Content uploaded by Felicia O. Akinyemi
Author content
All content in this area was uploaded by Felicia O. Akinyemi on Jul 18, 2016
Content may be subject to copyright.
This article was downloaded by: [Felicia Akinyemi]
On: 20 January 2013, At: 23:06
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
International Journal of Remote
Sensing
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/tres20
An assessment of land-use change in
the Cocoa Belt of south-west Nigeria
Felicia O. Akinyemi a
a Faculty of Architecture and Environmental Design, Kigali
Institute of Science and Technology, Kigali, Rwanda
Version of record first published: 18 Jan 2013.
To cite this article: Felicia O. Akinyemi (2013): An assessment of land-use change in the Cocoa Belt
of south-west Nigeria, International Journal of Remote Sensing, 34:8, 2858-2875
To link to this article: http://dx.doi.org/10.1080/01431161.2012.753167
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.
International Journal of Remote Sensing
Vol. 34, No. 8, 20 April 2013, 2858–2875
An assessment of land-use change in the Cocoa Belt
of south-west Nigeria
Felicia O. Akinyemi*
Faculty of Architecture and Environmental Design, Kigali Institute of Science and Technology,
Kigali, Rwanda
(Received 3 February 2012; accepted 14 August 2012)
The Nigerian government is reviving the agricultural sector to shift from its sole depen-
dence on crude oil for foreign exchange earnings. Thus, the Cocoa Belt (agro-ecological
region) of southwest Nigeria is important to the national economy. With the increas-
ing demand for land to grow export crops and to meet other needs such as settlement
expansion, land use is changing. Land-use data and mapping are essential inputs for the
process of formulating, implementing, and monitoring policy with the aim of reducing
the impact of land-cover/land-use (LCLU) change. Land-use types, their spatial extent
and dynamics over a 25 year period are examined from multispectral images of the
Landsat Thematic Mapper and Enhanced Thematic Mapper Plus. This study examines
the main drivers of LCLU change and the environmental impact. Results show that for-
est conversion to agricultural lands is the main trend, and cultivation is the main cause
of forest loss in the study area. The need to produce food for the teeming population,
coupled with the government’s policy to expand export crop production is resulting in
the loss of native forest, including areas designated as forest reserves. Results under-
score the need for deliberate land-use planning and management in this belt. This study
reveals the situation of unplanned and rapid changes to land use in the context of a
developing country where explicit policies to cater for such activities are absent.
1. Introduction
Embarking on export diversification, Nigeria adopted a trade liberalization policy during
the 1980s and 1990s. With the goal of expanding agricultural export production, the agri-
cultural sector is being revived. Agriculture is strategic to the Nigerian economy and plays
the key roles of supplying food for the population, raw materials for industries, earning for-
eign exchange, which is next only to that from crude oil. More than 70% of the population
derives their living from agriculture and agro-allied activities, with the sector contributing
about 41% of the gross domestic product, accounting for 5% of total exports and providing
88% of non-oil earnings (Ajayi et al. 2008). Attention is currently focused on increas-
ing the production of important export crops such as cocoa (Theobroma cacao), rubber
(Hevea brasilensis), oil palm (Elaeis guineesis), kola (Cola acuminata and Cola nitida)
and cassava (Manihot esculenta). Cocoa and rubber are the largest non-oil exports from
Nigeria.
The expansion of agricultural exports has, however, not been neutral to the environment
(UNEP 2002). Trade liberalization policies, as they affect agricultural commodities, often
*Email: felicia.akinyemi@gmail.com
ISSN 0143-1161 print/ISSN 1366-5901 online
© 2013 Taylor & Francis
http://dx.doi.org/10.1080/01431161.2012.753167
http://www.tandfonline.com
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
International Journal of Remote Sensing 2859
tend to concentrate on the increase in economic returns through increases in agricultural
production and output prices without giving due consideration to the environmental impact
(UNEP 2002). The debate about whether or not increasing yield eases pressure on land,
and forested land in particular, is an ongoing one. It is often assumed that intensifying agri-
culture will spare land for nature as, for a fixed demand, higher yields decrease the area
that needs to be cultivated (Angelsen and Kaimowitz 2001; Lambin and Meyfroidt 2011).
Angelsen (2010) studied the impact of different deforestation policies on agricultural pro-
duction and found that local yield increases tend to stimulate agricultural encroachment,
contrary to the logic of the global food equation, which suggests that yield increases take
pressure off forests. Land-use conflict ensues as land demand increases for cultivation and
settlement particularly.
Focusing on the Cocoa Belt of southwest Nigeria, this study quantifies the extent of
major land-use types and analyses land-cover/land-use (LCLU) change. This belt is a con-
ducive region for planting cocoa and is a hot spot of human activity. The natural vegetation
of the area is humid evergreen tropical rainforest, much of which has now been reduced to
secondary forest or replaced by perennial and annual crops (Salami, Ekanade, and Oyinloye
1999). Forest depletion and soil degradation are major consequences of land-use changes
in the belt (Akinyemi 2005). Nigeria lost 5% of its forest annually throughout the 1980s.
Nigeria’s forest area decreased from 14.9 million ha in 1980 to 10.1 million ha in 1990,
corresponding to an annual deforestation rate of approximately 3.5% (Forestry Research
Management, Evaluation and Coordination Unit – FORMECU 1985). The gross national
standing volume of timber tree species in Nigeria for the period between 1978 and 1988 was
estimated to be 181 million m3, while the total volume removed in that same period was
72 million m3(40% of standing volume), and this translates to a loss of 350,000–400,000 ha
of forest land per annum for the country. These trends do not support the government’s
policy of maintaining 20–25% of the land area under forest cover for the well-being of
the national, regional, and global environments (FME 2006). Neither does forest deple-
tion support the Millennium Development Goal’s target of minimizing forest loss in the
tropical regions of the world by 2015 (see United Nations – UN 2010). There is therefore
the need to evolve strategies to ensure the sustainable use of these natural resources in
Nigeria.
This study contributes to ongoing efforts by assessing and mapping land use as well as
analysing land-use dynamics using remote-sensing techniques. There is general agreement
that remote sensing is an adequate tool for producing reproducible and reliable informa-
tion on LCLU at different scales (Mulders 2001). Its applicability in studying tropical
forests has been demonstrated in various studies (e.g. Tanner et al. 1996; Oetter et al.
2000; Imbernon and Branthomme 2001; Tanaka and Sugimura 2001; Rikimaru, Roy, and
Miyatake 2002; Gillieson, Lawson, and Searle 2006; Huete and Saleska 2010). Our abil-
ity to frequently examine the status of LCLU and monitor changes in the major production
regions of the world is important, and remote-sensing techniques offer such a capability (Lo
2000).
For southwest Nigeria, the study of land use using satellite images is recent. The
earliest study mapped vegetation and land-use associations using 1:40,000 scale panchro-
matic aerial photographs (Adejuwon and Jeje 1973). More recently, Amamoo-Otchere
et al. (1998) revised base maps for human population census based on the use of satellite
images for differentiating between built-up and non-built-up land uses; Salami, Ekanade,
and Oyinloye (1999) quantified the extent of forest reserve incursion using aerial pho-
tographs and SPOT XS images; Orimogunje (2005) studied the Oluwa forest reserve;
Salami and Akinyede (2006) examined LCLU change in the region using images from
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
2860 F.O. Akinyemi
Landsat and the newly operational NigeriaSat-1 satellite of December 1986 and December
2004, respectively; Oates et al. (2008) examined options for the sustainable conservation of
tropical rainforest in Nigeria; Ogundele, Omotayo, and Odewunmi (2011) studied changes
in the spatial extent and types of forest in southwest Nigeria.
Several studies have attempted to quantify the extent of LCLU change in southwest
Nigeria but the statistics reported vary (see Table 1). The probable reasons are the dif-
ferences in study sites and the periods utilized for analysis. The main trends evident are
forest-cover depletion and expanding settlements.
The following are few examples of LCLU studies conducted in similar regions.
Nwadialor (2001) investigated deforestation in the Afaka Forest Reserve, Nigeria.
Yemefack, Bijker, and De Jong (2006) investigated spatial segregation of LCLU types
under shifting agriculture in southern Cameroon. Junge et al. (2010) studied the change
of LCLU and soil degradation in different agro-ecological zones of Nigeria and Benin;
Olayiwola, Eludoyin, and Ekecha (2011) studied change in land use in the Mezam Division
of the northwest province of Cameroon.
Furthermore, some factors influencing LCLU in southwest Nigeria were identified to be
the spread of rural settlements, expansion of urban settlement, evolution of road networks,
illegal logging, and government policy on agriculture (Mengistu and Salami 2007). The
long history of agricultural colonization of the area, together with increasing population
density, has further led to modification of the natural vegetation (Adejuwon 1971; Mengistu
and Salami 2007). It is a fact that land-use change is a result of socio-economic change,
which in turn affects the physical environment (Sader et al. 1985).
Sustainable land-use planning is essential whereby the ecological and socioeconomic
values of land are respected while simultaneously directing and finding the best-suited
areas that will fulfil the demands of development (Lambin et al. 2001; Fritsch and Hild
2002). Land-use mapping, change detection, and monitoring forms an integral part of
the regional land-use planning process whereby policies and strategic plans are made,
reviewed, and updated. These tasks typically involve the identification of emerging land-
use patterns, which are normally linked with other planning statistics such as employment,
housing, and population before the full significance of land-use change can be appar-
ent (Yaakup et al. 2002). There is therefore an urgent need for proper management
of land and the concomitant availability of detailed, accurate, and up-to-date land-use
information.
Deriving land-use information and mapping entails the use of an appropriate classi-
fication scheme. Unlike in the developed world where classification schemes suited to
peculiar environments are developed at various levels of details, Nigeria has no single
national land-use classification scheme. Consequently, Nigerian researchers are develop-
ing their own classification schemes making comparison of studies difficult. As there is
no continuity in the schemes used, it becomes practically impossible to use them from
one case study to another. Examples of such schemes are those that are developed by
the United States Geological Survey (Anderson et al. 1976), the Institute of Terrestrial
Ecology Land Classification System in use in Great Britain (Griffiths and Wooding 1989),
CORINE Land-Cover Nomenclature (European Commission 1993), and the Land Cover
Classification System (FAO 2000). A generalized one was developed by the AGRHYMET
and the EROS Data Center for West Africa (Tappan and Cushing 2004).
This work assesses LCLU change based on the premise that land-use change and impact
assessment could lead to the identification of generic trajectories and processes of change
(see Mertens and Lambin 1999).
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
International Journal of Remote Sensing 2861
Table 1. Studies on land use change in southwest Nigeria.
No Land-use type Image (years) *Change (%) Specific locality Study
1Forest/related cover
1.1 Exotic tree/plantation/forest
reserve
1963–1986 −23.26 07◦10N, 04◦33E Salami, Ekanade, and Oyinloye (1999)
1.2 High forest 1986 & 2004 −24.5 National Salami and Akinyede (2006)
1.3 High forest 1986 & 2002 −807
◦0238 –07
◦5556 N
04◦1452 –05
◦0734 E
Mengistu and Salami 2007
2 Agro-forestry
2.1 Agro-forestry 1963–1986 +377.72 07◦10N, 04◦33E Salami, Ekanade, and Oyinloye (1999)
2.2 Agro-forestry/secondary regrowth 1986 & 2002 −49.1 07◦35–08
◦00N,
04◦45–05
◦00E
Akinyemi (2005)
2.3 Derived savanna 1986 & 2002 −71.9 07◦0238 –07
◦5556 N
04◦1452 –05
◦0734 E
Mengistu and Salami 2007
3 Farmland
3.1 Arable farmland 1963–1986 −37.66 07◦10N, 04◦33E Salami, Ekanade, and Oyinloye (1999)
3.2 Shrubland/farmland complex 1986 & 2002 +413.6 07◦0238 –07
◦5556 N
04◦1452 –05
◦0734 E
Mengistu and Salami 2007
4 Settlement
4.1 Settlement/open space 1963–1986 +50 07◦10N, 04◦33E Salami, Ekanade, and Oyinloye (1999)
4.2 Settlement/bare surface 1986 & 2002 +192.4 07◦0238 –07
◦5556 N
04◦1452 –05
◦0734 E
Mengistu and Salami 2007
4.3 Built-up areas/roads 1986 & 2002 +88.41 07◦35–08
◦00N,
04◦45–05
◦00E
Akinyemi (2005)
5Others
5.1 Fallow land 1963–1986 +105.71 07◦10N, 04◦33E Salami, Ekanade, and Oyinloye (1999)
5.2 Derived savanna 1986 & 2004 −59.4 National Salami and Akinyede (2006)
Note: *Decline =–, *increase =+.
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
2862 F.O. Akinyemi
2. Study area
The study area traverses the part of Nigeria known as the Cocoa Belt with an approximate
size of 1735.298 km2. Occurring in the middle of the belt between 7◦27N–7
◦35N
and 4◦31E–4
◦39E, it is limited to a section of the Landsat scene (Path 190, Row
55) comprising Ile-Ife town and its surroundings (see Figure 1). The selection of the study
area was based on the author’s familiarity of the area, availability of auxiliary data and
maps, as well as its representativeness of the LCLU in the belt.
This belt lies within the humid evergreen tropical rainforest (Koppen’s AF climatic
classification, characterized by 1500–2200 mm year−1precipitation, mean monthly
temperature of 27◦C, and mean annual humidity of 76.6%). This is a transitional zone
between the freshwater mangrove swamps along the coast and the derived savannah
(around 1300 mm year−1). Both dry and wet seasons are experienced. The dry season
is short, lasting mainly from November/December to February (FOS 1988; Adejuwon,
Odekunle, and Omotayo 2007).
The study location is underlain by pre-Cambrian basement complex rocks with an undu-
lating topography (ranging from 6–12% inclination). Large sections of this forested zone
are now in patches due to human activities, such as cultivation, lumbering, and fuel wood
collection (UNECA 2007). Increasing population has often been cited as a major LCLU
change agent in southwest Nigeria, which has subjected the vast expanse of the tropical
rainforest zone and the unprotected landscape to intensive human colonization (see Salami,
Ekanade, and Oyinloye 1999).
3. Materials and methods
The availability of new very high resolution satellites such as Pléiades (http://www.
astrium-geo.com/pleiades/), frequent coverage, the possibility of digital image processing,
Figure 1. Location map of the study area.
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
International Journal of Remote Sensing 2863
and the recent free access to satellite images such as the Landsat series (see link
http://landsat.gsfc.nasa.gov/data/where.html) make satellite images available for natural
resources management and monitoring.
3.1. Objectives
This study investigates LCLU over time in order to better understand the process of change.
Are there factors driving the process? If yes, what are they? The main objectives are to
•identify major land-use types in the tropical rainforest zone,
•quantify the spatio-temporal variability of land-use change, and
•identify the driving factors of LCLU change.
This study investigates these salient issues for better appreciation of land-resource man-
agement in southwest Nigeria. LCLU change analysis allows for the identification of
major drivers of change and, by inference, the characterization of land-use dynamics (Pant,
Groten, and Roy 2000). We further examined the impacts of LCLU change, an element that
is missing in many Nigerian studies. This is because the lack of precise information about
LCLU changes and impacts is a major constraint for planning, development, and prudent
use of natural resources.
3.2. Land-use change detection procedure
This article assesses land-use change in the study area (see Figure 2). Figure 2 illustrates
the procedure developed and utilized for the assessment.
3.2.1. Data sources
Multispectral satellite images covering a 25 year period were analysed. To minimize the
influence of seasonal variations on the results, images of the same season were utilized (see
details in Table 2).
Landsat 7 Enhanced Thematic Mapper Plus (ETM+) SLC-off data has 22% of the data
missing from each scene due to a defect in the Scan Line Corrector (SLC), which caused
hardware failure. The deteriorated image quality resulting from SLC failure has become a
major obstacle for using Landsat ETM+data in applications since 2003 (Ju and Roy 2008;
Chen et al. 2011). For details of several methods developed to solve this problem, see USGS
(2004), Maxwell, Schmidt, and Storey (2007), Roy et al. (2008), Pringle, Schmidt, and Muir
(2009), and Chen et al. (2011). To minimize the gaps in the 2011 image, this study uses the
SLC-off to SLC-on (pre-2003 Landsat 7 image) method (see Scaramuzza, Micijevic, and
Chander 2004). It involves applying a local linear histogram matching in a moving window
of each missing pixel to derive the rescaling function. This rescaling function is then used to
convert the radiometric values of one input scene into equivalent radiometric values of the
scene being gap-filled, and the transformed data are used to fill the gaps of that scene. This
method was chosen because it is easy to implement and can resolve many of the problems
relating to missing data since the region of study is relatively homogenous, and the input
scene is of high quality (e.g. negligible cloud cover) and represents comparable seasonal
conditions (USGS 2004). Examples of studies that have utilized Landsat 7 ETM+SLC-off
data in applications are those of Bédard et al. (2008) who studied land-cover mapping and
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
2864 F.O. Akinyemi
Objectives of land-use planning and monitoring
Topographic map
Data
sources
Satellite image Field survey
Training parcel
selection
Georeferencing
Digital
image
processing
Control points
Land-use
change
analysis
Policy
Land-use units
gains and losses Land-use map Classified images
Impact analysis
Land use dynamics of
the Cocoa belt
False-colour
composite
image
01
03
02
Maximum-
likelihood
classifier
algorithm
Classified Data
Unclassified
Built-up
Built-up
0
16
1
2
0
1
20
Gallery
Water
Forest
Cultivated/Expo
Column Total
Figure 2. Integrated land-use assessment of the Nigerian Cocoa Belt.
Table 2. Metadata of images.
Image type Date/scene Product, quality Source
Landsat-TM 17 December,
1986/p190r55_5t861217
WRS-2, GeoCover,
EarthSat Ortho
Global Land Cover
Facility,
University of
Maryland. http://
glcf.umiacs.
umd.edu/
Landsat-ETM+3 January,
2002/p190r055_7t20020103
GeoCover,
EarthSat Ortho
Landsat-ETM+12 January,
2011/L71190055_05520110112
SLC-off, GeoTiff
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
International Journal of Remote Sensing 2865
visual assessment, Roy et al. (2010) who studied MODIS-Landsat data fusion, Potapov,
Turubanova, and Hansen (2011) who studied boreal forest-cover and change mapping, and
Potapov et al. (2012) who studied tropical forest-cover loss.
The collection of auxiliary data and ground-truthing was quite straightforward as the
author hails from the study region and is very familiar with the landscapes. Extensive field
visits were also conducted intermittently between 2002 and 2010 to several towns within
the Cocoa Belt.
3.2.2. Digital image classification
Several image classification methods exist in the literature. Some examples are (1) the
maximum likelihood classifier (MLC) algorithm (Tatsuoka 1971; Richards 1986; Aguiar,
Mascarenhas, and Shimabukuro 2000; Gorte 2002), (2) the decision tree classifier (Goetz
et al. 2003; Pal and Mather 2003; Yang et al. 2003), (3) neural networks (Joshi et al. 2006),
and (4) linear spectral mixture analysis (Smith et al. 1990; Theseira et al. 2003; Lu and
Weng 2009). Although each of these methods has its own advantages and disadvantages,
MLC is considered to give very accurate results in our case. Moreover, the landscape in
question is not as complex as in extensive urban environments, and the training sites were
selected to be representative of the Cocoa Belt. MLC assumes that the distribution of the
cluster of pixels forming each of the class training data are normally distributed. It quan-
titatively evaluates both the variance and covariance of the class spectral response patterns
to which an unknown pixel is classified.
For image classification, false-colour composite (FCC) images were created from
the Landsat-TM and ETM+bands 2, 3, and 4 displaying the blue, green, and red
colours, respectively. In classifying land-use types in the tropics, band 4 (near infrared
0.76–0.90 mm) responds most strongly to green vegetation cover and is best for vegetation-
type mapping and discriminating waterbodies; spectral band 3 (Red 0.63–0.69 mm) is good
for discriminating between plants, burnt area, exposed rocks with dark surfaces, and bare
and dry surfaces, e.g. concrete, red roofs, waterbodies, and soil rich in organic matter; spec-
tral band 2 (Green 0.52–0.60 mm) is good for detecting dirt roads (not tarred), settlement,
rock surfaces, paved surfaces, and bare soil.
For this study, a simple but suitable classification scheme was designed based on land-
use classes that were identified through field visits (see Figure 3). A classification scheme
serves as the key for differentiating LCLU types in an application.
On the basis of the classification scheme, LCLU maps were prepared for different
dates, and changes were detected using the post-classification comparison method. This
method entails comparing land-cover classifications between dates. The use of indepen-
dently produced classifications has the advantage of compensating for varied atmospheric
and phenological conditions between dates, or even the use of different sensors between
dates, because each classification is independently produced and mapped to a common
thematic reference (Loveland et al. 1999).
4. Results
The FCC images were classified into the following LCLU types, namely built-up area,
forest cover, water, cultivated or exposed land, and gallery vegetation. Although it was
not possible to further subdivide these land-use units because of the resolution of the
Landsat-TM and ETM+images used, the classification scheme was maintained during
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
2866 F.O. Akinyemi
Land consists approximately
of 80–100% buildings. These
are constructed with concrete
or mud and small open spaces
(comprising of small trees and
grasses) occur between
buildings
Built-up area
Definition
Level
Code
01
Forest reserve and secondary
(derived) forests such as
fallow regrowths
02
03
04
05 Gallery
vegetation
Gallery vegetation is a very
fragile ecosystem that occurs
on alluvial plains along water
courses (see arrow insert).
Cultivated/
exposed land
Water Open water surfaces with
>95% cover of water, e.g.
dam, river, etc.
Forested cover
Cultivaed land consists of
land used in growing export
and food crops. Exposed land
areas are cleared, with little or
no vegetation cover. Exposed
land and newly cultivated land
are combined because they both
have similar spectral
characteristics.
Water
Figure 3. Land-use classification scheme.
image classification. The land-use map for 1986 provided baseline data against which land-
use changes in the successive images were detected. FCC images and classified maps for
1986, 2002, and 2011 are shown in Figure 4.
The classification accuracies and kappa for the three comparable years (1986, 2002,
and 2011) are 87.33%, 72%, and 91.33% and 0.77%, 0.65%, and 0.84%, respectively (see
Tables 3 and 4).
4.1. Changes in LCLU (1986–2011)
The surface area covered by the LCLU types between 1986, 2002, and 2011 are presented
in Table 5, and the direction of the changes are summarized in Table 6. Table 5 shows
that the amount of surface area covered by each land-use type varied between 2002 and
2011. Until 1986, 63.09% of the total area of study was under forest cover (mature forest),
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
International Journal of Remote Sensing 2867
Ile-Ife Town
OAU Campus
Ile-Ife Town
OAU Campus
Ile-Ife Town
OAU Campu s
01
04
02
03
05
02
02
02
02
02
01
01
01
01
01
04
04
03
05
01 – Built-up (brown) 02 – Forest cover (dark green) 03 – Water (blue)
04 – Cultivated/Exposed land (light green) 05 – Gallery (red)
Ipetu-
modu Ipetu-
modu Ipetu-
modu
Bands 3,2,4 - Landsat TM (1986)
Bands 4,2,3 - Landsat ETM+ (2002)
Bands 3,2,4 - Landsat ETM+ (2011)
Land use (1986)
Land use (2002)
04
04
04
05
02
02
02
Land use classes
01
01
01
03
02
05
01
04
04
02
04
04 02
Land use (2011)
02
km
01326
Figure 4. False-colour composite images and classified land-use types.
whereas 30.96% was used for cultivation of tree and food crops or was an exposed land.
‘Exposed land’, in this study, was used for the classification of land without vegetation
cover. Built-up area, gallery vegetation, and water accounted for 6.37%, 2.54%, and 0.95%
of the study area, respectively.
The LCLU change transition matrix shows that there is considerable change over the
25 year period (28.6% of total area). Built-up area and gallery vegetation increased by
86.32% and 43.12%, respectively. Cultivated or exposed land and forest decreased by
16.91% and 0.96%, respectively. Further analysis indicated that of the 86.32% (96.52 km2)
increase recorded under built-up area, 42.30% was derived from cultivated or exposed land,
5.02% from forest, and 4.83% from gallery vegetation. This indicates the encroachment
of settlement into agricultural land. Tables 5 and 6 show that 1039.60 km2of land was
under forest cover in 1986, whereas it reduced to 1029.62 km2(80.99%) in 2011. Overall
assessment during the same period reveals that 181.95 km2of forest was converted into
agricultural land, whereas land under cultivation diminished only slightly (16.91%) from
543.80 km2to 451.87 km2, with 88.11 km2lost to the built-up land-use class, 180.50 km2
was restored to forest, and 33.66 km2was changed to gallery vegetation.
5. Discussion and conclusion
This article describes the examination of LCLU changes in the Cocoa Belt of southwest
Nigeria using the technologies of satellite remote sensing and a geographical information
system (GIS) from 1986 to 2011. This study further confirms the usefulness of satellite
remote sensing in examining LCLU changes over time in the context of a developing
country where explicit policies for managing such changes are absent. Information about
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
2868 F.O. Akinyemi
Table 3. Accuracy totals.
Class name
Reference
totals
Classified
totals
Number
correct
Number
wrong
Producer’s
accuracy (%)
User’s
accuracy (%)
1986 image
Built-up 9 10 7 3 77.78 70.00
Gallery 6 4 4 0 66.67 100
Water 211050 100
Forest 88 89 82 7 93.18 92.13
Cultivated/exposed
land
45 46 37 9 82.22 80.43
Totals 150 150 131 19
Overall classification accuracy =87.33%, Overall kappa coefficient =0.77
2002 image
Unclassified 1 5 0 5 −−
Built-up 20 20 16 4 80 80
Gallery 12 20 10 10 83.33 50
Water 12 15 12 3 100 80
Forest 24 20 17 2 70.83 85
Cultivated/exposed
land
31 20 17 3 54.84 85
Totals 100 100 72 27
Overall classification accuracy =72.00%, Overall kappa coefficient =0.65
2011 image
Built-up 17 16 16 0 94.12 100
Gallery 5 1 1 0 20 100
Water 0000−−
Forest 91 95 88 7 96.70 92.63
Cultivated/exposed
land
37 38 32 6 86.49 84.21
Totals 150 150 137 13
Note: Overall classification accuracy =91.33%, Overall kappa coefficient =0.84.
the direction, rhythm, and nature of LCLU changes and the major driving factors were
presented. Landsat TM and ETM+data proved particularly useful as they are now freely
available and cover enough large areas, which gave a synoptic view of forest-cover change
at a regional level (this study successfully utilized the ETM+SLC off data). In comparison,
some studies in Nigeria have used aerial photographs, but these are expensive to acquire,
the coverage is less, and the repeatability is infrequent.
Major drivers of land-use change and their impacts on the environment within the con-
text of the government trade liberalization policy were identified. This policy sought to
increase the production of export crops such as cocoa and rubber and boost their export as
a major source of non-oil foreign exchange earnings. The results of this study reveal that the
amount of forest cover gained from the cultivated or exposed land-use class almost equals
that lost to the same class. That is, forest cover is being converted to cultivated or exposed
land use and vice versa. The construction of a major highway to the north of Ile-Ife town
in the year 2002 further opened up prime forest areas, which to some extent contributes to
forest depletion and degradation as forests get fragmented. Forest degradation is charac-
terized by a reduction in forest quality and biomass by an opening up of the canopy. Some
re-forestation efforts have been initiated, such as those on the Obafemi Awolowo University
campus towards the north (refer back to Figure 4), where the authority employs a zoning
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
International Journal of Remote Sensing 2869
Table 4. Error matrix.
Classified data Built-up Gallery Water Forest Cultivated Row total
1986 image
Built-up 7 1 0 0 2 10
Gallery 0 4 0 0 0 4
Water 0 0 1 0 0 1
Forest 0 1 0 82 6 89
Cultivated/exposed
land
2 0 1 6 37 46
Column total 9 6 2 88 45 150
2002 image
Unclassified 0 0 0 1 4 5
Built-up 16 2 0 1 1 20
Gallery 1 10 0 3 6 20
Water 2 0 12 0 1 15
Forest 0 0 0 17 2 19
Cultivated/exposed
land
1 0 0 2 17 20
Column total 20 12 12 24 31 100
2011 image
Built-up 16 0 0 0 0 16
Gallery 0 1 0 0 0 1
Water 0 0 0 0 0 0
Forest 0 2 0 88 5 95
Cultivated/exposed
land
1 2 0 3 32 38
Column total 17 5 0 91 37 150
Table 5. Change in surface area covered by land-use types.
Surface area covered
1986 2002 2011
Land-use types (km2)(%)(km
2)(%)(km
2)(%)
Built-up 111.81 6.37 104.74 5.96 208.32 11.86
Cultivated/exposed
land
543.80 30.96 455.67 25.94 451.87 25.73
Forest 1039.60 59.19 1108.23 63.09 1029.62 58.62
Gallery 44.60 2.54 85.27 4.85 63.83 3.63
Water 16.67 0.95 2.57 0.15 2.84 0.16
Total surface area 1756.48 1756.48 1756.48
Note: 1986 is baseline data.
system on the campus that classifies forests and areas close to waterbodies as unsuitable
for agriculture. These efforts are to be emulated by the local government councils in charge
of developing Ile-Ife town and the surrounding settlements. Re-forestation is a worthwhile
effort because forest regeneration produces positive changes in the natural ecosystem.
The vast conversion of forest to cultivated land suggests that there is an increasing
need for agricultural land, which necessitates farmers to expand agricultural land into
forested areas. This has much to do with the shift in government policy to revive the
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
2870 F.O. Akinyemi
Tab le 6. L and u s e/cover change matrix, 1986–2011 (in km2).
2011 1986
1986 Built-up
Cultivated/exposed
land Gallery Forest Water Total
Built-up 94.10 10.37 5.69 1.32 0.32 111.80
Cultivated/exposed-
land
88.11 240.72 33.66 180.50 0.81 543.80
Gallery 10.07 16.28 10.14 8.08 0.04 44.60
Forest 10.46 181.95 12.70 833.86 0.64 1039.61
Water 5.59 2.56 1.65 5.85 1.03 16.68
2011 total 208.33 451.88 63.84 1029.61 2.84 1756.47
Change (km2) 96.52 −91.93 19.24 −9.98 −13.83
Change (%) 86.32 −16.91 43.12 −0.96 −82.98 28.60
agricultural sector. With the directive of the Nigerian government to increase its non-oil
foreign exchange earnings since the 1990s, the impacts of the rapid expansion of cocoa
production in the 1990s on the environment are noteworthy. The implication is an increase
in demand for land to cultivate crops for export and food. This finding that forested area
decreased tallies with that of Mengistu and Salami (2007) in the same study area. The
main trend in LCLU change is forest conversion into agricultural land and not so much
into built-up areas or settlements. This is in line with other studies that noted that most
of the tropical forests cleared each year are largely due to extensification of agriculture,
involving clearing the land of trees to plant crops (World Bank 1991; Bilsborrow 1994;
Salami, Ekanade, and Oyinloye 1999). Yemefack, Bijker, and De Jong (2006) also reported
the conversion of tropical rain forest to shifting cultivation in southern Cameroon. Gibbs
et al. (2010) found that across the tropics, between the years 1980–2000, more than 55%
of new agricultural land came at the expense of intact forests, and another 28% came from
disturbed forests. Other studies noted that urbanization is a main factor causing forest loss
in some contexts (Mertens and Lambin 1999; DeFries et al. 2010). Joshi et al. (2006) noted
that in Nepal, the relationship between forest land unit change and proximity to built-up
areas is a negative logarithmic function, with the frequency of deforestation reducing very
rapidly as movement is away from built-up areas.
Studies have identified the main implications of the increased production of export
crops as the loss of forest cover with consequences such as forest degradation, loss of the
gallery vegetation, biodiversity loss, and soil loss through expansion of cultivated areas
(USAID 2008; Savilaakso 2009). Echeverria et al. (2006) noted that the conversion of
forested land to other uses because of varied human activities is the greatest threat to
biological diversity as it results in habitat destruction and causes changes in species compo-
sition. For many species, the habitat degradation that accompanies selective, forest resource
exploitation, or which occurs in habitats next to cleared areas, can have serious conse-
quences: Many tropical forest birds, for instance, rely on pristine or near-pristine primary
forest and show low tolerance to selective logging (Birdlife International 2004). The defor-
estation of native vegetation, the subsequent re-forestation with non-native species and
changes in land use, human population pressure, and overharvesting can affect the stability
of local flora and fauna and lead to ecosystem disturbances (Wilson 2006). The depletion
and degradation of tropical forests worldwide due to clearing for agriculture, pastures, tim-
ber products, and infrastructure development is rapid, particularly so in the tropics, where
the local population depends more on exploiting forest resources for sustenance (FAO 2001;
Geist and Lambin 2002; Joshi et al. 2006).
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
International Journal of Remote Sensing 2871
The foregoing implies that despite gains made when some sections of land are converted
back to forest, other forested lands are still being opened up to cater for cultivation and set-
tlement needs. The main reason for built-up encroachment into cultivated or exposed land
is because of proximity of farmlands that occur next to settlements. Moreover, agriculture is
largely a subsistence activity, and farmers do not want to travel far to reach their farmlands,
hence the need to situate agricultural lands close to settlements. Consequently, upon con-
verting cultivated land into built-up areas, farmers are forced to move into nearby forest and
gallery vegetation, which leads to further opening up of forest canopy and depletion. These
are the major drivers of land-use change, which explain why the total surface area cov-
ered by forest and cultivated or exposed land is reduced, whereas built-up areas increased
over the study period. The identification of the driving forces of land-use changes and the
impacts of such changes are of utmost importance since these conditions are generally
replicated elsewhere (see Lambin et al. 2001). However, it must be borne in mind that fac-
tors underlying LCLU change may vary from place to place, implying that the same set of
underlying factors may yield various effects in different regions and at different geograph-
ical scales, leading to different patterns and processes of change (McDonald and Urban
2006; Weng, Rajasekar, and Hu 2011).
How then are the ensuing conflicts between land uses to be resolved as pressure on land
intensifies in the Cocoa Belt? How do we reconcile increased export crops and food produc-
tion and stemming the loss of forest cover? There is the need for effective land management
and monitoring of the belt. Some developing countries have managed a land-use transition
over recent decades that simultaneously increased their forest cover and agricultural pro-
duction. These countries have relied on various mixes of strategies such as agricultural
intensification, land-use zoning, forest protection, increased reliance on imported food and
wood products, creation of off-farm jobs, and foreign capital investments and remittances.
Sound policies and innovations can therefore reconcile forest preservation with food pro-
duction. Land uses that enhance food production while preserving ecosystems, especially
tropical forests, must be promoted (Lambin and Meyfroidt 2011). The impacts of land-
use changes such as increasing agricultural land on forest (ecosystem services) as well
as reduced deforestation on food production must be examined and findings should be
reflected in the formulation of appropriate, evidence-based policies (Angelsen 2010). Areas
with important functions for groundwater recharge, surface water runoff, and biotopes, for
instance, have to be identified and protected from impacts of land-use change (Fritsch and
Hild 2002). For environmental sustainability, environmental protection should be integrated
into socioeconomic, political, and health development initiatives. From the ecological point
of view, proper forest management as well as biodiversity and soil conservation are of great
importance in the Cocoa Belt.
Seeing that forest-cover loss in the study area is the main consequence of export crop
expansion and settlement growth, and as more land is being demanded, there is the need
for measures to curb the further loss of forests. Environmentally degrading practices such
as the practices of bush burning (a method commonly known as Slash and Burn employed
when cultivating or hunting) and cultivation on hill slopes are devastating, making slopes
susceptible to excessive soil erosion and sometimes, mud or landslide. These are at vari-
ance with sound natural resource management and the goals of sustainable development.
USAID (2002) noted that the ability of local communities to undertake land-use planning
and enforce zoning decisions is key to sustainable management. This is because not all uses
are compatible and not all users are responsible. Identifying and enforcing rules about use
is critical.
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
2872 F.O. Akinyemi
Acknowledgements
The comments from two anonymous reviewers are highly appreciated. The images used are from the
Global Land Cover Facility, University of Maryland.
References
Adejuwon, J. O. 1971. “Agricultural Colonization in Twentieth Century Western Nigeria.” Journal of
Tropical Geography 33: 1–8.
Adejuwon, J. O., and L. K. Jeje. 1973. Landuse Map of Ife Area. Lagos: Federal Surveys Department.
Adejuwon, J. O., T. O. Odekunle, and M. O. Omotayo. 2007. “Extended-range Weather
Forecasting in Sub-Saharan West Africa: Assessing a Potential Tool for Adapting Food
Production to Climate Variability and Climate Change.” Assessment of Impacts and Adaptation
to Climate Change Working Paper 46. Accessed January 31, 2012. www.aiaccproject.
org/working_papers/Working%20Papers/AIACC_WP46_Adejuwon.pdf.
Aguiar, A. P. D., N. D. A. Mascarenhas, and Y. E. Shimabukuro. 2000. “Use of Features Derived from
Proportions of Classes in a Pixel for the Multispectral Classification of Remote Sensing Images.”
Accessed July 20, 2012. http://www.isprs.org/proceedings/XXIX/congress/part3/259_XXIX-
part3.pdf.
Ajayi, S. A., W. T. Odofin, L. Denton, B. O. Ogunbodede, S. G. Ado, J. A. Morakinyo, D. O. Ladipo,
D. Dumet, Y. N. Ndirpaya, O. Adetula, S. E. Aladele, M. L. Umar, and A. A. Zaria. 2008. State
of Plant Genetic Resources for Food and Agriculture in Nigeria (1996–2008). Nigeria: Federal
Department of Agriculture/National Centre for Genetic Resources and Biotechnology.
Akinyemi, F. O. 2005. “Mapping Land Use Dynamics at a Regional Scale in SW Nigeria.” High-
Resolution Earth Imaging for Geospatial Information, Hannover, Germany, May 17–20, 1–7.
Amamoo-Otchere, E., R. O. Oyinloye, A. P. Ozah, and F. B. Agbo. 1998. “Automated Discrimination
between Built-up and Nonbuilt-up as a Settlement Detection Technique for Population Census
Base Map Revision.” International Archives of Photogrammetry and Remote Sensing (IAPRS)
32: 700–1.
Anderson, J. R., E. E. Hardy, J. T. Roach, and R. E. Witmer. 1976. A Land Use and Land Cover
Classification System for Use with Remote Sensor Data. Geological Survey Professional Paper
964, 1–41. Washington, DC: US Government Printing Office.
Angelsen, A. 2010. “Policies for Reduced Deforestation and Their Impact on Agricultural
Production.” Proceedings of the National Academy of Sciences (PNAS) 107: 19639–44.
Angelsen, A. and D. Kaimowitz, eds. 2001. Agricultural Technologies and Tropical Deforestation,
384 pp. Wallingford: CAB Intl.
Bedard, F., G. Reichert, R. Dobbins, and I. Trepanier. 2008. “Evaluation of Segment-Based Gap-
Filled Landsat ETM+SLC-Off Satellite Data for Land Cover Classification in Southern
Saskatchewan, Canada.” International Journal of Remote Sensing 29: 2041–54.
Bilsborrow, R. E.. 1994. “Population, Development and Deforestation: Some Recent Evidence.”
Proceedings of the United Nations Expert Group Meeting on Population, Environment and
Development, 117–34. New York: United Nations.
Birdlife International. 2004. “State of the World’s Bird: What Birds Tell Us about Problems.”
Pressure, 1–76. Accessed January 25, 2012. http://www.birdlife.org/datazone/userfiles/docs/
SOWB2004_en.pdf.
Chen, J., X. Zhu, J. E. Vogelmann, F. Gao, and J. Suming. 2011. “A Simple and Effective Method
for Filling Gaps in Landsat ETM+SLC-Off Images.” Remote Sensing of the Environment 115:
1053–64.
Defries, R. S., T. K. Rudel, M. Uriarte, and M. Hansen. 2010. “Deforestation Driven by Urban
Population Growth and Agricultural Trade in the Twenty-First Century.” Nature Geosciences
3: 178–81.
Echeverria, C., D. A. Coomes, J. Salas, J. M. Rey Benayas, A. Lara, and A. Newton. 2006. “Rapid
Deforestation and Fragmentation of Chilean Temperate Forests.” Biological Conservation 130:
481–94.
European Commission. 1993. CORINE Land Cover Technical Guide, Report EUR 12585EN, 144 pp.
Luxembourg: Office for Publications of the European Communities.
FAO (Food and Agriculture Organization). 2001. Global Forest Resources Assessment 2000. Main
Report, FAO Forestry Paper 140. Rome: FAO.
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
International Journal of Remote Sensing 2873
FAO (Food and Agriculture Organization). 2000. Land Cover Classification System: Classification
Concepts and User Manual, 208 pp. Rome: FAO.
FME (Federal Ministry of Environment). 2006. National Forest Policy.Abuja:FME.
FORMECU. 1985. Vegetation and Land-Use Maps of Nigeria. Abuja: Federal Department of
Forestry.
FOS (Federal Office of Statistics). 1988. Annual Abstract of Statistics. Lagos: FOS.
Fritsch, D., and H. Hild. 2002. “Generation of a Consistent Spatial Database for a Land Use Planning
Project in South Eastern China.” In Development and Technology Transfer in Geomatics for
Environmental and Resource Management, edited by E. P. Baltsavias, T. Boulocous, G. E. Mtalo,
J. Niedderost, and U. Rivett, IAPRS, vol. 34/6, 147–54.
Geist, H. J., and E. F. Lambin. 2002. “Proximate Causes and Underlying Driving Forces of Tropical
Deforestation.” Bioscience 52: 143–50.
Gibbs, H. K., A. S. Ruesch, F. Achard, M. K. Clayton, P. Holmgren, N. Ramankutty, and J. A. Foley.
2010. “Tropical Forests were the Primary Sources of New Agricultural Land in the 1980s and
1990s.” PNAS 107: 16732–7.
Gillieson, D., T. Lawson, and L. Searle. 2006. Applications of High Resolution Remote Sensing in
Rainforest Ecology and Management, 55 pp. Cairns: Cooperative Research Centre for Tropical
Rainforest Ecology and Management.
Goetz, S. J., R. K. Wright, A. J. Smith, E. Zinecker, and E. Schaub. 2003. “IKONOS Imagery for
Resource Management: Tree Cover, Impervious Surfaces and Riparian Buffer Analyses in the
Mid-Atlantic Region.” Remote Sensing of Environment 88: 195–208.
Gorte, B. 2002. “Supervised Image Classification.” In Spatial Statistics for Remote Sensing, edited
by A. Stein, F. Van der Meer, and B. Gorte., vol. 1, 153–63. The Netherlands: Kluwer Academic.
Griffiths, G. H., and M. G. Wooding. 1989. “Use of Satellite Data for the Preparation of Land Cover
Maps and Statistics.” SPI Division Working Paper 90, WP33, National Remote Sensing Centre,
Farnborough.
Huete, A. R., and S. R. Saleska. 2010. “Remote Sensing of Tropical Forest Phenology: Issues and
Controversies.” IAPRS 38: 539–41.
Imbernon, J., and A. Branthomme. 2001. “Characterization of Landscape Patterns of Deforestation
in Tropical Rain Forests.” International Journal of Remote Sensing 22: 1753–65.
Joshi, C., J. de Leeuw, A. K. Skidmore, I. C. van Duren, and H. Van Oosten. 2006. “Remotely Sensed
Estimation of Forest Canopy Density: A Comparison of the Performance of Four Methods.”
International Journal of Applied Earth Observation and Geoinformation (IJAG) 8: 84–95.
Ju, J. C., and D. P. Roy. 2008. “The Availability of Cloud-Free Landsat ETM Plus Data over the
Conterminous United States and Globally.” Remote Sensing of Environment 112: 1196–211.
Junge, B., T. Alabi, K. Sonder, S. Marcus, R. Abaidoo, D. Chikoye, and K. Stahr. 2010. “Use
of Remote Sensing and GIS for Improved Natural Resources Management: Case Study from
Different Agroecological Zones of West Africa.” International Journal of Remote Sensing 31:
6115–41.
Lambin, E. F., and P. Meyfroidt. 2011. “Global Land Use Change, Economic Globalization,
and the Looming Land Scarcity.” Proceedings of the National Academy of Sciences 108:
3465–72.
Lambin, E. F., B. L. Turner, II, H. Geist, S. Agbola, A. Angelsen, J. W. Bruce, O. Coomes, R. Dirzo,
G. Fischer, C. Folke, P. S. George, K. Homewood, J. Imbernon, R. Leemans, X. Li, E. F. Moran,
M. Mortimore, P. S. Ramakrishnan, J. F. Richards, H. Skanes, W. Steffen, G. D. Stone, U. Svedin,
T. Veldkamp, C. Vogel, and J. Xu. 2001. “Causes of Land-Use and Land-Cover Change: Moving
Beyond the Myths.” Global Environmental Change: Human and Policy Dimensions 11: 5–13.
Lo, C. P. 2000. “Mapping Atlanta’s Land Use/Cover Change and Its Impact on the Environment
Using Time Sequential Landsat Images.” IAPRS 33: 790–6.
Loveland, T. R., T. Sohl, K. Sayler, A. Gallant, J. Dwyer, J. Vogelmann, G. Zylstra, T. Wade,
C. Edmonds, D. Chaloud, and B. Jones. 1999. “Land Cover Trends: Rates, Causes, and
Consequences of Late-Twentieth Century U.S. Land Cover Change.” Research PlanEPA-
IAG Project No. DW14938108-01-0. Accessed June 11, 2012. http://www.epa.gov/esd/land-
sci/trends/pdf/trends_research_plan.pdf
Lu, D., and Q. Weng. 2009. “Extraction of Urban Impervious Surfaces from an IKONOS Image.”
International Journal of Remote Sensing 30: 1297–311.
Maxwell, S. K., G. L. Schmidt, and J. C. Storey. 2007. “A Multi-Scale Segmentation Approach to
Filling Gaps in Landsat ETM+SLC-Off Images.” International Journal of Remote Sensing 28:
5339–56.
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
2874 F.O. Akinyemi
McDonald, R. I., and D. L. Urban. 2006. “Spatially Varying Rules of Landscape Change: Lessons
from a Case Study.” Landscape & Urban Planning 74: 7–20.
Mengistu, D. A., and A. T. Salami. 2007. “Application of Remote Sensing and GIS in Land Use/Land
Cover Mapping and Change Detection in SW Nigeria.” African Journal of Environmental Science
and Technology 1: 99–109.
Mertens, B., and E. Lambin. 1999. “Modelling Land Cover Dynamics: Integration of Fine-Scale
Land Cover Data with Landscape Attributes.” International Journal of Applied Geoinformation
1: 48–51.
Mulders, M. A. 2001. “Advances in the Application of Remote Sensing and GIS for Surveying
Mountainous Land.” International Journal of Applied Geoinformation 3: 167–72.
Nwadialor, I. J. 2001. “An Assessment of Spatio-temporal Variabilities of Deforestation for
Sustainable Forestry Development: A Case Study of Afaka Forest Reserve.” International
Conference on Spatial Information for Sustainable Development Nairobi, Kenya, October 2–5.
Oates, J., R. Ikemeh, R. Bergl, and D. Ogunsesan. 2008. A Survey of Rain Forests in Ogun, Ondo
and Osun States in Southwestern Nigeria to Assess Options for Their Sustainable Conservation,
42 pp. Lagos: Nigerian Conservation Foundation.
Oetter, D. R., W. B. Cohen, M. Berterretche, T. K. Maiersperger, and R. E. Kennedy. 2000. “Land
Cover Mapping in an Agricultural Setting Using Multiseasonal Thematic Mapper Data.” Remote
Sensing of Environment 76: 39–155.
Ogundele, F. O., A. Omotayo, and S. G. Odewunmi. 2011. “Pattern and Problems of Deforestation in
South Western Nigeria.” International Journal of Academic Research 3: 641–55.
Olayiwola, A. M., A. O. Eludoyin, and A. E. Ekecha. 2011. “Land Use Change in the Mezam Division
of the North-West Province, Cameroon from Landsat Imageries (1988 and 2001).” Ethiopian
Journal of Environmental Studies and Management 4: 49–55.
Orimogunje, O. O. I. 2005. “The Impact of Land-use Dynamics on Oluwa Forest Reserve in
Southwestern Nigeria.” Unpublished PhD. thesis, Obafemi Awolowo University (OAU), Ile-Ife,
Nigeria, 175 pp.
Pal, M., and P. M. Mather. 2003. “An Assessment of the Effectiveness of Decision Tree Methods for
Land Cover Classification.” Remote Sensing of Environment 86: 554–65.
Pant, D. N., S. M. E. Groten, and P. S. Roy. 2000. “Forest Vegetation/Landuse Change Detection and
Impact Assessment in Part of Western Himalaya.” International Archives of Photogrammetry and
Remote Sensing 33: 1111–8.
Potapov, P. V., S. A. Turubanova, and M. C. Hansen. 2011. “Regional-Scale Boreal Forest Cover
and Change Mapping Using Landsat Data Composites for European Russia.” Remote Sensing of
Environment 115: 548–61.
Potapov, P. V., S. A. Turubanova, M. C. Hansen, B. Adusei, M. Broich, A. Altstatt, L. Mane, and C. O.
Justice. 2012. “Quantifying Forest Cover Loss in Democratic Republic of the Congo, 2000–2010,
with Landsat ETM+Data.” Remote Sensing of Environment. 122: 106–16.
Pringle, M. J., M. Schmidt, and J. S. Muir. 2009. “Geostatistical Interpolation of SLC-Off Landsat
ETM Plus Images.” ISPRS Journal of Photogrammetry and Remote Sensing 64: 654–64.
Richards, J. A. 1986. Remote Sensing Digital Image Analysis. Berlin: Springer.
Rikimaru, A., P. S. Roy, and S. Miyatake. 2002. “Tropical Forest Cover Density Mapping.” Tropical
Ecology 43: 39–47.
Roy, D. P., J. Ju, K. Kline, P. L. Scaramuzza, V. Kovalskyy, M. Hansen, T. R. Loveland, E. Vermote,
and C. Zhang. 2010. “Web-Enabled Landsat Data: Landsat ETM+Composited Mosaics of the
Conterminous United States.” Remote Sensing of Environment 114: 35–49.
Roy, D. P., J. Ju, P. Lewis, C. Schaaf, F. Gao, M. Hansen, and E. Lindquist. 2008. “Multi-
Temporal MODIS-Landsat Data Fusion for Relative Radiometric Normalization, Gap Filling,
and Prediction of Landsat Data.” Remote Sensing of Environment 112: 3112–30.
Sader, S. A., A. T. Joyce, R. B. Waide, and W. T. Lawrence. 1985. “Monitoring Tropical Forests
from Satellite and Aircraft Platforms: Some Limitations and New Approaches.” Paper presented
at Proceedings of Remote Sensing in Forest and Range Resource Management, Fort Collins, CO,
August 20–22, 473–82.
Salami, A. T., and J. Akinyede. 2006. “Space Technology for Monitoring and Managing Forest in
Nigeria.” International Symposium on Space and Forests, Vienna, Austria, June 7–16, 1–47.
Salami, A. T., O. Ekanade, and R. O. Oyinloye. 1999. “Detection of Forest Reserve Incursion
in South-Western Nigeria from a Combination of Multi-Date Aerial Photographs and High
Resolution Satellite Imagery.” International Journal of Remote Sensing 20: 1487–97.
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
International Journal of Remote Sensing 2875
Savilaakso, S. 2009. “The Influence of Disturbance on Lepidopteran Larvae and their Seasonality
on a Constant Resource in Kibale National Park, Uganda – Implications for Conservation.” PhD
diss., University of Joensuu, Finland.
Scaramuzza, P., E. Micijevic, and G. Chander. 2004. SLC Gap-Filled Products Phase
One Methodology, 1–5. Accessed June 18, 2012. http://landsat.usgs.gov/documents/
SLC_Gap_Fill_Methodology.pdf
Smith, M. O., S. L. Ustin, J. B. Adams, and A. R. Gillespie. 1990. “Vegetation in Deserts: A Regional
Measure of Abundance from Multispectral Images.” Remote Sensing of Environment 31: 1–26.
Tanaka, S., and T. Sugimura. 2001. “A New Frontier of Remote Sensing from IKONOS Images.”
International Journal of Remote Sensing 22: 1–5.
Tanner, E. V. J., V. Kapos, J. Adams, D. M. Newbery, H. H. T. Prins, and N. D. Brown. 1996. “Tropical
Forests: Spatial Pattern and Change with Time, as Assessed by Remote Sensing.” Proceedings
of the 37th Symposium of the British Ecological Society: Dynamics of Tropical Communities,
Cambridge University, 599–615.
Tappan, G., and M. Cushing. 2004. Use of SLC-Off Landsat Image Data for Monitoring Land
Use/Land Cover Trends in West Africa. Sioux Falls, SD: USGS EROS Data Center.
Tatsuoka, M. M. 1971. Multivariate Analysis: Techniques for Educational and Psychological
Research. New York: Wiley.
Theseira, M. A., G. Thomas, J. C. Taylor, F. Gemmell, and J. Varjo. 2003. “Sensitivity of Mixture
Modeling to Endmember Selection.” International Journal of Remote Sensing 24: 1559–75.
UN. 2010. The Millennium Development Goals Report. New York: UN, 80 p. Accessed
January 24, 2012. http://www.un.org/millenniumgoals/pdf/MDG%20Report%202010%20En%
20r15%20-low%20res%2020100615%20-.pdf
UNECA. 2007. Africa Review Report on Drought and Desertification. Accessed January 24, 2012.
http://www.un.org/esa/sustdev/csd/csd16/rim/eca_bg3.pdf .
UNEP. 2002. “The Export Crop Sector in Nigeria.” In Integrated Assessment of Trade Liberalization
and Trade-Related Policies: Synthesis Report of UNEP Country Projects-Round II.New
York/Geneva: UN. Accessed January 24, 2012. www.unep.ch/etu/publications/Synth_Nigeria.
PDF.
USAID. 2002. Nature, Wealth, and Power: Emerging Best Practice for Revitalizing Rural Africa.
Accessed July 17, 2012. http://pdf.usaid.gov/pdf_docs/PNACR288.pdf
USAID. 2008. Nigeria Biodiversity and Tropical Forestry Assessment: Maximizing Agricultural
Revenue in Key Enterprises for Targeted Sites (Markets), 98 pp. Washington, DC: Chemonics
International.
USGS. 2004. Phase 2 Gap-Fill Algorithm: SLC-off Gap-Filled Products Gap-Fill
Algorithm Methodology. Accessed June 18, 2012. http://landsat.usgs.gov/documents/
L7SLCGapFilledMethod.pdf
Weng, Q., U. Rajasekar, and X. Hu. 2011. “Modeling Urban Heat Islands and Their Relationship with
Impervious Surface and Vegetation Abundance by Using ASTER Images.” IEEE Transactions
on Geoscience and Remote Sensing 49: 4080–9.
Wilson, E. O. 2006. The Creation: An Appeal to Save Life on Earth, 175 pp. New York: Norton and
Company.
World Bank. 1991. Forest Sector Policy Paper. Washington, DC: World Bank.
Yaakup, A., S. Sulaiman, N. Musa, M. J. Ahmad, and M. Z. Ibrahim. 2002. “GIS as Tools for
Monitoring the Urban Development in Metropolitan Region: A Case of Klang Valley Region,
Peninsular Malaysia.” Paper presented at Proceedings of Governance and the Use of GIS in
Developing Countries, Enschede, The Netherlands, May 15–18, 1–13.
Yang, L., G. Xian, J. M. Klaver, and B. Deal. 2003. “Urban Land Cover Change Detection
through Sub-Pixel Imperviousness Mapping Using Remotely Sensed Data.” Photogrammetric
Engineering and Remote Sensing 69: 1003–10.
Yemefack, M., W. Bijker, and S. M. De Jong. 2006. “Investigating Relationships Between Landsat-
7 ETM+Data and Spatial Segregation of LULC Types under Shifting Agriculture in Southern
Cameroon.” International Journal of Applied Earth Observation and Geoinformation 8: 96–112.
International Journal of Remote Sensing 2013.34:2858-2875. downloaded from www.tandfonline.com
