Puyanáwa Institute
Land use policies

Land tenure in many parts of Brazil remains uncertain and contested.
These problems have recently been exacerbated by changes in the legal framework regulating protected areas and the land market. A particular challenge facing attempts to improve land tenure security and governance in Brazil is the lack of a single, integrated assessment of all types of land.
Here we address this problem and present a first integrated map of land tenure in Brazil that encompasses all official data sources pertaining to public and private land. Of the total (8.5 million km2), 36.1% of all land is public (with 6.4% officially undesignated), 44.2% is private and 16.6% is unregistered or of unknown tenure.
Surprisingly, overlaps between land tenure categories add up to 50% of Brazil’s registered territory. A clearer understanding of the uncertainties in land tenure, and the spatial distribution of those uncertainties can help guide research and public policies focused on minimising land conflicts and strengthening governance and territorial planning to improve the economic, environmental and social outcomes of land use in Brazil.
Despite the economic importance of agriculture to the Brazilian economy, systemic problems with the countries’ land tenure system present a major development challenge.
Such problems include widespread land disputes and conflicts, weak governance and highly unequal patterns of land ownership (Reydon et al., 2015; Lapola et al., 2014). Existing land tenure designations also remain highly unstable in the face of ongoing legislative changes by federal and state governments, increasing the risk of land market volatility, including through speculation and land grabbing. For example, the global phenomenon of degradation, downsizing and degazeting (PADDD) of protected areas increased markedly in the last five years in Brazil (Pack et al., 2016). In 2016, permission was given to sell public lands within agrarian reform settlements (MP 759/2016). Under the newly elected federal government, Brazil’s National Congress is currently reviewing the possibility of increasing the area of land that could be bought by foreigners.
The combination of these systemic problems and recent increases in land market uncertainty and volatility have serious implications for sustainability in Brazil, including efforts to reduce deforestation, especially in the Amazon. Indeed, the recent shift in Brazilian government policy towards reduced public protection and a less strictly enforced land tenure regime may help explain the upward trend in deforestation observed in the Amazon since 2012 (Soares-Filho and Rajão, 2018). As such land tenure problems may severely compromise Brazil’s ability to meet its Nationally Determined Contribution (NDC) to the Paris Agreement (Rochedo et al., 2018), including targets to end illegal deforestation in the Amazon and restore 12 million hectares of native vegetation by 2030.
Historical problems with Brazil’s land tenure system and the dilution of public protections by recent governments are exacerbated by the lack of a single, integrated assessment of all public and private lands classified with respect to different categories of use and tenure (Reydon et al., 2017). To help address this problem and contribute to a more transparent and effective territorial governance system for Brazil, we present a first integrated map of land tenure in Brazil that encompasses all official data sources pertaining to public and private land.
The map presented here builds on a preliminary version used by Freitas et al. (2018a), 2017 to model land use in Brazil and assess environmental impacts on different property types. The new map represents a major improvement in two main ways.
First, we have increased the coverage and accuracy of information on the distribution of private land made available through the Rural Environmental Registry (CAR in Portuguese acronym) of the Brazilian Forestry Law published in 2012 (Brazil, 2012). For this analysis, we used the version of the CAR dataset from 18 December 2018, a date close to the official deadline for landowners to register, which was 31 December. The CAR is a legal instrument to support environmental regulation and the implementation of the Forestry Law. It consists of a geodatabase where landowners self-declare property boundaries and land use designations in an electronic system to be validated by the state or federal government (Azevedo et al., 2017). The CAR system has been widely adopted by farmers and the database covers most private land in Brazil. Although CAR is not a land tenure classification in a legal sense, it provides the best available assessment of the distribution of private landholdings in Brazil, and is invaluable for identifying land tenure gaps.
Second, we combined the CAR data layer with 17 other official databases and assigned land into 14 tenure categories (Freitas et al., 2018b) describing the list of the 18 databases and 14 categories. To achieve this, an objective set of rules on land tenure hierarchies, i.e. which designation should have priority over others in the case of spatial overlaps, was developed in consultation with several leading scientists and land tenure experts. These rules take into account the legal basis for each land tenure category in the various datasets, differences in the quality and accuracy of official land tenure datasets, and the likelihood of future changes in land designation (Freitas et al. , 2018b).
For example, indigenous reserves have higher priority in the hierarchy than private lands under CAR because indigenous rights are established in the Brazilian Constitution. In addition, public and private lands formally registered and recognised by a public authority have higher priority than those that are not recognised as such, so that private lands certified by the agrarian reform agency (SIGEF/INCRA) have higher priority than lands that are only registered in the self-declaratory CAR system.
By merging CAR data with other public and private land tenure data, we provide the first wall-to-wall land tenure map for Brazil that combines all available official data sources on public and private land. The map is publicly available and can be consulted and downloaded at http://atlasagropecuario.imaflora.org/. In addition to providing much greater transparency of land ownership in Brazil, the map and accompanying estimates of the number and area of overlap between land tenure designations provide a new level of understanding of key gaps and inconsistencies in the spatial distribution of different land tenure classes. These data provide the basis for further research on land tenure and patterns of irregularity and non-compliance, and being publicly accessible can be used by both public and private institutions and civil society to channel improvements in land governance across Brazil. Increased transparency in land data can also support judicial decisions on land tenure, which can result in an increase in court cases being formally adjudicated.
a
We exclude APAs from the conservation unit category. APA (environmental protection area) is a type of sustainable use conservation unit that can occur in areas of public or private domain that allow human occupation and economic activities, including intensive agriculture, and total 44 million hectares. Its inclusion would confuse the interpretation of land ownership and overlap, as it does not imply expropriation of private land ownership and as such necessarily overlaps with other land tenure categories.
b
Public land that has not been designated for an end use. Our findings differ from the 65.5 million hectares of undesignated forest land in the Amazon found by Azevedo-Ramos and Moutinho (2018) due to the hierarchy rules adopted, where type B forests have a low level of priority and are classified as other categories. Type B forests are federal or state lands covered by forests whose final designations have not yet been decided. They are under the administration of the Brazilian Forest Service (SFB).
c
Cadastro Ambiental Rural (Rural Environmental Register).
d
Sistema de Gestão Fundiária – INCRA (INCRA land tenure management system).
Of the total Brazilian territory, we find that 36.1% is classified as public land and 44.2% as private land in 4,537,242 polygons of individual land units covering 682,513,148 ha (80.3%) of the country. An additional 3.1% of the country (26.3 million ha) is covered by urban and water areas. Public lands are mainly composed of protected areas (24.2% combining conservation units with indigenous reserves) and undesignated lands (6.4%) while agrarian reform settlements occupy 4.9% of the national territory. Private land is concentrated in large estates. Aggregating the properties found in the CAR and SIGEF registries, 97,000 large properties with an average size of 1876 ha cover 21.5% of the country. In comparison, indigenous reserves that provide livelihoods for 572,000 people cover 13% of Brazil (IBGE, 2010). Surprisingly, 16.6% of the entire Brazilian territory is not covered by any category and is not registered in any official database (Table 1). The distribution and share of land tenure categories varies substantially across regions and states in the country (Supporting Tables 1-7). Protected areas are concentrated in the northern region (95% in the Amazon biome) while private lands are concentrated in the south.
Fig. 1. Overlaps of land tenure categories.
Each bar represents the original area of each land tenure class, while the shaded area of each bar represents the area of that class that overlaps with one or more other land tenure classes. Overlapping areas were classified as a land tenure class type following our hierarchy of priority classifications. The area of overlap of each land tenure class with all other classes is given in Table 8).
The overlaps between the 14 land tenure designation categories amount to 354,601,858 ha of the area covered by known land (50%) or 41% of Brazil’s territory. Because several land tenure designations may overlap with each other in the same locality, this does not mean that 354 million hectares of land are under some kind of overlap (see Supporting Table 8). Overlaps on public land amount to 171 million ha (48% of all overlaps), while overlaps between public and private land tenure classes amount to 176 million ha (50%) and overlaps between different private land classes amount to 7 million ha (2%) (Fig. 1 and Supporting Table 8). The extent of overlap varies substantially across regions of the country with greater overlap and associated tenure uncertainty concentrated in the Northern region (Supporting Tables 9-13).
In providing the first wall-to-wall assessment of public and private land tenure for Brazil, we are able to identify four observations that are of great importance for territorial planning, land governance and research in the country.
First, we estimate that there is a minimum of 54.6 million hectares (6% of the total known area) of public land that remains undesignated, most of which is in the Amazon biome. This area of undesignated land is almost as large as the state of Minas Gerais, the fourth largest state in Brazil. The fact that there is substantial overlap between these undesignated lands and lands registered under public and private conditions highlights the pervasive uncertainty that characterises land temperature in the Amazon biome and the profound challenges this poses for forest conservation (see also Azevedo-Ramos and Moutinho, 2018). Azevedo-Ramos and Moutinho (2018) provided a less conservative estimate of the total area of undesignated land in the Brazilian Amazon at 65.5 million hectares that did not represent any hierarchy of land tenure designations for overlapping lands.
Secondly, one sixth (16.5%) of Brazil is not classified as an official land tenure register which represents a major challenge to efforts to improve land planning, increase legal compliance and reduce land conflicts such as land grabbing and ownership disputes between public and private actors, such as the recent situation in the Jamanxim National Forest. This public area has been illegally occupied by farmers in disputes that have resulted in deforestation and violence.
Third, the total area of overlap between at least two different land tenure areas comprises an area larger than that of any one land tenure category. Some overlaps between land boundaries may be due to errors in the mapping of individual public or private properties. Other overlaps may be the result of different official land records that are not properly updated by different state and federal agencies. In some cases, overlaps may be due to competing land claims and indicate the potential for land conflict. Regardless of the explanation, however, such widespread uncertainties in land tenure clearly illustrate the large extent of Brazilian land that is under legal dispute or at risk of being under legal dispute at different levels of policy and decision-making, from municipal to federal.
Finally, our assessment illustrates the unequal distribution of land ownership with large areas concentrated in relatively few private properties. In fact, almost half (48%) of all private land is concentrated in less than 100,000 properties, representing 2% of the total.
In conclusion, this study provides the first quantitative and spatially explicit assessment of the coverage, gaps and uncertainties in the land tenure status of the entire Brazilian territory. The data are organised at the most detailed property level, but allow for integration across the various levels of jurisdiction where land policy and decision-making occurs, from the municipal to the federal scale. Our assessment highlights the fragile state of land tenure in the country and provides an important step towards identifying priorities for public policy. Brazil is recognised as one of the most biodiverse countries in the world (Ferreira et al., 2014) and as one of the largest food producers (Sparovek et al., 2018). The success of efforts to ensure that agricultural development in Brazil is placed on a more sustainable and socially equitable basis depends critically on efforts to address the profound land tenure insecurity demonstrated by this study. More research is needed to help understand the reasons for the different types and levels of land tenure insecurity and the extent of land under undesignated or unregistered status in different parts of the country and in different land tenure regimes.