Transnational land deals: the
whereabouts in relation to the
agricultural land quality
Abstract -In this interdisciplinary research we argue that the economic theory of diminishing marginal returns leads to the hypothesis that foreign investors take the best agricultural land available in transnational land deals, leaving the local communities in developing countries to move to land of lesser quality. To prove this hypothesis, the concept of agricultural land quality has been extended so that it comprises not only earth sciences but also economic characteristics. With the use of Geo Information Systems (GIS), this broader concept has been used to find the best agricultural land in Madagascar, Ethiopia and DR Congo, after which these locations have been compared with the location of the land deals. We concluded that many land deals were not situated on ‘the best agricultural land’ as was beforehand expected and the hypothesis has therefore been rejected.
Students: Marrit Leenstra (Earth sciences –10095942) Agnethe Postema (Earth sciences –10658858) Supervisor: Dr. ir. Crelis Rammelt
Tutor: Koen van der Gaast MSc Word Count: 6305
For Interdiciplinary Project, a course of the BSc Future Planet Studies at the University of Amsterdam. 18th of December, 2015.
Conte
1. Introduction...3
2. Theoretical framework...4
2.1 Land deals...4
2.2 Economic theories...4
2.3 Agricultural land quality...5
3. Methods...6
3.1 Agricultural land quality...6
3.2 Data collection and storage...7
3.3 Data Analysis...7
4. Results...8
4.1 Agricultural land quality...8
4.2 National results...10
4.2.1 Madagascar...10
4.2.2 Ethiopia...12
4.2.3 DR Congo...14
4.3 The land deals...16
4.3.1 Madagascar...16
4.3.2 Ethiopia...17
4.3.3 DR Congo...17
4.4 Relations between land deals and quality of land...18
4.4.1 Madagascar...18
4.4.2 Ethiopia...19
4.4.3 DR Congo...20
5. Conclusion and discussion...21
1. Introduction
The 2007-2008 food crises led to an increase in public and scientific interest in transnational land deals (Anseeuw et al., 2012; Edelman, 2013). There are argues that it also led to an increase in land deals itself, since it encouraged countries to look for opportunities to enlarge its national food security (Deininger & Byerlee, 2011). In academic literature about land deals many emphasis has been given to the question what the effects of land deals are and whether or not land deals are good for the recipient countries (i.g. Cotula, Vermeulen, Leonard & Keeley, 2009; Kleemann & Thiele, 2015; Lavers, 2012). Nonetheless, globalization and thus foreign direct investment (FDI) is expected to continue and future prospects estimate a population growth of 40%, which would in combination with changes in consumption patterns, lead to the need of an increase of the food production by 70% (Deininger & Byerlee, 2011). Consequently, it is likely that more land deals will occur in the future. In this light, we argue that more research should focus on where the land deals are, or should be, located within a country and how a fair division of agricultural land can be made between local communities and foreign investors.
This research will start with this objective by identifying where existing land deals are located within three countries (Madagascar, Ethiopia and Democratic Republic (DR) Congo) and compare these locations with agricultural land quality data. The countries have been chosen based on the ranking of countries with the most reliable land deal data, see figure 1 (Anseeuw et al., 2012). However, the Philippines were ranked higher than the chosen countries, but unfortunately necessary Geo Information System (GIS) data was not found for this country.
Figure 1. Most targeted countries by land deals according to size of total reported acquisitions (Anseeuw et al., 2012).
The main question of the proposed research will be: to what extent do large scale land based investors take the best agricultural land available? Sub questions include “which characteristics determine the quality of agricultural land?” and “are suitable agricultural lands for crop growth and the occurrences of land deals related?”. The purpose of answering the questions is to proof or reject
the hypothesis that foreign investors take the best agricultural land available, leaving local communities to land of lesser quality.
An interdisciplinary approach is necessary, because the phenomenon is complex and affecting several disciplines (e.g. Anseeuw et al., 2012). Insights from only one discipline will likely lead to an inefficient allocation of land. The research will combine economic insights based on the law of diminishing marginal returns and the principle of political economy, because this offers an insight into efficient allocation of resources, with an earth sciences perspective on the quality of agricultural land, because it is expected that the quality of the agricultural land is mostly determined by earth scientific characteristics. However, an important limitation to this research is that the actors are not consulted. Hence, a sociological perspective could shed another light on the topic.
In the following section the theoretical framework for this research will be explained, after which the methodology will be described and the results will be presented. The research will close with a conclusion on what the results mean to the hypothesis and a brief discussion of these results.
2. Theoretical framework
2.1 Land deals
External financial flows have played an important role in financing Africa’s development. In the last decades, these external flows have shifted from aid-based to increasingly more foreign direct investment (FDI) (OECD, 2015). Transnational Large Scale Land Based investments or simply land deals are an example of this foreign direct investment. The estimates of the total agricultural land area of interest for foreign investors vary between 56 million hectares in one year (Deininger & Byerlee, 2011) to ‘nearly 230 million hectares since 2001’ (Kugelman, 2012 in Edelman, 2013). This difference highlights the problematic availability of reliable information, due to a lack of transparency in both the investor and recipient countries. Little information is publically available and as Edelman (2013) states; this matter might undermine the legitimacy of land deal research, which could lead to a decline in public interest. The Land Matrix database has tried to underpin this problem by assigning reliability ranks to the reported land deals. Of the 1217 deals reported until 2011, 625 were reported as reliable, covering a total area of 32.7 million hectares (Anseeuw et al., 2012). Edelman (2013) concludes that ‘the increase in land deals in recent years in doubtless real’ even though ‘the evidentiary basis for understanding it is frequently very weak’.
In general, there are two opinions about transnational land deals in developing countries; some see it is a development opportunity for rural communities whereas others argue that it further impedes the rural communities to reach a good food security status, since local communities often lose rights to land. Unfortunately, conclusive evidence has not yet been presented. The point of view for this research is that both theories can be true, depending on the exact whereabouts of the land deals within the target country.
2.2 Economic theories
The basis for this reasoning lies in economic theory, being the law of diminishing marginal returns of David Ricardo (1771-1823) and John Stuart Mill’s Principles of Political Economy (1857). The law of diminishing marginal returns assumes that when resources are scarce, expansion of the use of these resources will lead to lower returns per used unit of resource. Applied to agricultural production, this leads to the assumption that since the land of the best quality will have been taken first. Expanding the total area of agricultural production will lead to lower marginal returns, because the extra added land will be of lower quality (Hanley, Shogren, & White, 2001b). However, local communities often do not have formal registered rights over the land they use, especially in rural sparsely populated areas (Deininger & Byerlee, 2011). It is argued that as a result of a legal framework existing since the
colonial days, any unregistered land is presumed to be unused and thus available for land deals (Government of France, 2008, 2010 in: Deininger & Byerlee, 2011). Furthermore, Anseeuw et al. (2012) state that poor countries with weak land tenure schemes are among the most targeted countries. This implies that foreign investors seek to maximize their own profit by easy and cheap access to land. In this light the hypothesis ascends that, since the investors will target the best agricultural land available while trying to maximize their profit, they will displace the local
communities from their land. This hypothesis seems to be supported by data from the Land Matrix; the largest number of land deals (246 deals, 43%) target land that is already in use as cropland. Different cropping mosaics are mostly affected, indicating that the land is used by smallholders from the local communities. However, in surface terms, forests seem to me more targeted than croplands with 31% and 22% respectively (Anseeuw et al., 2012).
Ricardo however, made very broad simplifications and ignored other important factors. Mill (1806-1873) therefore expanded this economic theory in his Principles of Political Economy by stating that Ricardo’s diminishing marginal returns can be counteracted through technological progress, which increases production and therefore drives down the marginal production costs (Hanley, Shogren & White, 2001a). It is sometimes assumed that foreign companies have a higher technological standard and more knowledge. This assumption is justified by the premise that domestic companies have better knowledge of and access to the domestic market and foreign companies have to compensate for this disadvantage by higher productivity rates through higher technological standards and more technological knowledge (Graham & Krugman, 1991, in: Borensztein, De Gregorio & Lee, 1998). This seems even more likely in cases where investors from Europe, the US and Asia are targeting developing countries. Therefore, it can be also assumed that the foreign investors in land deals have better technologies and knowledge to outweigh the burden of marginal agricultural land. However, as the data above suggests, the foreign investors will not take the marginal agricultural land, but the best agricultural land available, whilst displacing the local communities. This situation seems highly unfavorable for the local communities. After all, they do not have the technology or the know-how to counteract the diminishing marginal returns and this could lead to a reduced local food security. However, evidence for this statement has not been conclusive, Andrianirina-Ratsialonana and Teyssier (2010) and Ullenberg (2009) for example reached opposing conclusions after analyzing one projects contract terms with regard to local food security (in: Cotula, 2011).
2.3 Agricultural land quality
Following section 2.2, the economic theories thus point towards a situation in which the foreign investors do get the best agricultural land available and the local communities are displaced to marginal agricultural land. However, theories can differ from real situations. To be able to check this theory empirically, firstly the concept best agricultural land has to be defined.
The best agricultural land is usually seen as agricultural land that produces the highest yields (Cassman, 1999). This represents the earth sciences perspective, since ‘land’ then comprises the physical environment, which includes climate, relief, soils, hydrology and vegetation as long as these influence the potential for land use. Whereas economic and social characteristics are usually not included in the concept (FAO, 1976). ‘Land quality’ is in this context a complex attribute of land which influences the suitability of land for a specific kind of use (FAO, 1976). The department of primary industries and the department of housing, local government and planning (DPI/DHLGP) of Queensland for example defines good quality agricultural land as “land which is capable of
sustainable use for agriculture, with a reasonable level of inputs, and without causing degradation of land or other natural resources”. The factors they use to assess the agricultural land quality are the soil and topographic and climatic limitations, again economic factors are explicitly not taken into account (DPI/DHLGP, 2013). Looking then at the best agricultural land quality for one specific crop,
the same line is often followed. Nguyen (1989) for example states that especially climate, topography, soil and water supply have been often used to classify wetlands for agricultural development and rice based production systems. However, she states that besides these factors also farming practices and socio-economic factors influence rice yields.
This last statement represents the economic perspective on agricultural land quality; not only is the physical environment important, other circumstances influence the yield and the profit that can be made on the land as well. Following Nguyen (1989) important (socio-) economic factors
influencing yields include land tenure, credit availability and market accessibility. Additionally, farming practices, their intensity and quality greatly influence yields and these farming practices are in turn in great part determined by the economic factors. Although farming practices, land tenure and credit availability influence yields, these characters cannot be used as agricultural land quality measures, because they are not attached to the land but to the people that are using it. Deininger & Byerlee (2011) assent with this economic view by stating that suitable agricultural land is land with at least 60% of the potential yield for a certain crop, located in an area with a population density less than 10 persons/km2and a travel distance of less than six hours to the next market based on available
transportation. Hence, agricultural land quality is not only determined by the physical environment, but also by the market accessibility.
3. Methods
3.1 Agricultural land quality
Agricultural land quality is one of the core concepts of the research, whereby the two different disciplines involved, earth sciences and economics, have different insights to the concept. As
described in the theoretical framework, the best agricultural land from an earth sciences perspective: the highest yielding land based on the physical environment. For this research the following physical environment characteristics will be used: annual average precipitation, annual average temperature, soil type and elevation. This selection is based on available environmental data (FAO, 1976; FAO, 2015a; Jenny, 1941 and Nguyen, 1989). From an economic perspective on the other hand, the best agricultural land has to be also close to infrastructure and needs to have the ability to access markets (e.g. Deininger & Byerlee, 2011). Deininger and Byerlee (2011) define ‘close to infrastructure’ as a travel distance of less than six hours to the next market, based on available transportation. However, as it is problematic to measure six hours travel distance, this measure has been converted to 5 km distance to the nearest road. This new measure is based on the assumption that walking 5 km will take approximately 1 hour and in the remaining 5 hours it will be possible to reach the nearest market (city) by public transportation. Common ground has been found by using Repko’s integrative technique of extension (Repko, 2012), hence the best agricultural land is viewed as the highest yielding land with no more than 5 km distance to the nearest road.
Since the highest yielding lands for agriculture can differ between crops, for each country the most produced crop in 2013 has been selected and the land quality has been determined using FAO data on these crops. For Madagascar this is rice, for Ethiopia maize and for DR Congo Cassava (FAO, 2015a).
Two categories are used to value the characteristics of the agricultural land quality: suitable and unsuitable. The suitable category is used when the condition for the specific parameter is allowing plant growth and the yields are economic viable. The unsuitable category is assigned when the condition for the specific parameter is preventing crop growth and economic viable agriculture is not possible. As the survey is interested in positive correlation between land deals and suitable
3.2 Data collection and storage
To allow organized storage and analysis of the land quality data in a GIS environment, a geographic database has been set up and a GIS layer has been found and stored for each of the agricultural land quality parameters of each country in online sources (e.g. ArcGIS online). To allow the combination and computation of the different layers, datasets have been reclassified to be marked as suitable or unsuitable in the attribute table, as described in paragraph 3.1. Unfortunately, not all data could be used for reclassification. For example, the precipitation and temperature data could not be altered, since the data was stored as a web map service (WMS). Consequently, the interpreting of this data became more difficult, as there are more categories displayed. For the soils map another problem occurred, as the data of all soils in the world was used. Like the data of the temperature and the precipitation, this data could also not be altered. Hence all soils are displayed in the legend, not only the soils that occur in the chosen country.
Additionally, a land deal spatial data layer has been created and stored in the database. Georeferenced data for the land deals was available at the Land Matrix. Unfortunately, there are some limitations to the Land Matrix data. First, the land deals are displayed as points, although in reality they should be polygons. Furthermore, land deals are notoriously un-transparent and the Land Matrix therefore relies extensively on unofficial sources (Land Matrix, 2015). However, the scope of this research does not allow verification of the data at the specific locations.
3.3 Data Analysis
For this survey qualitative and quantitative surveys have been made. Firstly, a literature study was made on the definition of land deals and relevance of land deals. After enough qualitative data was found, a quantitative survey could start on the occurrences of land deals in the world. Based on figure 1, Madagascar, Ethiopia and DR Congo were found at having high reliable data on land deals.
Subsequently after the countries where chosen, the highest crop production for each country was searched for. This included rice, maize and cassava. Next the suitable agricultural land quality
characteristics for each crop where chosen. This includes temperature, precipitation, altitude, soil and road distance.
Secondly continuation of the quantitative research based on the previous quantitative outcomes was made. The characteristics of the three chosen countries were searched, altered and reclassified when possible. An example of altering maps is displayed in figure 2, where the precipitation and road distance of 5 kilometers are displayed together with the occurrences of land deals.
Of course should be mentioned that during the whole survey, literature study was used to serve as background information and to decide what kind of data should be used. Another important aspect is the fact that ArcGIS data could be used for earth sciences and economic surveys. This increased the coherence of the survey as the results could easy be compared. As both authors have an earth sciences background, using ArcGIS as data source and analysis method was a natural decision and has increased the integration of our research insights. Hence, ArcGIS has clearly been used to create common ground for earth sciences and economy.
Figure 2. Screenshot of Ethiopia characteristics in ArcGIS. (ArcGIS, 2015).
4. Results
In this paragraph the results of qualitative and quantitative research earlier mentioned in the methods paragraph will be displayed and explained. First the results of the literature research to agricultural land quality requirements for suitable growth of rice, maize and cassava will be
presented. Hereafter, these results will be compared with the national maps of Madagascar, Ethiopia and DR Congo and the land deals within these countries. Based on these results a relation on the occurrences of land deals at suitable agricultural areas can be made.
4.1 Agricultural land quality
For each country the most produced crop in 2013 has been selected and the best agricultural land is determined using the growth conditions for these crops, in which the following characteristics are used: average annual precipitation, average annual temperature, soil type, elevation and distance to the road. The most produced crop for Madagascar is rice, for Ethiopia maize and for DR Congo Cassava (FAO, 2015a). Since rice cultivation often requires waterlogged soils (FAO, 2015d), the slope percentage is an important characteristic for rice cultivation and is therefore added to characteristics for rice. The quantitative data for rice, maize and cassava cultivation land quality is displayed in table 1. For all crops, a road distance of less than 5 km was set as suitable, as explained in section 3.1. Also explained in section 3.1 is the preference for only displaying the suitable category in table 1.
Parameter Rice (Madagascar) Maize (Ethiopia) Cassava (DR Congo)
Precipitation (mm) >1200 500-1500 500-2000
Temperature (°C) 12-35 10-45 10-45
Soils Arenosol, Vertisol, Solonetz, Chernozem, Solonchak, Nitisol, Gleysol, ferralsol, Kastanozem, Phaeozem, Gypsisol, Calcisol, Retisol, Acrisol, Lixisol, Alisol, Luvisol, Cambisol, Fluvisol, Stagnosol. Arenosol, Vertisol, Solonetz, Chernozem, Solonchak, Nitisol, Gleysol, ferralsol, Kastanozem, Phaeozem, Gypsisol, Calcisol, Retisol, Acrisol, Lixisol, Alisol, Luvisol, Cambisol, Fluvisol. Arenosol, Vertisol, Solonetz, Chernozem, Solonchak, Nitisol, Gleysol, ferralsol, Kastanozem, Phaeozem, Gypsisol, Calcisol, Retisol, Acrisol, Lixisol, Alisol, Luvisol, Cambisol, Fluvisol. Elevation (m) Slope (%) Distance to road (km) <1500 0-8 <5 <2000 <5 <1800 <5
Table 1. Suitable agricultural land characteristics necessary for the growth of Rice, Maize and Cassava (FAO, 2001; FAO, 2015b; FAO, 2015c; Nguyen, 1998).
As can be clearly seen in table 1, cultivation of rice requires a large amount of precipitation and a flat or only slightly sloping area, since the rice fields are often flooded. 1200mm of rainfall is
adequate, while the slope should not exceed 8% (Nguyen, 1998). Furthermore, a moderate to high temperature is preferred, with an annual average of 21°C as optimal temperature (FAO, 2015c). However, lower temperatures still allow rice growth although the temperature should not drop under 12°C and 35°C is during many stages of plant growth the critical maximum temperature. Rice can be grown on almost all types of soils, from sandy loam to heavy clay, as is displayed in table 1 (Nguyen, 1998).
Maize represents the second crop in table 1. This crop requires an annual average precipitation of minimal 500mm, although 1200-1500mm achieves the best yields. Maize has no tolerance to
flooding, so more precipitation will create unsuitable circumstances. The temperature should range within 10 - 45°C. Furthermore, the elevation must be below 2000, because only long-term varieties can survive in such high altitudes and flowering and maturing will then take more than a year, this is not considered economically viable. Lastly, maize cultivation requires a well-drained, fertile soil (FAO, 2015c). 20 options are enlisted in table 1 (FAO, 2015b).
Thirdly, the suitable characteristics for Cassava growth are enlisted in table 1. The annual minimum requirements for precipitation are 500mm, whereas the optimum lies within 1000-1500mm (FAO, 2001). A maximum amount was not found by literature research and therefore estimated on 2000mm, following the assumption that the suitable range lies 500mm under and above the optimum. The suitable temperature range lies within 10-45°C and cassava can grow on a wide range of soils. However, best soils for cassava are well-drained, light-textured, deep soils of intermediate fertility. Cassava can be grown on altitudes up to 1800m (FAO, 2001).
4.2 National results
4.2.1 Madagascar
The island of Madagascar is located at the south-east of Africa. It lies between 11057 and 250 29
South and 430 14 and 500 27 East. The island has an area of 587,041square kilometers. The following
Figure 3. a) Annual average precipitation of Madagascar. b) Annual average temperature of Madagascar. c) Soils of Madagascar. d) Elevation of Ethiopia. e) Slopes of Madagascar. f) 5 km distance from roads of Madagascar.
The first map (figure 3a) shows the annual average precipitation of Madagascar, which ranges from the classes 251-500 to 2501-3000mm. Most precipitation is received at the east coast of the island with some wet areas at the center as well, whereas the south is known for its low rainfall and long dry season from April to October. Unfortunately, the map did not support reclassification functions in ArcGIS, so the break value of 1200mm for suitable/unsuitable categories could not be showed. However, the green areas (1500-3000mm) are considered suitable, whereas orange (251-1000m) is unsuitable. The yellow category is in-between.
The temperature ranges from the classes 10.1-15°C in the highlands in the center of the country to 25.1-30°C at the north-west (figure 3b). Again reclassification was not possible. However, since the suitable category ranges between 12 and 35°C, only the yellow color (10.1-15°C) at the highlands is considered unsuitable for rice cultivation.
Figure 3c shows the soils of Madagascar. The largest area is comprised by Ferrasols in the east and center of the country, covering about 40% of the island. Subsequently, bands of Cambisols, Lixisols, Greyzems and Arenosols are found, cut by some minor soil categories. On all soils enable rice cultivation, so no unsuitable category is assigned in relation to soils.
Next, map 3d shows the elevation of Madagascar. This map is already divided in a suitable (<1500m) area and an unsuitable (>1500m) area. The unsuitable area is located at the highlands in the center of Madagascar and this area is partially corresponding to the unsuitable area in relation to the low temperature.
The slope map of figure 3e shows mostly the mountain ridge from the north to the south of the island. Although reclassification was not possible, it is possible to identify suitable areas in the green and yellow colors, orange and red colors represent unsuitable areas for rice cultivation.
The last map, figure 3f, shows the roads of Madagascar with a 5km distance buffer around them. All areas that are not covered with this buffer are unsuitable for agriculture. However, almost all the country is good accessible by road. The largest unsuitable area is located at the north of the country.
4.2.2 Ethiopia
Ethiopia is a country in the horn of Africa, bounded to the north by Eritrea, to the west by Sudan, to the south by Kenya and to the east by Somalia and Djibouti. It lies between 3°24’ and 14°53’ North; and 32°42’ and 48°12’ East. Ethiopia covers 1,120,000 square kilometers. The following maps show
the national results for Ethiopia.
Figure 4. a) Annual average precipitation of Ethiopia. b) Annual average temperature of Ethiopia. c) Soils of Ethiopia. d) Elevation of Ethiopia. e) 5 km distance from roads of Ethiopia.
In the first map (figure 4a) the annual average precipitation of Ethiopia ranges from 95 to 2226 mm. The area facing the most precipitation is located at the west part of Ethiopia, which is closest to South Sudan. The area in the southern east side, located next to Somalia, endures the least
precipitation. Due to the map format, reclassification of the precipitation map was once more not possible. Suitable areas for maize cultivation range between a precipitation of 500-1500mm. Yellow and light green areas (south-east) therefore receive too less precipitation and are unsuitable and in the blue areas the precipitation is too large (the center of the country). Therefore, only the green areas are considered suitable for maize cultivation.
The annual average temperature of Ethiopia (figure 4b) ranges from 4°Cin the middle,
mountainous region of the country to 31°C in the north eastern part of the country. Whereby should be mentioned that at all corners of the country small areas of high temperatures occur. The
temperature is too low in the center, mountain region of the country (blue area in figure 4b), and therefore this area is considered unsuitable for maize cultivation.
In figure 4c the soils of Ethiopia are displayed: Leptosol, Fluvisol, Regosol, Vertisol, Cambisol, Luvisol, Nitisol, Phaeozem, Andosol, Acrisol, Gypsisol. All soil types of Ethiopia occur at several places throughout the country. Leptosols and Andosols are unsuitable for maize cultivation. Leptosols cover about 15% of the country, mostly in the north east, whereas Andosols are less common.
The altitude map (4d) has already been reclassified in areas which have an altitude lower than 2000 meters, and areas with an altitude higher than 2000 meters above sea level, representing respectively the suitable area and the unsuitable area. The unsuitable area based on altitude, corresponds with the low temperature area and the area with too much precipitation.
The last map (figure 4e) shows the areas with a distance of 5km or less to the roads. The largest road density is in the north and south of the country. All areas that are not covered with the orange color are unsuitable for agriculture.
4.2.3 DR Congo
DR Congo is located at central Africa and borders Republic of the Congo, the Central African Republic and South Sudan to the north, Uganda,
Rwanda, Burundi and Tanzania to the east and Zambia and Angola to the south. In the west DR Congo is bordered to the Atlantic Ocean. DR Congo covers 2,345,409 square kilometers.
Figure 5. a) Annual average precipitation of DR Congo. b) Annual average temperature of DR Congo. c) Soils of DR Congo. d) Elevation of DR Congo. e) 5 km distance from roads of DR Congo.
Figure 5a shows the annual average precipitation of DR Congo. The precipitation ranges from the classes 751-1000mm to 2001-2500mm. The south-east is relatively dry compared to the north-west
of the country. However, only the dark green spots in the category 2001-2500mm are considered unsuitable for cassava cultivation, because these areas are too wet.
The temperature in DR Congo ranges from the classes 10.1-15°C to 25.1-30°C (figure 5b). The coldest area is the mountain ridge in the east of the country. However, the annual average temperature does not go under the critical minimum temperature of 10°C for cassava. Therefore, there are no unsuitable areas for cassava cultivation based on the temperature.
Figure 5c shows the soil type distribution in DR Congo. The largest area is comprised by Ferrasols. The map shows furthermore Acrisols and Cambisols in the east, Gleysols in the west, Nitisols in the north and Arenosols in the south, combined with some minor soil classes. All major soil types of DR Congo are suitable for cassava cultivation.
Next, map 5d shows the elevation of DR Congo. The elevation map is again already divided in a suitable area (<1800m) and an unsuitable area (>1800m). The unsuitable area is really small, with only some outcrops at the east of the country.
The last map, figure 5e, shows the roads of DR Congo with a 5km distance buffer around them. All areas that are not covered with this buffer are unsuitable for agriculture for economic reasons. The unsuitable areas are mostly located in the northern and southern part of the country, whereas the road density is rather high in the middle band of the country.
4.3 The land deals
4.3.1 Madagascar
Although Anseeuw et al. (2012) reported that Madagascar was the country with the second most reliable land deal data and approximately 41 land deals, at the moment only 9 land deals were recorded in the Land Matrix (Land Matrix, 2015). The land deals are well spread throughout the country, as displayed in figure 6.
4.3.2 Ethiopia
According to Anseeuw et al. (2012) was Ethiopia the country with the third most reliable land deal data and the country with the highest number of reported land deals. 60 deals were recorded in the land matrix database (Land Matrix, 2015). The deals are displayed in figure 7. The land deals seem to be concentrated at the west side and central part of the country.
Figure 7. Land deals in Ethiopia.
4.3.3 DR Congo
DR Congo is the country with the fourth most reliable land matrix data, although the reported number of land deals is quite low (Anseeuw et al., 2012). 12 land deals have been reported in the land matrix (Land Matrix, 2015), however, 3 deals are located at multiple locations and therefore 18 locations have been placed on the map (figure 8). The land deals are mostly located at the center of the country, with a concentration of several land deals along the largest river, le Congo.
4.4 Relations between land deals and quality of land
In the following paragraphs, the maps with the land quality characteristics and the maps with the land deals will be combined, so that conclusions towards the suitability of the agricultural land for the land deals can be drawn.
4.4.1 Madagascar
As is displayed in figure 9a, only two of the land deals are situated in wet enough areas for rice growth (green area). Whereas four more land deals are situated in transition zones (yellow). The temperature map 9b, shows that all the land deals are located at suitable temperature levels for rice growth. In map 9c, all suitable soils for rice growth are colored green, this reveals that all nine land deals are situated at suitable soils for rice growth. The next map (9d) displaying the elevation shows that all land deals are below 1500 meters, which is again suitable for rice growth. Zooming into the slope map (9e) reveals that all the land deals are situated on the green area, indicating a slope of maximum 4.7 degrees. As rice can endure slopes until 8 degrees, all land deals are in suitable areas. Lastly all the land deals are situated at a distance of 5 kilometer or less from a road (9e).
Figure 9. a) Relation of precipitation and land deals in
Madagascar, b) temperature and land deals, c) suitable soils and land deals d) elevation and land deals, e) slope and land deals, and f) distance of <5km from roads and land deals.
So apart from the precipitation, all land deals are situated on suitable agricultural land for rice growth. However, precipitation is an important limiting factor for plant growth; hence 22% of the land deals are situated on unsuitable agricultural land for rice growth and 44% of the land deals or located in the transition zone.
It can be concluded that the hypothesis that foreign investors take the best lands, has to be rejected based on the production of rice because of the precipitation map. This might indicate that foreign investors grow other crops than rice, and therefore do compete less with the local
communities for the best agricultural ‘rice’ land.
4.4.2 Ethiopia
Figure 10. a) Relation of precipitation and land deals in Ethiopia, b)
temperature and land deals, c) suitable soils and land deals d) elevation and land deals, and e) distance of <5 km from roads and land deals.
Figure 10a shows that 20 of the 60 land deals are situated in to wet (blue) or to dry (yellow) areas for maize growth. However, the tension of land deals in the west of the country seem to correspond with
the precipitation map, where the west is the area with
more precipitation. In the temperature map (10b), all land deals in the cold, blue areas are located at land unsuitable for maize growth. The soil map (10c) shows that 10 of the 60land deals are situated on white areas. While only the green areas are suitable soils for maize growth. Figure 10d reveals that 19 of the land deals the altitude is too high for maize growth. The road map (10e) shows a lot of 5 kilometers distance to road areas, with the largest road density in the north and south of the country. If land deals are compared with the road map, 7 of the 60 land deals are situated more than 5 kilometers from a road. These seven land deals are spread over the country. Hence the center of Ethiopia and the east part are unsuitable for maize growth and some unsuitable areas exist throughout the country.
Therefore, for Ethiopia it is not easy to confirm of reject the hypothesis that foreign investors take the best agricultural land. However, a third of the land deals are located at unsuitable agricultural land for maize growth, so the hypothesis will be rejected.
4.4.3 DR Congo
Figure 11ashows that only the darker green areas receive more than 2000 mm of precipitation annually, which is unsuitable for cassava growth. Only one of the land deals is situated in a too wet environment. In figure 11b the temperature ranges from 10 to 32 degrees, which is suitable for cassava growth. Figure 11c, the suitable soils are colored green. The map shows that most of the land deals are not situated on suitable lands for cassava growth. Namely they are located near water bodies (rivers and/or lakes). However, these areas are marked unsuitable because ‘water bodies’ itself is not a soil, so they can be considered as areas with ‘no data’. However, the land deals will probably be situated next to the water bodies and not in the water bodies, unfortunately, the maps are not detailed enough to see which soils are existing there. As there are no unsuitable marked soils (apart from water bodies) near the land deals, it will be assumed that the land deals are situated on suitable land. The next map (11d) shows that all land deals are situated at an elevation lower than 1800 meters, which is suitable for cassava growth. Lastly, the 5 km road map (11e) shows that all land deals are occurring within 5 kilometers of a road.
Hence 94% of all land deals are situated on suitable lands if water bodies are neglected as unsuitable soils. This means that the hypothesis that the foreign investors take the best land for cassava growth can be confirmed. However, cassava is a crop that can grow in many different conditions and is even often used on marginal land. Therefore, the conclusion for this statement based on cassava should be taken with some caution.
Figure 11.a) Relation of precipitation and land deals in DR Congo, b) temperature and land deals, c) suitable soils and land deals d) elevation and land deals, e) (distance of <5km from) roads and land deals.
5. Conclusion and discussion
The purposes of this report was to proof or reject the hypothesis that foreign investors take the best land available, leaving the local communities to move to land of lesser quality. In order to survey this hypothesis, the following main question needed to be answered: “to what extent do large scale land based investors take the best agricultural land available?”. In order to answer this main question two sub questions were made: “which characteristics determine the quality of agricultural land?” and “are suitable agricultural lands for specific crop growth and the occurrences of land deals related”.
Based on the results the questions can be answered. The first question “which characteristics determine the quality of agricultural land?” has been answered by using Repko’s integrative
technique of extension (Repko, 2012), in which the earth sciences definition has been extended with the economic view. The quality of land is then determined by the amount of precipitation,
temperature, elevation, soil type, slope for rice and road distance. There has been chosen to measure the first 5 factors based on the most produced crop in the targeted country, in this manner the best agricultural land has been viewed from a local perspective. The last factor (road distance) has been classified as high quality land with a travel distance of less than six hours to the next market based on available transportation. For measuring purposes, this has been translated into less than 5 km distance from the nearest road.
The second question “are suitable agricultural lands for specific crop growth and the occurrences of land deals related” cannot be undisputedly answered, since the results were varying.
In Madagascar, 7 of the 9 land deals are situated at unsuitable or transition areas for rice growth, due to a lack of precipitation. Hence, based on the precipitation, suitable areas for rice growth in Madagascar are situated in the middle and north part of the country. The other factors show a positive relation between land deals and suitable lands. However, since all factors have to be suitable, it can be concluded that most of the land deals are not situated at suitable land for rice growth.
In Ethiopia, the central mountainous region and the dry eastern parts are unsuitable for maize growth. The suitable areas are situated at all sides of the central mountain region of the country, where the temperature, precipitation, elevation and road distance are suitable for maize growth. The dispersion of suitable soils for maize growth is throughout the whole country. Hence no conclusion on suitability of soils can be made. One third of the land deals are negatively correlated with the climatic factors. Based on these outcomes can be concluded that a major part of the land deals is not situated at suitable land for maize growth.
For cassava growth in DR Congo, almost no unsuitable agricultural land exists. Taken this in account, the conclusion that the land deals are indeed situated in suitable land for cassava growth should be taken with caution. Furthermore, should be mentioned that water bodies (which is no soil) has been neglected as an unsuitable soil and classified as ‘no data’ areas. There has been chosen to mark this ‘no data’ as suitable land, since no unsuitable soils have been found near the water bodies on the map.
There are several explanations for this result. Firstly, the best agricultural land was viewed from the local perspective, hence based on the nationally most produced crop, because there has been tried to find out if the land deals were displacing local communities to agricultural land of lesser quality. However, foreign investors often use other crops with different characteristics and therefore the best agricultural land for local communities is not automatically the best agricultural land for foreign investors. The competition for land between investors and local communities could be therefore smaller than was expected. Secondly, as foreign investors are capable of buying land, they have probably also the monetary means to use irrigation for their crops. If true, this could explain why in Madagascar most of the land deals are situated in areas with a lack of precipitation. However,
this does not explain why the land deals in Ethiopia are situated in unsuitable areas, as there are also land deals in to wet areas. This might again be explained by the use of different crops, with higher tolerance to high amounts of precipitation. Further research to the crop use of the foreign investors is therefore recommended.
One factor showed a remarkable 96% positive correlation with land deals: road distance. It seems naturally that foreign investors choose areas within a good infrastructure, since in the first place, they want to make profit in these areas and therefore need the roads to transport their products to the market or their home country.
Based on these sub questions, the main question: “to what extent do large scale land based investors take the best agricultural land available?” can be answered.
In Madagascar only 22% of the land deals are situated in suitable lands. In Ethiopia 33% of the land deals are situated in suitable agricultural. Only DR Congo has a remarkable 94% of the land deals occurring in suitable lands.
Based on this results it can be said that investors thus not take the best agricultural land available. However, as noted before, ‘the best agricultural land available’ is seen from the local perspective. This conclusion therefore means that the competition for land between the local communities and foreign investors is smaller than at forehand expected and local communities are not displaced from their ground, as long as the land deals area does not get to large. It is therefore advised that national policies according land deals should set a maximum to the land deal area, so that the local communities have enough land to sustain food security.
6. References
Anseeuw, W., Boche, M., Breu, T., Giger, M., Lay, J., Messerli, P., & Nolte, K. (2012). Transnational land deals for agriculture in the global South.
ArcGIS (2015). ArcGIS Online.
Borensztein, E., De Gregorio, J., & Lee, J. W. (1998). How does foreign direct investment affect economic growth?. Journal of international Economics, 45(1), 115-135.
Cassman, K. G. (1999). Ecological intensification of cereal production systems: Yield potential, soil quality, and precision agriculture, PNAS 96(11), 5952-5959.
Cotula, L. (2011). Land deals in Africa: What is in the contracts?. IIED.
Cotula, L., Vermeulen, S., Leonard, R., Keeley, J. (2009). Land grab or development opportunity?: agricultural investment and international land deals in Africa. Iied.
Deininger, K. W., &Byerlee, D. (2011). Rising global interest in farmland: can it yield sustainable and equitable benefits?. World Bank Publications.
DPI/DHLGP. (1993). Planning guidelines: The identification of good quality agricultural land. Queensland: CSIRO Publications
Edelman, M. (2013). Messy hectares: questions about the epistemology of land grabbing data. Journal of Peasant Studies, 40(3), 485-501.
FAO. (1976). A framework for land evaluation. Soil Bulletin 32. Food and Agricultural Organization of the United Nations, Rome.
FAO (2001). Strategic Environmental Assessment. An assessment of the Impact of Cassava Production and Processing on the Environment and Biodiversity. Volume 5. Rome. Food and Agriculture
Organization of the United Nations, International fund for agricultural development. FAO. (2015a). FAOSTAT. Retrieved 1 October 2015 from:
http://faostat.fao.org/site/485/default.aspx#ancor
FAO (2015b). World Reference Base for soil resources 2014: International soil classification system for naming soils and creating legends for soil maps, Food and Agriculture Organization of the United Nations.
FAO. (2015c). Grassland species profiles. Retrieved 1st October 2015: http://www.fao.org/ag/agp/AGPC/doc/Gbase/data/pf000342.htm
FAO. (2015d). Rice productions systems.Retrieved 2th Decembre 2015 from Climate-Smart Agriculture: http://www.fao.org/climatechange/climatesmartpub/66245/en
Hanley, N., Shogren, J. F., & White, B. (2001a). §6.1 Economic Growth and Development. In Introduction to Environmental Economics (pp. 120-124). New York: Oxford University Press. Hanley, N., Shogren, J. F., & White, B. (2001b). §6.2 Predictions from the Past. In Introduction to Environmental Economics (pp. 125-129). New York: Oxford University Press.
Jenny, H. (1941). Factors of Soil Formation: A System of Quantitative Pedology. Dover Publications, INC. New York.
Kleemann, L., & Thiele, R. (2015). Rural welfare implications of large-scale land acquisitions in Africa: A theoretical framework. Economic Modelling, 51, 269-279.
Land Matrix (2015). Retrieved from landmatrix.org at Novembre 2015: http://www.landmatrix.org/en
Lavers, T. (2012) ‘Land grab’ as development strategy? The political economy of agricultural investment in Ethiopia, The Journal of Peasant Studies, 39:1, 105-132, DOI:
10.1080/03066150.2011.652091
Nguyen, V. N. (1998). Factors affecting wetland rice production and the classification of wetlands for agricultural production. Wetlands characterization and classification for sustainable agricultural development. Harare: Food and Agriculture Organization of the United Nations, Sub-Regional Office for East and Southern Africa.
OECD. (2015). African Economic Outlook 2015: Regional Development and Spatial Inclusion. OECD Publishers.
