Journal of Agricultural, Food and Environmental Sciences Vol 75 No 1 (2021) 32-47

Original scientific paper

LAND SUITABILITY FOR RICE CROP FARMING IN KWARA STATE

USING GIS-BASED MULTI-CRITERIA DECISION ANALYSIS
Ayo Babalola, Mohammed O. Idrees *, Ruth K. Aniyikaye, Hossein A. Ahmadu,
Oyedapo A. Ipadeola

Department of Surveying and Geoinformatics, Faculty of Environmental Sciences, University

of Ilorin, P.M.B. 1515, Ilorin, Nigeria
∗corresponding author: mohammed.oi@unilorin.edu.ng dare.idrees@gmail.com

ABSTRACT
This study employs GIS-based multi-criteria decision approach to identify suitable areas for
cultivating rice crop in Kwara State, Nigeria, using essential climatic, soil, terrain and
environmental variables selected based on FAO framework for land evaluation. Weights
indicating the relative importance of each variable was determined using Analytical
Hierarchical Process (AHP). The criteria, their weights and constraints were integrated in GIS
environment to produce suitability map, classified into five levels of suitability (Very highly
suitable, highly suitable, moderately suitable, low suitable and not suitable) using weighted
overlay operation. The result indicates that 9.7% (343803.75 ha) of the total land area is
unsuitable for cultivating rice while 14.6% (516169.46 ha) is classified as low suitable area.
The moderately suitable, highly suitable and very highly suitable classes occupy 30.8%
(1091145.20 ha), 40.56% (1436504.55 ha) and 4.4% (154408.94 ha), respectively. Quantitative
assessment of the work yields overall accuracy (area under the ROC curve) of 0.97 (97%).
Based on the findings of this study, we recommend that the state land use planning agency
review zoning mechanism, incorporates grassroots participatory land use planning policy and
evaluate suitable land for other essential crops by incorporating GIS in order to sufficiently
allocate lands for optimal utilization.

Key words: Remote sensing, food security, agriculture, AHP, Nigeria, land use planning.

INTRODUCTION
All crops require water, suitable topography and geomorphology, fertile soil, and favorable
climate. As many of the underlying processes influence the suitability of crop cultivation,
spatially heterogeneous soil chemistry and land features can be easily identified using remote
sensing and GIS technologies (Hao et al., 2019). The problem of selecting appropriate land for
cultivation of a particular crop has long been an empirical issue with many researchers and
institutions developing framework for optimal agriculture land use. This notwithstanding,
many agricultural land uses are currently below optimal capacity in most part of the world. The
classification of land into different suitability classes is dependent on particular set of soil
characteristics, climatic condition, topographic and other attributes of land specific to a crop
type or a family of crops (Kihoro et al., 2013; Sinha et al., 2014). This has made precision in
land utilization planning a necessity.
As the world population continue to grow geometrically, with many living in urban areas,
agricultural productivity and food security are issues of discuss across the globe. The impact
of unsustainable use of land resources and absence of utilization of land according to potential
suitability portends serious danger to food production and supply in developing countries,
particularly rice crop. Rice (Oryza sativa) is a staple food for nearly half of the world’s
population, and also a key source of employment and income for rural populace (Khattak &

32
 Ayo Babalola et al.

Shabbir, 2012; Oriola & Olabode, 2014). It is rapidly becoming a major food crop in most of
sub-Saharan Africa. Rice can be grown as a dry land crop, but it is by origin and by preference
of most farmers a wetland crop. Demand for rice is driven by population growth; however, it
is produced in much smaller quantities far below local demand.
Nigeria is the most populous black nation in the world whose citizens largely depend on
agriculture at the household level (Merem et al., 2017). Nonetheless, more than 90% of the rice
consumed is imported. The demand for rice in recent time has motivated massive investment
in its cultivation to keep pace with the rising local and international market needs. However,
potentially rich (sub humid and sub arid) rainfed farm land is experiencing high rate of nutrient
depletion caused by climate change (Kumar & Patel, 2020; Oriola & Olabode, 2014). Other
factors limiting rice production and yield is lack of sufficient information on soil characteristics
as well as poor agricultural practices (Ceballos-Silva & López-Blanco, 2003).
Kwara State, located in the Nigeria North-Central zone, is one of the hubs of rice production
in the country. In recent years, the state has consistently record low rice production, resulting
to increase in rice importations to meet local consumption (Babatunde et al., 2019; Falola et
al., 2014). The State has abundant unexploited land resources that can be utilized to increase
rice production. However, no land evaluation studies have been done to determine how well
the qualities of land units in the region match rice crop growing and yield requirements. One
of the efforts of the state to reposition agriculture is the development of Agricultural Master
Plan for efficient management and optimum utilization of land resources
(http://www.kwarappp.gov.ng/kamp.html). Utilization of agriculture land in past decades
regardless of land suitability has impacted negatively on food production capacity (Abah &
Petja, 2016; Merem et al., 2017). This study utilizes multi-criteria decision evaluation (MDE)
technique to map suitable land for rice cultivation in Kwara State using topographical, soil and
climatic factors.

MATERIALS AND METHODS

Study Area and Dataset

Kwara, located in the North-central zone, is one of the States in Nigeria. Geographically,
the State lies between longitude 08o 30’N to 10o 08’N and latitude of 05o 00’E to 06o 12’E
covering approximate area of 36,835 Km2 and altitude that range between 13 m and 657 m
above msl. Kwara state is bordered in the North by Niger state, in the East by Kogi and Ekiti
States, Osun and Oyo states in the South while in the West by Republic of Benin (Figure 1).
The population of Kwara state is reported to be about 2,365,353, based on the 2006 census
(NPC, 2006). The state is situated in the transitional zone within the forest and the guinea
savannah regions of Nigeria. The vegetation is typically of tropical savannah species with
riparian forest along the river bank. In terms of climate, the state is classified as tropical
savannah under the influence of the two trade winds prevailing over the country. The state
experiences two seasons, rainy season and dry season. The rainy season occurs between March
and November with annual rainfall that varies between 800 mm to 1200 mm, with the peak
between August and September while dry season is experienced from November and March.
Also, the mean monthly temperature is generally high throughout the year with daily average
temperatures vary from 22.5 ºC and 27.5 ºC (Babatunde et al., 2019; Falola et al., 2014; Sadiq
et al., 2017).
In this study, the process of assessing site suitability for rice cultivation in Kwara state
involves a number of datasets categorized under four subgroups: topographical factors, soil
characteristics, climate and environmental factors. The topographical factors which include
altitude, slope and slope aspect were derived from 30 m resolution SRTM digital elevation
model (DEM) downloaded from the United States Geological Survey (USGS) data archive

33
 Journal of Agricultural, Food and Environmental Sciences Vol 75 No 1 (2021) 32-47

(http://srtm.csi.cgiar.org/). Soil data was also utilized in this study. Soil characteristics, texture,
soil drainage, depth, nitrogen, potassium, and phosphorus were obtained from the Harmonized
World Soil Database (http://www.isricl.com) while 250 m grid carbon and soil PH data were
downloaded from SOILGRID data depository (https://soilgrids.org). Also, the climatic data,
temperature and precipitation, were gotten from the World Climate data depository
(http://www.worldclim.org) whereas the environmental factors, administrative boundary, road
network and river, and land use and land cover were acquired from DIVA-GIS
(http://www.diva-gis.org/) and Landsat-8 imagery downloaded from the USGS data depository
(http://earthexplorer.usgs.gov).

Figure 1. Location of the study area

Preparation of Data Layers

From the four data subgroups, 16 conditioning factors that influence rice cultivation were
generated. The conditioning factors are chosen based on thorough examination of the variables
reported in literatures (e.g. Abah & Petja, 2016; Ceballos-Silva & López-Blanco, 2003; Hao et
al., 2019; Kihoro et al., 2013, 2013; Victor & Samson, 2019). Most of the data layers come in
geographic coordinate system and in different spatial resolution. Therefore, the various dataset
collected were preprocessed prior to generating input map layers. Upon clipping to the study
area, all the data acquired were reprojected to the Universal Transverse Mercator (UTM),
WGS84 coordinate system using nearest neighbor resampling method. From the 1 arc second
(30 m resolution) SRTM global DEM, three topographical factors, altitude, slope and aspect,
were derived. The resolution of the soil data is coarse (250 m). All the eight soil characteristics
(texture, drainage, soil depth, nitrogen, potassium, phosphorus, carbon and soil PH) were also
resampled to 30 m resolution raster map layers. Similarly, 1 km resolution climatic factors,
temperature and precipitation were reprojected and resampled to match the other dataset. The
land use/cover map was generated from the Landsat-8 satellite imagery using the supervised
classification method (Ceballos-Silva & López-Blanco, 2003; Jeevalakshmi et al., 2016) while

34
 Ayo Babalola et al.

the distance road and distance from river map layers were produced using the Euclidean
distance tool in ArcGIS 10.4.

35
Journal of Agricultural, Food and Environmental Sciences Vol 75 No 1 (2021) 32-47

36
 Ayo Babalola et al.

Figure 2. Topographical, soil, climatic and environmental variables influencing rice growth;
(a) elevation, (b) slope, (c) slope aspect, (d) soil texture, (e) soil drainage, (f) soil depth, (g)
soil nitrogen, (h) potassium (i) phosphorus, (j) soil carbon (k) soil PH, (l) temperature, (m)
precipitation, (n) LULC, (o) distance from road, and (p) distance from river.

Topographical Factors. Elevation, Slope and Slope aspects (Figure 2a, 2b and 2c) are
topographical factors that determine the position of points relative to reference datum, the rate
of change of the vertical distance between two points and their distance (gradient) and direction
the slope faces. Rice crop is generally said to thrive well in low altitude, slope less than 15o
and flat north facing slope (Kihoro et al., 2013). Flatter terrain corresponds with lower slope
value, while steeper terrain corresponds with higher slope value. Flat fields having smooth
surface are reported to be better for rice cultivation because it facilitates proportionate
distribution of water (Kaaya et al., 2019; Victor & Samson, 2019). In addition, high slope and
elevation increase the cost of operation and maintenance of machinery where mechanized
farming is practiced.

Soil Factors. Soil physical (texture, drainage and depth) and Chemical (nitrogen, potassium,
phosphorus, carbon and Ph) properties vary from place to place. Thus, the knowledge of soil
chemical and physical characteristics is vital to crop growth and yield. Soil texture (Figure 2d)
is described by the relative amount of sand, silt, and clay contents, which affect the workability,
aeration, drainage, root penetrability and water-releasing capacity of soil (Ahmed et al., 2017).
The United State Department of Agriculture (USDA) categorizes soil texture into twelve
classes; namely, sand, loamy sand, sandy loam, silt loam, loam, sandy clay loam, silty clay
loam, clay loam, clay, sandy clay, silty clay, and silt clay loam. Of these classes, clay and clay
loam are the most suitable for rice growth (Ahmed et al., 2017). Soil drainage (Figure 2e) is
also a characteristic of texture.

37
 Journal of Agricultural, Food and Environmental Sciences Vol 75 No 1 (2021) 32-47

According to the United State Department of Agriculture (USDA), soil drainage could be
well drained, moderately well drained, imperfect, poor, and very poor. For optimum growth
and productivity, the very poor and poor drainage soil is better for rice farming to guarantee
steady availability of water in the soil. Soil depth (Figure 2f) describes the thickness of topsoil
above the subsoil such as consolidated rock or cemented materials like gravel restricting root
soil penetration (Ahmed et al., 2017). Most annual crops have a rooting depth of about 50cm,
unlike tree crops that can reach beyond 150cm. Studies have shown that effective soil depth of
about 100cm is considered best for most annual crops because they produce good yields at this
soil depth (Ahmed et al., 2017).
Rice absorbs large quantities of nitrogen, phosphorus and potassium (N.P.K) to enhance
growth, yield and grain quality (see Figures 2g, 2h and 2i). Nitrogen is present in the soil, but
often in insufficient quantities. Rice needs Nitrogen almost throughout the vegetative cycle,
particularly at tillering and panicle initiation stages (Samanta et al., 2011). The higher the
percentages of Nitrogen available in soil, the higher the yield of the rice. Deficiency in nitrogen
content in soil manifest in the rice leave as uniform yellowish appearance. Phosphorus is an
important soil chemical element that controls the physiological development of rice plant. It is
vital to healthy root development, tillering, early flowering, pollination and ripening.
Phosphorus in sufficient quantity accelerates rice development, reduces the period of maturity
and accelerates plant recovery after stress resulting from rodent attack, cold, water shortage,
etc. Potassium is a coordinating element regulates the biochemical activity of nitrogen and
phosphorus in plant. It regulates photosynthetic activities and its transformation into
carbohydrates (Daniel, 2011; Khattak & Shabbir, 2012).
Other parameters of importance to rice growth and yield are soil PH and organic carbon
matter (Figure 2j and 2k), both of which affect soil fertility. Soil pH expresses the concentration
of hydrogen ions (H+) measured in terms of the soil solution’s (soil water and its dissolved
substances) acidity and alkalinity usually determined on a scale from 0 to 14. Acidic solutions
have a pH less than 7, while basic or alkaline solutions have a pH greater than 7 (McCauley et
al., 2009). Rice grows optimally in slightly acid soils of Ph value between 6 to 7 (Samanta et
al., 2011). Soil organic matter is the combination of plant and animal residues at various stages
of decomposition and cells and tissues of soil organisms. Soil organic matter controls soil
nutrient and organic carbon management. Soil carbon determines nutrient availability and the
capacity of the soil to improve plant development and also to deliver ecosystem services such
as reduction in greenhouse gas emissions and pollutant (Daniel, 2011; Kaaya et al., 2019;
McCauley et al., 2009). Specifically, soil carbon improves rice yield and crop adaptation to
environmental changes.
Climatic Factor. Precipitation and temperature (Figure 2l and 2m) are the two most important
climatic elements considered in this study. Rice is a tropical and sub-tropical crop normally
grown at a fairly high temperature and high rainfall regime between 20oC to 40°C and 1250
mm to 2000 mm of annual rainfall. Both climatic factors are favorable for rice cultivation in
most parts of the study area (Khattak & Shabbir, 2012; Kihoro et al., 2013).
Environmental factors. Human settlement and activities are largely controlled by natural
phenomena such as the physical landscape, water availability, vegetation/forest distribution,
and fertile soil for crop production (Ceballos-Silva & López-Blanco, 2003; Maddahi et al.,
2017). Here, the land use map of the study area was classified into five classes: water body,
forest, grassland, bare earth and built-up area (Figure 2n). The land use map allows assessment
of the distribution of land use/cover to determine available land for inclusion or exclusion in
the decision-making process.
Water availability is essential for rice cultivation to support growth and irrigation (Samanta
et al., 2011; Sinha et al., 2014). Thus, distance from river is included as a variable for
assessment of suitable land to farm rice. The closer the land is to the river, the more likely the

38
 Ayo Babalola et al.

soil water content particularly at lower elevation and slope. Similarly, movement of farm
machinery, seedlings and produce in and out of the farm requires accessibility. Using the
Euclidean Distance tool, distance from road, distance from river were generated (Figure 2o and
2p). Euclidean Distance analysis allows quantifying the spatial relationship between the factors
and the suitable location in linear distance (Anees et al., 2020; Mukti et al., 2016).

Multi Criteria Evaluation (MCE)

Determination of rice crop requirements and definition of sub criteria is based on expert
opinion and the knowledge gained from previous studies (as cited in Table 1). Since it is
unscientific to assume that all the criteria contribute to rice growth and yield equally, the weight
of each criterion was determined using analytical hierarchical process (AHP). The AHP
introduced by Saaty (Saaty, 2004) is multi-criteria decision analysis (MCDA) that has been
widely utilized to objectively determine the relative importance decision elements in many
studies (Abach & Ngigi, 2016; Ahmed et al., 2017; Hao et al., 2019; Kihoro et al., 2013, 2013;
Kumar & Patel, 2020; Park et al., 2019; Tien et al., 2018). In AHP, a measure of how significant
the input parameters are is examined through the process of pairwise comparison in which the
degree of importance of each pair of criteria is valued on a scale of 1 – 9, where 1 indicates
“equal importance”, 3 represents “Moderate importance”, 5 denotes “Strong importance”,
while 7 and 9 signify “Very strong- and Extremely importance” respectively (Taherdoost,
2018). One of the advantages of AHP is that it provides mechanism for verifying the reliability
of the final decision of the pairwise comparison process through assessment of the consistency
ratio - CR (Equation 1) which indicates whether the matrix ratings were randomly generated
or not (Saaty, 2013). In principle, CR less than or equal to 0.1 (10%) indicates that the decision
is acceptable but where CR >10%, the pairwise matrix need to be revised.
𝐶𝐼
𝐶𝑅 = 𝑅𝐼 (1)

Where CI= (λmax - n)/(n-1), RI = Random Index, n = number of criteria, λ = average of

consistency vector and λmax = priority vector multiplied by each column total.
For each of the input factors, suitability levels were defined on a scale of 1 to 5 (Table 1);
5 (very high suitability), 4 (high suitability), 3 (medium suitability), 2 (low suitability), and 1
(very low suitability). Apart from defining criteria levels of importance, the process also
satisfies the requirements for all input raster to be integer which was done using the
Reclassification Tools in ArcGIS environment.

Suitability Evaluation and Validation

The final suitability map was produced using weighted overlay analysis. Weighted overlay
analysis combines the input raster layers and their assigned weights indicating relative
importance or percent influence. Overlaying two or more datasets is much easier with raster
data than vector data since the former does not require any topological operation (Bagheri et
al., 2012). The weights of all the input variables must integer and sums up to 100; where
decimal value is obtained; it must be rounded up to the appropriate whole number. For areas
required to be excluded in the analysis, such as water body and urbanized or developed areas,
restricted value was used in the land use and land cover map. The process compares the
different criteria in each layer and their respective degree of relevance to arrive at an objective
suitability decision for the entire area. The resulting suitability map was validated using the
receiver operating characteristic’s (ROC) area under the curve (AUC) measure. This was done
by analyzing sampled points of rice farms obtained from the Kwara State Ministry of
Agriculture combined with field data collected using handheld GPS and negative points
randomly selected based on the knowledge of the authors of the study area.

39
 Journal of Agricultural, Food and Environmental Sciences Vol 75 No 1 (2021) 32-47

Table 1. Rice crop requirements and sub-criteria

Parameters Level of suitability Reference

NS (1) LS (2) MS (3) HS (4) VHS (5)
Elevation (m) 528 – 399- 271 – 399 142 – 13 – 142 Ujoh et al.,
657 528 271 2019
Slope (%) 30-72% 25-30% 20-25% 15-20% 0-15% Ujoh et al.,
2019
Aspect West, South East, North Flat, Ujoh et al.,
South West South East East, North 2019
North
West
Soil Texture Loamy Loam Sandy/cla Clayloa Clay Suhairi et al.,
sandy y/loam m, loam 2018
Soil Drainage Well Modera Imperfect Poor Very Suhairi et al.,
drain te poor 2018
Soil Depth 153-164 165- 175-184 185-194 195-200 Suhairi et al.,
(cm) 174 2018
Nitrogen 141-481 482- 601-759 760- 1007- Suhairi et al.,
(100g/kg) 600 1006 1959 2018
Potassium 45-101 102- 120-140 141-213 214-462 Suhairi et al.,
(100g/kg) 119 2018
Phosphorus 84-262 263- 321-393 394-538 539- Suhairi et al,.
(100g/kg) 320 1135 2018
Carbon (%) 3-25% 26-53% 54-101% 102- 249- Olaleye et
248% 441% al., 2008
Soil PH - <5.0 5.3-5.8 5.8-5.9 6.0-6.8 Olaleye et
al., 2008
Temperature - - - - 200C –
0
C 300C
Precipitation - - - 800- 1000-
(mm) 1000 1500
Distance to 14144- 10608- 7072- 3537- 0-3536 Abah &
water (m) 17678 14143 10607 7071 Petja, 2016;
Sinha et al.,
2014
Distance to 44380- 33284- 22190- 11096- 0-11095 Merem et al.,
Road (m) 55473 44379 33284 22189 2017
LULC Built-up Barren Forest Grass Wet land Victor &
area and Land land Samson,
Water 2019
NS – Not suitable, LS – low suitable, MS – moderately suitable, HS – highly suitable, VHS – very highly
suitable

RESULTS AND DISCUSSION

Relaibility of the outcome of any multicriteria decision process with AHP depends on the
suitability of the pairwise comparison (Table 2) and consequently the reliability measured
through the consistency ratio (Table 3).

40
 Ayo Babalola et al.

Table 2. Pairwise comparisons matrix of the input variables

Phosphor
Drainage

to Rivers
Precipita

Nitrogen
Tempera

to Roads
Distance

Distance
Potassiu

Elevatio
Soil PH
Texture

Carbon

Aspect
Depth

Slope
Soil

Soil

Soil
tion

ture

us

n
Precipitation 1 2 3 1 2 3 3 3 3 3 2 2 2 1 5
Temperature 1/2 1 3 2 2 2 2 2 2 2 3 3 3 2 5
Soil Texture 1/3 1/3 1 2 2 2 2 2 2 2 2 2 2 2 5
Soil
Drainage 1.0 1/2 1/2 1 2 2 2 2 2 2 3 3 3 1 5
Soil Depth 1/2 1/2 1/2 1/2 1 2 2 2 2 2 3 3 3 2 4
Nitrogen 1/3 1/2 1/2 1/2 1/2 1 1 1 1 1 3 3 3 2 5
Phosphorus 1/3 1/2 1/2 1/2 1/2 1 1 1 1 1 2 2 2 2 5
Potassium 1/3 1/2 1/2 1/2 1/2 1 1 1 1 1 2 2 2 2 5
Carbon 1/3 1/2 1/2 1/2 1/2 1 1 1 1 1 2 2 2 2 5
Soil PH 1/3 1/2 1/2 1/2 1/2 1 1 1 1 1 2 2 2 2 5
Elevation 1/2 1/3 1/2 1/3 1/3 1/3 1/2 1/2 1/2 1/2 1 1 1 2 3
Slope 1/2 1/3 1/2 1/3 1/3 1/3 1/2 1/2 1/2 1/2 1 1 1 2 5
Aspecct 1/2 1/3 1/2 1/3 1/3 1/3 1/2 1/2 1/2 1/2 1 1 1 3 2
Distance to
Rivers 1 1/2 1/2 1 1/2 1/2 1/2 1/2 1/2 1/2 1/2 1/2 1/3 1 4
Distance to
Roads 1/5 1/5 1/5 1/5 1/4 1/5 1/5 1/5 1/5 1/5 1/3 1/5 1/5 1/4 1
Weight 0.134 0.120 0.095 0.097 0.086 0.065 0.058 0.058 0.058 0.058 0.037 0.039 0.039 0.042 0.015
16.077 0.048

41
 Journal of Agricultural, Food and Environmental Sciences Vol 75 No 1 (2021) 32-47

Once the matrix pairwise comparison has been obtained, factor weights were calculated
and the computed weight were (Table 3): precipitation (0.134), temperature (0.120), soil
texture (0.095), soil drainage (0.097), soil depth (0.086), nitrogen (0.065). Others are
phosphorus, potassium, carbon, soil PH (0.058, each), elevation (0.037), slope and aspect
(0.039 each), distance to rivers (0.042), and distance to roads (0.015). The consistency ratio
obtain was 4.8% (0.048), which is considered acceptable (Abach & Ngigi, 2016; Saaty, 2004).
Also, the computed consistency ratio for each of the factors varies between 0.5% and 6.5%, all
of which are within the acceptable limit of <=10%. Parameter ranking identified precipitation,
temperature, soil drainage, soil texture and soil depth as the first, second, third and fifth best
variables, respectively. Next to these variables are nitrogen in the sixth position while
phosphorous, potassium, carbon and soil PH occupied the seventh position. At the base of the
rating are Distance to river, slope, aspect, elevation and distance to roads from 11th to 15th
position.

Table 3. Variable weight, rank and consistency check

Variable Priority Rank Consistency check

1 Precipitation 13.4% 1 6.5%
2 Temperature 12.0% 2 5.0%
3 Soil Texture 9.5% 4 4.0%
4 Soil Drainage 9.7% 3 3.3%
5 Soil Depth 8.6% 5 2.7%
6 Nitrogen 6.5% 6 2.4%
7 Phosphorus 5.8% 7 1.2%
8 Potassium 5.8% 7 1.2%
9 Carbon 5.8% 7 1.2%
10 Soil PH 5.8% 7 1.2%
11 Elevation 3.7% 14 1.5%
12 Slope 3.9% 12 1.7%
13 Aspect 3.9% 13 2.4%
14 Dist. to Rivers 4.2% 11 3.0%
15 Dist. to Roads 1.5% 15 0.5%
Consistency Ratio CR = 4.8%; Principal eigen value = 16.08 and
Eigenvector solution: 6 iterations

Figure 3 presents the suitability map categorized into five classes: restricted area, low
suitable, moderately suitable, highly suitable and very highly suitable, utilizing the natural
break classification method (Ouri et al., 2020; Shabani et al., 2020). Kwara state has a total
landmass of 3542031.89 hectares classified into five landcover classes; grassland/shrubs,
forest, built-up areas, bare earth/cultivated land, and water body. The grassland/shrubs and
forest cover types occupy about 77% of the land (42.68% and 34.41%, respectively) while
built-up areas, bare earth/cultivated land, and water body occupy 7.57% (268128.29 ha),
14.86% (526481.40 ha) and 0.47% (16762.69 ha). According to Figure 3, the number of
hectares available to each suitability class are: very highly suitable 154408.94 ha (4.36%),
highly suitable 1436504.55 ha (40.56%), moderately suitable 1091145.20 ha (30.81%), and
low suitable 516169.46 ha (14.57%), while the remaining 10% comprised of water body and
developed areas are excepted from the analysis. The validation result yields area under the
ROC curve value of 0.98 (98%).

42
 Ayo Babalola et al.

Figure 3. Map of suitable land for Rice crop farming in Kwara State

Generally, land suitability for crop production is dependent on a number of factors

including climatic condition, soil physical and chemical properties, and the landscape of the
area. In this study, through the combination of factors that favour rice crop cultivation (Figure
2), their respective weight produced using MCE technique (Table 3) and associated constraints
(based on land use/cover types), areas suitable for the rice crop cultivation were delineated
(Figure 3). Fundamentally, the mean annual temperature and precipitation receive in the study
area favour rice crop growth and yield (Merem et al., 2017; Sadiq et al., 2017); so, both factors
are not critical determinants of land suitability as the other variables, particularly the soil
properties. Distance to river and road are environmental variables that are not directly linked
to crop growth but support logistical requirement such as irrigation and ease of transportation
of farm machinery and produce to the desire market (Babatunde et al., 2019; Kaaya et al.,
2019).
Cross-sectional analysis between the map of suitable areas and the soil variables provides
useful information on how each variable contributes to the final output. In terms of the soil
physical property, the entire study area has high soil depth in the range of 195 cm – 200 cm.
The very high a nd high suitability areas are located in areas with abundant clayloam,
sandyclayloam and loaming soil associated with imperfect and well-moderately drained soil,
whereas the moderate and low suitability classes are found on moderately drained sandyloam
soil which dominates the study area. The distribution of the different suitability classes supports
the finding of Daniel (2011), Kihoro et al. (2013) and Merem et al. (2017) that sorghum and
rice are hydromorphic crops that grow well in silt and some combination of loamy soil, clayey
and sand that develop within terraces.
In terms of the soil chemical properties, the study area possesses moderate to mild soil
acidity with soil PH value between 5.3 – 6.8. High soil PH trends northwest – southeast in low
land close to river channels and decreases southward (Figure 2k). The very high and high

43
 Journal of Agricultural, Food and Environmental Sciences Vol 75 No 1 (2021) 32-47

suitability classes fall within mildly acidic soil with PH value between 6 – 6.8 that is rich in
nitrogen (500 – 1959 g/kg), potassium (120 – 462 g/kg), soil organic carbon (26% - 53%) but
low in phosphorus (84 – 262 g/kg). These areas h a v e moderately shallow soil depth
relatively rich in plant nutrients to support productive rice cultivation with minimum
tillage and careful erosion control (McCauley et al., 2009; Sadiq et al., 2017). Conversely, the
moderate and low suitability classes are associated with moderate soil acidity (5.3 – 6.0), low
nitrogen, potassium, and soil carbon in the range of 141 – 481 g/kg, 45 – 119 g/kg, and 3% -
25%, respectively, and high phosphorus between 263 and 534 g/kg. These fall in irrigated land
largely affected by salinity, poor drainage and nutrients due to exposure to severe erosion
influenced by topography. Our result has shown that Kwara state has very good climate and
soil characteristics for rice production. Crossing the suiatbility map with the NDVI image
depicts the present reality with respect to the land availability and vegetation amount and
distribution. Based on the acreage currently been used for rice cultivation, the potential acreage
available is underutilized. Our findings also indicated that majority of the area classed at very
high and high suitable in the south-eastern area constitute forest while the north and central
areas classed moderately suitable are fast becoming degraded lands perhaps due to nutrient
depletion, urbanization and desertification.

CONCLUSION
This study demonstrates the application of remote sensing and spatial analysis by
integrating various datasets of different spatial and temporal resolutions in a GIS environment.
MCE of climatic factors, soil characteristics, relief and environmental variables useful to
delineate suitable areas for rice crops production have been utilized. In this way, a thematic
map that takes into account specific rice crop growth characteristics to indicate varying levels
of suitability areas was generated. Absence of complete, reliable and up to date data from
relevant local and national government agencies have negatively impacted on ability to
provides information at a regional scale that could be used by local farmers to select their crop
pattern. We found the freely available remote sensing data crucial to obtaining useful
information in a precise and relatively fast way.
We applied the MCE technique to identify areas for rice crop farming within a GIS
environment using the selected climate, soil, topographic and environmental variables that
support rice farming. The soils are well drained and the topography relatively flat with few
rock outcrops in the northern side of the state. Across the state, the soil have sufficient nutrients
of high agricultural potentials in compliance with the USDA (1951) recommendation of soil
with sand composition > 50% and clay of 15% to allow the soil cohere and mould fairly easily
when mist without being sticky. This is evidenced with moderate, high and very high sutable
classes occuping ~75% of the landmass.
This study has shown that combination of MCE–GIS provide rational and objective basis
to making decisions in rice crop farming in Kwara state and the methodology can be replicated
in other states of the Federation for future specific studies on rice crop. However, we know that
decision-making process to select adequate land for rice crop could be based not only on the
information provided in this study, but also on other issues such agricultural policies of the
local and federal governments, crop production priority, and supports, in addition to local
cultural farming traditions.

44
 Ayo Babalola et al.

REFERENCES
Abach, R. O., & Ngigi, M. M. (2016). Land suitability study for rice growing in Kisumu
county. International Journal of Geomatics and Geosciences, 7(1), 33–42.
Abah, R. C., & Petja, B. M. (2016). Crop Suitability Mapping for Rice, Cassava, and Yam in
North Central Nigeria. Journal of Agricultural Science, 9(1), 96.
https://doi.org/10.5539/jas.v9n1p96
Ahmed, G. B., Shariff, A. R. M., Idrees, M. O., Balasundram, S. K., & Bin Abdullah, A. F.
(2017). Gis-based land suitability mapping for rubber cultivation in seremban, malaysia.
International Journal of Applied Engineering Research, 12(20), 9420–9433.
Anees, M. M., Mann, D., Sharma, M., Banzhaf, E., & Joshi, P. K. (2020). Assessment of urban
dynamics to understand spatiotemporal differentiation at various scales using remote sensing
and geospatial tools. Remote Sensing, 12(8). https://doi.org/10.3390/RS12081306
Babatunde, R., Omoniwa, A., & Aliyu, J. (2019). Post-Harvest Losses along the Rice Value
Chain in Kwara State, Nigeria: An Assessment of Magnitude and Determinants. Cercetari
Agronomice in Moldova, 52(2), 141–150. https://doi.org/10.2478/cerce-2019-0014
Bagheri, M., Sulaiman, W. N. A., & Vaghefi, N. (2012). Land use suitability analysis using
multi criteria decision analysis method for coastal management and planning: a case study of
Malaysia. Journal of Environmental Science and Technology, 5(5), 364–372.
https://doi.org/10.3923/jest.2012.364.372
Ceballos-Silva, A., & López-Blanco, J. (2003). Delineation of suitable areas for crops using a
Multi-Criteria Evaluation approach and land use/cover mapping: A case study in Central
Mexico. Agricultural Systems, 77(2), 117–136. https://doi.org/10.1016/S0308-
521X(02)00103-8
Daniel, O. A. (2011). Determining Rice Productivity Level for Sustainable Agricultural
Development in. Journal of Sustainable Development in Africa, 13(5), 125–135.
Falola, A., Animashaun, J. O., & Olorunfemi, O. D. (2014). Determinants of Commercial
Production of Rice in Rice-Producing Areas of Kwara State , Nigeria. Albanian Journal of
Agricultural Science, 13(2), 59–65.
Hao, N. H., Van, P. Van, & Ha, K. M. (2019). Applying AHP method and GIS to evaluate land
suitability for paddy rice crop in Quang Xuong district, Thanh Hoa province. Can Tho
University Journal of Science, Vol.11(3)(3), 1. https://doi.org/10.22144/ctu.jen.2019.032
Jeevalakshmi, D., Reddy, S. N., & Manikiam, B. (2016). Land cover classification based on
NDVI using LANDSAT8 time series: A case study Tirupati region. International Conference
on Communication and Signal Processing, ICCSP 2016, 560056, 1332–1335.
https://doi.org/10.1109/ICCSP.2016.7754369
Kaaya, A. K., Winowiecki, L., Slater, B. K., Sciences, G., & Resources, N. (2019). Multi-
criteria Land Evaluation for Rice Production using GIS and Analytic Hierarchy Process in
Kilombero Valley, Tanzania. 18(2), 88–98.
Khattak, P., & Shabbir, R. (2012). Temporal Analysis of Wheat Yield and Climatic Trends in
Pakistan. Elixir Agriculture, 52(10), 11598–11603.
https://doi.org/10.5829/idosi.wasj.2013.24.10.1714
Kihoro, J., Bosco, N. J., & Murage, H. (2013). Suitability analysis for rice growing sites using
a multicriteria evaluation and GIS approach in great Mwea region, Kenya. SpringerPlus, 2(1),
1–9. https://doi.org/10.1186/2193-1801-2-265
Kumar, R., & Patel, G. R. (2020). Assessment of Agro-Land Suitability for Rice (Oryza sativa
L.) in Bhal Area of Gujarat Using GIS and Remote Sensing. International Journal of Current
Microbiology and Applied Sciences, 9(4), 1207–1214.
https://doi.org/10.20546/ijcmas.2020.904.143
Maddahi, Z., Jalalian, A., Zarkesh, M. M. K., & Honarjo, N. (2017). Land suitability analysis
for rice cultivation using a GIS-based fuzzy multi-criteria decision making approach: Central

45
 Journal of Agricultural, Food and Environmental Sciences Vol 75 No 1 (2021) 32-47

part of amol district, Iran. Soil and Water Research, 12(1), 29–38.
https://doi.org/10.17221/1/2016-SWR
McCauley, A., Jones, C., & Jacobsen, J. (2009). Soil pH and Organic Matter. Nutrient
Management Module No. 8, 8, 1–12.
Merem, E. C., Twumasi, Y., Wesley, J., Isokpehi, P., Shenge, M., Fageir, S., Crisler, M.,
Romorno, C., Hines, A., Hirse, G., Ochai, S., Leggett, S., & Nwagboso, E. (2017). Analyzing
Rice Production Issues in the Niger State Area of Nigeria’s Middle Belt. Food and Public
Health, 7(1), 7–22. https://doi.org/10.5923/j.fph.20170701.02
Mukti, A., Prasetyo, L. B., & Rushayati, S. B. (2016). Mapping of Fire Vulnerability in Alas
Purwo National Park. Procedia Environmental Sciences, 33, 290–304.
https://doi.org/10.1016/j.proenv.2016.03.080
NPC. (2006). Nigeria Population Census - 2006. Nigeria Data Portal.
https://nigeria.opendataforafrica.org/ifpbxbd/state-population-2006
Olaleye, A. O., Akinbola, G. E., Marake, V. M., Molete, S. F., & Mapheshoane, B. (2008). Soil
in suitability evaluation for irrigated lowland rice culture in southwestern nigeria: Management
implications for sustainability. Communications in Soil Science and Plant Analysis, 39(19–20),
2920–2938. https://doi.org/10.1080/00103620802432824
Oriola, E., & Olabode, A. (2014). Geodatabase for Sustainable Rice Production in Kwara State.
International Journal of Advanced Biological Research, 4(1), 36–47.
Ouri, A. E., Golshan, M., Janizadeh, S., Cerdà, A., & Melesse, A. M. (2020). Soil Erosion
Susceptibility Mapping in Kozetopraghi. Land, 1–18. https://doi.org/doi:10.3390/land9100368
Park, S., Im, J., Jang, E., Rhee, J., & Basin, P. R. (2019). Fuzzy AHP Integrated with GIS
Analyses for Drought Risk Assessment : A Case Study from Upper. Agricultural and Forest
Meteorology, 216, 157–169. http://dx.doi.org/10.1016/j.agrformet.2015.10.011
Saaty, T. L. (2004). Decision making — the Analytic Hierarchy and Network Processes
(AHP/ANP). Journal of Systems Science and Systems Engineering, 13(1), 1–35.
https://doi.org/10.1007/s11518-006-0151-5
Sadiq, M. S., Singh, I. ., & Yakubu, G. . (2017). Global Warming Potentials of Lowland Paddy
Rice Production in Kwara State of Nigeria : Lessons for Sustainable Agriculture. International
Journal of Research in Agriculture and Forestry, 4(9), 20–32.
Samanta, S., Pal, B., & Pal, D. K. (2011). Land Suitability Analysis for Rice Cultivation Based
on Multi-Criteria Decision Approach through GIS. International Journal of Science &
Emerging Technologies, 1, 12–20.
Shabani, S., Reza, H., & Blaschke, T. (2020). Forest stand susceptibility mapping during
harvesting using logistic regression and boosted regression tree machine learning models.
Global Ecology and Conservation, 22, e00974. https://doi.org/10.1016/j.gecco.2020.e00974
Sinha, D. D., Singh, A. N., & Singh, U. S. (2014). Site suitability analysis for dissemination of
salt-tolerant rice varieties in Southern Bangladesh. International Archives of the
Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives, 40(8),
961–966. https://doi.org/10.5194/isprsarchives-XL-8-961-2014
Suhairi, T. A. S. T. M., Jahanshiri, E., & Nizar, N. M. M. (2018). Multicriteria land suitability
assessment for growing underutilised crop, bambara groundnut in Peninsular Malaysia. IOP
Conference Series: Earth and Environmental Science, 169(1). https://doi.org/10.1088/1755-
1315/169/1/012044
Taherdoost, H. (2018). Decision Making Using the Analytic Hierarchy Process ( AHP ); A
Step by Step Approach. International Journal of Economics and Management System,
February.
Tien, D., Le, H. Van, & Hoang, N. (2018). Ecological Informatics GIS-based spatial prediction
of tropical forest fi re danger using a new hybrid machine learning method. Ecological
Informatics, 48(April), 104–116. https://doi.org/10.1016/j.ecoinf.2018.08.008

46
 Ayo Babalola et al.

Ujoh, F., Igbawua, T., & Ogidi Paul, M. (2019). Suitability mapping for rice cultivation in
Benue State, Nigeria using satellite data. Geo-Spatial Information Science, 22(4), 332–344.
https://doi.org/10.1080/10095020.2019.1637075
Victor, O. K., & Samson, A. O. (2019). An Application of GIS-Based Multi-Criteria Decision
Making Approach for Land Evaluation and Suitability Mapping for Rice Cultivation in Oye-
Ekiti, Nigeria. Journal of Agriculture and Environmental Sciences, 8(1), 0–9.
https://doi.org/10.15640/jaes.v8n1a3

47