Cerrado monitoring systems are important (MMA 2018; Inpe 2018;
208 Fearnside 2016), but improved legal enforcement will not limit most clearing for soy, which is within
209 legal limits. Moreover, nearly half of the remaining vegetation that can be cleared legally is on land with
210 potential for soy production, but private-sector policies that further restrict clearing by suppliers to soy
211 traders could protect additional vegetation. In fact, such policies are likely the only way for companies
212 to meet commitments to eliminate deforestation from their supply chains by 2020 (United Nations
213 2014; Consumer Goods Forum 2014). Zero-deforestation policies could result in minimal constraints on
214 current soy farms because most (83%) have no remaining soy-suitable areas left to clear legally. Instead,
215 market restrictions could provide important limits on speculative acquisition and clearing of new
216 properties for soy production, and encourage the use of ample already-cleared and suitable areas.
217 Additional public or private policy measures may be needed to support use of already cleared
218 areas, which could involve higher costs due to competition with other activities and to distance from
9 219 established soy farms. However, even moderate intensification in the cattle sector, such as increasing
220 the number of animals per hectare, could make room for soy expansion (Strassburg et al 2017;
221 Strassburg et al 2014; Cohn et al 2014); indeed, conversion of pasture was the pathway for more than
222 half of soy expansion between 2001 and 2014 (Rudorff et al 2015). Increased integration of soy
223 production with cattle ranching could also support expansion of soy on cleared areas (Gil et al 2016;
224 Rausch & Gibbs 2016). Similarly, policies supporting improved efficiency in soy farming could increase
225 production while reducing the need for expansion (Battisti et al 2018; Rada 2013). However, these
226 types of “land-sparing” approaches must be accompanied by a policy mix that limits clearing in other
227 sectors to avoid perverse outcomes such as rebound-effects (Arima et al 2011).
228 Private sector efforts to eliminate conversion for soy in the Cerrado will face challenges that
229 were not present when the Soy Moratorium implemented in the Amazon. First, surveys of soy farmers
230 and buyers indicate that the influence of large companies on producer land management decisions is
231 more limited in the Cerrado, where as much as 40% of the soy produced is consumed domestically,
232 largely as animal feed. When the Soy Moratorium was implemented, the 28 signatory trading
233 companies purchased 90 percent of Amazon soy directly from farmers, largely for export (Godar et al
234 2015; Gibbs et al 2014; Brannstrom et al 2012). The same companies purchase large quantities of soy in
235 the Cerrado, but their combined market share is lower than it was in the Amazon a decade ago. Unlike
236 in the Amazon, traders in the Cerrado also purchase from smaller, local firms that act as intermediaries
237 with producers; these firms are not subject to the same consumer pressure as traders due to their lack
238 of global presence [table 4]. However, strategies that involve local buyers and smaller traders may be
239 necessary to avoid leakage, as has been observed in the Amazon under the zero deforestation cattle
240 agreements (Gibbs et al 2015).
10 241 Second, simply monitoring land use change in the Cerrado is more difficult than in the Amazon,
242 as pasture and natural grasslands can appear similar in satellite imagery, which has complicated
243 previous efforts to distinguish prior land use and land cover in areas of soy expansion (Müller et al 2014;
244 Sano et al 2010; Brannstrom et al 2008). However, the newly-released PRODES Cerrado maps, based on
245 the well-established PRODES system for the Brazilian Amazon, are a promising development (MMA
246 2018).
247 Conclusion
248 The spotlight that has long been on the environmental consequences of agricultural production
249 in the Amazon is now widening to include Brazil’s Cerrado. We show that curtailing soy-driven
250 conversion would eliminate a major incentive for clearing, and that current cleared land could
251 accommodate considerable future expansion. Solutions, however, will require a mix of policy and
252 economic incentives to offset the potential higher costs of already cleared land. Additional research is
253 urgently needed to better quantify the climatic costs of continued conversion, the opportunity costs of
254 leaving native vegetation intact, and the economic costs of alternate production strategies.
255 Soy expansion was limited in the most recent crop year (2017) and newly released PRODES data
256 suggest that Cerrado clearing has slowed since 2015 (MMA 2018; Conab 2018). These developments
257 should be seen as an opportunity for ongoing efforts to eliminate soy expansion as a driver of Cerrado
258 clearing. As shown with the Amazon Soy Moratorium a decade ago, declining trends in conversion may
259 signal an opportune moment to shift away from clearing for soy. However, ending conversion from soy
260 expansion will not be enough; pasture expansion must also be addressed to avoid spillover effects.
11 261 References
262 Aguiar, A. P. D., Vieira, I. C. G., Assis, T. O., Dalla-Nora, E. L., Toledo, P. M., Oliveira Santos-Junior, R. A., 263 Batistella, M., Coelho, A. S., Savaget, E. K., Aragão, L. E. O. C., Nobre, C. A. and Ometto, J. P. H. (2016), 264 Land use change emission scenarios: anticipating a forest transition process in the Brazilian Amazon. 265 Glob Change Biol, 22: 1821–1840. doi:10.1111/gcb.13134.
266 Arante, A.E., L. G. Ferreira, M.T. Coe. The seasonal carbon and water balances of the Cerrado 267 environment of Brazil: Past, present, and future influences of land cover and land use. ISPRS Journal of 268 Photogrammy and Remote Sensing 117, 66 (2016).
269 Arima, E.Y., Richards, P., Walker, R., Caldas, M.M. Statistical confirmation of indirect land use change in 270 the Brazilian Amazon. Environmental Research Letters 6 (2011).
271 Battisti, R., P.C., Sentelhas, J.A.L. Pascoalino, H. Sako, J.P. d. S. Dantas, M. F. Morares. Soybean yield gap 272 in the areas of yield contest in Brazil. International Journal of Plant Production 273 https://doi.org/10.1007/s42106-018-0016-0 (2018)
274 Brannstrom, C., Rausch, L., Brown, J.C., Andrade, R.M.T.d., Miccolis, A. Compliance and market exclusion 275 in Brazilian agriculture: Analysis and implications for “soft” governance. Land Use Policy 29, 357 (2012).
276 Brannstrom, C., W. Jepson, A.M. Filippi, D. Redo, Z. Xu, S. Ganesh, Land change in the Brazilian Savanna 277 (Cerrado), 1986-2002: Comparative analysis and implications for land-use policy. Land Use Policy. 25, 278 579 (2008).
279 Cardoso da Silva, J.M., J.M. Bates, Biogeographic patterns and conservation in the South American 280 Cerrado: a tropical savanna hotspot. BioScience 52, 225 (2002).
281 Carranza, T., A. Balmford, V. Kapos, A. Manica, Protected area effectiveness in reducing conversion in a 282 rapidly vanishing ecosystem: the Brazilian Cerrado. Conservation Letters 7, 216 (2014).
283 Coe, M.T., P. M. Brando, L.A. Deegan, M.N. Macedo, C. Neill, D.V. Silvério, The forests of the Amazon 284 and Cerrado moderate regional climate and are the key to the future. Tropical Conservation Science 10 285 (2017)
286 Cohn, A., Mosiner, A., Havliík, P., Valin, H., Herrero, M., Schmid, E., O’Hare, M., Obersteiner, M. Cattle 287 ranching intensification in Brazil can reduce global greenhouse gas emissions by sparing land from 288 deforestation. PNAS 111, 7236 (2014).
289 Conab (Companhia Nacional de Abastecimento) Série histórica das safras – soja. (Available online: 290 https://www.conab.gov.br/info-agro/safras/serie-historica-das-safras?start=20) (2018).
291 Consumer Goods Forum, Deforestation Resolution 2010 (available at: 292 https://www.theconsumergoodsforum.com/press_releases/consumer-goods-industry-announces- 293 initiatives-on-climate-protection/)
294 Consumer Goods Forum, Statement of support for the objectives of the Cerrado Manifesto (available at 295 http://www.theconsumergoodsforum.com/files/resources/sustainability/deforestation/Letter-of- 296 business-support-for-Cerrado-Manifesto-2017.pdf) (2017).
297 Fearnside, P.M. Brazilian politics threaten environmental policies. Science 353, 746 (2016).
12 298 Freitas, F.L.M.d, Sparovek, G., Mörtenbert, U., Silveira, S., Klug, I., Berndes, G.. Offsetting legal deficits of 299 native vegetation among Brazilian landholders: Effects on nature protection and socioeconomic 300 development. Land Use Policy 68, 189 (2017).
301 Gibbs, H. K., J. Munger, J. L’Roe, P. Barreto, R. Pereira, M. Christie, T. Amaral, N.F. Walker, Did ranchers 302 and slaughterhouses respond to zero-deforestation agreements in the Brazilian Amazon? Conservation 303 Letters 9, 32 (2015).
304 Gibbs, H. K., L. Rausch, J. Munger, I. Schelly, D.C. Morton, P. Noojipady, B. Soares-Filho, P. Barreto, L. 305 Micol, N.F. Walker, Brazil’s Soy Moratorium. Science 347, 377 (2015a)
306 Gil, J., R. Garrett, R. Berger. Determinants of crop-livestock integration in Brazil: Evidence from the 307 household and regional levels. Land Use Policy 59, 557 (2016).
308 Godar, J., U. M. Persson, E.J. Tizado, P. Meyfroidt. Towards more accurate and policy relevant footprint 309 analyses: Tracing fine-scale socio-environmental impacts of production to consumption. Ecological 310 Economics 112, 25 (2015).
311 INPE (Instituto Nacional de Pesquisa Espaciál) Deter Cerrado – Alerta de Desmatamento (available at: 312 http://www.terrabrasilis.dpi.inpe.br) (2018).
313 Jepson, W., A disappearing biome? Reconsidering land-cover change in the Brazilian savanna. The 314 Geographical Journal 171, 99 (2005).
315 Klink, C.A., R.B. Machado, Conservation of the Brazilian Cerrado. Conservation Biology 19, 707 (2005).
316 Lambin, E. F., H.K. Gibbs, L. Ferreira, R. Grau, P. Mayaux, P. Meyfroidt, D.C. Morton, T.K. Rudel, I. 317 Gasparri, J. Munger, Estimating the world’s potentially available cropland using a bottom-up approach. 318 Global Environmental Change 23, 892 (2013).
319 Lapola, D.M., L. A. Martinelli, C.A. Peres, J.P.H.B. Ometto, M.E. Ferreira, C.A. Nobre, A.P.D. Aguilar, 320 M.M.C. Bustamante, M.F. Cardoso, M.H. Costa, C.A. Joly, C.C. Leite, P. Moutinho, G. Sampaio, B.B.N. 321 Strassburg, I.C.G. Vieira, Pervasive transition of the Brazilian land-use system. Nature Climate Change 4, 322 27 (2014).
323 Macedo, M.N. R.S. DeFries, D.C. Morton, C.M. Stickler, G.L. Galford, Y.E. Shimabukuro, Decoupling of 324 deforestation and soy production in the southern Amazon during the late 2000s. Proc. Natl. Acad. Sci. 325 U.S.A. 109, 1341 (2012).
326 MMA (Ministério do Mieo Ambiente) Governo divulga desmatamento no Cerrado. (available at: 327 http://www.mma.gov.br/index.php/comunicacao/agencia-informma?view=blog&id=3066) (2018).
328 Müller, H., P. Rufin, P. Griffiths, A.J.B. Siqueira, P. Hostert, Mining dense Landsat time series for 329 separating cropland and pasture n a heterogenous Brazilian savanna landscape. Remote Sensing of the 330 Environment 156, 490 (2014).
331 Noojipady, P., D.C. Morton, M.N. Macedo, D.C. Victoria, C. Huang, H.K. Gibbs, E.L. Bolfe, Forest carbon 332 emissions from cropland expansion in the Brazilian Cerrado biome. Environmental Research Letters 12, 333 DOI: 10.1088/1748-9326/aa5986 (2017).
13 334 Pires, G.F., Abrahão, G.M., Brumatti, L.M., Oliveira, L.J.C., Costa, M.H., Liddicoat, S., Kato, E., Ladle, R.J. 335 Increased climate risk in Brazilian double cropping agriculture systems: Implications for land use in 336 Northern Brazil. Agricultural and Forest Meteorology 228-229, 286-298 (2016).
337 Rada, N., Assessing Brazil’s Cerrado agricultural miracle. Food Policy 38, 146 (2013).
338 Rausch, L., H. K. Gibbs. Property arrangements and soy governance in the Brazilian state of Mato Grosso: 339 Implications for deforestation-free production. Land 5(2) (2016).
340 Rudorff B, Risso J, Aguiar D et al. (2015) Análise Geoespacial da Dinâmica das Culturas Anuais no Bioma 341 Cerrado: 2000 a 2014, pp. 28. Florianópolis, Santa Catarina, Brasil. Available at: 342 http://biomas.agrosatelite.com.br/img/Analise_geoespacial_da_dinamica_das_culturas_anuais_no_bio 343 ma_Cerrado_2000a2014.pdf.
344 Sano, E.E., R. Rosa, J.L.S. Brito, L.G. Ferreira, Land cover mapping of the tropical savanna region in Brazil, 345 Environmental Monitoring Assessment 166, 113 (2010).
346 Soares-Filho, B., Rajão, R., Macedo, M., Carneiro, A., Costa, W., Coe, M., Rodrigues, H. and Alencar, A. 347 Cracking Brazil's forest code. Science. 344(6182). 363-364. (2014)
348 Sparovek, G., Berndes, G., Barretto, A.G.d.O.P., Kulg, I.L.F. The revision of the Brazilian Forest Act: 349 increased deforestation or a historic step towards balancing agricultural development and nature 350 conservation. Environmental Science & Policy 16, 65 (2012).
351 Spera, S.A., G.L. Galford, M.T. Coe, M.N. Macedo, J.F. Mustard, Land-use change affects water recycling 352 in Brazil’s last agricultural frontier. Global Change Biology 2, 3405 (2016).
353 Strassburg, B.B.N., A. E. Latawiec, L.G. Barioni, C.A. Nobre, V.P. da Silva, J.F. Valentim, M. Vianna, E.D. 354 Assad, When enough should be enough: Improving the use of current agricultural lands could meet 355 production demands and spare natural habitats in Brazil. Global Environmental Change 28, 84 (2014).
356 Strassburg, B.B.N., T. Brooks, R. Feltran-Barbieri, A. Iribarrem, R. Crouzeilles, R. Loyola, A. E. Latawiec, 357 F.J.B. Oliveira Filho, C.A. de M. Scaramuzza, F.R. Scarano, B. Soares-Filho, A. Balmford, Moment of truth 358 for the Cerrado hotspot. Nature Ecology & Evolution 1, DOI: 10.1038/s41559-017-0099 (2017).
359 United Nations, Declaration on Forests (available at: http://www.un.org/climatechange/summit/wp- 360 content/uploads/sites/2/2014/07/New-York-Declaration-on-Forest-%E2%80%93-Action-Statement-and- 361 Action-Plan.pdf) (2014).
362 Vieira, R.R.S., B.R. Ribeiro, F.M. Resende, F.T. Brum, N. Machado, L. P. Sales, L. Macedo, B. Soares- 363 Filho, R. Loyola, Compliance to Brazil’s Forest Code will not protect biodiversity and ecosystem services. 364 Diversity and Distributions 24, 434 (2017).
14 365
366 Fig. 1. The role of soy in annual Cerrado clearing. Soy expansion contributed at least 22% of the recent 367 clearing in the Cerrado, 2003-2014, and was largely within the legal clearing limits prescribed by the 368 forest code (blue line). The bars show the percent of annual clearing for soy (based on soy expansion
369 within three years (yeari – yeari+3) onto areas deforested in the year indicated), and for other purposes 370 (the remainder of deforestation in that year). For years with an asterisk (2012, 2013, and 2014), annual 371 clearing for soy is underestimated because they include fewer than three years following clearing due to 372 data limitations.
373
374
375
15 376
377 378 Fig. 2. Vegetated and cleared land highly suitable for soy in the Cerrado. For each microwatershed, we 379 mapped the dominant land cover types that are highly suitable for soy (18). Cleared areas suitable for 380 soy are concentrated in the Southern Cerrado. Native vegetation suitable for soy is concentrated in 381 Mato Grosso and in Matopiba; the Forest Code limits conversion of some of this vegetation.
382
16 Table 1. Annual Clearing for Soy
ENTIRE CERRADO MATOPIBA MATO GROSSO SOUTHERN CERRADO
% of Total % of Total % of Total % of Total Year of Total Total Total Total Annual Annual Annual Annual Deforestation from Clearing, Clearing, Clearing, Clearing, Clearing Clearing Clearing Clearing LAPIG ha ha ha ha for Soy for Soy for Soy for Soy 2003 762,047 32% 239,303 46% 342,802 33% 179,941 11% 2004 853,027 26% 383,066 37% 289,303 22% 180,658 10% 2005 460,922 19% 183,108 36% 165,226 9% 112,588 7% 2006 335,333 16% 134,084 34% 76,642 6% 124,607 3% 2007 422,114 19% 209,280 33% 63,636 10% 149,197 3% 2008 362,662 32% 240,935 45% 48,667 13% 73,061 3% 2009 274,828 27% 136,425 45% 59,017 17% 79,386 4% 2010 358,806 17% 253,469 22% 26,682 11% 78,655 2% 2011 697,426 22% 376,344 34% 107,362 17% 213,720 4% 2012* 735,094 21% 523,317 26% 60,450 21% 151,328 2% 2013* 407,424 17% 241,897 24% 77,858 13% 87,669 4% 2014* 439,183 2% 279,242 2% 71,131 1% 88,810 1% TOTAL 6,108,866 22% 3,200,471 31% 1,388,775 19% 1,519,620 5% * Clearing for soy is underestimated in the years 2012, 2013, and 2014. Estimates in 2003-2011 include the area planted to soy within three years of conversion.
17 Table 2. Legality of Cerrado clearing through time. We assumed that areas not identified as cleared were covered with native vegetation. Legality estimates for all years were based on the percent of the property or the watershed that could be cleared under the rules established by the 2012 Forest Code and did not consider whether a license was obtained or if areas of permanent preservation were intact. Property boundaries covered 77% of soy area. Watersheds cover 100% of soy area.
2003* 2004* 2005* 2006* 2007* 2008* 2009 2010 2011 2012 2013 2014 2015 TOTAL
Area of clearing (ha) 555,962 607,607 308,337 229,667 286,090 262,619 189,023 233,112 463,168 489,203 288,928 309,591 226,198 4,449,505
Property % legal 78% 82% 80% 88% 87% 88% 87% 88% 86% 88% 87% 87% 86% 85% Boundaries % illegal 22% 18% 20% 12% 13% 12% 13% 12% 14% 12% 13% 13% 14% 15% Area of clearing (ha) 719,407 786,647 410,555 311,949 388,724 340,599 250,769 324,117 644,562 677,258 373,827 401,424 284,843 5,914,681
EntireCerrado Watersheds % legal 74% 80% 77% 83% 86% 86% 86% 88% 86% 90% 87% 90% 89% 84% % illegal 26% 21% 23% 17% 14% 14% 14% 12% 14% 10% 13% 10% 11% 16% Area of clearing (ha) 151,370 238,265 115,630 76,987 118,491 153,251 84,263 121,767 200,084 262,757 155,625 189,868 172,241 2,040,600 Property 90% 90% 92% 91% 89% 90% 92% 85% 87% 86% 88% 86% 87% 88%
% legal Boundaries % illegal 10% 10% 8% 9% 11% 10% 8% 16% 13% 14% 12% 14% 13% 12%
Area of clearing (ha) 220,913 339,606 165,412 121,596 183,772 224,381 123,310 221,581 344,267 480,867 216,901 253,528 219,746 3,115,881 Matopiba Watersheds % legal 93% 96% 94% 95% 92% 93% 94% 90% 91% 92% 92% 93% 92% 93% % illegal 7% 4% 6% 5% 8% 7% 6% 10% 9% 8% 8% 7% 8% 7%
Area of clearing (ha) 250,923 208,149 104,489 39,429 31,818 22,477 29,783 15,757 50,012 27,850 44,256 38,719 19,637 883,299 Property % legal 62% 63% 56% 74% 66% 62% 66% 73% 68% 67% 74% 77% 69% 65% Boundaries % illegal 38% 37% 44% 26% 34% 38% 34% 27% 32% 33% 26% 23% 31% 35% Area of clearing (ha) 322,944 271,984 139,945 69,867 59,691 46,112 53,491 26,022 95,050 55,107 74,227 63,746 35,564 1,313,751
Mato Grosso Mato Watersheds % legal 73% 56% 56% 53% 69% 64% 57% 67% 63% 62% 66% 70% 75% 68% % illegal 27% 44% 44% 47% 31% 36% 43% 33% 37% 38% 34% 30% 25% 32% Area of clearing (ha) 133,449 127,762 74,769 87,540 106,807 54,399 52,812 47,527 141,934 93,728 60,689 54,659 21,272 1,057,346 Property % legal 89% 92% 92% 90% 91% 92% 91% 93% 90% 91% 92% 92% 94% 91% Boundaries % illegal 11% 8% 8% 10% 9% 8% 9% 7% 10% 9% 8% 8% 6% 9% Area of clearing (ha) 175,551 175,057 105,198 120,486 145,261 70,106 73,968 76,514 205,245 141,284 82,698 84,149 29,533 1,485,049
Watersheds % legal 81% 84% 81% 78% 86% 86% 86% 90% 87% 92% 88% 90% 88% 86% Southern Cerrado Southern % illegal 19% 16% 19% 22% 14% 14% 14% 10% 13% 8% 12% 10% 12% 14% * excess clearing prior to 2008 on properties <4FM was forgiven under the 2012 update to the Forest Code. Estimates presented here, however, consider this deforestation on small properties to be illegal clearing, because it was illegal at the time it happened.
18 Table 3. Area and location of land suitable for soy that was previously cleared and that could be cleared legally under the Forest Code (“surplus”). Only areas that are privately owned or eligible for private ownership are included.
Entire Mato Southern
Cerrado Matopiba Grosso Cerrado Cleared area highly suitable for soy 22,975,496 1,954,731 2,547,852 18,472,913
Additional cleared area potentially (ha) suitable for soy 15,220,975 513,330 2,237,221 12,470,425
Cleared area Cleared Maximum cleared area suitable for soy 38,196,471 2,468,060 4,785,073 30,943,338
Vegetation on areas highly
suitable for soy that could be legally cleared 8,002,563 3,840,642 1,110,328 3,051,593
Surplus vegetation on areas
potentially suitable for soy 6,563,552 1,522,445 1,177,537 3,863,570 (ha) Maximum vegetation area
suitable for soy 14,566,115 5,363,087 2,287,865 6,915,163 Legally clearable Legally Natural Vegetation Natural Surplus vegetation on areas not
suitable for soy 18,014,922 12,318,386 1,263,254 4,433,282
Vegetation on highly suitable area 9,540,411 2,464,349 2,965,747 4,110,315
(ha) Vegetation on potentially Ineligible
for clearing for suitable area 11,275,201 865,312 4,234,671 6,175,218
19 Table 4. Properties with land highly suitable for soy that could be legally cleared under the Forest Code requirements
Number of soy properties % soy properties with Number of non-soy % of non-soy properties with suitable and legally suitable and legally properties with suitable and with suitable and legally clearable area clearable area legally clearable area clearable area
Total non- Total soy Region >10ha >25ha >100ha >10ha >25ha >100ha soy >10ha >25ha >100ha >10ha >25ha >100ha properties properties Mato Grosso 10,170 1,168 886 413 11.48% 8.71% 4.06% 28,754 3,790 2,808 1,350 13.18% 9.77% 4.69% MATOPIBA 5,902 1,458 1,113 569 24.70% 18.86% 9.64% 105,043 11,149 7,786 3,523 10.61% 7.41% 3.35% Southern 25,806 4,531 2,866 847 17.56% 11.11% 3.28% 408,486 26,054 13,757 3,882 6.38% 3.37% 0.95% Cerrado Total non- Total soy State >10ha >25ha >10ha >25ha >100ha soy >10ha >25ha >100ha >10ha >25ha >100ha properties >100ha properties Bahia 2,296 567 444 244 10.63% 19.34% 10.63% 21,964 1,010 906 703 4.60% 4.12% 3.20%
Distrito Federal 0.00% 0.00% 0.00% 120 0.00% 0.00% 0.00% ------Goiás 12,721 2,561 1,608 481 3.78% 12.64% 3.78% 101,030 11,909 6,348 1,699 11.79% 6.28% 1.68% Marahnão 1,205 352 274 138 11.45% 22.74% 11.45% 36,459 4,037 2,588 1,014 11.07% 7.10% 2.78% Mato Grosso 10,170 1,168 886 413 4.06% 8.71% 4.06% 28,754 3,790 2,808 1,350 13.18% 9.77% 4.69% Mato Grosso 2,185 538 408 164 7.51% 18.67% 7.51% 17,130 2,885 2,113 931 16.84% 12.34% 5.43% do Sul Minas Gerais 7,863 1,229 749 182 2.31% 9.53% 2.31% 208,765 9,616 4,685 1,151 4.61% 2.24% 0.55%
Parana 138 27 13 1 0.72% 9.42% 0.72% 1,063 24 5 2.26% 0.47% 0.00% - Piaui 780 113 78 51 6.54% 10.00% 6.54% 12,928 508 354 203 3.93% 2.74% 1.57% Sao Paulo 2,899 176 88 19 0.66% 3.04% 0.66% 80,378 1,620 606 101 2.02% 0.75% 0.13% Tocantins 1,621 426 317 136 8.39% 19.56% 8.39% 33,692 5,594 3,938 1,603 16.60% 11.69% 4.76% Total 41,878 7157 4,865 1,829 4.37% 11.62% 4.37% 542,283 40,993 24,351 8,755 7.56% 4.49% 1.61%
20
21