This article was downloaded by: [KU Leuven University Library]
On: 04 January 2013, At: 00:40
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
African Geographical Review
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/rafg20
More people, more trees in South
Eastern Tanzania: local and global
drivers of land-use/cover changes
Andrew Kabanza
a
, Stefaan Dondeyne
b
, John Tenga
a
, Didas
Kimaro
c
, Jean Poesen
b
, Elly Kafiriti
b
& Jozef Deckers
a
a
Naliendele Agricultural Research Institute, Mtwara, Tanzania
b
Department of Earth and Environmental Sciences, Catholic
University, Heverlee, Belgium
c
Department of Agricultural Engineering and Land Planning,
Sokoine University of Agriculture, Morogoro, Tanzania
Version of record first published: 04 Jan 2013.
To cite this article: Andrew Kabanza , Stefaan Dondeyne , John Tenga , Didas Kimaro , Jean
Poesen , Elly Kafiriti & Jozef Deckers (2013): More people, more trees in South Eastern
Tanzania: local and global drivers of land-use/cover changes, African Geographical Review,
DOI:10.1080/19376812.2012.746093
To link to this article: http://dx.doi.org/10.1080/19376812.2012.746093
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-
conditions
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation
that the contents will be complete or accurate or up to date. The accuracy of any
instructions, formulae, and drug doses should be independently verified with primary
sources. The publisher shall not be liable for any loss, actions, claims, proceedings,
demand, or costs or damages whatsoever or howsoever caused arising directly or
indirectly in connection with or arising out of the use of this material.
RESEARCH ARTICLE
More people, more trees in South Eastern Tanzania: local and global
drivers of land-use/cover changes
Andrew Kabanza
a
, Stefaan Dondeyne
b
*, John Tenga
a
, Didas Kimaro
c
, Jean Poesen
b
, Elly
Kafiriti
b
and Jozef Deckers
a
a
Naliendele Agricultural Research Institute, Mtwara, Tanzania;
b
Department of Earth and
Environmental Sciences, Catholic University, Heverlee, Belgium;
c
Department of Agricultural
Engineering and Land Planning, Sokoine University of Agriculture, Morogoro, Tanzania
(Received 28 April 2012; accepted 29 October 2012)
Land degradation in South Eastern Tanzania, the country’s major cashew producing area,
has been attributed to deforestation. By comparing land-use/cover maps derived from aerial
photographs of 1965 with maps derived from satellite images of 2002, we assessed how
land-use changed in six villages, and relate these to local and global drivers. Land-use/cover
changes are complex processes, which we analyzed by determining the relative net changes,
losses, persistence and gains of each land-use/cover categories. Widespread planting of
cashew trees only started in the 1960s; while the ‘villagisation’ program in the 1970s,
altered settlement patterns as centrally planned villages were created. Population growth and
rural development policies were major local drivers for land-use/cover change; international
trade and technological innovations were principal global drivers. Though population
increase led to a reduction of natural vegetation, the spread of cashew trees resulted in a
case of ‘more people, more trees.’ How far the ensuing deforestation affected the
biodiversity of the area and how sustainable the production of cashew nuts actually is,
remains yet unresolved questions.
Keywords: cashew nut; deforestation; villagisation; population growth; land degradation
Introduction
General concern about land-use change arose from the realization that transformation in land-
use/cover influences, surface hydrology, soil erosion, climate and biodiversity, hence the inter-
est in processes leading to deforestation, desertification and other changes in natural
vegetation (Lambin, Geist, and Lepers, 2003). It is commonly accepted that the growing
human population in the world incre asingly needs more arable land, which results into a
decline of world’s forests, grasslands and woodlands (Meyfroidt, Rudel, and Lambin, 2010).
However, while land-use changes are often attributed to a single factor, such as shifting
cultivation or population growth, Lambin et al. (2003) argued that land-use change is always
caused by multiple interacting factors, with demographic, economic, technological, policy and
cultural factors as the major underlying causes.
South Eastern Tanzania is the country’s principal production area of cashew nuts
(Anacardium occidentale L.), one of Tanzania’s major export commodities (Topper and
Kasuga, 2003; CBT, 2008); in the period from 2003 to 2008 Tanzania ranked between the
*Corresponding author. Email: stefaan.dondeyne@ees.kuleuven.be
African Geographical Review, 2013
http://dx.doi.org/10.1080/19376812.2012.746093
Ó 2012 The African Specialty Group of the Association of American Geographers
Downloaded by [KU Leuven University Library] at 00:40 04 January 2013
2
nd
and 8
th
world biggest cashew nut producing countries (FAO, 2011). Most of the cashew
nuts are exported as raw nuts to India, where they are processed before being re-exported to
Europe or North America (URT, 2008). Paradoxically, while So uth Eastern Tanzania is
embedded in the world trade through the cashew production, due to the poor road infrastruc-
ture has been isolated from the rest of the country and has often been regarded a backward
periphery (Seppälä and Koda, 1998). Widespread planting of cashew trees, which are mostly
grown by smallholders, only took hold in the 1960s (Topper and Kasuga, 2003). At that time
people in South Eastern Tanzania were living scattered over the area. In the 1970s, they were
resettled into newly created villages as part of the ‘villagisation’ program. Figure 1 illustrates
the impact of the introduction of cashew trees and villagisation on land-use/cover in South
Eastern Tanzania. The aim of the villagisation policy was to achieve rural transformation
through village settlement schemes as had been recommended by the World Bank in 1960
(Lugoe, 2008). By concentrating people in villages it was expected not only to facilitate the
provision of public services, but also to allow for the possibility of large-scale farming and,
doing so, to increase productivity. Villagisation was implemented more systematically in
South Eastern Tanzania than in other parts of the country (Coulson, 1977).
The Makonde plateau is the most important cashew growing area in South Eastern
Tanzania and where land degradation processe s, such as soil erosion (Achten et al., 2008), soil
acidification (Ngatunga et al 2001; 2003) and losses of soil organic carbon (Rossi et al., 2009),
Figure 1. Aerial photographs of 1965 and of 1981 illustrating land-use/cover changes around Lipalwe,
at the southern edge of the Makonde plateau: (a) in 1965 people were living scattered and annual crops
were dominating the agricultural landscape (light coloured fields) which were in rotation with a
bushfallow (dark coloured fields); (b) in 1981 people were living in large villages (arrows); fields on the
escarpment, which had annual crops in 1962 had reverted to woodland in 1981 (frame 1), while on the
plateau cashew trees, appearing as dark dots, dominate the landscape (frame 2).
2 A. Kabanza et al.
Downloaded by [KU Leuven University Library] at 00:40 04 January 2013
have been linked to land-use/cover changes. As the FAO statistics show (FAO, 2011), follow-
ing a steady increase in the 1960s cashew nut production reached a peak of 145000 MT in
1973. Thereafter, it went into a catastrophic decline reaching a minimum in 1989 of 17000
MT. As resettled people were bound to abandon their fields, this decline has partly been attrib-
uted to the villagisation policy. Moreover, during this period, the government took control over
the marketing system leading to producer prices dropping from 70% of the export price in
1972 to 24% of it in 1980. Production was further constrained in the 1970s by the outbreak of
powdery mildew disease (Oidium anacardii Noack), a fungus whose spread was favored by
the many untended trees (Topper and Kasuga, 2003). Cashew nut production only recovered
with the market liberalization in the 1990s, as farmers were getting favorable prices again
(Rweyemamu, 2002). Meanwhile, internationally supported research found that sulfur is
effective to control the disease. The multi-donor funded Cashew Improvement Program
promoted the use of sulfur together with other improved crop husbandry practices, and so
contributed to the recovery of the cashew nut production (Topper and Kas uga, 2003).
Understanding the patterns and processes of land-use/cover changes has become a funda-
mental goal in studies when investigating the complex interactions between humans and the
environment (Aldwaik and Pontius, 2012). The standard method for comparing land-use/cover
status at two points in time is to determine a cross-tabulation matrix comparing swaps and
overall changes in land-use/cover categories. To gain more detailed insights into the land-use/
dynamics Pontius, Shusas, and McEachern (2004) proposed to analyze the net change, loss,
persistence and gain of land-use/cover categories.
Given the resettlement program, population growth and the increased cashew production
in South Eastern Tanzania, we wanted to assess how these changes may have affected land-
use/cover changes. By focusing on in six villages, we particularly wanted to get better
insights into the forces driving these changes. We therefore not only determined the overall
changes but also investigated how various potential drives may have affected the net change,
loss, persistence and gain of areas allocated to settlement, annual crops, cashew orchards and
natural vegetation.
Materials and methods
Description of the study area
The Makonde plateau (38° 03’–40° 30’ E and 10° 05’–11° 25’ S) reaches 120 km inland and
is separated from the coast by a narrow plain (Figure 2). Sloping from east to west, its alti-
tude ranges from 80 to 927 m above sea level. The eastern, lower part has sharply incised
valleys and is mapped as the ‘Makonde Dissected Plateau’; the western, higher part is
mapped as the ‘Makonde High Plateau’ where a 300 metre high escarpment forms the
western edge of the plateau.
Natural vegetation on the Makonde plateau is part of the Eastern African coastal forests,
which is considered a global biodiversity hotspot (Timberlake et al., 2011). It consists primar-
ily of woodlands, wooded grasslands and bushlands. Woodlands and pockets of evergreen
forests also occur on the escarpment. The traditional cropping system is a fallow system: after
slashing and burning of forest or fallow vegetation, fields are cultivated for four years with
maize, upland rice, sorghum and cassava. Though cashew trees produce well on the plateau,
they only produce below 800 meters above sea level, as at higher altitude the climate is too
cold for fruit setting (Dondeyne et al., 2003a).
We studied land-use/cover changes in six villages: two on the Makonde Dissected Plateau
(Naliendele, Nachunyu), two on the Makonde Hi gh Plateau (Mahuta bondeni, Nambunga)
and two at the base of the escarpment (Chiwambo, Lipalwe) (Figure 2).
African Geographical Review 3
Downloaded by [KU Leuven University Library] at 00:40 04 January 2013
Land-use/cover maps
Land-use cover maps were made based on panchromatic aerial photographs of 1965 (scale
approximate scale of 1:50,000, Survey and Mapping Division of Tanzania) and which were
compared with land-use/cover maps deriv ed from Landsat TM images from 2002. Aerial
photographs were interpreted using mirror stereoscope. Image interpretatio n elements and
characteristics such as relief, drainage pattern, vegetation cover, land-use, association of
objects and features such as footpath and tracks were taken as key attributes for land cover
analysis following procedures as outlined by Dent and Young (1981) and Lillesand and
Kiefer (2000). The interpreted features were then used as a base for establishing broad classes
of land cover units, which were delineated, digitised and converted into a GIS database
(ESRI, 1995) to produce land-use/cover maps of the study sites for 1965 period. These maps
were geo-referenced to Universal Transversal Mercator (UTM) Coordinates zone 37 south,
using the topographic maps as reference (sheets: Mtwara 296/3, Tandahimba 307/1, Newala
306/4, Namikupa 307/3, Lulindi 306/3).
Landsat TM images of 2002 had been obtained from the US Geological Service.
1
These
images were digital enhanced with the ERDAS software progra m (ERDAS, 2003). To rein-
force the visual interpretability of the image, a color composite (Landsat TM bands 4 5 3)
was prepared and its contrast was stretched using a Gaussian distribution function. Further-
more, a 3 3 high pass filter was applied to the color composite to further enhance visual
interpretability of linear features, e.g. rivers, and patterns such as cultivation. To ensure
accurate ident ification of land-use/cover classes and geometric compatibility with infor mation
obtained from aerial photo interpretation, the Landsat TM images were also geo-referenced to
the co-ordinate system of the national topographic maps (UTM37s). Subsequently, land units
that could visually be identified as homogenous land- use/cover classes were digitized on
screen using ArcView software (ESRI, 1995) to obtain maps comparable to, and compatible
with, the ones derived from the aerial photographs.
The preliminary made maps were subsequently verified and corrected after cross checking
in the field in 2004. GPS was used to locate observation sites. Local people were involved to
give additional information on land-use/cover.
Figure 2. Location of six study sites on the Makonde plateau in South Eastern Tanzania. Terrain
image derived from SRTMv4 data (available at http://srtm.csi.cgiar.org).
4 A. Kabanza et al.
Downloaded by [KU Leuven University Library] at 00:40 04 January 2013