�������� ����� ��

Spatial mapping of selected soil fertility parameters and by kriging andco-kriging models in yam-based-cropping systems of Nigeria

Martin Jemo, Olumuyiwa J. Jayeoba, Tunrayo Alabi, Antonio LopezMontes

PII: S2352-0094(14)00025-XDOI: doi: 10.1016/j.geodrs.2014.10.001Reference: GEODRS 16

To appear in:

Received date: 11 May 2014Revised date: 5 October 2014Accepted date: 9 October 2014

Please cite this article as: Jemo, Martin, Jayeoba, Olumuyiwa J., Alabi, Tunrayo,Montes, Antonio Lopez, Spatial mapping of selected soil fertility parameters and bykriging and co-kriging models in yam-based-cropping systems of Nigeria, (2014), doi:10.1016/j.geodrs.2014.10.001

This is a PDF file of an unedited manuscript that has been accepted for publication.As a service to our customers we are providing this early version of the manuscript.The manuscript will undergo copyediting, typesetting, and review of the resulting proofbefore it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers thatapply to the journal pertain.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

1

Spatial mapping of selected soil fertility parameters and by kriging

and co-kriging models in yam-based-cropping systems of Nigeria

Authors: Martin Jemo1*; Olumuyiwa J. Jayeoba2, Tunrayo Alabi1, and Antonio Lopez

Montes1

Authors’ affiliations:

1International Institute of Tropical Agriculture (IITA), Oyo Road, PMB 5320, Ibadan, Nigeria

2Agronomy Department, Faculty of Agriculture, Nasarawa State University, Keffi, Lafia Campus

* Corresponding author

Running title: Spatial mapping of selected soil fertility parameters

Pages: 16

Tables 4: Figures 4.

Corresponding author’s address:

IITA, Oyo Road, PMB 5320, Ibadan, Nigeria

Tel: +234-02-7517472 Fax: +234-02-2412221

E-mail: M.jemo@cgiar.org

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

2

Abstract

Regardless of the plant type to be grown in a particular soil, the required nutrients (i.e., N, P, K, etc.)

must be supplied for growth and wellbeing. However, in the context of the small farming systems of

the tropics, many soils lack the capacity to adequately supply such nutrients amount for the plants to

achieve their optimal and/or potential yield. An extensive survey covering the area 5° 21'E, lat. 9°

22' N and the 9° 55'E, 5° 41'N was conducted in the fields of over 385 yam grower farmers to

spatially map selected soil fertility parameters, N, P, and K, under yam based cropping systems of

Nigeria. Using geo-statistical mapping by spatial prediction tools, we aimed to determine the spatial

distribution of soil fertility levels and develop recommendations for a broader soil fertility

intervention in the targeted areas. Spatial analysis of the data showed that the best fit semi-variogram

models for nutrients N and K were Rational Quadratic and Hole Effect and for available P an

ordinary kriging technique. Relative Nugget Effect (RNE), nugget-to-sill ratio (Co/Co + C) of the

total N concentration in soil was 70%, under the yam based cropping systems of Nigeria 45% for

exchangeable K. The available P in soil (Bray-P) was weakly spatially dependent with RNE of 83%

and a range of 1.8 km. The results of this study demonstrated the need for site-specific

recommendations for P and K while recommendations for N management could be done at a

regional scale.

Keywords: Integrated soil fertility management, nutrient concentration, site-specific nutrient

management, spatial distribution and modeling.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

3

1. Introduction

Yams (Dioscorea spp.) are monocotyledonous plants of economic importance to many subsistence

farmers in West and Central Africa (FAO, 2005). The tuber is an important staple food for millions

of people living in the growing areas (Suja et al. 2012; Tchabi et al. 2008). Additional qualities in the

crop include their high and rich carbohydrate content, accounting for about 50–80% starch on dry

weight basis. Under the subsistence conditions of many smallholders in the crop growing areas, the

current yield is low compared with potentially achievable yields, being curtailed by various biotic and

abiotic constraints including pests and diseases, weed competition, and deteriorating soil fertility

problems (Amusa et al. 2003; Bridge et al. 2005; Coyne et al. 2005; Tchabi et al. 2010).

In the traditional cropping system, yam is generally the first crop planted after the conversion of

long-term fallows into agricultural land, mostly because of the crop’s high nutrient requirement and

demand (Diby et al. 2011; Carsky et al. 2001). Ample evidence of declining and poor soil fertility and

nutrients depletion are frequently observed for many crops growing in African soils, including yam

due to the shortening duration of fallow, nutrient mining, the highly fragile nature of many of these

soils, the accelerated degradation process, and the lack of adapted agricultural practices to restore

soil fertility (Stoorvogel and Smalling, 1993; Sanchez, 2002; Schlecht et al. 2006; Frossard et al.

2007).

The major macronutrients are indispensable for the optimal growth and wellbeing of plants. Their

bioavailability to roots is of considerable economic importance because they are the major plant

nutrients derived from the soil in organic or in inorganic forms (Marschner et al. 1995). With respect

to N, it is a key building block of the protein molecules; thus it is an indispensable component of the

protoplasm of plants, animals, and soil chemicals. The major role of P is the accumulation and

release of energy during cellulose metabolism. N deficiency in plants causes chlorosis. Plants become

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

4

stunted and poor growth pattern in general are observed (Ibia and Udo, 2009). Deficiency of P in

plants results in dull green to purple coloration of leaves; deficiency of K is expressed as yellow to

brown discoloration, followed by necrosis of the lower leaves.

At the farm scale, various research outputs on nutrient recommendations; if exist in many countries

are to treated fields as homogenous areas and develop the fertilizer requirements on a whole field

basis. However, large spatial heterogeneity in soil fertility, nutrient resource allocation and

distributions in smallholder agricultural land are frequently observed, with high variability within a

single farm field (Tittonel et al. 2013). Flowers et al. (2005), Santra et al. (2008) reported that at least

70 sampled fields were not homogeneous and sampling techniques to describe field variability have

been recommended in India.

The broad expansion of the geo-statistical tools in the recent years and the introduction of the

mapping by spatial prediction combined with Global Positioning Systems (GPS), and Geographic

Information Systems (GIS) have accelerated modeling and predictions of landscape study. This has

contributed to delineate potential areas of low or high nutrients distribution (Burrough, and

McDonnell, 1998). Geo-statistical mapping is also used in producing soil fertility maps, to

demarcate areas that express low nutrient availability and define threshold values that could be

important to measure the recovery of such lands, and also to helping in designing trials with the

objective of formulating fertilizer recommendation at a regional or country scale (Goovaerts, 1997;

Webster and Oliver, 2001; Nielsen and Wendroth, 2003). These technologies allow fields to be

mapped accurately and allow complex spatial relationships between soil fertility factors to be

computed at the very high and detailed robust resolution scale (Patil et al. 2011). The kriging

technique has been used for many decades as a synonym for geo-statistical interpolation and has

been proved as sufficiently robust for estimating values at not un-sampled locations based on the

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

5

sampled data (Yost et al. 1982; Trangmar et al. 1987; Miller et al. 1988; Voltz and Webster, 1990;

Chien et al. 1997; Lark, 2002).

With yam, there is a paucity of information governing the plant’s nutrient acquisition; the whole soil

fertility problem is still unsolved, with a very limited knowledge of crop physiology related to uptake

efficiency and use (Hgaza et al. 2012). No adequate or balanced site-specific fertilizer

recommendations have so far been developed or proposed for many subsistence farmers growing

yam in the study area. The purpose of this present study was to characterize the spatial variability in

concentrations of N, available-P, and exchangeable K under the yam based systems in Nigeria, to

allow appropriate site-specific recommendations of generalized fertilizers profitable at the regional

scale to be designed.

2. Material and Methods

2.1. Site description

Initial selection of potential sampling sites was made using Ar-GIS 10.0, to determine areas of low

yam yield production, high poverty index, poor resource allocation, and low soil fertility levels, as

defined in the project proposal funded by the Ministry of Agriculture, Fishery and Forestry of Japan

(MAFF). The preliminary studies led to the following States of Nigeria: Benue, Nasarawa, Kogi, and

Ebonyi, being identified as representing the yam belt production areas. The extent is Long. 5° 21'E,

Lat. 9° 22' N and 9° 55'E, 5° 41'N in the north central area and part of southeast Nigeria. The agro-

ecology of the study predominantly lies within the derived savanna zone (Fig.1), with annual rainfall

1200 –1800 mm and length of growing period 210 –240 days. The mean relative humidity is over

70% in the morning and falls to between 50 and 70% in the afternoon. The mean annual

temperature is 27 °C and the annual temperature range is 8 °C.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

6

2.2. Soil sampling

Soil samples were collected from 385 selected yam field sites from the beginning of September to

the end of October 2012. Fields were established and managed by the farmers. In each farmer’s

field, an area of 10 × 10 m2 was delineated. Soil was sampled from a depth of 0 –30 cm around 20

yams mounds and bulked to obtain about 5 kg. Fresh samples were returned to the laboratory at the

Soil Microbiology Laboratory at the International Institute of Tropical Agriculture (IITA), Nigeria.

Upon collection, geographical coordinates using Garmin, Etrex 20, GPS Unit and the chlorophyll

content of 20 leaves from each plant were recorded. In the laboratory, the fresh samples were

divided in two pools, one kept at 4 0C for the further microbial and molecular analyses, the other

used for physical and chemical analyses.

2.3. Soil physical and chemical properties analyses

Physical and chemical properties of the soils were assessed at the normal room temperature after the

samples were oven-dried at 60 0C for 2 days. The soil was sieved to pass a 2-mm sieve. Soil texture

was assessed using a sedimentation-pipetting method following hydrogen peroxide treatment. The

soil pH (H2O) was measured in 1:2.5 (v:v) water to soil suspension. Total N and P concentrations in

soil were measured after samples were digested in sulphuric acid-selenium mixture at 350 0C for 8 hr.

Total N concentration in the extract was determined by the Kjeldahl method (Bremner and

Mulvaney, 1982). The available-P was extracted using the Bray I method (Bray and Kurtz, 1945)

using the following extractants: 0.025 M HCl; 0.03 M NH4F. The concentration was measured

calorimetrically according to the procedure of Motomizu et al. (1983). Cation exchange capacity was

measured using BaCl2 extraction of exchangeable cations and their respective concentration was

determined in the extracts using the flame test method (Robertson et al. 1996).

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

7

Plant N and P concentration

AMF observation and identification

2.4. Statistical analysis

Exploratory data analysis was performed with SPSS (version 16) software. The data distributions

were analyzed by classical statistics (mean maximum, minimum, standard deviation, skewness,

kurtosis, and coefficient of variation). Histograms with normal curves were plotted for all the

fertility variables for the possible outliers and extraneous values.

Spatial analysis of the classified soil fertility data was performed using ArcGIS 10.0 and ArcGIS

Geo-statistical Analyst Extension software packages. The prediction of the spatial process at non-

sampling sites using geo-statistics requires a theoretical semi-variogram. We thus constructed a

theoretical variogram based on the experimental variogram and tested various models for the

appropriate one that could efficaciously estimate spatial statistics as each model yields different

values for nugget variance and range, which are essential for geo-statistical analyses (Trangmar,

1985). Model selections for semi-variograms were done on the basis of a goodness of model fit

criterion. An interpolation technique called ordinary kriging was used to produce the spatial

distribution of the soil parameters for the study area. The data were checked for normality and

transformed as appropriate. Spatial analyses of the classified total N, available P, and exchangeable

K were performed in a GIS environment and experimental semi-variograms were calculated using

equation (1):

(1)

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

8

Where γ(h) is the semi-variance for the lag distance h. N (h) is the number of sample pairs separated

by the lag distance h, z(xα) is the measured value at αth sample location and z(xα+h) is the measured

value at point α+hth sample location.

Various semi-variogram models (Circular, Spherical, Tetraspherical, Pentaspherical, Exponential,

Gaussian, Rational Quadratic, Hole effect, K-Bessel, J-Bessel, Stable) were tested for each set of soil

parameter data to identify which could be appropriately used to explain the variability of the dataset:

a) Linear semi-variogram model: The simplest model that can be fitted in one dimension is linear.

It has a slope w an intercept or nugget variance Co and it is given by:

(2)

(3)

It has no sill. If w = 0, then the semi-variogram is said to show a pure nugget effect.

b) Spherical model: Spherical model can be expressed as:

(4)

(5)

(6)

Where, a is the range, C0 + C is the sill and C0 is the nugget variance.

c) Exponential model: The formula for the exponential model is given by:

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

9

(7)

Where, r is a distance parameter controlling the spatial extent of the function γ(h).

Other theoretical models (Rational, Quadratic, Hole effect, K-Bessel or Gaussian) were fitted to

experimental semi-variograms.

Since statistically abnormal distribution of datasets can have an adverse impact on semi-variograms

and further interpolation, an elementary knowledge of soil fertility data was conducted before semi-

variogram analysis.

3. Results

The total N data across the study area ranged from 1 to 5 g N kg-1 of soil; available P ranged from

0.05 to 63.15 mg P kg-1 soil and exchangeable-K ranged from 0.03 to 4.97 meq K/100 g soil. The

data were highly skewed with high levels of kurtosis, 33.4 for total N and 59.9 for available P,

compared with 3.95 in the data for K (Table 1).

The frequency distributions and normal curves of the raw data of total N, available-P, and

exchangeable K are presented in Figure 2. With N concentration in soil, the observed frequency

indicated that, in up to 70 % of the soil analyzed, the N content ranged between 0.0 and 1.0 g N kg

of soil, which corresponds to a very low level of N in yam based cropping systems as shown in our

guide data interpretation (Table 2 and Fig. 2a). With respect to available-P, various soil samples

analyzed showed a concentration that ranged from 0.01 to 20 mg P kg-1 soil, with the mean situated

at 70% (Fig. 2b). The frequencies for available K ranged 0 to 1.0 meq K/100 g soil.

The frequency distributions and normal curves of the adjusted data of total N and available P after

the removal of the outliers showed a very similar logic correlation with those of the data collected

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

10

from the field (Figs 2d, 2e, 2f). A linear correlation was observed from the expected normal value of

the data distribution against the observed values (Figs 3a, 3b&3c).

The semi-variance statistics of measured soil properties are shown in Table 3. Several models were

fitted to the semi-variograms and the Rational Quadratic model was obtained as the best fit for total

N and exchangeable K. The best-fit model for available P was the Hole effect. Anisotropy was not

evident in the directional semi-variograms for any of the properties. Therefore, isotropic models

were fitted. All the semi-variogram models displayed positive nugget effects (N=0.136, P=20.48,

and K=0.45), which may be as a result of sampling error, random inherent variability, or short-range

variability.

To identify the spatial distribution patterns of soil properties in the study area, it is necessary to

present the data in the form of a map. For this purpose, distribution maps of N and P were obtained

by the ordinary kriging based on the Rational Quadratic model; that for P was from the Hole effect

model.

Spatial distribution maps of surface (0–30cm) soil total N, available P, and exchangeable K of

selected fields in the yam growing belt of Nigeria are presented in Figure 4 a, b&c).

Areas and percentage fertility classes are presented in Table 4. The spatial interpolation of total N in

the surface soil in the study area revealed that the total N content is very low in 84,091km2 (90.7%)

of the total land area (<0.1%); the remaining 8,673 km2 (9.3%) of the land area are in the low range.

Spatial distribution of soil available P in the study area at surface (0–30 cm) depth showed that

46,633 km2 of the area (80.3%) are low in available-P while 43,450 km2 (46.8%) of the study area are

in the medium range. The spatial distribution further revealed that only 2,681 km2 (2.9%) of the

study area are high in available-P. Figure 4 shows the spatial distribution of exchangeable K at the

surface (0–30cm) depth.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

11

Unlike the spatial pattern observed in total N and available-P, the map revealed 61,366 km2 of the

study area (66.2%) to be high in exchangeable K content and 4,210 km2 (4.5%) to be very high. Only

14,234 km2 (15.3%) are low and 12,954 km2 (14.0%) are medium in exchangeable K content. Spatial

distribution maps of the soil fertility status as shown indicated a very low N concentration in 90.7%

of all the samples analyzed against only 9.7% with low N concentrations.

With respect to P concentrations in field soil samples, low (50.3%), medium (46.8%), and high

(2.9%) levels were observed at different percentages. With K levels, four ranges of availability were

differentiated: low (15.3%), medium (14.0%), high (66.2%), and very high (4.5%).

4. Discussion

Sustainable yam production requires well-drained and fertile soil. Soil nutrient replenishment from

organic and mineral sources is a prerequisite for the continuous cultivation of such soils, particularly

under intensive production. Application of the appropriate type, quantity, and rates of fertilizer is

essential for profitable and sustainable farming enterprises. For the course of this study, we used the

ordinary kriging method for making optimal, unbiased estimates of regionalized variables at non-

sampled locations using the structural properties of the semi-variogram and the initial set of

measured data. A useful feature of kriging is that an error term expressing the estimation variance or

uncertainty in estimation is calculated for each interpolated value. Kriging always produces an

estimate equal to the measured value if it is interpolating at a location where a measurement is

obtained (Burrough, 1997; Stein and Bouma, 1997; Jayeoba et al. 2012). Distinct classes of spatial

dependence for the soil properties were obtained by the ratio of the nugget to sill. If the ratio was

<25%, the variable was considered strongly spatially dependent, between 25 and 75%, moderately,

dependent and >75%, weakly dependent (Cambardella et al. 1994). In accordance with the nugget

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

12

effect, the semi-variogram increased until the variance of the data called sill C was reached. Under

this semi-variogram value, the regionalized variables at the sampling locations are spatially

correlated. The sill value represents total experimental errors (Ersoy, 2004). If the distance between

two pairs increases, the variogram of those two pairs will also increase. Eventually, the increase of

the distance cannot cause an increase in the variogram. The distance, which causes the variogram to

reach the plateau, is called the range. In other words, the range is considered as the distance beyond

which observations are not spatially dependent (Gallardo, 2003). The range of spatial dependence

varied from 8,616 m for K, 1,800 m for available P, to 623 m for total N. Total N showed the lowest

range, indicating spatial correlation within a smaller distance among the studied soil properties. This

indicated that samples should be taken at comparatively shorter distances (within 1 km). N is a very

dynamic element. It not only exists on earth in many forms but also undergoes many

transformations in and out of the soil. Soil N content is spatially dependent on other properties,

such as soil texture, soil temperature, soil moisture, soil organic matter, and plant residues. These

properties are variable within a short distance. Relative nugget effect (ratio of the nugget to sill) of

total N was 70% (Table 3) showing a moderately spatial dependence (Cambardella et al. 1994).

Exchangeable K showed a high spatial correlation to a greater distance among the soil properties

studied with a range of 8.6 km. Unlike N, soil K content is relatively stable over the years because

most of the K found in the soil exists as a mineral such as feldspar and mica. The transfer of mineral

K to other states in the K cycle is a very slow process. Essentially, mineral K is not available for

plant uptake during a single growing season. Relative nugget effect of exchangeable K was 45%

(Table 3) showing a moderately spatial dependence, due to soil mineralogy and soil formation

processes (Cambardella et al. 1994; Shukla et al. 2004). However available P was weakly spatially

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

13

dependent (relative nugget effect = 83%) with a range of 1.8 km, probably due the nature of the

inner and parental rock material and the lack of P inputs supply to the crops.

Based on spatial fertility distribution as presented above, a blanket fertilizer recommendation for

yam in the study area may be adequate for P, but suboptimal. Optimal levels for N and excess for K

are to be recommended depending in the initial level of these nutrients at the soil site. Applying the

right amount of inputs in the right place and at the right time benefits crops, soils, and groundwater,

and thus the entire crop cycle.

5. Conclusion

Geo-statistical characterization of the spatial variability through semi-variograms or correlograms

generally brings new insight into the way soil attributes are influenced by the environment, such as

the geographical distribution of soil types or topography. The traditional approach to soil fertility

management has been to treat fields as homogenous areas and to calculate fertilizer requirements on

a whole field basis. The predicted maps obtained could be helpful to the farmers and soil

management experts to design land management and fertilizer recommendations taking into account

the spatial heterogeneity of soil fertility for particular nutrients. The results of this study

demonstrated the need for site-specific nutrient management as required in precision agriculture.

6. Acknowledgements

The research work was jointly supported by the Ministry of Agriculture, Fishery and Forestry of

Japan though a Research Grant to the International Institute of Tropical Agriculture (IITA), and

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

14

The Nasarawa State University, Keffi, Department of Agronomy, Lafia Campus, with the

commitment of their staff on the collection part of the data used for this study. We acknowledge the

field and laboratory support from Mr Eniola Moses (Soil Microbiology) and Mr Peters Muwiya

(formerly in Soil Microbiology) for their role in processing the soil samples and GPS coordinates

recording during field-work.

7. References

Alexander, M., 1977. Introduction to Soil Microbiology. John Wiley & Sons, Inc New York.

Amusa, N.A., Adegbite, A.A., Muhammed, S., Baiyewu, R.A., 2003. Yam diseases and its

application in predicting surface map of hydraulic parameters in an agricultural farm. African

Journal of Biotechnology 2: 497-502.

Bray, R. H., Kurtz, L. T., 1945. Determination of total, and available forms of phosphorus in soils.

Soil Sci. 59: 39- 45.

Bremner, J.M., Mulvaney, C. S., 1982. Total-Nitrogen. In: Methods of Soil analysis, Part 2. A.L.

Page, R.H. Mille, and D.R. Kecney (eds). American Society of Agronomy Madison, W1,

pp 595-624.

Bridge J, Coyne, D., Kwoseh, C. K., 2005. Nematode Parasites of tropical Root and Tuber

Crops (Excluding potatoes). In: Luc M, Sikora RA and Bridge J (eds) plant parasitic

nematodes in subtropical and tropical agriculture. 2nd Edition, CAB International

Publishing. pp. 221-258.

Burrough, P.A., 1997. Principles of Geographical Information Systems, pp. 132-137. University

Press, Oxford, UK.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

15

Burrough, P.A., McDonnell, R.A., 1998. Principles of Geographical Information Systems. Oxford

University Press, Oxford, UK.

Cambardella, C.A., Moorman, T.B., Nowak, J.M., Parkin, T.B., Karlen, D.L., Turco, R.F., Konopka,

A.E., 1994. Field-scale variability of soil properties in central Iowa soils. Soil Science Society

of America Journal 58:1501-1511.

Carsky, R.J., Wolo, N., Manyong, V.M., Tian, G., 2001. Nutrient balance model for design of

sustainable soil fertility in yam cropping systems. In: Akoroda, M.O. and Ngeve, J.M. (eds)

Tropical Root Crops. Proceedings of the 7th Triennal Symposium of the International

Society for Tropical Root Crops—Africa Branch, Cotonou, Bénin, pp 198-209.

Chien, Y.J., Lee, D.Y., Guo, H.Y., Houng, K.H., 1997. Geostatistical analysis of soil properties of

mid-west Taiwan soils. Soil Science 162, pp. 291–298.

Coyne, D.L., Tchabi, A., Baimey, H., Labuschagne, N, Rotifa, I., 2005. Distribution and prevalence

of nematodes (Scutellonema bradys and Meloidogyne spp.) on marketed yam (Dioscorea spp.) in

West Africa: Field Crops Research 96: 142-150.

Diby, L.N., Hgaza, V.K., Tié, T.B., Assa, A., Carsky, R., Girardin, O., 2011. How does soil fertility

affect yam growth? Acta Agriculturae Scandinavica Section B 61:448-457.

Ersoy, A., Yunsel, T.Y., Cetin, M., 2004. Characterization of land contaminated by heavy metal

mining using geostatistical methods. Archives of Environmental Contamination and

Toxicology 46, 162-175.

FAO data., 2005. http://www.fao.org/docrep/T1696E/t1696e07.htm). Food and Agriculture

Organization Statistic of the United Nation, Rome, Italy.

Flowers, M., Weisz, R., White, J.G., 2005. Yield-based management zones and grid sampling large

land areas. I. Semi-variograms. Soil Science Society of America Journal 46: 1028-1032.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

16

Frossard, E., Bunemann, E.K., Carsky, R., 2007. Integrated nutrient management as a tool to

combat soil degradation in sub-Saharan Africa. In: Bearth, T.B.B,. Kappel, R,. Kruger, G.,

Pfister, R., (eds). Afrika im Wandel. Zuricher, Hochschulforum Band 40, Zurich,

Switzerland; 2007. pp 137-146.

Gallardo, M., 2003. Spatial variability of soil properties in a floodplain forest in northwestern Spain.

Ecosystems 6:564 – 576.

Goovaerts, P., 1997. Geostatistics for Natural Resources Evaluation. Oxford University Press,

Oxford, UK.

Hgaza, V.K., Diby, L.N., Oberson, A., Tschannen, A., Tié, B.T., Sangakkara, U.R., Aké, S., Frossard,

E., 2012. Nitrogen Use by Yam as Affected by Mineral Fertilizer Application, Agronomy

Journal 104:1558-1568.

Ibia, T.O., Udo, E.J., 2009. Guide to fertilizer use for crops in Akwa Ibom State, Nigeria. Sibon

Books Ltd, Festac, Lagos. 104 pp.

Jayeoba, O.J., Sangari, D.U., Alabi, R.T., 2012. Geo-statistical Analyses of the Spatial Variation of

Soil Reaction (pH) in Nasarawa State, North Central Nigeria International Journal of Science

and Advanced Technology (ISSN 2221-8386).

Lark, R.M., 2002. Optimized spatial sampling of soil for estimation of the variogram by maximum

likelihoods. Geoderma 105, pp. 49-80.

Marschner, H., 1995. Mineral nutrition of higher plants, 2nd eds. Academic Press, London, UK.

Miller, M.P., Singer, P.M.J., and Nielsen, D.R., 1988. Spatial variability of wheat yields and soil

properties on complex hills. Soil Science Society of America Journal 52, pp. 1133-1141.

Motomizu, S., Wakimoto, P., Toei, K., 1983. Spectrophotometric determination of phosphate in

river waters with molybdate and malachite green. Analyst (London) 108, 361-367.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

17

Nielsen, D.R., Wendroth, O., 2003. Spatial and temporal statistics: Sampling field soils and

their vegetation. Reiskirchen, Catena Verlag, 398 pp

Patil, S.S., Patil, V.C., Al-Gaadi, K.A., 2011. Spatial Variability in Fertility Status of Surface

predicting soil properties from sample information. Journal of Soil Science 41, 473-490.

Robertson, G.P., Coleman, D.C., Bledsoe, C., Solins, P., 1996. Standard soil method for long-term

ecological research. Oxford University Press, Oxford, UK.

Sanchez, P.A., 2002. Soil fertility and hunger in Africa. Science 295: 2019-2020.

Santra, P., Chropra, V.K., Chakraborty, D., 2008. Spatial variability of soil properties and its

application in predicting surface map of hydraulic parameters in an agricultural farm.

Current Science 95(7): 937-945.

Schlecht, E., Buerkert, A., Tielkes, E., Bationo, A., 2006. A critical analysis of challenges and

opportunities for soil fertility restoration in Sudano-Sahelian West Africa. Nutrient Cycling

in Agroecosystems 76:109-136.

Shukla, M.K., Slater, B.K., Lal, R., Cepuder, P., 2004. Spatial variability of soil properties

and rice yield on recently cleared land. Soil Science Society of America Journal 51, 668-674.

Stein, A. Bouma, J.L., 1997. Methods for comparing spatial variability patterns of millet

yield and soil data. Soil Science Society of America Journal 61, 861-870.

Stoorvogel, J.J., Smaling, E.M.A, Janssen, B.H., 1993. Calculating soil nutrient balances in Africa at

different scales. I. Supra-national scale. Nutrient Cycling in Agroecosystems 35: 227-235.

Suja, G., Sundaresan, S. Susan, K.J., Sreekumar, J. Misra, R.S., 2012. Higher yield, profit and soil

quality from organic farming of elephant foot yam. Agronomy for Sustainable

Development 32:755-764.

Tchabi, A., Coyne, D., Hountondji, F., Lawouin, L., Wiemken, A., Oehl, F., 2010. Efficacy of

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

18

indigenous arbuscular mycorrhizal fungi for promoting white yam (Dioscorea rotundata)

growth in West Africa. Applied Soil Ecology 45: 92-100.

Tchabi, A., Coyne, D., Hountondji, F., Lawouin, L., Wiemken, A., Oehl, F., 2008. Arbuscular

mycorrhizal fungal communities in sub-Saharan savannas of Bénin, West Africa, as affected

by agricultural land use intensity and ecological zone. Mycorrhiza 18, 181-195.

Tittonel, P., Muriuki, A., Klpwijk, C.J., Sherperd, K.D., Coe, R., Vanlauwe, B. 2013. Soil heterogeneity

and soil fertility gradients in smallholder agricultural system of the East African highlands.

Soil Science Society of America Journal. 77: 525 – 538.

Trangmar, B.B., Yost, R.S.,Uehara, G., 1985. Application of geostatistics to spatial studies of soil

properties. Advances in Agronomy 38:45-94.

Trangmar, B.B., Yost, R.S., Wade, M.K., Uehara, G., Sudjadi, M., 1987. Spatial Variation of

Soil Properties and Rice Yield on Recently Cleared Land. Soil Science Society of

America Journal 51: 668-674.

Voltz, M., Webster, R., 1990. A comparison of kriging, cubic splines and classification for

predicting soil properties from sample information. Journal of Soil Science 41, 473-490.

Webster, R. Oliver, M.A. 2001. Statistical methods in Soil and Land resource survey. Oxford

University Press, Oxford, UK.

Yost, R.S., Uehara, G., Fox, R.L., 1982. Geostatistical analysis of soil chemical properties of large

land areas. I. Semi-variograms. Soil Science Society of America Journal 46: 1028-1032.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

19

Legend to Figures

Figure 1. Boundaries of the Republic of Nigeria showing the areas in the yam belt targeted and sampled for the study. Figure 2. Histograms of normal observations (2a,b,c) and the revised (2d,e, f) data from selected soil

fertility parameters analyzed.

Figure 3. Normal Q-Q plot of nitrogen (g N kg-1 soil), of (b) phosphorus (mg P kg-1 soil), and

potassium (meq K/100g soil).

Figure 4. Map of ordinary kriging prediction showing representation of the selected soil fertility

parameters, (a) nitrogen (g N kg-1 soil), available phosphorus (mg P kg-1 soil), and available

potassium (meq K/100g soil) at the landscape level in the yam belt target areas of Nigeria.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

20

Figure 1.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

21

Figure 2.

(a)

(b)

(c)

(d)

(e)

(f)

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

22

Figure 3.

(a)

(b)

(c)

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

23

Figure 4

(a)Nitrogen content

(b) Available phosphorus

(c) Exchangeable Potassium

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

24

Table 1. Selected soil properties across sampling sites conducted in 2012.

pH

(H20)

TN OC

Bray-P

K Ca Mg Na

Ex_acid ECEC SAND SILT CLAY

[g kg-1] [mg P kg-1] [meq/100g soil] [%]

N 385 385 385 385 385 385 385 385 385 385 385 385 385

Mean 6.03 0.88 5.9 4.7 0.75 3.4 1.73 0.45 0.37 6.9 67.32 15.9 16.8

Std. Error

of Mean 0.03 0.03 0.22 0.29 0.04 0.1 0.08 0.02 0.03 0.18 0.77 0.49 0.44

Std.

Deviation 0.54 0.48 4.06 5.4 0.82 1.88 1.4 0.33 0.5 3.27 14.15 9.02 8.07

Variance 0.29 0.24 16.49 29.7 0.67 3.53 1.97 0.11 0.26 10.7 200.16 81.36 65.1

Skewness -

0.42 4.43 0.62 6.81 1.86 1.67 1.89 1.57 1.85 0.56 -1.04 2.64 0.74

Std. Error

of

Skewness

0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13

Kurtosis -

0.47 32.26 1.93 60.3 3.62 5.5 3.64 2.44 4.5 -0.11 1.21 8.21 0.72

Std. Error

of

Kurtosis

0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26

Range 2.6 4.96 22.47 62.0 4.94 15.09 6.78 1.7 2.92 17.4 70.7 56 41

Minimum 4.5 0.04 0.03 1.2 0.03 0.48 0.18 0.08 0 0 18.8 3.2 3

Maximum 7.1 5 22.5 63.1 4.97 15.57 6.96 1.78 2.92 17.4 89.5 59.2 44

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

25

Table 2. Interpretation guide for evaluating analytical data of selected soil fertility parameters.

Total N [g N kg-1]

Bray 1-P [mg P kg-1]

Exch. K+

[meq K/100g soil ]

< 1 Very low < 0.1 Very low

1 – 2 Low < 8 Low 0.1 – 0.3 Low

2 – 5 Medium 8 – 20 Medium 0.3 – 0.7 Moderate

5 – 10 High > 20 High 0.7 – 2 High

>10 Very high > 2 Very high

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

26

Table 3. Parameter values of different model fittings of semi-variogram of selected soil properties analyzed in the yam belt target areas for the study.

Properties

Model No of Lag

Lag size Nugget (Co)

Partial sill (C)

Range (h)/ m

Sill (Co + C)

Ratio Co/(Co + C) (m)

Total Nitrogen [g kg-1]

Rational Quadratic

12 51.9 0.14 0.06 623.2 0.2 0.70

Available Phosphorus [mg kg-1]

Hole Effect

12 150 20.5 4.25 1800 24.73 0.83

Exchangeable Potassium [meq/100 g soil]

Rational Quadratic

15 574.3 0.45 0.56 8,615.7 1.01 0.45

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

27

Table 4. Fertility status, area (km), and percentage of land areas covered by each selected soil fertility parameter at the landscape level in the targeted yam belt of Nigeria.

Fertility Class

N Available-P Exchangeable-K

Area (km2) [%] Area (km2) [%] Area (km2) [%]

Very Low 84,091 90.7

- 0

- 0

Low 8,673 9.3

46,633 50.3

14,234 15.3

Medium - 0

43,450 46.8

12,954 14

High - 0

2,681 2.9

61,366 66.2

Very High - 0

- 0

4,210 4.5

Total 92,764 100 92,764 100 92,764 100

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

28

Spatial mapping of selected soil fertility parameters by ordinary kriging model in yam-based-cropping systems of Nigeria

Author: Mjemo et al. Submitted to Geoderma Regional

Highlight

Soil fertility status of selected parameters determined under yam growing areas of

Nigeria;

Ordinary kriging model used to model selected soil fertility parameters;

Site-specific requirement for N and P while K advised for regional scale application;

Designed soil fertility intervention for yam proposed in the investigated areas.