Research ArticleVariability of Soil Physical Properties in a Clay-Loam Soil andIts Implication on Soil Management Practices
Samuel I Haruna and Nsalambi V Nkongolo
Center of Excellence for Geospatial Information Sciences Department of Agriculture and Environmental Science Lincoln UniversityJefferson City MO 65102-0029 USA
Correspondence should be addressed to Nsalambi V Nkongolo nkongololincolnuedu
Received 26 August 2013 Accepted 17 November 2013
Academic Editors R K Kolka and H O Liechty
Copyright copy 2013 S I Haruna and N V Nkongolo This is an open access article distributed under the Creative CommonsAttribution License which permits unrestricted use distribution and reproduction in any medium provided the original work isproperly cited
We assessed the spatial variability of soil physical properties in a clay-loam soil cropped to corn and soybean The study wasconducted at Lincoln University in Jefferson City Missouri Soil samples were taken at four depths 0ndash10 cm 10ndash20 20ndash40 and40ndash60 cm and were oven dried at 105∘C for 72 hours Bulk density (BDY) volumetric (VWC) and gravimetric (GWC) watercontents volumetric air content (VAC) total pore space (TPS) air-filled (AFPS) and water-filled (WFPS) pore space the relativegas diffusion coefficient (DIFF) and the pore tortuosity factor (TORT) were calculated Results showed that in comparison todepth 1 means for AFPS Diff TPS and VAC decreased in Depth 2 Opposingly BDY Tort VWC andWFPS increased in depth 2Semivariogram analysis showed that GWC VWC BDY and TPS in depth 2 fitted to an exponential variogram model The rangeof spatial variability (119860
0
) for BDY TPS VAC WFPS AFPS DIFF and TORT was the same (2577m) in depths 1 and 4 suggestingthat these soil properties can be sampled together at the same distance The analysis also showed the presence of a strong (le25)to weak (gt75) spatial dependence for soil physical properties
1 Introduction
Characterizing the spatial variability and distribution of soilproperties is important in predicting the rates of ecosys-tem processes with respect to natural and anthropogenicfactors [1] and in understanding how ecosystems and theirservices work [2] In agriculture studies of the effects of landmanagement on soil properties have shown that cultivationgenerally increases the potential for soil degradation due tothe breakdown of soil aggregates and the reduction of soilcohesion water content and nutrient holding capacity [34] Cultivation especially when accompanied by tillage hasbeen reported to have significant effects on topsoil structureand thus the ability of soil to fulfill essential soil functionsand services in relation to root growth gas and watertransport and organic matter turnover [5ndash7] Soil propertiesvary considerably under different crops tillage type andintensity fertilizer types and application rates Consequentlythe physical properties of the soil are also affected by manyfactors that change vertically with depth laterally across fields
and temporally in response to climate and human activity [8]Since this variability affects plant growth nutrient dynamicsand other soil processes knowledge of the spatial variabilityof soil physical properties is therefore necessary To study thespatial distribution of soil properties techniques such as clas-sical statistics and geostatistics have been widely applied [9ndash11] Geostatistics provides the basis for the interpolation andinterpretation of the spatial variability of soil properties [912ndash14] Information on the spatial variability of soil propertiesleads to better management decisions aimed at correctingproblems and at least maintaining productivity and sustain-ability of the soils and thus increasing the precision of farmingpractices [1 15] A better understanding of the spatial variabil-ity of soil properties would enable refining agricultural man-agement practices by identifying sites where remediation andmanagement are needed This promotes sustainable soil andland use and also provides a valuable base against which sub-sequent future measurements can be proposed [14] Despitethe importance of this topic in agriculture the literature isnot abundant on the variability of soil physical properties in
Figure 1 Study area (Lincoln Universityrsquos Freeman farm) showingthe plots
CentralMissouri Furthermore existing studies on the spatialvariability of soil properties have focused on the top soil (0ndash20 cm) with less or no studies at deeper soil depths (30ndash100 cm)The objective of this study was therefore to assess thespatial variability of soil physical properties at various depths(0ndash10 cm 10ndash20 20ndash40 and 40ndash60 cm) in a clay-loam soilcropped to corn and soybean and determine how knowledgeon this variability can affect soil management practices
2 Materials and Methods
21 Experimental Site The study was conducted at LincolnUniversityrsquos Freeman farm in Jefferson City Missouri Thegeographic coordinates of the study site are 38∘58101584011610158401015840Nlatitude and 92∘1010158405310158401015840WlongitudeThe soil of the experimentsite is a Waldron clay-loam (Fine smectitic calcareousmesic Aeric Fluvaquents) The study area is almost flat withan average slope of 2 The experimental field was madeof 48 plots of 1219m width by 2134m length each The48 plots were arranged in a grid of 4 plots in the widthby 12 plots in length as shown in Figure 1 One half ofthe plots was planted to corn (Zea mays) while the otherhalf was planted to soybean (Glycine max) Soybean andcorn plots all received 2631 kgha of nitrogen 6725 kghaof phosphorus and 8967 kgha of potassium Corn plotsreceived 20175 kgha of additional nitrogen in the form ofurea
22 Soil Sampling Soil samples were collected in themiddle of each plot after planting and full seeds emergenceCylindrical cores of 315 cm radius and 10 or 20 cm heightwere used to collect soil samples at four depths 0ndash10 cm10ndash20 20ndash40 and 40ndash60 cm corresponding to depths 1 2 3and 4 respectively The cylinders of 10 cm height were usedfor soil samples collection at depths 1 and 2 while the 20 cmheight cylinders used for sampling at depths 3 and 4 A totalof 576 soil samples were collected as follows 48 plots times 4depths times 3 replicates (at the middle of each plot) Collectedsamples were taken to the laboratory where they wereweighed (fresh weight of sample FWS) then oven dried at105∘C for 72 hrs The weight was taken after oven drying (dryweight of soil DWS) Soil physical properties were calculatedas follows Soil bulk density (BDY gsdotcmminus3) = (DWSV)where DWS is the dry weight of soil and 119881 the volume ofcylinder (total volume of soil) Volumetric water content(VWC cm3sdotcmminus3) = (FWS minus DWS)119881) with FWS being the
fresh weight of soil gravimetric water content (GWC gsdotgminus1)= [(FWS minus DWS)DWS] where FWS is the fresh weight ofsoil total pore space (TPS cm3sdotcmminus3) = 1 minus (BDYPDY)where PDY is the soil particle density (taken as 265 g cmminus3)volumetric air content (VAC cm3sdotcmminus3) = TPS minus VWCwater-filled pore space (WFPS ) = 100 lowast (VWCTPS)air-filled pore space (AFPS ) = 100 lowast (VACTPS) relativegas diffusion coeffient (Diff cm2sminus1sdotcmminus2sdots) = (VAC)2 porespace tortuosity (Tort msdotmminus1) = (1VAC) [16]
23 Statistical andGeospatial Analysis After calculation dataon soil physical properties was first transferred to Statistix90 to compute summaries of simple statistics then to GS+(Geostatistics for environmental science) 70 for semivari-ogram analysis A semivariogram (a measure of the strengthof statistical correlation as a function of distance) is definedby the following equation [17]
120574 (ℎ) =1
2119898 (ℎ)
119898(ℎ)
sum
119894=1
[119911 (119909119894+ ℎ) minus 119911 (119909
119894)]2
(1)
where 120574(ℎ) is the experimental semivariogram value at adistance interval ℎ 119898(ℎ) is number of sample value pairswithin the distance interval ℎ and 119885(119883
119894) and 119885(119883
119894+ ℎ)
are sample values at two points separated by the distance ℎExponential and spherical models were the empirical semi-variograms The stationary models that is exponential (2)and spherical model (3) that fitted to experimental semivari-ograms were defined in the following equations [18]
120574 (ℎ) = 1198620+ 1198621[1 minus expminus(ℎ
119886)] (2)
120574 (ℎ) = 1198620+ 1198621[(
3ℎ
2119886) minus (
ℎ3
21198863)] when ℎ le 119886
= 1198620+ 1198621 when ℎ ge 119886
(3)
where 1198620is the nugget 119862
1is the partial sill and 119886 is the
range of spatial dependence to reach the sill (1198620+ 1198621)
The ratio 1198620(1198620+ 1198621) and the range are the parameters
that characterize the spatial structure of a soil property The1198620(1198620+ 1198621) relation is the proportion in the dependence
zone and the range defines the distance over which the soilproperty values are correlated with each other [19] A lowvalue for the 119862
0(1198620+ 1198621) ratio and a high range generally
indicate that high precision of the property can be obtained byParfitt et al [19] The classification proposed by Cambardellaet al [14] which considers the degree of spatial dependence(DSD = 119862
0(1198620+ 1198621) times 100) as strong when DSD le 25
moderate when 25 lt DSD le 75 and weak when DSD gt
75 was used in this study to classify the degree of spatialdependence of each soil property
3 Results and Discussion
31 Summaries of Statistics for Soil Physical Properties Over-all descriptive statistics for soil properties in this studyshowedmoderate to high skewness for some of the properties
ISRN Soil Science 3
Table 1 Descriptive statistics for soil physical properties at four depths in a clay-loam soil
AFPS air-filled pore space () BDY soil bulk density (gcmminus3) DIFF relative gas diffusion coefficient (m2 sminus1mminus2 s) GWC gravimetric water content of soil(gsdotgminus1) TPS total pore spaces (cm3 cmminus3) TORT Pore tortuosity factor (msdotmminus1) VAC volumetric air content (cm3 cmminus3) VWC volumetric water content(cm3 cmminus3) WFPS water-filled pore space ()
(Table 1) The highly skewed soil parameters included soilbulk density (BDY) diffusivity (DIFF) and volumetric watercontent (VWC) whereas total pore space (TPS) was moder-ately skewed Air-filled pore space (AFPS) had a low skew-ness Highly skewed parameters indicate that these propertieshave a local distribution that is high values were found forthese properties at some points but most values were low[20] The other soil physical properties were approximatelynormally distributed on the field The underlying reason forsoil properties being normally or nonnormally distributedmay be associated with differences in management practicesland use vegetation cover and topographic effects on thevariability of soil erosion across the landscape of the fieldThese factors can be the sources for a large or very smallvariation of soil properties in some of the samples whichleads to the nonnormal distribution [21] A wide range ofspatial variability was observed for soil physical properties(Table 1) For instance soil bulk density (BDY) ranged from101 to 123 g cmminus3 for depth 1 115 to 146 g cmminus3 for depth 2096 to 119 g cmminus3 and 104 to 118 g cmminus3 for depths 3 and 4respectively (Figure 2) Soil bulk densitywas also significantlyhigher in the second depth (14 g cmminus3) than all the other 3depths where it varied between 118 g cmminus3 and 124 g cmminus3Themean value of AFPSwas significantly lower in the seconddepth (265 cm3sdotcmminus3) than in all other 3 depths whereit varied from 3934 to 457 cm3sdotcmminus3 Soil pore tortuosityfactor (TORT) and water-filled pore space (WFPS) werealso significantly higher in the second depth (1246 cmsdotcmminus1and 7346 resp) However the relative gas diffusion
coefficient (DIFF) gravimetric water content (GWC) totalpore space (TPS) and volumetric air content (VAC) weresignificantly lower in the second depth (002m2sminus1mminus2s021 gsdotgminus1 042 cm3sdotcmminus3 and 012 cm3sdotcmminus3 resp) (Table 1)The variability in soil physical properties is understandablesince the soil of this site has a smectite layer (claypan) inthe 10ndash20 cm which corresponded to our second samplingdepthThis layer of smectite is hard and compact with verylow pore space high mass-volume ratio (bulk density) andhigh water retention capability (because of their large surfacearea) As a consequence of the presence of this smectite layerin depth 2 the mean of water-filled pore space (WFPS) wasslightly lower in the first depth (54) than in all four depthsIn fact air predominates the pore space in the first depthand cultivation loosened the soil thereby allowing the watertrapped in the pore space to evaporate Higher GWC VWCand TPS at the lower depths (20ndash60 cm) mean that crops(especially corn and soybean grown in the field) were able toaccess water and dissolved nutrients through their roots Infact despite the claypan layer (10ndash20 cm) it has been reportedby various researchers that crop roots were able to penetrateinto and through this layer of smectitic clay [22ndash24] and thatroot growth may increase within the claypan layer [23] asa result of plant adaptation to water-limited soil layers Ingeneral the use of the coefficient of variation (CV) is a com-mon procedure to assess variability in soil properties since itallows comparison among properties with different units ofmeasurement Overall the coefficient of variation for all soilphysical properties in the four depths ranged from 483 to
4 ISRN Soil Science
0
05
1
15
2
D1 D2 D3 D4Depth
Soil
bulk
den
sity
(gmiddotcm
minus3)
(a)
0
01
02
03
04
D1 D2 D3 D4
Gra
vim
etric
wat
erco
nten
t (g
g)
Depth
(b)
0
01
02
03
04
05
06
07
D1 D2 D3 D4
Tota
l por
e spa
ce (c
m3
cmminus3)
Depth
(c)
0
005
01
015
02
025
03
035
04
D1 D2 D3 D4Depth
Volu
met
ric w
ater
cont
ent
(cm
3cm
minus3)
(d)
Figure 2 Variation of soil bulk density gravimetric water content total pore space and volumetric water content with depth
9161 (Table 1) The pore tortuosity factor (TORT) showedthe highest variation while soil bulk density (BDY) showedthe least variation The CV indicated that there was a strongspatial variability of the soil properties investigated Howeverto have a better assessment of such spatial variability acrossthe entire field a geostatistical analysis was used
32 Spatial Variability of Soil Properties After computingsummaries of simple statistics with Statistix 90 data onsoil physical properties was transferred to GS+ (geostatisticsfor environmental science) 70 for semivariogram analysisSemivariogrammodel fitwas determined from the coefficientof determination (1198772) values which range from 0 (very poormodel fit) to 1 (very good model fit) Table 2 shows soilphysical properties which mainly responded to exponentialand linear variogram models with the exponential modelproviding the best fit In the 10ndash20 cm depth exponentialmodel provided the best fit for BDY (119877
2
= 093) withthe spherical model providing very poor model fit Poretortuosity also responded to an exponential variogrammodelin the 20ndash40 cm depth (1198772 = 057) although spherical modelwas noticed Linear and exponential models were observed
in the 40ndash60 cm depth for TPS (1198772 = 046) with linearmodel providing a better fit (Table 2) In general for alldepths model fit was not very strong with the exception ofgravimetric water content and bulk density in the seconddepth Overall the exponential model provided the best fitwith about 65 of the physical properties fitting this modelIn geostatistical theory the range of the spatial variabilityof the semivariogram is the distance between correlatedmeasurements (the minimum lateral distance between twopoints before the change in property is noticed) and can be aneffective criterion for the evaluation of sampling design andmapping of soil propertiesThe value that the semivariogrammodel attains at the range (the value on the 119910-axis) is calledthe sill The partial sill is the sill minus the nugget [2526] Theoretically at zero separation distance (lag = 0) thesemivariogram value is zero However at an infinitesimallysmall separation distance the semivariogram often exhibitsa nugget effect (the apparent discontinuity at the beginningof many semivariogram graphs) which is some value greaterthan zeroThenugget effect can be attributed tomeasurementerrors or spatial sources of variation at distances smallerthan the sampling interval (or both) Measurement erroroccurs because of the error inherent inmeasuring devices To
ISRN Soil Science 5
Table 2 Variogram parameters for soil physical properties at four depths in a clay-loam soil
DSD degree of spatial dependence strong DSD (DSD le 25) moderate DSD (25 lt DSD le 75) and weak DSD (DSD gt 75) according to Cambardella etal (1994) [14]
6 ISRN Soil Science
eliminate this error multiple samples were taken from eachsampling point Natural phenomena can vary spatially over arange of scales Variation atmicroscales smaller than the sam-pling distances will appear as a part of the nugget effect Table2 shows that the spatial correlation (range) of soil propertieswidely varied from 1m for volumetric water content (VWC)in depth four to 64m for gravimetric water content (GWC)in depth 2 However for the first and second depth (whichare agriculturally more important) the range of spatialcorrelation varied from 3m for volumetric air content (VAC)in depth 2 to 64m for GWC in depth 2 Beyond these rangesthere is no spatial dependence (autocorrelation) The spatialdependence can indicate the level of similarity or disturbanceof the soil condition According to Lopez-Granados et al [27]and Ayoubi et al [17] a large range indicates that the mea-sured soil property value is influenced by natural and anthro-pogenic factors over greater distances than parameters whichhave smaller ranges Thus a range of about 64m for GWCin this study indicates that the measured GWC values canbe influenced in the soil over greater distances as comparedto the soil parameters having smaller range (Table 2) Thismeans that soil variables with smaller range such asVWCandVACare good indicators of themore disturbed soils (themoredisturbed a soil is the more variable some soil propertiesbecome) The more variable properties have a shorter rangeof correlation The different ranges of the spatial dependenceamong the soil properties may be attributed to differencesin response to the erosionmdashdeposition factors land use-cover parent material and human interferences in the studyarea The nugget which is an indication of microvariabilitywas significantly higher for water-filled pore space (WFPS)and air-filled pore space (AFPS) when compared to theothersThis can be explained by our sampling distance whichcould not capture well their spatial dependence The lowestnugget was for GWC (Table 2) This indicates that GWChad low spatial variability within small distances Knowledgeof the range of influence for various soil properties allowsone to construct independent accurate datasets for similarareas in future soil sampling design to perform statisticalanalysis [17] This aids in determining where to resampleif necessary and design future field experiments that avoidspatial dependence Therefore for future studies aimed atcharacterizing the spatial dependency of soil properties inthe study area andor a similar area it is recommendedthat the soil properties be sampled at distances shorterthan the range found in this study Cambardella et al [14]established the classification of degree of spatial dependence(DSD) between adjacent observations of soil property gt 75to correspond to weak spatial structure In this study thesemivariograms indicated strong spatial dependence (DSDle 25) for soil physical properties such as bulk densitygravimetric water content volumetric water content totalpore space and diffusivity The rest of the soil physicalproperties (water-filled pore space Air-filled pore space andtortuosity)measured exhibited very weak spatial dependence(DSD gt 75) (Table 2) The strong spatial dependence of thesoil properties may be controlled by intrinsic variations insoil characteristics such as texture and mineralogy whereasextrinsic variations such as tillage and other soil and water
management practices may also control the variability of theweak spatially dependent parameters [14]
33 Spatial Distribution of Soil Properties across the FieldInterpolated maps portraying the distribution of soil physicalproperties in various depths are shown in Figure 3 for soilgravimetric (GWC) and volumetric (VWC) contents andwater-filled pore space (WFPS) Gravimetric water contentshowed a good spatial distribution across the field withthe highest values located around the southwestern portionof the field Volumetric water content also showed goodspatial distribution across the field with high values locatedin the northern central and southwestern portions of thefield Water-filled pore has a distribution similar to that ofvolumetric water contentThe other soil properties howevershowed very poor spatial distribution in the field This ismost probably due to their poor sill (119862
0+ 119862) model fit and
coefficient of determination (1198772
) Even though the spatialvariability was not very pronounced there were areas onthe field that had slightly higher values of these physicalproperties than the rest of the field In general bulk densitytotal pore space volumetric air content Air-filled pore spacediffusivity and tortuosity were very high in the field eventhough they did not exhibit very distinguishable variabilityThis lack of visible spatial variability is supported because thesampling distance (range) is 26m for these properties
34 Implications of Spatial Variability of Soil Physical Prop-erties on Soil Management Results of this study indicatedthat the spatial variability of soil water content (GWC andVWC) was high This can be explained among many otherreasons by soil type (clay-loam) which was able to hold morewater But with intensive tillage this soil water content couldbe adversely affected Studies have shown that tillage prac-tices can alter soil physical properties and consequently thehydrological behavior of agricultural fields especially whena similar tillage system has been practiced for a long period[15 28ndash31] Tillage intensity has also considerable effectson spatial structure and spatial variability of soil properties[15 30]Therefore this study can help determine site-specificsoil management and decision making To do so the spatialvariability of soil properties developed through kriging willbe an important tool Different ranges of spatial dependencewere noticed in the field The different ranges of the spatialdependence among the soil properties may be attributed todifferences in response to the erosionmdashdeposition factorsland use-cover parent material and human interferences inthe study areaThe different ranges can also be used in futurestudies to determine the sampling distance of different soilphysical properties on the field Also the sill (119862
0+119862) can help
determine where the variability or change in soil propertystops This will be useful especially for the irrigation pur-poses Generally with farmers facing the decision of whetheror not to till and the intensity of tillage a spatial variabilitystudy can help in this decision making Maps producedin this study can also be used for irrigation purposes asthey can clearly indicate which portion of the field needsirrigation (soil water content) To do this soil water content
Figure 3 Spatial distribution of gravimetric (GWC) and volumetric (VWC) water contents in 0ndash10 cm depth and that of water-filled porespace (WFPS) in 20 cm depth
information can be collected and analyzed geospatially toproduce field maps The process can be repeated frequentlyto obtain up-to-date soil water content information To avoidfrequent destructive sampling for water content analysisequipment that allows insitu measurements such as TDRmethods and water mark sensors can be used Since a differ-ent range of spatial dependence among soil properties showsdifferences in response to human interferences and land use-cover this will help reduce human activities that increase soilbulk density and cause soil compaction like the use of heavyequipment It can also serve as a reference for the type of cropto be grown (cover crops for erosion susceptible areas)
4 Conclusion
We assessed the spatial variability of soil physical propertiesin a clay-loam soil cropped to corn and soybean Resultsshowed that soil physical properties either decreased orincreased sharply in the second depth (due to the presence ofa smectite layer) before leveling up or dropping off but with-out reaching the first depth value in either case In additiondepending on soil physical property maps produced by krig-ing showed either goodor poor spatial distributionThe semi-variogram analysis showed the presence of a strong (le25)to weak (gt75) spatial dependence of soil properties Ourunderstanding of the behavior of soil properties in this studyprovides new insights for soil site-specific management in
addressing issues such as ldquowhere to place the proper interven-tionsrdquo (tillage irrigation and crop type to be grown)
Acknowledgment
This research is part of a regional collaborative projectsupported by the USDA-NIFA Award no 2011-68002-30190Cropping Systems Coordinated Agricultural Project ClimateChangeMitigation andAdaptation inCorn-basedCroppingSystems (Project website httpsustainablecornorg)
References
[1] D Schimel J Melillo H Tian et al ldquoContribution of increasingCO2
and climate to carbon storage by ecosystems in the UnitedStatesrdquo Science vol 287 no 5460 pp 2004ndash2006 2000
[2] C Kosmas S Gerontidis and M Marathianou ldquoThe effect ofland use change on soils and vegetation over various lithologicalformations on Lesvos (Greece)rdquoCatena vol 40 no 1 pp 51ndash682000
[3] J Igbal J A Thomasson J N Jenkins P R Owens and F DWhisler ldquoSpatial variability analysis of soil physical propertiesof alluvial soilsrdquo Soil Science Society of America Journal vol 69pp 1ndash14 2005
[4] S Zhang X Zhang T Huffman X Liu and J Yang ldquoInfluenceof topography and land management on soil nutrients variabil-ity in Northeast ChinardquoNutrient Cycling in Agroecosystems vol89 no 3 pp 427ndash438 2011
8 ISRN Soil Science
[5] L J Munkholm A Garbout and S B Hansen ldquoTillage effectson topsoil structural quality assessed using X-ray CT soil coresand visual soil evaluationrdquo Soil amp Tillage Research vol 128 pp104ndash109 2013
[6] A J Franzluebbers ldquoSoil organicmatter stratification ratio as anindicator of soil qualityrdquo Soil and Tillage Research vol 66 no 2pp 95ndash106 2002
[7] L J Munkholm P Schjoslashnning K J Rasmussen and KTanderup ldquoSpatial and temporal effects of direct drilling onsoil structure in the seedling environmentrdquo Soil and TillageResearch vol 71 no 2 pp 163ndash173 2003
[8] A Swarowsky R A Dahlgren K W Tate J W Hopmans andA T OrsquoGeen ldquoCatchment-scale soil water dynamics in a medi-terranean-type oak woodlandrdquoVadose Zone Journal vol 10 no3 pp 800ndash815 2011
[9] RWebster andMAOliver StatisticalMethods in Soil and LandResource Survey Oxford University Press Oxford UK 1990
[10] A Saldana A Stein and J A Zinck ldquoSpatial variability of soilproperties at different scaleswithin three terraces of theHenaresriver (Spain)rdquo Catena vol 33 no 3-4 pp 139ndash153 1998
[11] Z M Wang K S Song B Zhang et al ldquoSpatial variabilityand affecting factors of soil nutrients in croplands of NortheastChina a case study in Dehui countyrdquo Plant Soil and Environ-ment vol 55 no 3 pp 110ndash120 2009
[12] R Webster ldquoQuantitative spatial analysis of soil in the fieldrdquoAdvances in Soil Science vol 3 pp 1ndash70 1985
[13] H Pohlmann ldquoGeostatisticalmodelling of environmental datardquoCatena vol 20 no 1-2 pp 191ndash198 1993
[14] C A Cambardella T B Moorman J M Novak et al ldquoField-scale variability of soil properties in central Iowa soilsrdquo Soil Sci-ence Society of America Journal vol 58 no 5 pp 1501ndash1511 1994
[15] E Ozgoz ldquoLong term conventional tillage effect on spatial vari-ability of some soil physical propertiesrdquo Journal of SustainableAgriculture vol 33 no 2 pp 142ndash160 2009
[16] N V Nkongolo R Hatano and V Kakembo ldquoDiffusivity mod-els and greenhouse gases fluxes from a forest pasture grasslandand corn field in Northern Hokkaido Japanrdquo Pedosphere vol20 no 6 pp 747ndash760 2010
[17] S H Ayoubi S M Zamani and F Khomali ldquoSpatial variabilityof some soil properties for site-specific farming in northernIranrdquo International Journal of Plant Production vol 2 pp 225ndash236 2007
[18] T M Burgess and R Webster ldquoOptimal interpolation andisarithmic mapping of soil properties I The semi-variogramand punctual krigingrdquo Journal of Soil amp Science vol 31 no 2pp 315ndash331 1980
[19] J M B Parfitt L C Timm E A Pauletto et al ldquoSpatial vari-ability of the chemical physical and biological properties inlowland cultivated with irrigated ricerdquo Revista Brasileira deCiencia do Solo vol 33 no 4 pp 819ndash830 2009
[20] C R Grego S R Vieira and A L Lourencao ldquoSpatial distri-bution of Pseudaletia sequax Franclemlont in triticale under no-till managementrdquo Scientia Agricola vol 63 no 4 pp 321ndash3272006
[21] G B Tesfahunegn L Tamene and P L G Vlek ldquoCatchment-scale spatial variability of soil properties and implications onsite-specific soil management in northern Ethiopiardquo Soil andTillage Research vol 117 pp 124ndash139 2011
[22] S J Grecu M B Kirkham E T Kanemasu D W Sweeney LR Stone and G A Milliken ldquoRoot growth in a claypan witha perennial-annual rotationrdquo Soil Science Society of AmericaJournal vol 52 no 2 pp 488ndash494 1988
[23] D B Myers N R Kitchen K A Sudduth R E Sharp andR J Miles ldquoSoybean root distribution related to claypan soilproperties and apparent soil electrical conductivityrdquo Crop Sci-ence vol 47 no 4 pp 1498ndash1509 2007
[24] P Jiang N R Kitchen S H Anderson E J Sadler and KA Sudduth ldquoEstimating palnt available water using the simpleinverse yield model for claypan landscapesrdquo Agronomy Journalvol 100 pp 1ndash7 2008
[25] A Utset M E Ruiz J Herrera and D P De Leon ldquoA geo-statistical method for soil salinity sample site spacingrdquo Geo-derma vol 86 no 1-2 pp 143ndash151 1998
[26] W Fu H Tunney and C Zhang ldquoSpatial variation of soilnutrients in a dairy farm and its implications for site-specificfertilizer applicationrdquo Soil and Tillage Research vol 106 no 2pp 185ndash193 2010
[27] F Lopez-Granados M Jurado-Exposito S Atenciano AGarcıa-Ferrer M Sanchez De La Orden and L Garcıa-TorresldquoSpatial variability of agricultural soil parameters in southernSpainrdquo Plant and Soil vol 246 no 1 pp 97ndash105 2002
[28] R L Hill ldquoLong-term conventional and no-tillage effects onselected soil physical propertiesrdquo Soil Science Society of AmericaJournal vol 54 no 1 pp 161ndash166 1990
[29] D E Buschiazzo J L Panigatti and PW Unger ldquoTillage effectson soil properties and crop production in the subhumid andsemiarid Argentinean Pampasrdquo Soil and Tillage Research vol49 no 1-2 pp 105ndash116 1998
[30] T Tsegaye and R L Hill ldquoIntensive tillage effects on spatialvariability of soil physical propertiesrdquo Soil Science vol 163 no2 pp 143ndash154 1998
[31] J A Gomez J V Giraldez M Pastor and E Fereres ldquoEffects oftillage method on soil physical properties infiltration and yieldin an olive orchardrdquo Soil and Tillage Research vol 52 no 3-4pp 167ndash175 1999
Figure 1 Study area (Lincoln Universityrsquos Freeman farm) showingthe plots
CentralMissouri Furthermore existing studies on the spatialvariability of soil properties have focused on the top soil (0ndash20 cm) with less or no studies at deeper soil depths (30ndash100 cm)The objective of this study was therefore to assess thespatial variability of soil physical properties at various depths(0ndash10 cm 10ndash20 20ndash40 and 40ndash60 cm) in a clay-loam soilcropped to corn and soybean and determine how knowledgeon this variability can affect soil management practices
2 Materials and Methods
21 Experimental Site The study was conducted at LincolnUniversityrsquos Freeman farm in Jefferson City Missouri Thegeographic coordinates of the study site are 38∘58101584011610158401015840Nlatitude and 92∘1010158405310158401015840WlongitudeThe soil of the experimentsite is a Waldron clay-loam (Fine smectitic calcareousmesic Aeric Fluvaquents) The study area is almost flat withan average slope of 2 The experimental field was madeof 48 plots of 1219m width by 2134m length each The48 plots were arranged in a grid of 4 plots in the widthby 12 plots in length as shown in Figure 1 One half ofthe plots was planted to corn (Zea mays) while the otherhalf was planted to soybean (Glycine max) Soybean andcorn plots all received 2631 kgha of nitrogen 6725 kghaof phosphorus and 8967 kgha of potassium Corn plotsreceived 20175 kgha of additional nitrogen in the form ofurea
22 Soil Sampling Soil samples were collected in themiddle of each plot after planting and full seeds emergenceCylindrical cores of 315 cm radius and 10 or 20 cm heightwere used to collect soil samples at four depths 0ndash10 cm10ndash20 20ndash40 and 40ndash60 cm corresponding to depths 1 2 3and 4 respectively The cylinders of 10 cm height were usedfor soil samples collection at depths 1 and 2 while the 20 cmheight cylinders used for sampling at depths 3 and 4 A totalof 576 soil samples were collected as follows 48 plots times 4depths times 3 replicates (at the middle of each plot) Collectedsamples were taken to the laboratory where they wereweighed (fresh weight of sample FWS) then oven dried at105∘C for 72 hrs The weight was taken after oven drying (dryweight of soil DWS) Soil physical properties were calculatedas follows Soil bulk density (BDY gsdotcmminus3) = (DWSV)where DWS is the dry weight of soil and 119881 the volume ofcylinder (total volume of soil) Volumetric water content(VWC cm3sdotcmminus3) = (FWS minus DWS)119881) with FWS being the
fresh weight of soil gravimetric water content (GWC gsdotgminus1)= [(FWS minus DWS)DWS] where FWS is the fresh weight ofsoil total pore space (TPS cm3sdotcmminus3) = 1 minus (BDYPDY)where PDY is the soil particle density (taken as 265 g cmminus3)volumetric air content (VAC cm3sdotcmminus3) = TPS minus VWCwater-filled pore space (WFPS ) = 100 lowast (VWCTPS)air-filled pore space (AFPS ) = 100 lowast (VACTPS) relativegas diffusion coeffient (Diff cm2sminus1sdotcmminus2sdots) = (VAC)2 porespace tortuosity (Tort msdotmminus1) = (1VAC) [16]
23 Statistical andGeospatial Analysis After calculation dataon soil physical properties was first transferred to Statistix90 to compute summaries of simple statistics then to GS+(Geostatistics for environmental science) 70 for semivari-ogram analysis A semivariogram (a measure of the strengthof statistical correlation as a function of distance) is definedby the following equation [17]
120574 (ℎ) =1
2119898 (ℎ)
119898(ℎ)
sum
119894=1
[119911 (119909119894+ ℎ) minus 119911 (119909
119894)]2
(1)
where 120574(ℎ) is the experimental semivariogram value at adistance interval ℎ 119898(ℎ) is number of sample value pairswithin the distance interval ℎ and 119885(119883
119894) and 119885(119883
119894+ ℎ)
are sample values at two points separated by the distance ℎExponential and spherical models were the empirical semi-variograms The stationary models that is exponential (2)and spherical model (3) that fitted to experimental semivari-ograms were defined in the following equations [18]
120574 (ℎ) = 1198620+ 1198621[1 minus expminus(ℎ
119886)] (2)
120574 (ℎ) = 1198620+ 1198621[(
3ℎ
2119886) minus (
ℎ3
21198863)] when ℎ le 119886
= 1198620+ 1198621 when ℎ ge 119886
(3)
where 1198620is the nugget 119862
1is the partial sill and 119886 is the
range of spatial dependence to reach the sill (1198620+ 1198621)
The ratio 1198620(1198620+ 1198621) and the range are the parameters
that characterize the spatial structure of a soil property The1198620(1198620+ 1198621) relation is the proportion in the dependence
zone and the range defines the distance over which the soilproperty values are correlated with each other [19] A lowvalue for the 119862
0(1198620+ 1198621) ratio and a high range generally
indicate that high precision of the property can be obtained byParfitt et al [19] The classification proposed by Cambardellaet al [14] which considers the degree of spatial dependence(DSD = 119862
0(1198620+ 1198621) times 100) as strong when DSD le 25
moderate when 25 lt DSD le 75 and weak when DSD gt
75 was used in this study to classify the degree of spatialdependence of each soil property
3 Results and Discussion
31 Summaries of Statistics for Soil Physical Properties Over-all descriptive statistics for soil properties in this studyshowedmoderate to high skewness for some of the properties
ISRN Soil Science 3
Table 1 Descriptive statistics for soil physical properties at four depths in a clay-loam soil
AFPS air-filled pore space () BDY soil bulk density (gcmminus3) DIFF relative gas diffusion coefficient (m2 sminus1mminus2 s) GWC gravimetric water content of soil(gsdotgminus1) TPS total pore spaces (cm3 cmminus3) TORT Pore tortuosity factor (msdotmminus1) VAC volumetric air content (cm3 cmminus3) VWC volumetric water content(cm3 cmminus3) WFPS water-filled pore space ()
(Table 1) The highly skewed soil parameters included soilbulk density (BDY) diffusivity (DIFF) and volumetric watercontent (VWC) whereas total pore space (TPS) was moder-ately skewed Air-filled pore space (AFPS) had a low skew-ness Highly skewed parameters indicate that these propertieshave a local distribution that is high values were found forthese properties at some points but most values were low[20] The other soil physical properties were approximatelynormally distributed on the field The underlying reason forsoil properties being normally or nonnormally distributedmay be associated with differences in management practicesland use vegetation cover and topographic effects on thevariability of soil erosion across the landscape of the fieldThese factors can be the sources for a large or very smallvariation of soil properties in some of the samples whichleads to the nonnormal distribution [21] A wide range ofspatial variability was observed for soil physical properties(Table 1) For instance soil bulk density (BDY) ranged from101 to 123 g cmminus3 for depth 1 115 to 146 g cmminus3 for depth 2096 to 119 g cmminus3 and 104 to 118 g cmminus3 for depths 3 and 4respectively (Figure 2) Soil bulk densitywas also significantlyhigher in the second depth (14 g cmminus3) than all the other 3depths where it varied between 118 g cmminus3 and 124 g cmminus3Themean value of AFPSwas significantly lower in the seconddepth (265 cm3sdotcmminus3) than in all other 3 depths whereit varied from 3934 to 457 cm3sdotcmminus3 Soil pore tortuosityfactor (TORT) and water-filled pore space (WFPS) werealso significantly higher in the second depth (1246 cmsdotcmminus1and 7346 resp) However the relative gas diffusion
coefficient (DIFF) gravimetric water content (GWC) totalpore space (TPS) and volumetric air content (VAC) weresignificantly lower in the second depth (002m2sminus1mminus2s021 gsdotgminus1 042 cm3sdotcmminus3 and 012 cm3sdotcmminus3 resp) (Table 1)The variability in soil physical properties is understandablesince the soil of this site has a smectite layer (claypan) inthe 10ndash20 cm which corresponded to our second samplingdepthThis layer of smectite is hard and compact with verylow pore space high mass-volume ratio (bulk density) andhigh water retention capability (because of their large surfacearea) As a consequence of the presence of this smectite layerin depth 2 the mean of water-filled pore space (WFPS) wasslightly lower in the first depth (54) than in all four depthsIn fact air predominates the pore space in the first depthand cultivation loosened the soil thereby allowing the watertrapped in the pore space to evaporate Higher GWC VWCand TPS at the lower depths (20ndash60 cm) mean that crops(especially corn and soybean grown in the field) were able toaccess water and dissolved nutrients through their roots Infact despite the claypan layer (10ndash20 cm) it has been reportedby various researchers that crop roots were able to penetrateinto and through this layer of smectitic clay [22ndash24] and thatroot growth may increase within the claypan layer [23] asa result of plant adaptation to water-limited soil layers Ingeneral the use of the coefficient of variation (CV) is a com-mon procedure to assess variability in soil properties since itallows comparison among properties with different units ofmeasurement Overall the coefficient of variation for all soilphysical properties in the four depths ranged from 483 to
4 ISRN Soil Science
0
05
1
15
2
D1 D2 D3 D4Depth
Soil
bulk
den
sity
(gmiddotcm
minus3)
(a)
0
01
02
03
04
D1 D2 D3 D4
Gra
vim
etric
wat
erco
nten
t (g
g)
Depth
(b)
0
01
02
03
04
05
06
07
D1 D2 D3 D4
Tota
l por
e spa
ce (c
m3
cmminus3)
Depth
(c)
0
005
01
015
02
025
03
035
04
D1 D2 D3 D4Depth
Volu
met
ric w
ater
cont
ent
(cm
3cm
minus3)
(d)
Figure 2 Variation of soil bulk density gravimetric water content total pore space and volumetric water content with depth
9161 (Table 1) The pore tortuosity factor (TORT) showedthe highest variation while soil bulk density (BDY) showedthe least variation The CV indicated that there was a strongspatial variability of the soil properties investigated Howeverto have a better assessment of such spatial variability acrossthe entire field a geostatistical analysis was used
32 Spatial Variability of Soil Properties After computingsummaries of simple statistics with Statistix 90 data onsoil physical properties was transferred to GS+ (geostatisticsfor environmental science) 70 for semivariogram analysisSemivariogrammodel fitwas determined from the coefficientof determination (1198772) values which range from 0 (very poormodel fit) to 1 (very good model fit) Table 2 shows soilphysical properties which mainly responded to exponentialand linear variogram models with the exponential modelproviding the best fit In the 10ndash20 cm depth exponentialmodel provided the best fit for BDY (119877
2
= 093) withthe spherical model providing very poor model fit Poretortuosity also responded to an exponential variogrammodelin the 20ndash40 cm depth (1198772 = 057) although spherical modelwas noticed Linear and exponential models were observed
in the 40ndash60 cm depth for TPS (1198772 = 046) with linearmodel providing a better fit (Table 2) In general for alldepths model fit was not very strong with the exception ofgravimetric water content and bulk density in the seconddepth Overall the exponential model provided the best fitwith about 65 of the physical properties fitting this modelIn geostatistical theory the range of the spatial variabilityof the semivariogram is the distance between correlatedmeasurements (the minimum lateral distance between twopoints before the change in property is noticed) and can be aneffective criterion for the evaluation of sampling design andmapping of soil propertiesThe value that the semivariogrammodel attains at the range (the value on the 119910-axis) is calledthe sill The partial sill is the sill minus the nugget [2526] Theoretically at zero separation distance (lag = 0) thesemivariogram value is zero However at an infinitesimallysmall separation distance the semivariogram often exhibitsa nugget effect (the apparent discontinuity at the beginningof many semivariogram graphs) which is some value greaterthan zeroThenugget effect can be attributed tomeasurementerrors or spatial sources of variation at distances smallerthan the sampling interval (or both) Measurement erroroccurs because of the error inherent inmeasuring devices To
ISRN Soil Science 5
Table 2 Variogram parameters for soil physical properties at four depths in a clay-loam soil
DSD degree of spatial dependence strong DSD (DSD le 25) moderate DSD (25 lt DSD le 75) and weak DSD (DSD gt 75) according to Cambardella etal (1994) [14]
6 ISRN Soil Science
eliminate this error multiple samples were taken from eachsampling point Natural phenomena can vary spatially over arange of scales Variation atmicroscales smaller than the sam-pling distances will appear as a part of the nugget effect Table2 shows that the spatial correlation (range) of soil propertieswidely varied from 1m for volumetric water content (VWC)in depth four to 64m for gravimetric water content (GWC)in depth 2 However for the first and second depth (whichare agriculturally more important) the range of spatialcorrelation varied from 3m for volumetric air content (VAC)in depth 2 to 64m for GWC in depth 2 Beyond these rangesthere is no spatial dependence (autocorrelation) The spatialdependence can indicate the level of similarity or disturbanceof the soil condition According to Lopez-Granados et al [27]and Ayoubi et al [17] a large range indicates that the mea-sured soil property value is influenced by natural and anthro-pogenic factors over greater distances than parameters whichhave smaller ranges Thus a range of about 64m for GWCin this study indicates that the measured GWC values canbe influenced in the soil over greater distances as comparedto the soil parameters having smaller range (Table 2) Thismeans that soil variables with smaller range such asVWCandVACare good indicators of themore disturbed soils (themoredisturbed a soil is the more variable some soil propertiesbecome) The more variable properties have a shorter rangeof correlation The different ranges of the spatial dependenceamong the soil properties may be attributed to differencesin response to the erosionmdashdeposition factors land use-cover parent material and human interferences in the studyarea The nugget which is an indication of microvariabilitywas significantly higher for water-filled pore space (WFPS)and air-filled pore space (AFPS) when compared to theothersThis can be explained by our sampling distance whichcould not capture well their spatial dependence The lowestnugget was for GWC (Table 2) This indicates that GWChad low spatial variability within small distances Knowledgeof the range of influence for various soil properties allowsone to construct independent accurate datasets for similarareas in future soil sampling design to perform statisticalanalysis [17] This aids in determining where to resampleif necessary and design future field experiments that avoidspatial dependence Therefore for future studies aimed atcharacterizing the spatial dependency of soil properties inthe study area andor a similar area it is recommendedthat the soil properties be sampled at distances shorterthan the range found in this study Cambardella et al [14]established the classification of degree of spatial dependence(DSD) between adjacent observations of soil property gt 75to correspond to weak spatial structure In this study thesemivariograms indicated strong spatial dependence (DSDle 25) for soil physical properties such as bulk densitygravimetric water content volumetric water content totalpore space and diffusivity The rest of the soil physicalproperties (water-filled pore space Air-filled pore space andtortuosity)measured exhibited very weak spatial dependence(DSD gt 75) (Table 2) The strong spatial dependence of thesoil properties may be controlled by intrinsic variations insoil characteristics such as texture and mineralogy whereasextrinsic variations such as tillage and other soil and water
management practices may also control the variability of theweak spatially dependent parameters [14]
33 Spatial Distribution of Soil Properties across the FieldInterpolated maps portraying the distribution of soil physicalproperties in various depths are shown in Figure 3 for soilgravimetric (GWC) and volumetric (VWC) contents andwater-filled pore space (WFPS) Gravimetric water contentshowed a good spatial distribution across the field withthe highest values located around the southwestern portionof the field Volumetric water content also showed goodspatial distribution across the field with high values locatedin the northern central and southwestern portions of thefield Water-filled pore has a distribution similar to that ofvolumetric water contentThe other soil properties howevershowed very poor spatial distribution in the field This ismost probably due to their poor sill (119862
0+ 119862) model fit and
coefficient of determination (1198772
) Even though the spatialvariability was not very pronounced there were areas onthe field that had slightly higher values of these physicalproperties than the rest of the field In general bulk densitytotal pore space volumetric air content Air-filled pore spacediffusivity and tortuosity were very high in the field eventhough they did not exhibit very distinguishable variabilityThis lack of visible spatial variability is supported because thesampling distance (range) is 26m for these properties
34 Implications of Spatial Variability of Soil Physical Prop-erties on Soil Management Results of this study indicatedthat the spatial variability of soil water content (GWC andVWC) was high This can be explained among many otherreasons by soil type (clay-loam) which was able to hold morewater But with intensive tillage this soil water content couldbe adversely affected Studies have shown that tillage prac-tices can alter soil physical properties and consequently thehydrological behavior of agricultural fields especially whena similar tillage system has been practiced for a long period[15 28ndash31] Tillage intensity has also considerable effectson spatial structure and spatial variability of soil properties[15 30]Therefore this study can help determine site-specificsoil management and decision making To do so the spatialvariability of soil properties developed through kriging willbe an important tool Different ranges of spatial dependencewere noticed in the field The different ranges of the spatialdependence among the soil properties may be attributed todifferences in response to the erosionmdashdeposition factorsland use-cover parent material and human interferences inthe study areaThe different ranges can also be used in futurestudies to determine the sampling distance of different soilphysical properties on the field Also the sill (119862
0+119862) can help
determine where the variability or change in soil propertystops This will be useful especially for the irrigation pur-poses Generally with farmers facing the decision of whetheror not to till and the intensity of tillage a spatial variabilitystudy can help in this decision making Maps producedin this study can also be used for irrigation purposes asthey can clearly indicate which portion of the field needsirrigation (soil water content) To do this soil water content
Figure 3 Spatial distribution of gravimetric (GWC) and volumetric (VWC) water contents in 0ndash10 cm depth and that of water-filled porespace (WFPS) in 20 cm depth
information can be collected and analyzed geospatially toproduce field maps The process can be repeated frequentlyto obtain up-to-date soil water content information To avoidfrequent destructive sampling for water content analysisequipment that allows insitu measurements such as TDRmethods and water mark sensors can be used Since a differ-ent range of spatial dependence among soil properties showsdifferences in response to human interferences and land use-cover this will help reduce human activities that increase soilbulk density and cause soil compaction like the use of heavyequipment It can also serve as a reference for the type of cropto be grown (cover crops for erosion susceptible areas)
4 Conclusion
We assessed the spatial variability of soil physical propertiesin a clay-loam soil cropped to corn and soybean Resultsshowed that soil physical properties either decreased orincreased sharply in the second depth (due to the presence ofa smectite layer) before leveling up or dropping off but with-out reaching the first depth value in either case In additiondepending on soil physical property maps produced by krig-ing showed either goodor poor spatial distributionThe semi-variogram analysis showed the presence of a strong (le25)to weak (gt75) spatial dependence of soil properties Ourunderstanding of the behavior of soil properties in this studyprovides new insights for soil site-specific management in
addressing issues such as ldquowhere to place the proper interven-tionsrdquo (tillage irrigation and crop type to be grown)
Acknowledgment
This research is part of a regional collaborative projectsupported by the USDA-NIFA Award no 2011-68002-30190Cropping Systems Coordinated Agricultural Project ClimateChangeMitigation andAdaptation inCorn-basedCroppingSystems (Project website httpsustainablecornorg)
References
[1] D Schimel J Melillo H Tian et al ldquoContribution of increasingCO2
and climate to carbon storage by ecosystems in the UnitedStatesrdquo Science vol 287 no 5460 pp 2004ndash2006 2000
[2] C Kosmas S Gerontidis and M Marathianou ldquoThe effect ofland use change on soils and vegetation over various lithologicalformations on Lesvos (Greece)rdquoCatena vol 40 no 1 pp 51ndash682000
[3] J Igbal J A Thomasson J N Jenkins P R Owens and F DWhisler ldquoSpatial variability analysis of soil physical propertiesof alluvial soilsrdquo Soil Science Society of America Journal vol 69pp 1ndash14 2005
[4] S Zhang X Zhang T Huffman X Liu and J Yang ldquoInfluenceof topography and land management on soil nutrients variabil-ity in Northeast ChinardquoNutrient Cycling in Agroecosystems vol89 no 3 pp 427ndash438 2011
8 ISRN Soil Science
[5] L J Munkholm A Garbout and S B Hansen ldquoTillage effectson topsoil structural quality assessed using X-ray CT soil coresand visual soil evaluationrdquo Soil amp Tillage Research vol 128 pp104ndash109 2013
[6] A J Franzluebbers ldquoSoil organicmatter stratification ratio as anindicator of soil qualityrdquo Soil and Tillage Research vol 66 no 2pp 95ndash106 2002
[7] L J Munkholm P Schjoslashnning K J Rasmussen and KTanderup ldquoSpatial and temporal effects of direct drilling onsoil structure in the seedling environmentrdquo Soil and TillageResearch vol 71 no 2 pp 163ndash173 2003
[8] A Swarowsky R A Dahlgren K W Tate J W Hopmans andA T OrsquoGeen ldquoCatchment-scale soil water dynamics in a medi-terranean-type oak woodlandrdquoVadose Zone Journal vol 10 no3 pp 800ndash815 2011
[9] RWebster andMAOliver StatisticalMethods in Soil and LandResource Survey Oxford University Press Oxford UK 1990
[10] A Saldana A Stein and J A Zinck ldquoSpatial variability of soilproperties at different scaleswithin three terraces of theHenaresriver (Spain)rdquo Catena vol 33 no 3-4 pp 139ndash153 1998
[11] Z M Wang K S Song B Zhang et al ldquoSpatial variabilityand affecting factors of soil nutrients in croplands of NortheastChina a case study in Dehui countyrdquo Plant Soil and Environ-ment vol 55 no 3 pp 110ndash120 2009
[12] R Webster ldquoQuantitative spatial analysis of soil in the fieldrdquoAdvances in Soil Science vol 3 pp 1ndash70 1985
[13] H Pohlmann ldquoGeostatisticalmodelling of environmental datardquoCatena vol 20 no 1-2 pp 191ndash198 1993
[14] C A Cambardella T B Moorman J M Novak et al ldquoField-scale variability of soil properties in central Iowa soilsrdquo Soil Sci-ence Society of America Journal vol 58 no 5 pp 1501ndash1511 1994
[15] E Ozgoz ldquoLong term conventional tillage effect on spatial vari-ability of some soil physical propertiesrdquo Journal of SustainableAgriculture vol 33 no 2 pp 142ndash160 2009
[16] N V Nkongolo R Hatano and V Kakembo ldquoDiffusivity mod-els and greenhouse gases fluxes from a forest pasture grasslandand corn field in Northern Hokkaido Japanrdquo Pedosphere vol20 no 6 pp 747ndash760 2010
[17] S H Ayoubi S M Zamani and F Khomali ldquoSpatial variabilityof some soil properties for site-specific farming in northernIranrdquo International Journal of Plant Production vol 2 pp 225ndash236 2007
[18] T M Burgess and R Webster ldquoOptimal interpolation andisarithmic mapping of soil properties I The semi-variogramand punctual krigingrdquo Journal of Soil amp Science vol 31 no 2pp 315ndash331 1980
[19] J M B Parfitt L C Timm E A Pauletto et al ldquoSpatial vari-ability of the chemical physical and biological properties inlowland cultivated with irrigated ricerdquo Revista Brasileira deCiencia do Solo vol 33 no 4 pp 819ndash830 2009
[20] C R Grego S R Vieira and A L Lourencao ldquoSpatial distri-bution of Pseudaletia sequax Franclemlont in triticale under no-till managementrdquo Scientia Agricola vol 63 no 4 pp 321ndash3272006
[21] G B Tesfahunegn L Tamene and P L G Vlek ldquoCatchment-scale spatial variability of soil properties and implications onsite-specific soil management in northern Ethiopiardquo Soil andTillage Research vol 117 pp 124ndash139 2011
[22] S J Grecu M B Kirkham E T Kanemasu D W Sweeney LR Stone and G A Milliken ldquoRoot growth in a claypan witha perennial-annual rotationrdquo Soil Science Society of AmericaJournal vol 52 no 2 pp 488ndash494 1988
[23] D B Myers N R Kitchen K A Sudduth R E Sharp andR J Miles ldquoSoybean root distribution related to claypan soilproperties and apparent soil electrical conductivityrdquo Crop Sci-ence vol 47 no 4 pp 1498ndash1509 2007
[24] P Jiang N R Kitchen S H Anderson E J Sadler and KA Sudduth ldquoEstimating palnt available water using the simpleinverse yield model for claypan landscapesrdquo Agronomy Journalvol 100 pp 1ndash7 2008
[25] A Utset M E Ruiz J Herrera and D P De Leon ldquoA geo-statistical method for soil salinity sample site spacingrdquo Geo-derma vol 86 no 1-2 pp 143ndash151 1998
[26] W Fu H Tunney and C Zhang ldquoSpatial variation of soilnutrients in a dairy farm and its implications for site-specificfertilizer applicationrdquo Soil and Tillage Research vol 106 no 2pp 185ndash193 2010
[27] F Lopez-Granados M Jurado-Exposito S Atenciano AGarcıa-Ferrer M Sanchez De La Orden and L Garcıa-TorresldquoSpatial variability of agricultural soil parameters in southernSpainrdquo Plant and Soil vol 246 no 1 pp 97ndash105 2002
[28] R L Hill ldquoLong-term conventional and no-tillage effects onselected soil physical propertiesrdquo Soil Science Society of AmericaJournal vol 54 no 1 pp 161ndash166 1990
[29] D E Buschiazzo J L Panigatti and PW Unger ldquoTillage effectson soil properties and crop production in the subhumid andsemiarid Argentinean Pampasrdquo Soil and Tillage Research vol49 no 1-2 pp 105ndash116 1998
[30] T Tsegaye and R L Hill ldquoIntensive tillage effects on spatialvariability of soil physical propertiesrdquo Soil Science vol 163 no2 pp 143ndash154 1998
[31] J A Gomez J V Giraldez M Pastor and E Fereres ldquoEffects oftillage method on soil physical properties infiltration and yieldin an olive orchardrdquo Soil and Tillage Research vol 52 no 3-4pp 167ndash175 1999
AFPS air-filled pore space () BDY soil bulk density (gcmminus3) DIFF relative gas diffusion coefficient (m2 sminus1mminus2 s) GWC gravimetric water content of soil(gsdotgminus1) TPS total pore spaces (cm3 cmminus3) TORT Pore tortuosity factor (msdotmminus1) VAC volumetric air content (cm3 cmminus3) VWC volumetric water content(cm3 cmminus3) WFPS water-filled pore space ()
(Table 1) The highly skewed soil parameters included soilbulk density (BDY) diffusivity (DIFF) and volumetric watercontent (VWC) whereas total pore space (TPS) was moder-ately skewed Air-filled pore space (AFPS) had a low skew-ness Highly skewed parameters indicate that these propertieshave a local distribution that is high values were found forthese properties at some points but most values were low[20] The other soil physical properties were approximatelynormally distributed on the field The underlying reason forsoil properties being normally or nonnormally distributedmay be associated with differences in management practicesland use vegetation cover and topographic effects on thevariability of soil erosion across the landscape of the fieldThese factors can be the sources for a large or very smallvariation of soil properties in some of the samples whichleads to the nonnormal distribution [21] A wide range ofspatial variability was observed for soil physical properties(Table 1) For instance soil bulk density (BDY) ranged from101 to 123 g cmminus3 for depth 1 115 to 146 g cmminus3 for depth 2096 to 119 g cmminus3 and 104 to 118 g cmminus3 for depths 3 and 4respectively (Figure 2) Soil bulk densitywas also significantlyhigher in the second depth (14 g cmminus3) than all the other 3depths where it varied between 118 g cmminus3 and 124 g cmminus3Themean value of AFPSwas significantly lower in the seconddepth (265 cm3sdotcmminus3) than in all other 3 depths whereit varied from 3934 to 457 cm3sdotcmminus3 Soil pore tortuosityfactor (TORT) and water-filled pore space (WFPS) werealso significantly higher in the second depth (1246 cmsdotcmminus1and 7346 resp) However the relative gas diffusion
coefficient (DIFF) gravimetric water content (GWC) totalpore space (TPS) and volumetric air content (VAC) weresignificantly lower in the second depth (002m2sminus1mminus2s021 gsdotgminus1 042 cm3sdotcmminus3 and 012 cm3sdotcmminus3 resp) (Table 1)The variability in soil physical properties is understandablesince the soil of this site has a smectite layer (claypan) inthe 10ndash20 cm which corresponded to our second samplingdepthThis layer of smectite is hard and compact with verylow pore space high mass-volume ratio (bulk density) andhigh water retention capability (because of their large surfacearea) As a consequence of the presence of this smectite layerin depth 2 the mean of water-filled pore space (WFPS) wasslightly lower in the first depth (54) than in all four depthsIn fact air predominates the pore space in the first depthand cultivation loosened the soil thereby allowing the watertrapped in the pore space to evaporate Higher GWC VWCand TPS at the lower depths (20ndash60 cm) mean that crops(especially corn and soybean grown in the field) were able toaccess water and dissolved nutrients through their roots Infact despite the claypan layer (10ndash20 cm) it has been reportedby various researchers that crop roots were able to penetrateinto and through this layer of smectitic clay [22ndash24] and thatroot growth may increase within the claypan layer [23] asa result of plant adaptation to water-limited soil layers Ingeneral the use of the coefficient of variation (CV) is a com-mon procedure to assess variability in soil properties since itallows comparison among properties with different units ofmeasurement Overall the coefficient of variation for all soilphysical properties in the four depths ranged from 483 to
4 ISRN Soil Science
0
05
1
15
2
D1 D2 D3 D4Depth
Soil
bulk
den
sity
(gmiddotcm
minus3)
(a)
0
01
02
03
04
D1 D2 D3 D4
Gra
vim
etric
wat
erco
nten
t (g
g)
Depth
(b)
0
01
02
03
04
05
06
07
D1 D2 D3 D4
Tota
l por
e spa
ce (c
m3
cmminus3)
Depth
(c)
0
005
01
015
02
025
03
035
04
D1 D2 D3 D4Depth
Volu
met
ric w
ater
cont
ent
(cm
3cm
minus3)
(d)
Figure 2 Variation of soil bulk density gravimetric water content total pore space and volumetric water content with depth
9161 (Table 1) The pore tortuosity factor (TORT) showedthe highest variation while soil bulk density (BDY) showedthe least variation The CV indicated that there was a strongspatial variability of the soil properties investigated Howeverto have a better assessment of such spatial variability acrossthe entire field a geostatistical analysis was used
32 Spatial Variability of Soil Properties After computingsummaries of simple statistics with Statistix 90 data onsoil physical properties was transferred to GS+ (geostatisticsfor environmental science) 70 for semivariogram analysisSemivariogrammodel fitwas determined from the coefficientof determination (1198772) values which range from 0 (very poormodel fit) to 1 (very good model fit) Table 2 shows soilphysical properties which mainly responded to exponentialand linear variogram models with the exponential modelproviding the best fit In the 10ndash20 cm depth exponentialmodel provided the best fit for BDY (119877
2
= 093) withthe spherical model providing very poor model fit Poretortuosity also responded to an exponential variogrammodelin the 20ndash40 cm depth (1198772 = 057) although spherical modelwas noticed Linear and exponential models were observed
in the 40ndash60 cm depth for TPS (1198772 = 046) with linearmodel providing a better fit (Table 2) In general for alldepths model fit was not very strong with the exception ofgravimetric water content and bulk density in the seconddepth Overall the exponential model provided the best fitwith about 65 of the physical properties fitting this modelIn geostatistical theory the range of the spatial variabilityof the semivariogram is the distance between correlatedmeasurements (the minimum lateral distance between twopoints before the change in property is noticed) and can be aneffective criterion for the evaluation of sampling design andmapping of soil propertiesThe value that the semivariogrammodel attains at the range (the value on the 119910-axis) is calledthe sill The partial sill is the sill minus the nugget [2526] Theoretically at zero separation distance (lag = 0) thesemivariogram value is zero However at an infinitesimallysmall separation distance the semivariogram often exhibitsa nugget effect (the apparent discontinuity at the beginningof many semivariogram graphs) which is some value greaterthan zeroThenugget effect can be attributed tomeasurementerrors or spatial sources of variation at distances smallerthan the sampling interval (or both) Measurement erroroccurs because of the error inherent inmeasuring devices To
ISRN Soil Science 5
Table 2 Variogram parameters for soil physical properties at four depths in a clay-loam soil
DSD degree of spatial dependence strong DSD (DSD le 25) moderate DSD (25 lt DSD le 75) and weak DSD (DSD gt 75) according to Cambardella etal (1994) [14]
6 ISRN Soil Science
eliminate this error multiple samples were taken from eachsampling point Natural phenomena can vary spatially over arange of scales Variation atmicroscales smaller than the sam-pling distances will appear as a part of the nugget effect Table2 shows that the spatial correlation (range) of soil propertieswidely varied from 1m for volumetric water content (VWC)in depth four to 64m for gravimetric water content (GWC)in depth 2 However for the first and second depth (whichare agriculturally more important) the range of spatialcorrelation varied from 3m for volumetric air content (VAC)in depth 2 to 64m for GWC in depth 2 Beyond these rangesthere is no spatial dependence (autocorrelation) The spatialdependence can indicate the level of similarity or disturbanceof the soil condition According to Lopez-Granados et al [27]and Ayoubi et al [17] a large range indicates that the mea-sured soil property value is influenced by natural and anthro-pogenic factors over greater distances than parameters whichhave smaller ranges Thus a range of about 64m for GWCin this study indicates that the measured GWC values canbe influenced in the soil over greater distances as comparedto the soil parameters having smaller range (Table 2) Thismeans that soil variables with smaller range such asVWCandVACare good indicators of themore disturbed soils (themoredisturbed a soil is the more variable some soil propertiesbecome) The more variable properties have a shorter rangeof correlation The different ranges of the spatial dependenceamong the soil properties may be attributed to differencesin response to the erosionmdashdeposition factors land use-cover parent material and human interferences in the studyarea The nugget which is an indication of microvariabilitywas significantly higher for water-filled pore space (WFPS)and air-filled pore space (AFPS) when compared to theothersThis can be explained by our sampling distance whichcould not capture well their spatial dependence The lowestnugget was for GWC (Table 2) This indicates that GWChad low spatial variability within small distances Knowledgeof the range of influence for various soil properties allowsone to construct independent accurate datasets for similarareas in future soil sampling design to perform statisticalanalysis [17] This aids in determining where to resampleif necessary and design future field experiments that avoidspatial dependence Therefore for future studies aimed atcharacterizing the spatial dependency of soil properties inthe study area andor a similar area it is recommendedthat the soil properties be sampled at distances shorterthan the range found in this study Cambardella et al [14]established the classification of degree of spatial dependence(DSD) between adjacent observations of soil property gt 75to correspond to weak spatial structure In this study thesemivariograms indicated strong spatial dependence (DSDle 25) for soil physical properties such as bulk densitygravimetric water content volumetric water content totalpore space and diffusivity The rest of the soil physicalproperties (water-filled pore space Air-filled pore space andtortuosity)measured exhibited very weak spatial dependence(DSD gt 75) (Table 2) The strong spatial dependence of thesoil properties may be controlled by intrinsic variations insoil characteristics such as texture and mineralogy whereasextrinsic variations such as tillage and other soil and water
management practices may also control the variability of theweak spatially dependent parameters [14]
33 Spatial Distribution of Soil Properties across the FieldInterpolated maps portraying the distribution of soil physicalproperties in various depths are shown in Figure 3 for soilgravimetric (GWC) and volumetric (VWC) contents andwater-filled pore space (WFPS) Gravimetric water contentshowed a good spatial distribution across the field withthe highest values located around the southwestern portionof the field Volumetric water content also showed goodspatial distribution across the field with high values locatedin the northern central and southwestern portions of thefield Water-filled pore has a distribution similar to that ofvolumetric water contentThe other soil properties howevershowed very poor spatial distribution in the field This ismost probably due to their poor sill (119862
0+ 119862) model fit and
coefficient of determination (1198772
) Even though the spatialvariability was not very pronounced there were areas onthe field that had slightly higher values of these physicalproperties than the rest of the field In general bulk densitytotal pore space volumetric air content Air-filled pore spacediffusivity and tortuosity were very high in the field eventhough they did not exhibit very distinguishable variabilityThis lack of visible spatial variability is supported because thesampling distance (range) is 26m for these properties
34 Implications of Spatial Variability of Soil Physical Prop-erties on Soil Management Results of this study indicatedthat the spatial variability of soil water content (GWC andVWC) was high This can be explained among many otherreasons by soil type (clay-loam) which was able to hold morewater But with intensive tillage this soil water content couldbe adversely affected Studies have shown that tillage prac-tices can alter soil physical properties and consequently thehydrological behavior of agricultural fields especially whena similar tillage system has been practiced for a long period[15 28ndash31] Tillage intensity has also considerable effectson spatial structure and spatial variability of soil properties[15 30]Therefore this study can help determine site-specificsoil management and decision making To do so the spatialvariability of soil properties developed through kriging willbe an important tool Different ranges of spatial dependencewere noticed in the field The different ranges of the spatialdependence among the soil properties may be attributed todifferences in response to the erosionmdashdeposition factorsland use-cover parent material and human interferences inthe study areaThe different ranges can also be used in futurestudies to determine the sampling distance of different soilphysical properties on the field Also the sill (119862
0+119862) can help
determine where the variability or change in soil propertystops This will be useful especially for the irrigation pur-poses Generally with farmers facing the decision of whetheror not to till and the intensity of tillage a spatial variabilitystudy can help in this decision making Maps producedin this study can also be used for irrigation purposes asthey can clearly indicate which portion of the field needsirrigation (soil water content) To do this soil water content
Figure 3 Spatial distribution of gravimetric (GWC) and volumetric (VWC) water contents in 0ndash10 cm depth and that of water-filled porespace (WFPS) in 20 cm depth
information can be collected and analyzed geospatially toproduce field maps The process can be repeated frequentlyto obtain up-to-date soil water content information To avoidfrequent destructive sampling for water content analysisequipment that allows insitu measurements such as TDRmethods and water mark sensors can be used Since a differ-ent range of spatial dependence among soil properties showsdifferences in response to human interferences and land use-cover this will help reduce human activities that increase soilbulk density and cause soil compaction like the use of heavyequipment It can also serve as a reference for the type of cropto be grown (cover crops for erosion susceptible areas)
4 Conclusion
We assessed the spatial variability of soil physical propertiesin a clay-loam soil cropped to corn and soybean Resultsshowed that soil physical properties either decreased orincreased sharply in the second depth (due to the presence ofa smectite layer) before leveling up or dropping off but with-out reaching the first depth value in either case In additiondepending on soil physical property maps produced by krig-ing showed either goodor poor spatial distributionThe semi-variogram analysis showed the presence of a strong (le25)to weak (gt75) spatial dependence of soil properties Ourunderstanding of the behavior of soil properties in this studyprovides new insights for soil site-specific management in
addressing issues such as ldquowhere to place the proper interven-tionsrdquo (tillage irrigation and crop type to be grown)
Acknowledgment
This research is part of a regional collaborative projectsupported by the USDA-NIFA Award no 2011-68002-30190Cropping Systems Coordinated Agricultural Project ClimateChangeMitigation andAdaptation inCorn-basedCroppingSystems (Project website httpsustainablecornorg)
References
[1] D Schimel J Melillo H Tian et al ldquoContribution of increasingCO2
and climate to carbon storage by ecosystems in the UnitedStatesrdquo Science vol 287 no 5460 pp 2004ndash2006 2000
[2] C Kosmas S Gerontidis and M Marathianou ldquoThe effect ofland use change on soils and vegetation over various lithologicalformations on Lesvos (Greece)rdquoCatena vol 40 no 1 pp 51ndash682000
[3] J Igbal J A Thomasson J N Jenkins P R Owens and F DWhisler ldquoSpatial variability analysis of soil physical propertiesof alluvial soilsrdquo Soil Science Society of America Journal vol 69pp 1ndash14 2005
[4] S Zhang X Zhang T Huffman X Liu and J Yang ldquoInfluenceof topography and land management on soil nutrients variabil-ity in Northeast ChinardquoNutrient Cycling in Agroecosystems vol89 no 3 pp 427ndash438 2011
8 ISRN Soil Science
[5] L J Munkholm A Garbout and S B Hansen ldquoTillage effectson topsoil structural quality assessed using X-ray CT soil coresand visual soil evaluationrdquo Soil amp Tillage Research vol 128 pp104ndash109 2013
[6] A J Franzluebbers ldquoSoil organicmatter stratification ratio as anindicator of soil qualityrdquo Soil and Tillage Research vol 66 no 2pp 95ndash106 2002
[7] L J Munkholm P Schjoslashnning K J Rasmussen and KTanderup ldquoSpatial and temporal effects of direct drilling onsoil structure in the seedling environmentrdquo Soil and TillageResearch vol 71 no 2 pp 163ndash173 2003
[8] A Swarowsky R A Dahlgren K W Tate J W Hopmans andA T OrsquoGeen ldquoCatchment-scale soil water dynamics in a medi-terranean-type oak woodlandrdquoVadose Zone Journal vol 10 no3 pp 800ndash815 2011
[9] RWebster andMAOliver StatisticalMethods in Soil and LandResource Survey Oxford University Press Oxford UK 1990
[10] A Saldana A Stein and J A Zinck ldquoSpatial variability of soilproperties at different scaleswithin three terraces of theHenaresriver (Spain)rdquo Catena vol 33 no 3-4 pp 139ndash153 1998
[11] Z M Wang K S Song B Zhang et al ldquoSpatial variabilityand affecting factors of soil nutrients in croplands of NortheastChina a case study in Dehui countyrdquo Plant Soil and Environ-ment vol 55 no 3 pp 110ndash120 2009
[12] R Webster ldquoQuantitative spatial analysis of soil in the fieldrdquoAdvances in Soil Science vol 3 pp 1ndash70 1985
[13] H Pohlmann ldquoGeostatisticalmodelling of environmental datardquoCatena vol 20 no 1-2 pp 191ndash198 1993
[14] C A Cambardella T B Moorman J M Novak et al ldquoField-scale variability of soil properties in central Iowa soilsrdquo Soil Sci-ence Society of America Journal vol 58 no 5 pp 1501ndash1511 1994
[15] E Ozgoz ldquoLong term conventional tillage effect on spatial vari-ability of some soil physical propertiesrdquo Journal of SustainableAgriculture vol 33 no 2 pp 142ndash160 2009
[16] N V Nkongolo R Hatano and V Kakembo ldquoDiffusivity mod-els and greenhouse gases fluxes from a forest pasture grasslandand corn field in Northern Hokkaido Japanrdquo Pedosphere vol20 no 6 pp 747ndash760 2010
[17] S H Ayoubi S M Zamani and F Khomali ldquoSpatial variabilityof some soil properties for site-specific farming in northernIranrdquo International Journal of Plant Production vol 2 pp 225ndash236 2007
[18] T M Burgess and R Webster ldquoOptimal interpolation andisarithmic mapping of soil properties I The semi-variogramand punctual krigingrdquo Journal of Soil amp Science vol 31 no 2pp 315ndash331 1980
[19] J M B Parfitt L C Timm E A Pauletto et al ldquoSpatial vari-ability of the chemical physical and biological properties inlowland cultivated with irrigated ricerdquo Revista Brasileira deCiencia do Solo vol 33 no 4 pp 819ndash830 2009
[20] C R Grego S R Vieira and A L Lourencao ldquoSpatial distri-bution of Pseudaletia sequax Franclemlont in triticale under no-till managementrdquo Scientia Agricola vol 63 no 4 pp 321ndash3272006
[21] G B Tesfahunegn L Tamene and P L G Vlek ldquoCatchment-scale spatial variability of soil properties and implications onsite-specific soil management in northern Ethiopiardquo Soil andTillage Research vol 117 pp 124ndash139 2011
[22] S J Grecu M B Kirkham E T Kanemasu D W Sweeney LR Stone and G A Milliken ldquoRoot growth in a claypan witha perennial-annual rotationrdquo Soil Science Society of AmericaJournal vol 52 no 2 pp 488ndash494 1988
[23] D B Myers N R Kitchen K A Sudduth R E Sharp andR J Miles ldquoSoybean root distribution related to claypan soilproperties and apparent soil electrical conductivityrdquo Crop Sci-ence vol 47 no 4 pp 1498ndash1509 2007
[24] P Jiang N R Kitchen S H Anderson E J Sadler and KA Sudduth ldquoEstimating palnt available water using the simpleinverse yield model for claypan landscapesrdquo Agronomy Journalvol 100 pp 1ndash7 2008
[25] A Utset M E Ruiz J Herrera and D P De Leon ldquoA geo-statistical method for soil salinity sample site spacingrdquo Geo-derma vol 86 no 1-2 pp 143ndash151 1998
[26] W Fu H Tunney and C Zhang ldquoSpatial variation of soilnutrients in a dairy farm and its implications for site-specificfertilizer applicationrdquo Soil and Tillage Research vol 106 no 2pp 185ndash193 2010
[27] F Lopez-Granados M Jurado-Exposito S Atenciano AGarcıa-Ferrer M Sanchez De La Orden and L Garcıa-TorresldquoSpatial variability of agricultural soil parameters in southernSpainrdquo Plant and Soil vol 246 no 1 pp 97ndash105 2002
[28] R L Hill ldquoLong-term conventional and no-tillage effects onselected soil physical propertiesrdquo Soil Science Society of AmericaJournal vol 54 no 1 pp 161ndash166 1990
[29] D E Buschiazzo J L Panigatti and PW Unger ldquoTillage effectson soil properties and crop production in the subhumid andsemiarid Argentinean Pampasrdquo Soil and Tillage Research vol49 no 1-2 pp 105ndash116 1998
[30] T Tsegaye and R L Hill ldquoIntensive tillage effects on spatialvariability of soil physical propertiesrdquo Soil Science vol 163 no2 pp 143ndash154 1998
[31] J A Gomez J V Giraldez M Pastor and E Fereres ldquoEffects oftillage method on soil physical properties infiltration and yieldin an olive orchardrdquo Soil and Tillage Research vol 52 no 3-4pp 167ndash175 1999
Figure 2 Variation of soil bulk density gravimetric water content total pore space and volumetric water content with depth
9161 (Table 1) The pore tortuosity factor (TORT) showedthe highest variation while soil bulk density (BDY) showedthe least variation The CV indicated that there was a strongspatial variability of the soil properties investigated Howeverto have a better assessment of such spatial variability acrossthe entire field a geostatistical analysis was used
32 Spatial Variability of Soil Properties After computingsummaries of simple statistics with Statistix 90 data onsoil physical properties was transferred to GS+ (geostatisticsfor environmental science) 70 for semivariogram analysisSemivariogrammodel fitwas determined from the coefficientof determination (1198772) values which range from 0 (very poormodel fit) to 1 (very good model fit) Table 2 shows soilphysical properties which mainly responded to exponentialand linear variogram models with the exponential modelproviding the best fit In the 10ndash20 cm depth exponentialmodel provided the best fit for BDY (119877
2
= 093) withthe spherical model providing very poor model fit Poretortuosity also responded to an exponential variogrammodelin the 20ndash40 cm depth (1198772 = 057) although spherical modelwas noticed Linear and exponential models were observed
in the 40ndash60 cm depth for TPS (1198772 = 046) with linearmodel providing a better fit (Table 2) In general for alldepths model fit was not very strong with the exception ofgravimetric water content and bulk density in the seconddepth Overall the exponential model provided the best fitwith about 65 of the physical properties fitting this modelIn geostatistical theory the range of the spatial variabilityof the semivariogram is the distance between correlatedmeasurements (the minimum lateral distance between twopoints before the change in property is noticed) and can be aneffective criterion for the evaluation of sampling design andmapping of soil propertiesThe value that the semivariogrammodel attains at the range (the value on the 119910-axis) is calledthe sill The partial sill is the sill minus the nugget [2526] Theoretically at zero separation distance (lag = 0) thesemivariogram value is zero However at an infinitesimallysmall separation distance the semivariogram often exhibitsa nugget effect (the apparent discontinuity at the beginningof many semivariogram graphs) which is some value greaterthan zeroThenugget effect can be attributed tomeasurementerrors or spatial sources of variation at distances smallerthan the sampling interval (or both) Measurement erroroccurs because of the error inherent inmeasuring devices To
ISRN Soil Science 5
Table 2 Variogram parameters for soil physical properties at four depths in a clay-loam soil
DSD degree of spatial dependence strong DSD (DSD le 25) moderate DSD (25 lt DSD le 75) and weak DSD (DSD gt 75) according to Cambardella etal (1994) [14]
6 ISRN Soil Science
eliminate this error multiple samples were taken from eachsampling point Natural phenomena can vary spatially over arange of scales Variation atmicroscales smaller than the sam-pling distances will appear as a part of the nugget effect Table2 shows that the spatial correlation (range) of soil propertieswidely varied from 1m for volumetric water content (VWC)in depth four to 64m for gravimetric water content (GWC)in depth 2 However for the first and second depth (whichare agriculturally more important) the range of spatialcorrelation varied from 3m for volumetric air content (VAC)in depth 2 to 64m for GWC in depth 2 Beyond these rangesthere is no spatial dependence (autocorrelation) The spatialdependence can indicate the level of similarity or disturbanceof the soil condition According to Lopez-Granados et al [27]and Ayoubi et al [17] a large range indicates that the mea-sured soil property value is influenced by natural and anthro-pogenic factors over greater distances than parameters whichhave smaller ranges Thus a range of about 64m for GWCin this study indicates that the measured GWC values canbe influenced in the soil over greater distances as comparedto the soil parameters having smaller range (Table 2) Thismeans that soil variables with smaller range such asVWCandVACare good indicators of themore disturbed soils (themoredisturbed a soil is the more variable some soil propertiesbecome) The more variable properties have a shorter rangeof correlation The different ranges of the spatial dependenceamong the soil properties may be attributed to differencesin response to the erosionmdashdeposition factors land use-cover parent material and human interferences in the studyarea The nugget which is an indication of microvariabilitywas significantly higher for water-filled pore space (WFPS)and air-filled pore space (AFPS) when compared to theothersThis can be explained by our sampling distance whichcould not capture well their spatial dependence The lowestnugget was for GWC (Table 2) This indicates that GWChad low spatial variability within small distances Knowledgeof the range of influence for various soil properties allowsone to construct independent accurate datasets for similarareas in future soil sampling design to perform statisticalanalysis [17] This aids in determining where to resampleif necessary and design future field experiments that avoidspatial dependence Therefore for future studies aimed atcharacterizing the spatial dependency of soil properties inthe study area andor a similar area it is recommendedthat the soil properties be sampled at distances shorterthan the range found in this study Cambardella et al [14]established the classification of degree of spatial dependence(DSD) between adjacent observations of soil property gt 75to correspond to weak spatial structure In this study thesemivariograms indicated strong spatial dependence (DSDle 25) for soil physical properties such as bulk densitygravimetric water content volumetric water content totalpore space and diffusivity The rest of the soil physicalproperties (water-filled pore space Air-filled pore space andtortuosity)measured exhibited very weak spatial dependence(DSD gt 75) (Table 2) The strong spatial dependence of thesoil properties may be controlled by intrinsic variations insoil characteristics such as texture and mineralogy whereasextrinsic variations such as tillage and other soil and water
management practices may also control the variability of theweak spatially dependent parameters [14]
33 Spatial Distribution of Soil Properties across the FieldInterpolated maps portraying the distribution of soil physicalproperties in various depths are shown in Figure 3 for soilgravimetric (GWC) and volumetric (VWC) contents andwater-filled pore space (WFPS) Gravimetric water contentshowed a good spatial distribution across the field withthe highest values located around the southwestern portionof the field Volumetric water content also showed goodspatial distribution across the field with high values locatedin the northern central and southwestern portions of thefield Water-filled pore has a distribution similar to that ofvolumetric water contentThe other soil properties howevershowed very poor spatial distribution in the field This ismost probably due to their poor sill (119862
0+ 119862) model fit and
coefficient of determination (1198772
) Even though the spatialvariability was not very pronounced there were areas onthe field that had slightly higher values of these physicalproperties than the rest of the field In general bulk densitytotal pore space volumetric air content Air-filled pore spacediffusivity and tortuosity were very high in the field eventhough they did not exhibit very distinguishable variabilityThis lack of visible spatial variability is supported because thesampling distance (range) is 26m for these properties
34 Implications of Spatial Variability of Soil Physical Prop-erties on Soil Management Results of this study indicatedthat the spatial variability of soil water content (GWC andVWC) was high This can be explained among many otherreasons by soil type (clay-loam) which was able to hold morewater But with intensive tillage this soil water content couldbe adversely affected Studies have shown that tillage prac-tices can alter soil physical properties and consequently thehydrological behavior of agricultural fields especially whena similar tillage system has been practiced for a long period[15 28ndash31] Tillage intensity has also considerable effectson spatial structure and spatial variability of soil properties[15 30]Therefore this study can help determine site-specificsoil management and decision making To do so the spatialvariability of soil properties developed through kriging willbe an important tool Different ranges of spatial dependencewere noticed in the field The different ranges of the spatialdependence among the soil properties may be attributed todifferences in response to the erosionmdashdeposition factorsland use-cover parent material and human interferences inthe study areaThe different ranges can also be used in futurestudies to determine the sampling distance of different soilphysical properties on the field Also the sill (119862
0+119862) can help
determine where the variability or change in soil propertystops This will be useful especially for the irrigation pur-poses Generally with farmers facing the decision of whetheror not to till and the intensity of tillage a spatial variabilitystudy can help in this decision making Maps producedin this study can also be used for irrigation purposes asthey can clearly indicate which portion of the field needsirrigation (soil water content) To do this soil water content
Figure 3 Spatial distribution of gravimetric (GWC) and volumetric (VWC) water contents in 0ndash10 cm depth and that of water-filled porespace (WFPS) in 20 cm depth
information can be collected and analyzed geospatially toproduce field maps The process can be repeated frequentlyto obtain up-to-date soil water content information To avoidfrequent destructive sampling for water content analysisequipment that allows insitu measurements such as TDRmethods and water mark sensors can be used Since a differ-ent range of spatial dependence among soil properties showsdifferences in response to human interferences and land use-cover this will help reduce human activities that increase soilbulk density and cause soil compaction like the use of heavyequipment It can also serve as a reference for the type of cropto be grown (cover crops for erosion susceptible areas)
4 Conclusion
We assessed the spatial variability of soil physical propertiesin a clay-loam soil cropped to corn and soybean Resultsshowed that soil physical properties either decreased orincreased sharply in the second depth (due to the presence ofa smectite layer) before leveling up or dropping off but with-out reaching the first depth value in either case In additiondepending on soil physical property maps produced by krig-ing showed either goodor poor spatial distributionThe semi-variogram analysis showed the presence of a strong (le25)to weak (gt75) spatial dependence of soil properties Ourunderstanding of the behavior of soil properties in this studyprovides new insights for soil site-specific management in
addressing issues such as ldquowhere to place the proper interven-tionsrdquo (tillage irrigation and crop type to be grown)
Acknowledgment
This research is part of a regional collaborative projectsupported by the USDA-NIFA Award no 2011-68002-30190Cropping Systems Coordinated Agricultural Project ClimateChangeMitigation andAdaptation inCorn-basedCroppingSystems (Project website httpsustainablecornorg)
References
[1] D Schimel J Melillo H Tian et al ldquoContribution of increasingCO2
and climate to carbon storage by ecosystems in the UnitedStatesrdquo Science vol 287 no 5460 pp 2004ndash2006 2000
[2] C Kosmas S Gerontidis and M Marathianou ldquoThe effect ofland use change on soils and vegetation over various lithologicalformations on Lesvos (Greece)rdquoCatena vol 40 no 1 pp 51ndash682000
[3] J Igbal J A Thomasson J N Jenkins P R Owens and F DWhisler ldquoSpatial variability analysis of soil physical propertiesof alluvial soilsrdquo Soil Science Society of America Journal vol 69pp 1ndash14 2005
[4] S Zhang X Zhang T Huffman X Liu and J Yang ldquoInfluenceof topography and land management on soil nutrients variabil-ity in Northeast ChinardquoNutrient Cycling in Agroecosystems vol89 no 3 pp 427ndash438 2011
8 ISRN Soil Science
[5] L J Munkholm A Garbout and S B Hansen ldquoTillage effectson topsoil structural quality assessed using X-ray CT soil coresand visual soil evaluationrdquo Soil amp Tillage Research vol 128 pp104ndash109 2013
[6] A J Franzluebbers ldquoSoil organicmatter stratification ratio as anindicator of soil qualityrdquo Soil and Tillage Research vol 66 no 2pp 95ndash106 2002
[7] L J Munkholm P Schjoslashnning K J Rasmussen and KTanderup ldquoSpatial and temporal effects of direct drilling onsoil structure in the seedling environmentrdquo Soil and TillageResearch vol 71 no 2 pp 163ndash173 2003
[8] A Swarowsky R A Dahlgren K W Tate J W Hopmans andA T OrsquoGeen ldquoCatchment-scale soil water dynamics in a medi-terranean-type oak woodlandrdquoVadose Zone Journal vol 10 no3 pp 800ndash815 2011
[9] RWebster andMAOliver StatisticalMethods in Soil and LandResource Survey Oxford University Press Oxford UK 1990
[10] A Saldana A Stein and J A Zinck ldquoSpatial variability of soilproperties at different scaleswithin three terraces of theHenaresriver (Spain)rdquo Catena vol 33 no 3-4 pp 139ndash153 1998
[11] Z M Wang K S Song B Zhang et al ldquoSpatial variabilityand affecting factors of soil nutrients in croplands of NortheastChina a case study in Dehui countyrdquo Plant Soil and Environ-ment vol 55 no 3 pp 110ndash120 2009
[12] R Webster ldquoQuantitative spatial analysis of soil in the fieldrdquoAdvances in Soil Science vol 3 pp 1ndash70 1985
[13] H Pohlmann ldquoGeostatisticalmodelling of environmental datardquoCatena vol 20 no 1-2 pp 191ndash198 1993
[14] C A Cambardella T B Moorman J M Novak et al ldquoField-scale variability of soil properties in central Iowa soilsrdquo Soil Sci-ence Society of America Journal vol 58 no 5 pp 1501ndash1511 1994
[15] E Ozgoz ldquoLong term conventional tillage effect on spatial vari-ability of some soil physical propertiesrdquo Journal of SustainableAgriculture vol 33 no 2 pp 142ndash160 2009
[16] N V Nkongolo R Hatano and V Kakembo ldquoDiffusivity mod-els and greenhouse gases fluxes from a forest pasture grasslandand corn field in Northern Hokkaido Japanrdquo Pedosphere vol20 no 6 pp 747ndash760 2010
[17] S H Ayoubi S M Zamani and F Khomali ldquoSpatial variabilityof some soil properties for site-specific farming in northernIranrdquo International Journal of Plant Production vol 2 pp 225ndash236 2007
[18] T M Burgess and R Webster ldquoOptimal interpolation andisarithmic mapping of soil properties I The semi-variogramand punctual krigingrdquo Journal of Soil amp Science vol 31 no 2pp 315ndash331 1980
[19] J M B Parfitt L C Timm E A Pauletto et al ldquoSpatial vari-ability of the chemical physical and biological properties inlowland cultivated with irrigated ricerdquo Revista Brasileira deCiencia do Solo vol 33 no 4 pp 819ndash830 2009
[20] C R Grego S R Vieira and A L Lourencao ldquoSpatial distri-bution of Pseudaletia sequax Franclemlont in triticale under no-till managementrdquo Scientia Agricola vol 63 no 4 pp 321ndash3272006
[21] G B Tesfahunegn L Tamene and P L G Vlek ldquoCatchment-scale spatial variability of soil properties and implications onsite-specific soil management in northern Ethiopiardquo Soil andTillage Research vol 117 pp 124ndash139 2011
[22] S J Grecu M B Kirkham E T Kanemasu D W Sweeney LR Stone and G A Milliken ldquoRoot growth in a claypan witha perennial-annual rotationrdquo Soil Science Society of AmericaJournal vol 52 no 2 pp 488ndash494 1988
[23] D B Myers N R Kitchen K A Sudduth R E Sharp andR J Miles ldquoSoybean root distribution related to claypan soilproperties and apparent soil electrical conductivityrdquo Crop Sci-ence vol 47 no 4 pp 1498ndash1509 2007
[24] P Jiang N R Kitchen S H Anderson E J Sadler and KA Sudduth ldquoEstimating palnt available water using the simpleinverse yield model for claypan landscapesrdquo Agronomy Journalvol 100 pp 1ndash7 2008
[25] A Utset M E Ruiz J Herrera and D P De Leon ldquoA geo-statistical method for soil salinity sample site spacingrdquo Geo-derma vol 86 no 1-2 pp 143ndash151 1998
[26] W Fu H Tunney and C Zhang ldquoSpatial variation of soilnutrients in a dairy farm and its implications for site-specificfertilizer applicationrdquo Soil and Tillage Research vol 106 no 2pp 185ndash193 2010
[27] F Lopez-Granados M Jurado-Exposito S Atenciano AGarcıa-Ferrer M Sanchez De La Orden and L Garcıa-TorresldquoSpatial variability of agricultural soil parameters in southernSpainrdquo Plant and Soil vol 246 no 1 pp 97ndash105 2002
[28] R L Hill ldquoLong-term conventional and no-tillage effects onselected soil physical propertiesrdquo Soil Science Society of AmericaJournal vol 54 no 1 pp 161ndash166 1990
[29] D E Buschiazzo J L Panigatti and PW Unger ldquoTillage effectson soil properties and crop production in the subhumid andsemiarid Argentinean Pampasrdquo Soil and Tillage Research vol49 no 1-2 pp 105ndash116 1998
[30] T Tsegaye and R L Hill ldquoIntensive tillage effects on spatialvariability of soil physical propertiesrdquo Soil Science vol 163 no2 pp 143ndash154 1998
[31] J A Gomez J V Giraldez M Pastor and E Fereres ldquoEffects oftillage method on soil physical properties infiltration and yieldin an olive orchardrdquo Soil and Tillage Research vol 52 no 3-4pp 167ndash175 1999
DSD degree of spatial dependence strong DSD (DSD le 25) moderate DSD (25 lt DSD le 75) and weak DSD (DSD gt 75) according to Cambardella etal (1994) [14]
6 ISRN Soil Science
eliminate this error multiple samples were taken from eachsampling point Natural phenomena can vary spatially over arange of scales Variation atmicroscales smaller than the sam-pling distances will appear as a part of the nugget effect Table2 shows that the spatial correlation (range) of soil propertieswidely varied from 1m for volumetric water content (VWC)in depth four to 64m for gravimetric water content (GWC)in depth 2 However for the first and second depth (whichare agriculturally more important) the range of spatialcorrelation varied from 3m for volumetric air content (VAC)in depth 2 to 64m for GWC in depth 2 Beyond these rangesthere is no spatial dependence (autocorrelation) The spatialdependence can indicate the level of similarity or disturbanceof the soil condition According to Lopez-Granados et al [27]and Ayoubi et al [17] a large range indicates that the mea-sured soil property value is influenced by natural and anthro-pogenic factors over greater distances than parameters whichhave smaller ranges Thus a range of about 64m for GWCin this study indicates that the measured GWC values canbe influenced in the soil over greater distances as comparedto the soil parameters having smaller range (Table 2) Thismeans that soil variables with smaller range such asVWCandVACare good indicators of themore disturbed soils (themoredisturbed a soil is the more variable some soil propertiesbecome) The more variable properties have a shorter rangeof correlation The different ranges of the spatial dependenceamong the soil properties may be attributed to differencesin response to the erosionmdashdeposition factors land use-cover parent material and human interferences in the studyarea The nugget which is an indication of microvariabilitywas significantly higher for water-filled pore space (WFPS)and air-filled pore space (AFPS) when compared to theothersThis can be explained by our sampling distance whichcould not capture well their spatial dependence The lowestnugget was for GWC (Table 2) This indicates that GWChad low spatial variability within small distances Knowledgeof the range of influence for various soil properties allowsone to construct independent accurate datasets for similarareas in future soil sampling design to perform statisticalanalysis [17] This aids in determining where to resampleif necessary and design future field experiments that avoidspatial dependence Therefore for future studies aimed atcharacterizing the spatial dependency of soil properties inthe study area andor a similar area it is recommendedthat the soil properties be sampled at distances shorterthan the range found in this study Cambardella et al [14]established the classification of degree of spatial dependence(DSD) between adjacent observations of soil property gt 75to correspond to weak spatial structure In this study thesemivariograms indicated strong spatial dependence (DSDle 25) for soil physical properties such as bulk densitygravimetric water content volumetric water content totalpore space and diffusivity The rest of the soil physicalproperties (water-filled pore space Air-filled pore space andtortuosity)measured exhibited very weak spatial dependence(DSD gt 75) (Table 2) The strong spatial dependence of thesoil properties may be controlled by intrinsic variations insoil characteristics such as texture and mineralogy whereasextrinsic variations such as tillage and other soil and water
management practices may also control the variability of theweak spatially dependent parameters [14]
33 Spatial Distribution of Soil Properties across the FieldInterpolated maps portraying the distribution of soil physicalproperties in various depths are shown in Figure 3 for soilgravimetric (GWC) and volumetric (VWC) contents andwater-filled pore space (WFPS) Gravimetric water contentshowed a good spatial distribution across the field withthe highest values located around the southwestern portionof the field Volumetric water content also showed goodspatial distribution across the field with high values locatedin the northern central and southwestern portions of thefield Water-filled pore has a distribution similar to that ofvolumetric water contentThe other soil properties howevershowed very poor spatial distribution in the field This ismost probably due to their poor sill (119862
0+ 119862) model fit and
coefficient of determination (1198772
) Even though the spatialvariability was not very pronounced there were areas onthe field that had slightly higher values of these physicalproperties than the rest of the field In general bulk densitytotal pore space volumetric air content Air-filled pore spacediffusivity and tortuosity were very high in the field eventhough they did not exhibit very distinguishable variabilityThis lack of visible spatial variability is supported because thesampling distance (range) is 26m for these properties
34 Implications of Spatial Variability of Soil Physical Prop-erties on Soil Management Results of this study indicatedthat the spatial variability of soil water content (GWC andVWC) was high This can be explained among many otherreasons by soil type (clay-loam) which was able to hold morewater But with intensive tillage this soil water content couldbe adversely affected Studies have shown that tillage prac-tices can alter soil physical properties and consequently thehydrological behavior of agricultural fields especially whena similar tillage system has been practiced for a long period[15 28ndash31] Tillage intensity has also considerable effectson spatial structure and spatial variability of soil properties[15 30]Therefore this study can help determine site-specificsoil management and decision making To do so the spatialvariability of soil properties developed through kriging willbe an important tool Different ranges of spatial dependencewere noticed in the field The different ranges of the spatialdependence among the soil properties may be attributed todifferences in response to the erosionmdashdeposition factorsland use-cover parent material and human interferences inthe study areaThe different ranges can also be used in futurestudies to determine the sampling distance of different soilphysical properties on the field Also the sill (119862
0+119862) can help
determine where the variability or change in soil propertystops This will be useful especially for the irrigation pur-poses Generally with farmers facing the decision of whetheror not to till and the intensity of tillage a spatial variabilitystudy can help in this decision making Maps producedin this study can also be used for irrigation purposes asthey can clearly indicate which portion of the field needsirrigation (soil water content) To do this soil water content
Figure 3 Spatial distribution of gravimetric (GWC) and volumetric (VWC) water contents in 0ndash10 cm depth and that of water-filled porespace (WFPS) in 20 cm depth
information can be collected and analyzed geospatially toproduce field maps The process can be repeated frequentlyto obtain up-to-date soil water content information To avoidfrequent destructive sampling for water content analysisequipment that allows insitu measurements such as TDRmethods and water mark sensors can be used Since a differ-ent range of spatial dependence among soil properties showsdifferences in response to human interferences and land use-cover this will help reduce human activities that increase soilbulk density and cause soil compaction like the use of heavyequipment It can also serve as a reference for the type of cropto be grown (cover crops for erosion susceptible areas)
4 Conclusion
We assessed the spatial variability of soil physical propertiesin a clay-loam soil cropped to corn and soybean Resultsshowed that soil physical properties either decreased orincreased sharply in the second depth (due to the presence ofa smectite layer) before leveling up or dropping off but with-out reaching the first depth value in either case In additiondepending on soil physical property maps produced by krig-ing showed either goodor poor spatial distributionThe semi-variogram analysis showed the presence of a strong (le25)to weak (gt75) spatial dependence of soil properties Ourunderstanding of the behavior of soil properties in this studyprovides new insights for soil site-specific management in
addressing issues such as ldquowhere to place the proper interven-tionsrdquo (tillage irrigation and crop type to be grown)
Acknowledgment
This research is part of a regional collaborative projectsupported by the USDA-NIFA Award no 2011-68002-30190Cropping Systems Coordinated Agricultural Project ClimateChangeMitigation andAdaptation inCorn-basedCroppingSystems (Project website httpsustainablecornorg)
References
[1] D Schimel J Melillo H Tian et al ldquoContribution of increasingCO2
and climate to carbon storage by ecosystems in the UnitedStatesrdquo Science vol 287 no 5460 pp 2004ndash2006 2000
[2] C Kosmas S Gerontidis and M Marathianou ldquoThe effect ofland use change on soils and vegetation over various lithologicalformations on Lesvos (Greece)rdquoCatena vol 40 no 1 pp 51ndash682000
[3] J Igbal J A Thomasson J N Jenkins P R Owens and F DWhisler ldquoSpatial variability analysis of soil physical propertiesof alluvial soilsrdquo Soil Science Society of America Journal vol 69pp 1ndash14 2005
[4] S Zhang X Zhang T Huffman X Liu and J Yang ldquoInfluenceof topography and land management on soil nutrients variabil-ity in Northeast ChinardquoNutrient Cycling in Agroecosystems vol89 no 3 pp 427ndash438 2011
8 ISRN Soil Science
[5] L J Munkholm A Garbout and S B Hansen ldquoTillage effectson topsoil structural quality assessed using X-ray CT soil coresand visual soil evaluationrdquo Soil amp Tillage Research vol 128 pp104ndash109 2013
[6] A J Franzluebbers ldquoSoil organicmatter stratification ratio as anindicator of soil qualityrdquo Soil and Tillage Research vol 66 no 2pp 95ndash106 2002
[7] L J Munkholm P Schjoslashnning K J Rasmussen and KTanderup ldquoSpatial and temporal effects of direct drilling onsoil structure in the seedling environmentrdquo Soil and TillageResearch vol 71 no 2 pp 163ndash173 2003
[8] A Swarowsky R A Dahlgren K W Tate J W Hopmans andA T OrsquoGeen ldquoCatchment-scale soil water dynamics in a medi-terranean-type oak woodlandrdquoVadose Zone Journal vol 10 no3 pp 800ndash815 2011
[9] RWebster andMAOliver StatisticalMethods in Soil and LandResource Survey Oxford University Press Oxford UK 1990
[10] A Saldana A Stein and J A Zinck ldquoSpatial variability of soilproperties at different scaleswithin three terraces of theHenaresriver (Spain)rdquo Catena vol 33 no 3-4 pp 139ndash153 1998
[11] Z M Wang K S Song B Zhang et al ldquoSpatial variabilityand affecting factors of soil nutrients in croplands of NortheastChina a case study in Dehui countyrdquo Plant Soil and Environ-ment vol 55 no 3 pp 110ndash120 2009
[12] R Webster ldquoQuantitative spatial analysis of soil in the fieldrdquoAdvances in Soil Science vol 3 pp 1ndash70 1985
[13] H Pohlmann ldquoGeostatisticalmodelling of environmental datardquoCatena vol 20 no 1-2 pp 191ndash198 1993
[14] C A Cambardella T B Moorman J M Novak et al ldquoField-scale variability of soil properties in central Iowa soilsrdquo Soil Sci-ence Society of America Journal vol 58 no 5 pp 1501ndash1511 1994
[15] E Ozgoz ldquoLong term conventional tillage effect on spatial vari-ability of some soil physical propertiesrdquo Journal of SustainableAgriculture vol 33 no 2 pp 142ndash160 2009
[16] N V Nkongolo R Hatano and V Kakembo ldquoDiffusivity mod-els and greenhouse gases fluxes from a forest pasture grasslandand corn field in Northern Hokkaido Japanrdquo Pedosphere vol20 no 6 pp 747ndash760 2010
[17] S H Ayoubi S M Zamani and F Khomali ldquoSpatial variabilityof some soil properties for site-specific farming in northernIranrdquo International Journal of Plant Production vol 2 pp 225ndash236 2007
[18] T M Burgess and R Webster ldquoOptimal interpolation andisarithmic mapping of soil properties I The semi-variogramand punctual krigingrdquo Journal of Soil amp Science vol 31 no 2pp 315ndash331 1980
[19] J M B Parfitt L C Timm E A Pauletto et al ldquoSpatial vari-ability of the chemical physical and biological properties inlowland cultivated with irrigated ricerdquo Revista Brasileira deCiencia do Solo vol 33 no 4 pp 819ndash830 2009
[20] C R Grego S R Vieira and A L Lourencao ldquoSpatial distri-bution of Pseudaletia sequax Franclemlont in triticale under no-till managementrdquo Scientia Agricola vol 63 no 4 pp 321ndash3272006
[21] G B Tesfahunegn L Tamene and P L G Vlek ldquoCatchment-scale spatial variability of soil properties and implications onsite-specific soil management in northern Ethiopiardquo Soil andTillage Research vol 117 pp 124ndash139 2011
[22] S J Grecu M B Kirkham E T Kanemasu D W Sweeney LR Stone and G A Milliken ldquoRoot growth in a claypan witha perennial-annual rotationrdquo Soil Science Society of AmericaJournal vol 52 no 2 pp 488ndash494 1988
[23] D B Myers N R Kitchen K A Sudduth R E Sharp andR J Miles ldquoSoybean root distribution related to claypan soilproperties and apparent soil electrical conductivityrdquo Crop Sci-ence vol 47 no 4 pp 1498ndash1509 2007
[24] P Jiang N R Kitchen S H Anderson E J Sadler and KA Sudduth ldquoEstimating palnt available water using the simpleinverse yield model for claypan landscapesrdquo Agronomy Journalvol 100 pp 1ndash7 2008
[25] A Utset M E Ruiz J Herrera and D P De Leon ldquoA geo-statistical method for soil salinity sample site spacingrdquo Geo-derma vol 86 no 1-2 pp 143ndash151 1998
[26] W Fu H Tunney and C Zhang ldquoSpatial variation of soilnutrients in a dairy farm and its implications for site-specificfertilizer applicationrdquo Soil and Tillage Research vol 106 no 2pp 185ndash193 2010
[27] F Lopez-Granados M Jurado-Exposito S Atenciano AGarcıa-Ferrer M Sanchez De La Orden and L Garcıa-TorresldquoSpatial variability of agricultural soil parameters in southernSpainrdquo Plant and Soil vol 246 no 1 pp 97ndash105 2002
[28] R L Hill ldquoLong-term conventional and no-tillage effects onselected soil physical propertiesrdquo Soil Science Society of AmericaJournal vol 54 no 1 pp 161ndash166 1990
[29] D E Buschiazzo J L Panigatti and PW Unger ldquoTillage effectson soil properties and crop production in the subhumid andsemiarid Argentinean Pampasrdquo Soil and Tillage Research vol49 no 1-2 pp 105ndash116 1998
[30] T Tsegaye and R L Hill ldquoIntensive tillage effects on spatialvariability of soil physical propertiesrdquo Soil Science vol 163 no2 pp 143ndash154 1998
[31] J A Gomez J V Giraldez M Pastor and E Fereres ldquoEffects oftillage method on soil physical properties infiltration and yieldin an olive orchardrdquo Soil and Tillage Research vol 52 no 3-4pp 167ndash175 1999
eliminate this error multiple samples were taken from eachsampling point Natural phenomena can vary spatially over arange of scales Variation atmicroscales smaller than the sam-pling distances will appear as a part of the nugget effect Table2 shows that the spatial correlation (range) of soil propertieswidely varied from 1m for volumetric water content (VWC)in depth four to 64m for gravimetric water content (GWC)in depth 2 However for the first and second depth (whichare agriculturally more important) the range of spatialcorrelation varied from 3m for volumetric air content (VAC)in depth 2 to 64m for GWC in depth 2 Beyond these rangesthere is no spatial dependence (autocorrelation) The spatialdependence can indicate the level of similarity or disturbanceof the soil condition According to Lopez-Granados et al [27]and Ayoubi et al [17] a large range indicates that the mea-sured soil property value is influenced by natural and anthro-pogenic factors over greater distances than parameters whichhave smaller ranges Thus a range of about 64m for GWCin this study indicates that the measured GWC values canbe influenced in the soil over greater distances as comparedto the soil parameters having smaller range (Table 2) Thismeans that soil variables with smaller range such asVWCandVACare good indicators of themore disturbed soils (themoredisturbed a soil is the more variable some soil propertiesbecome) The more variable properties have a shorter rangeof correlation The different ranges of the spatial dependenceamong the soil properties may be attributed to differencesin response to the erosionmdashdeposition factors land use-cover parent material and human interferences in the studyarea The nugget which is an indication of microvariabilitywas significantly higher for water-filled pore space (WFPS)and air-filled pore space (AFPS) when compared to theothersThis can be explained by our sampling distance whichcould not capture well their spatial dependence The lowestnugget was for GWC (Table 2) This indicates that GWChad low spatial variability within small distances Knowledgeof the range of influence for various soil properties allowsone to construct independent accurate datasets for similarareas in future soil sampling design to perform statisticalanalysis [17] This aids in determining where to resampleif necessary and design future field experiments that avoidspatial dependence Therefore for future studies aimed atcharacterizing the spatial dependency of soil properties inthe study area andor a similar area it is recommendedthat the soil properties be sampled at distances shorterthan the range found in this study Cambardella et al [14]established the classification of degree of spatial dependence(DSD) between adjacent observations of soil property gt 75to correspond to weak spatial structure In this study thesemivariograms indicated strong spatial dependence (DSDle 25) for soil physical properties such as bulk densitygravimetric water content volumetric water content totalpore space and diffusivity The rest of the soil physicalproperties (water-filled pore space Air-filled pore space andtortuosity)measured exhibited very weak spatial dependence(DSD gt 75) (Table 2) The strong spatial dependence of thesoil properties may be controlled by intrinsic variations insoil characteristics such as texture and mineralogy whereasextrinsic variations such as tillage and other soil and water
management practices may also control the variability of theweak spatially dependent parameters [14]
33 Spatial Distribution of Soil Properties across the FieldInterpolated maps portraying the distribution of soil physicalproperties in various depths are shown in Figure 3 for soilgravimetric (GWC) and volumetric (VWC) contents andwater-filled pore space (WFPS) Gravimetric water contentshowed a good spatial distribution across the field withthe highest values located around the southwestern portionof the field Volumetric water content also showed goodspatial distribution across the field with high values locatedin the northern central and southwestern portions of thefield Water-filled pore has a distribution similar to that ofvolumetric water contentThe other soil properties howevershowed very poor spatial distribution in the field This ismost probably due to their poor sill (119862
0+ 119862) model fit and
coefficient of determination (1198772
) Even though the spatialvariability was not very pronounced there were areas onthe field that had slightly higher values of these physicalproperties than the rest of the field In general bulk densitytotal pore space volumetric air content Air-filled pore spacediffusivity and tortuosity were very high in the field eventhough they did not exhibit very distinguishable variabilityThis lack of visible spatial variability is supported because thesampling distance (range) is 26m for these properties
34 Implications of Spatial Variability of Soil Physical Prop-erties on Soil Management Results of this study indicatedthat the spatial variability of soil water content (GWC andVWC) was high This can be explained among many otherreasons by soil type (clay-loam) which was able to hold morewater But with intensive tillage this soil water content couldbe adversely affected Studies have shown that tillage prac-tices can alter soil physical properties and consequently thehydrological behavior of agricultural fields especially whena similar tillage system has been practiced for a long period[15 28ndash31] Tillage intensity has also considerable effectson spatial structure and spatial variability of soil properties[15 30]Therefore this study can help determine site-specificsoil management and decision making To do so the spatialvariability of soil properties developed through kriging willbe an important tool Different ranges of spatial dependencewere noticed in the field The different ranges of the spatialdependence among the soil properties may be attributed todifferences in response to the erosionmdashdeposition factorsland use-cover parent material and human interferences inthe study areaThe different ranges can also be used in futurestudies to determine the sampling distance of different soilphysical properties on the field Also the sill (119862
0+119862) can help
determine where the variability or change in soil propertystops This will be useful especially for the irrigation pur-poses Generally with farmers facing the decision of whetheror not to till and the intensity of tillage a spatial variabilitystudy can help in this decision making Maps producedin this study can also be used for irrigation purposes asthey can clearly indicate which portion of the field needsirrigation (soil water content) To do this soil water content
Figure 3 Spatial distribution of gravimetric (GWC) and volumetric (VWC) water contents in 0ndash10 cm depth and that of water-filled porespace (WFPS) in 20 cm depth
information can be collected and analyzed geospatially toproduce field maps The process can be repeated frequentlyto obtain up-to-date soil water content information To avoidfrequent destructive sampling for water content analysisequipment that allows insitu measurements such as TDRmethods and water mark sensors can be used Since a differ-ent range of spatial dependence among soil properties showsdifferences in response to human interferences and land use-cover this will help reduce human activities that increase soilbulk density and cause soil compaction like the use of heavyequipment It can also serve as a reference for the type of cropto be grown (cover crops for erosion susceptible areas)
4 Conclusion
We assessed the spatial variability of soil physical propertiesin a clay-loam soil cropped to corn and soybean Resultsshowed that soil physical properties either decreased orincreased sharply in the second depth (due to the presence ofa smectite layer) before leveling up or dropping off but with-out reaching the first depth value in either case In additiondepending on soil physical property maps produced by krig-ing showed either goodor poor spatial distributionThe semi-variogram analysis showed the presence of a strong (le25)to weak (gt75) spatial dependence of soil properties Ourunderstanding of the behavior of soil properties in this studyprovides new insights for soil site-specific management in
addressing issues such as ldquowhere to place the proper interven-tionsrdquo (tillage irrigation and crop type to be grown)
Acknowledgment
This research is part of a regional collaborative projectsupported by the USDA-NIFA Award no 2011-68002-30190Cropping Systems Coordinated Agricultural Project ClimateChangeMitigation andAdaptation inCorn-basedCroppingSystems (Project website httpsustainablecornorg)
References
[1] D Schimel J Melillo H Tian et al ldquoContribution of increasingCO2
and climate to carbon storage by ecosystems in the UnitedStatesrdquo Science vol 287 no 5460 pp 2004ndash2006 2000
[2] C Kosmas S Gerontidis and M Marathianou ldquoThe effect ofland use change on soils and vegetation over various lithologicalformations on Lesvos (Greece)rdquoCatena vol 40 no 1 pp 51ndash682000
[3] J Igbal J A Thomasson J N Jenkins P R Owens and F DWhisler ldquoSpatial variability analysis of soil physical propertiesof alluvial soilsrdquo Soil Science Society of America Journal vol 69pp 1ndash14 2005
[4] S Zhang X Zhang T Huffman X Liu and J Yang ldquoInfluenceof topography and land management on soil nutrients variabil-ity in Northeast ChinardquoNutrient Cycling in Agroecosystems vol89 no 3 pp 427ndash438 2011
8 ISRN Soil Science
[5] L J Munkholm A Garbout and S B Hansen ldquoTillage effectson topsoil structural quality assessed using X-ray CT soil coresand visual soil evaluationrdquo Soil amp Tillage Research vol 128 pp104ndash109 2013
[6] A J Franzluebbers ldquoSoil organicmatter stratification ratio as anindicator of soil qualityrdquo Soil and Tillage Research vol 66 no 2pp 95ndash106 2002
[7] L J Munkholm P Schjoslashnning K J Rasmussen and KTanderup ldquoSpatial and temporal effects of direct drilling onsoil structure in the seedling environmentrdquo Soil and TillageResearch vol 71 no 2 pp 163ndash173 2003
[8] A Swarowsky R A Dahlgren K W Tate J W Hopmans andA T OrsquoGeen ldquoCatchment-scale soil water dynamics in a medi-terranean-type oak woodlandrdquoVadose Zone Journal vol 10 no3 pp 800ndash815 2011
[9] RWebster andMAOliver StatisticalMethods in Soil and LandResource Survey Oxford University Press Oxford UK 1990
[10] A Saldana A Stein and J A Zinck ldquoSpatial variability of soilproperties at different scaleswithin three terraces of theHenaresriver (Spain)rdquo Catena vol 33 no 3-4 pp 139ndash153 1998
[11] Z M Wang K S Song B Zhang et al ldquoSpatial variabilityand affecting factors of soil nutrients in croplands of NortheastChina a case study in Dehui countyrdquo Plant Soil and Environ-ment vol 55 no 3 pp 110ndash120 2009
[12] R Webster ldquoQuantitative spatial analysis of soil in the fieldrdquoAdvances in Soil Science vol 3 pp 1ndash70 1985
[13] H Pohlmann ldquoGeostatisticalmodelling of environmental datardquoCatena vol 20 no 1-2 pp 191ndash198 1993
[14] C A Cambardella T B Moorman J M Novak et al ldquoField-scale variability of soil properties in central Iowa soilsrdquo Soil Sci-ence Society of America Journal vol 58 no 5 pp 1501ndash1511 1994
[15] E Ozgoz ldquoLong term conventional tillage effect on spatial vari-ability of some soil physical propertiesrdquo Journal of SustainableAgriculture vol 33 no 2 pp 142ndash160 2009
[16] N V Nkongolo R Hatano and V Kakembo ldquoDiffusivity mod-els and greenhouse gases fluxes from a forest pasture grasslandand corn field in Northern Hokkaido Japanrdquo Pedosphere vol20 no 6 pp 747ndash760 2010
[17] S H Ayoubi S M Zamani and F Khomali ldquoSpatial variabilityof some soil properties for site-specific farming in northernIranrdquo International Journal of Plant Production vol 2 pp 225ndash236 2007
[18] T M Burgess and R Webster ldquoOptimal interpolation andisarithmic mapping of soil properties I The semi-variogramand punctual krigingrdquo Journal of Soil amp Science vol 31 no 2pp 315ndash331 1980
[19] J M B Parfitt L C Timm E A Pauletto et al ldquoSpatial vari-ability of the chemical physical and biological properties inlowland cultivated with irrigated ricerdquo Revista Brasileira deCiencia do Solo vol 33 no 4 pp 819ndash830 2009
[20] C R Grego S R Vieira and A L Lourencao ldquoSpatial distri-bution of Pseudaletia sequax Franclemlont in triticale under no-till managementrdquo Scientia Agricola vol 63 no 4 pp 321ndash3272006
[21] G B Tesfahunegn L Tamene and P L G Vlek ldquoCatchment-scale spatial variability of soil properties and implications onsite-specific soil management in northern Ethiopiardquo Soil andTillage Research vol 117 pp 124ndash139 2011
[22] S J Grecu M B Kirkham E T Kanemasu D W Sweeney LR Stone and G A Milliken ldquoRoot growth in a claypan witha perennial-annual rotationrdquo Soil Science Society of AmericaJournal vol 52 no 2 pp 488ndash494 1988
[23] D B Myers N R Kitchen K A Sudduth R E Sharp andR J Miles ldquoSoybean root distribution related to claypan soilproperties and apparent soil electrical conductivityrdquo Crop Sci-ence vol 47 no 4 pp 1498ndash1509 2007
[24] P Jiang N R Kitchen S H Anderson E J Sadler and KA Sudduth ldquoEstimating palnt available water using the simpleinverse yield model for claypan landscapesrdquo Agronomy Journalvol 100 pp 1ndash7 2008
[25] A Utset M E Ruiz J Herrera and D P De Leon ldquoA geo-statistical method for soil salinity sample site spacingrdquo Geo-derma vol 86 no 1-2 pp 143ndash151 1998
[26] W Fu H Tunney and C Zhang ldquoSpatial variation of soilnutrients in a dairy farm and its implications for site-specificfertilizer applicationrdquo Soil and Tillage Research vol 106 no 2pp 185ndash193 2010
[27] F Lopez-Granados M Jurado-Exposito S Atenciano AGarcıa-Ferrer M Sanchez De La Orden and L Garcıa-TorresldquoSpatial variability of agricultural soil parameters in southernSpainrdquo Plant and Soil vol 246 no 1 pp 97ndash105 2002
[28] R L Hill ldquoLong-term conventional and no-tillage effects onselected soil physical propertiesrdquo Soil Science Society of AmericaJournal vol 54 no 1 pp 161ndash166 1990
[29] D E Buschiazzo J L Panigatti and PW Unger ldquoTillage effectson soil properties and crop production in the subhumid andsemiarid Argentinean Pampasrdquo Soil and Tillage Research vol49 no 1-2 pp 105ndash116 1998
[30] T Tsegaye and R L Hill ldquoIntensive tillage effects on spatialvariability of soil physical propertiesrdquo Soil Science vol 163 no2 pp 143ndash154 1998
[31] J A Gomez J V Giraldez M Pastor and E Fereres ldquoEffects oftillage method on soil physical properties infiltration and yieldin an olive orchardrdquo Soil and Tillage Research vol 52 no 3-4pp 167ndash175 1999
Figure 3 Spatial distribution of gravimetric (GWC) and volumetric (VWC) water contents in 0ndash10 cm depth and that of water-filled porespace (WFPS) in 20 cm depth
information can be collected and analyzed geospatially toproduce field maps The process can be repeated frequentlyto obtain up-to-date soil water content information To avoidfrequent destructive sampling for water content analysisequipment that allows insitu measurements such as TDRmethods and water mark sensors can be used Since a differ-ent range of spatial dependence among soil properties showsdifferences in response to human interferences and land use-cover this will help reduce human activities that increase soilbulk density and cause soil compaction like the use of heavyequipment It can also serve as a reference for the type of cropto be grown (cover crops for erosion susceptible areas)
4 Conclusion
We assessed the spatial variability of soil physical propertiesin a clay-loam soil cropped to corn and soybean Resultsshowed that soil physical properties either decreased orincreased sharply in the second depth (due to the presence ofa smectite layer) before leveling up or dropping off but with-out reaching the first depth value in either case In additiondepending on soil physical property maps produced by krig-ing showed either goodor poor spatial distributionThe semi-variogram analysis showed the presence of a strong (le25)to weak (gt75) spatial dependence of soil properties Ourunderstanding of the behavior of soil properties in this studyprovides new insights for soil site-specific management in
addressing issues such as ldquowhere to place the proper interven-tionsrdquo (tillage irrigation and crop type to be grown)
Acknowledgment
This research is part of a regional collaborative projectsupported by the USDA-NIFA Award no 2011-68002-30190Cropping Systems Coordinated Agricultural Project ClimateChangeMitigation andAdaptation inCorn-basedCroppingSystems (Project website httpsustainablecornorg)
References
[1] D Schimel J Melillo H Tian et al ldquoContribution of increasingCO2
and climate to carbon storage by ecosystems in the UnitedStatesrdquo Science vol 287 no 5460 pp 2004ndash2006 2000
[2] C Kosmas S Gerontidis and M Marathianou ldquoThe effect ofland use change on soils and vegetation over various lithologicalformations on Lesvos (Greece)rdquoCatena vol 40 no 1 pp 51ndash682000
[3] J Igbal J A Thomasson J N Jenkins P R Owens and F DWhisler ldquoSpatial variability analysis of soil physical propertiesof alluvial soilsrdquo Soil Science Society of America Journal vol 69pp 1ndash14 2005
[4] S Zhang X Zhang T Huffman X Liu and J Yang ldquoInfluenceof topography and land management on soil nutrients variabil-ity in Northeast ChinardquoNutrient Cycling in Agroecosystems vol89 no 3 pp 427ndash438 2011
8 ISRN Soil Science
[5] L J Munkholm A Garbout and S B Hansen ldquoTillage effectson topsoil structural quality assessed using X-ray CT soil coresand visual soil evaluationrdquo Soil amp Tillage Research vol 128 pp104ndash109 2013
[6] A J Franzluebbers ldquoSoil organicmatter stratification ratio as anindicator of soil qualityrdquo Soil and Tillage Research vol 66 no 2pp 95ndash106 2002
[7] L J Munkholm P Schjoslashnning K J Rasmussen and KTanderup ldquoSpatial and temporal effects of direct drilling onsoil structure in the seedling environmentrdquo Soil and TillageResearch vol 71 no 2 pp 163ndash173 2003
[8] A Swarowsky R A Dahlgren K W Tate J W Hopmans andA T OrsquoGeen ldquoCatchment-scale soil water dynamics in a medi-terranean-type oak woodlandrdquoVadose Zone Journal vol 10 no3 pp 800ndash815 2011
[9] RWebster andMAOliver StatisticalMethods in Soil and LandResource Survey Oxford University Press Oxford UK 1990
[10] A Saldana A Stein and J A Zinck ldquoSpatial variability of soilproperties at different scaleswithin three terraces of theHenaresriver (Spain)rdquo Catena vol 33 no 3-4 pp 139ndash153 1998
[11] Z M Wang K S Song B Zhang et al ldquoSpatial variabilityand affecting factors of soil nutrients in croplands of NortheastChina a case study in Dehui countyrdquo Plant Soil and Environ-ment vol 55 no 3 pp 110ndash120 2009
[12] R Webster ldquoQuantitative spatial analysis of soil in the fieldrdquoAdvances in Soil Science vol 3 pp 1ndash70 1985
[13] H Pohlmann ldquoGeostatisticalmodelling of environmental datardquoCatena vol 20 no 1-2 pp 191ndash198 1993
[14] C A Cambardella T B Moorman J M Novak et al ldquoField-scale variability of soil properties in central Iowa soilsrdquo Soil Sci-ence Society of America Journal vol 58 no 5 pp 1501ndash1511 1994
[15] E Ozgoz ldquoLong term conventional tillage effect on spatial vari-ability of some soil physical propertiesrdquo Journal of SustainableAgriculture vol 33 no 2 pp 142ndash160 2009
[16] N V Nkongolo R Hatano and V Kakembo ldquoDiffusivity mod-els and greenhouse gases fluxes from a forest pasture grasslandand corn field in Northern Hokkaido Japanrdquo Pedosphere vol20 no 6 pp 747ndash760 2010
[17] S H Ayoubi S M Zamani and F Khomali ldquoSpatial variabilityof some soil properties for site-specific farming in northernIranrdquo International Journal of Plant Production vol 2 pp 225ndash236 2007
[18] T M Burgess and R Webster ldquoOptimal interpolation andisarithmic mapping of soil properties I The semi-variogramand punctual krigingrdquo Journal of Soil amp Science vol 31 no 2pp 315ndash331 1980
[19] J M B Parfitt L C Timm E A Pauletto et al ldquoSpatial vari-ability of the chemical physical and biological properties inlowland cultivated with irrigated ricerdquo Revista Brasileira deCiencia do Solo vol 33 no 4 pp 819ndash830 2009
[20] C R Grego S R Vieira and A L Lourencao ldquoSpatial distri-bution of Pseudaletia sequax Franclemlont in triticale under no-till managementrdquo Scientia Agricola vol 63 no 4 pp 321ndash3272006
[21] G B Tesfahunegn L Tamene and P L G Vlek ldquoCatchment-scale spatial variability of soil properties and implications onsite-specific soil management in northern Ethiopiardquo Soil andTillage Research vol 117 pp 124ndash139 2011
[22] S J Grecu M B Kirkham E T Kanemasu D W Sweeney LR Stone and G A Milliken ldquoRoot growth in a claypan witha perennial-annual rotationrdquo Soil Science Society of AmericaJournal vol 52 no 2 pp 488ndash494 1988
[23] D B Myers N R Kitchen K A Sudduth R E Sharp andR J Miles ldquoSoybean root distribution related to claypan soilproperties and apparent soil electrical conductivityrdquo Crop Sci-ence vol 47 no 4 pp 1498ndash1509 2007
[24] P Jiang N R Kitchen S H Anderson E J Sadler and KA Sudduth ldquoEstimating palnt available water using the simpleinverse yield model for claypan landscapesrdquo Agronomy Journalvol 100 pp 1ndash7 2008
[25] A Utset M E Ruiz J Herrera and D P De Leon ldquoA geo-statistical method for soil salinity sample site spacingrdquo Geo-derma vol 86 no 1-2 pp 143ndash151 1998
[26] W Fu H Tunney and C Zhang ldquoSpatial variation of soilnutrients in a dairy farm and its implications for site-specificfertilizer applicationrdquo Soil and Tillage Research vol 106 no 2pp 185ndash193 2010
[27] F Lopez-Granados M Jurado-Exposito S Atenciano AGarcıa-Ferrer M Sanchez De La Orden and L Garcıa-TorresldquoSpatial variability of agricultural soil parameters in southernSpainrdquo Plant and Soil vol 246 no 1 pp 97ndash105 2002
[28] R L Hill ldquoLong-term conventional and no-tillage effects onselected soil physical propertiesrdquo Soil Science Society of AmericaJournal vol 54 no 1 pp 161ndash166 1990
[29] D E Buschiazzo J L Panigatti and PW Unger ldquoTillage effectson soil properties and crop production in the subhumid andsemiarid Argentinean Pampasrdquo Soil and Tillage Research vol49 no 1-2 pp 105ndash116 1998
[30] T Tsegaye and R L Hill ldquoIntensive tillage effects on spatialvariability of soil physical propertiesrdquo Soil Science vol 163 no2 pp 143ndash154 1998
[31] J A Gomez J V Giraldez M Pastor and E Fereres ldquoEffects oftillage method on soil physical properties infiltration and yieldin an olive orchardrdquo Soil and Tillage Research vol 52 no 3-4pp 167ndash175 1999
[5] L J Munkholm A Garbout and S B Hansen ldquoTillage effectson topsoil structural quality assessed using X-ray CT soil coresand visual soil evaluationrdquo Soil amp Tillage Research vol 128 pp104ndash109 2013
[6] A J Franzluebbers ldquoSoil organicmatter stratification ratio as anindicator of soil qualityrdquo Soil and Tillage Research vol 66 no 2pp 95ndash106 2002
[7] L J Munkholm P Schjoslashnning K J Rasmussen and KTanderup ldquoSpatial and temporal effects of direct drilling onsoil structure in the seedling environmentrdquo Soil and TillageResearch vol 71 no 2 pp 163ndash173 2003
[8] A Swarowsky R A Dahlgren K W Tate J W Hopmans andA T OrsquoGeen ldquoCatchment-scale soil water dynamics in a medi-terranean-type oak woodlandrdquoVadose Zone Journal vol 10 no3 pp 800ndash815 2011
[9] RWebster andMAOliver StatisticalMethods in Soil and LandResource Survey Oxford University Press Oxford UK 1990
[10] A Saldana A Stein and J A Zinck ldquoSpatial variability of soilproperties at different scaleswithin three terraces of theHenaresriver (Spain)rdquo Catena vol 33 no 3-4 pp 139ndash153 1998
[11] Z M Wang K S Song B Zhang et al ldquoSpatial variabilityand affecting factors of soil nutrients in croplands of NortheastChina a case study in Dehui countyrdquo Plant Soil and Environ-ment vol 55 no 3 pp 110ndash120 2009
[12] R Webster ldquoQuantitative spatial analysis of soil in the fieldrdquoAdvances in Soil Science vol 3 pp 1ndash70 1985
[13] H Pohlmann ldquoGeostatisticalmodelling of environmental datardquoCatena vol 20 no 1-2 pp 191ndash198 1993
[14] C A Cambardella T B Moorman J M Novak et al ldquoField-scale variability of soil properties in central Iowa soilsrdquo Soil Sci-ence Society of America Journal vol 58 no 5 pp 1501ndash1511 1994
[15] E Ozgoz ldquoLong term conventional tillage effect on spatial vari-ability of some soil physical propertiesrdquo Journal of SustainableAgriculture vol 33 no 2 pp 142ndash160 2009
[16] N V Nkongolo R Hatano and V Kakembo ldquoDiffusivity mod-els and greenhouse gases fluxes from a forest pasture grasslandand corn field in Northern Hokkaido Japanrdquo Pedosphere vol20 no 6 pp 747ndash760 2010
[17] S H Ayoubi S M Zamani and F Khomali ldquoSpatial variabilityof some soil properties for site-specific farming in northernIranrdquo International Journal of Plant Production vol 2 pp 225ndash236 2007
[18] T M Burgess and R Webster ldquoOptimal interpolation andisarithmic mapping of soil properties I The semi-variogramand punctual krigingrdquo Journal of Soil amp Science vol 31 no 2pp 315ndash331 1980
[19] J M B Parfitt L C Timm E A Pauletto et al ldquoSpatial vari-ability of the chemical physical and biological properties inlowland cultivated with irrigated ricerdquo Revista Brasileira deCiencia do Solo vol 33 no 4 pp 819ndash830 2009
[20] C R Grego S R Vieira and A L Lourencao ldquoSpatial distri-bution of Pseudaletia sequax Franclemlont in triticale under no-till managementrdquo Scientia Agricola vol 63 no 4 pp 321ndash3272006
[21] G B Tesfahunegn L Tamene and P L G Vlek ldquoCatchment-scale spatial variability of soil properties and implications onsite-specific soil management in northern Ethiopiardquo Soil andTillage Research vol 117 pp 124ndash139 2011
[22] S J Grecu M B Kirkham E T Kanemasu D W Sweeney LR Stone and G A Milliken ldquoRoot growth in a claypan witha perennial-annual rotationrdquo Soil Science Society of AmericaJournal vol 52 no 2 pp 488ndash494 1988
[23] D B Myers N R Kitchen K A Sudduth R E Sharp andR J Miles ldquoSoybean root distribution related to claypan soilproperties and apparent soil electrical conductivityrdquo Crop Sci-ence vol 47 no 4 pp 1498ndash1509 2007
[24] P Jiang N R Kitchen S H Anderson E J Sadler and KA Sudduth ldquoEstimating palnt available water using the simpleinverse yield model for claypan landscapesrdquo Agronomy Journalvol 100 pp 1ndash7 2008
[25] A Utset M E Ruiz J Herrera and D P De Leon ldquoA geo-statistical method for soil salinity sample site spacingrdquo Geo-derma vol 86 no 1-2 pp 143ndash151 1998
[26] W Fu H Tunney and C Zhang ldquoSpatial variation of soilnutrients in a dairy farm and its implications for site-specificfertilizer applicationrdquo Soil and Tillage Research vol 106 no 2pp 185ndash193 2010
[27] F Lopez-Granados M Jurado-Exposito S Atenciano AGarcıa-Ferrer M Sanchez De La Orden and L Garcıa-TorresldquoSpatial variability of agricultural soil parameters in southernSpainrdquo Plant and Soil vol 246 no 1 pp 97ndash105 2002
[28] R L Hill ldquoLong-term conventional and no-tillage effects onselected soil physical propertiesrdquo Soil Science Society of AmericaJournal vol 54 no 1 pp 161ndash166 1990
[29] D E Buschiazzo J L Panigatti and PW Unger ldquoTillage effectson soil properties and crop production in the subhumid andsemiarid Argentinean Pampasrdquo Soil and Tillage Research vol49 no 1-2 pp 105ndash116 1998
[30] T Tsegaye and R L Hill ldquoIntensive tillage effects on spatialvariability of soil physical propertiesrdquo Soil Science vol 163 no2 pp 143ndash154 1998
[31] J A Gomez J V Giraldez M Pastor and E Fereres ldquoEffects oftillage method on soil physical properties infiltration and yieldin an olive orchardrdquo Soil and Tillage Research vol 52 no 3-4pp 167ndash175 1999
