Spatial heterogeneity of soil physico-chemical properties in contrasting wetland soils in two agro-ecological zones of Lesotho.
Wetlands are complex and fragile ecosystems with considerable spatial variability (Reddy 1993). They are heterogeneous environments which can exhibit substantial spatial and temporal variability in soil properties (Lyons et al. 1998; Stolt et al. 2001; Bruland and Richardson 2004; Grunwald et al. 2006; Cohen et al. 2008). This variability can have important consequences for wetland biota and biogeochemistry. Understanding of the soil spatial patterns in wetland ecosystems is made difficult by the number of factors that affect soil properties and which operate at different spatial and temporal scales. Climatic factors such as precipitation and temperature, land form, topography, and complex interactions among physical, chemical, and biological processes play key roles in configuring soil properties (Sinowski and Auerswald 1999; Johnson et al. 2000). Human activities affect self-organisation in wetlands, which in turn affects spatial patterns of soil properties such as pH, nutrient concentrations, and organic matter content. The principal morphological differences between wetland soils and upland soils can be attributed to the fact that, in wetlands, specific biogeochemical processes take place because soil water frequently fills the soil pores and void spaces, resulting in saturated conditions (Hurt and Carlisle 2005).
Research on soil variability in the tropics has commonly used conventional statistics (e.g. coefficient of variation) to assess soil variability. Such statistics assume that variation is randomly distributed within mapping units. However, soil properties frequently exhibit spatial dependency. Generally, samples collected close to one another are more similar than samples collected at greater distances. Parametric statistics are inadequate for analysis of spatially dependent variables because they assume that measured observations are independent of their distribution in space (Kravchenko et al. 2006). Soil variability has been investigated by many researchers in the temperate zone (e.g. Wilding and Drees 1983; Cahn et al. 1994; Cambardella and Karlen 1999; Eltaib et al. 2002); however, there has been limited research on tropical soils (Ogunkunle 1993; Adderley et al. 1997; Wuddivira et al. 2000; Okae-Anti 2001; Akinbola et al. 2006) or in various countries (Cassel et al. 2000; Gaston et al. 2001; Corstanje et al. 2006; Iqbal et al. 2005). To our knowledge, there are no reports of studies on soil variability in the upland and wetland soils of Lesotho, Southern Africa.
It has been reported that human activity has adversely affected global carbon (C), nitrogen (N), and phosphorus (P) cycles, and this has contributed to climate alteration that will generate discernible feedbacks to all organisms and ecosystems on the Earth (Mitsch and Gosselink 2000; Wang et al. 2006; He et al. 2008). Understanding the stores and storage potential of C, N, and P has helped to discern how ecosystems will respond to natural and anthropogenic disturbances under different management strategies (He et al. 2008).
There has been migration of population, along with livestock, from other agro-ecological zones of Lesotho to the lowlands. As a result, many wetlands are being converted to grazing land. The rapid expansion of livestock numbers and the conversion of wetlands to cropped lands have caused significant degradation of these fragile ecosystems, accelerating peat decomposition and C loss, and causing a sharp decline in wetland biodiversity, despite some measures (e.g. banning pasture) to restore and protect these wetlands. Studies on the effects of human activities following wetland conversion on soil physico-chemical properties and on spatial patterns of soil properties are limited. Information on these effects will support decision making on best management of such wetlands for present and future generations. Brouwer et al. (1993) reported that knowledge of spatial variability within a farmer's field may help to reduce risk of severe and irreversible environmental degradation by revealing vulnerable areas. Similarly, in wetland soils, knowledge of the soil variability within and between fields can reduce the risk of irreversible environmental degradation.
In wetlands where spatial patterns in soil properties have been estimated, the information has proved useful. For example, in the case of the soil P spatial patterns in the Everglades, USA, the information provided insight on the extent and severity of P enrichment zones and processes that govern the enrichment processes (DeBusk et al. 1994, 2001; Grunwald et al. 2006). Spatial variability is also important when considering the environmental and ecological functions of a wetland (Stolt et al. 2001). Some of these functions, such as production of vegetal materials for handicraft and building components, fishing and hunting, cattle grazing (especially in summer), filtration of nutrients and other chemicals from river water, tourism, and absorbing temporary large quantities of water to release slowly, cannot be estimated directly.
Spatial variability in wetland soils has also provided insight into underlying ecosystem processes and may itself indicate wetland condition (Cohen et al. 2008; Bai et al. 2010). As such, it is important to distinguish systematic variability from random variability and determine the relative importance of each (Stolt et al. 2001). Systematic spatial relationships in wetland soils are the result of differences in parent material, elevation, erosional or depositional environment, frequency of flooding, vegetation, pedogenic effects, and hydrology (Johnston et al. 1984; Hayati and Proctor 1990; Gaston et al. 1990; Farrish 1991; Reese and Moorhead 1996). However, random effects are attributed to unrecognised differences in these parameters, as well as differences due to sampling and laboratory error (Wilding and Drees 1983). These random effects often obscure or confound soil-elevation, soil-vegetation, or soil-hydrology relationships; therefore, to understand spatial relationships in wetland soils, random variability needs to be recognised and separated from systematic variability (Stolt et al. 2001).
Numerous processes can influence soil spatial patterns at a particular location (Stolt et al. 2001), and Bruland and Richardson (2004) illustrate that wetland type may strongly affect spatial pattern. The uneven distribution of soil characteristics such as nutrient availability, organic content, and mineral content implicitly reflects the processes that occur within the larger ecosystem (Corstanje et al. 2006). Although the importance of spatial heterogeneity is well recognised in wetlands, the scale or extent to which it occurs and how it might affect coexistence and diversity of species is poorly understood in Lesotho. There is little information on the effects of human activities (i.e. livestock grazing and watering) on soil physico-chemical properties in Lesotho or their spatial variability. Most studies on soil spatial variability in other regions have been used to provide accurate information for site-specific recommendations. Furthermore, there is little published scientific information on the status of spatial variability of soil physico-chemical properties of wetlands in Lesotho. As such, the information presented in this study may be used to design efficient monitoring and management schemes by wetland managers. Bai et al. (2010) reported that the distributions and dynamics of P forms in soil, especially in the land/inland-water ecotone, can be significantly impacted by various biogeochemical and environmental factors (i.e. soil moisture, soil organic matter, and clay content). Thus, this study was conducted specifically to evaluate spatial variability of soil physico-chemical properties in two wetlands in two different agro-ecological zones of Lesotho with different levels of anthropogenic impacts.
Materials and methods
Two locations were selected for this study: Thaba-Putsoa, 70 km south-east of the capital Maseru (27[degrees]58.234'E, 29[degrees]25.798'S; elevation 2638m); and T'sakholo, ~76km south of Masem (27[degrees]10.360'E, 29[degrees]40.469'S; elevation 1570m).
Thaba-Putsoa is in the Mountain agro-ecological zone (AEZ), while T'sakholo is in the Lowland AEZ. Lesotho is generally considered a semi-arid country with unusually high levels of climate variability. Thaba-Putsoa is characterised by very low temperatures in winter, ranging between -8[degrees]C and 7[degrees]C with frequent occurrences of snow. Mean annual temperatures range between -8[degrees]C and 30[degrees]C, and the highest annual rainfall (1000-1400mm). The area is represented by a sequence of clastic sedimentary formations (Burgersdorp formation, Molteno formation, Elliot formation, and Clarens formation) overlain by a laterally continuous section of basalt up to 1600 m thick. Lithosols are the major soil group in this area. T'sakholo is characterised by maximum temperatures varying between 32[degrees]C in summer and -7[degrees]C in winter; average temperatures are 25[degrees]C and 15[degrees]C, respectively. Rainfall occurs predominantly between October and April, ranging between 600 and 900 mm annually. The geology of the site consists of rocks belonging to the Burgersdorp formation, which underlies the western part of the country, with highly erodible duplex soils characterising the area. General site descriptions are given in Table 1.
The level of impact on wetlands in these two AEZs was characterised as low, medium, or high, based on local land-use characteristics (Hughes 1995; Teels and Adamus 2001) or the intensity of anthropogenic pressures such as mining, smelting, industrial pollution, and livestock grazing/watering. Low-level impacted wetlands has little (<5%) or no agricultural activity within 150 m of the wetland boundary (Chipps et al. 2006), whereas high-level impacted wetlands had agricultural activities within 10m of at least one-third (33%) of the wetland boundary, and the medium-level impacted wetlands had agricultural activities within 10 m of 5-32% of the wetland boundary. The wetland at T'sakholo is characterised as high-level impacted, whereas that at Thaba-Putsoa is low-level impacted. The land-use types in the two wetlands are grazing and livestock watering.
Soils were sampled along two transects running across each wetland. At T'sakholo, sampling was from the wetland centre and extended outwards into the stream-banks. At Thaba-Putsoa, sampling was at the wetland centre and extended outwards into the surrounding upland. At T'sakholo, samples were also collected along the stream-bank. In general, the transects ranged between 250 and 700m and soil samples were collected at intervals of 50 m in mini-pits dug to a depth of [less than or equal to]0.5 m. Data recorded on the morphology of the sites included drainage, vegetation, land use, and degree of erosion. The soil properties described included colour, horizon boundary, structure, consistency, texture, mottles, and roots density. Soil samples were taken from the last horizon up to the first horizon in each profile pit, and the maximum depth of soil examination was different for each pit. In total, 76 (T'sakholo wetlands) and 27 (Thaba-Putsoa) soil samples were used. The soils were classified using USDA Soil Taxonomy (Soil Survey Staff 1999).
In each wetland, samples were collected, labelled, bagged, and transported to the laboratory for routine soil analysis. The soil samples were air-dried for 48 h and crushed to pass through a 2-mm sieve, and analysed for the following parameters: particle size by the hydrometer method (Bouyoucus 1962); soil pH in water, using a glass electrode pH meter at 1 : 2.5 soil : water ratio; soil organic carbon (SOC) by the method of Walkley and Black (1934); available P by Bray and Kurtz No. 1 method (Bray and Kurtz 1945); cation exchange capacity (CEC) using ammonium acetate at pH 7. Base cations ([Ca.sup.2+], [Na.sup.+], [Mg.sup.2+], [K.sup.+]) were extracted using 1 N N[H.sub.4]OAc and the filtered extracts were all determined with a flame atomic absorption spectrometer (AAnalysed 200, PerkinElmer Inc., Waltham, MA). The SOC pool (C-pool, kg C [m.sup.-2]) was calculated using a relationship as given by Wairiu and Lal (2003):
C - pool = d x BD x C content
where d is soil layer thickness (m), BD is bulk density (kg [m.sup.-3]), and C content units are g [g.sup.-1].
Statistical analyses and mapping
Classical statistics provides the overall variability of the soil properties; however, it does not provide the spatial trend. Data collected from these two wetland were subjected to summary statistics (mean, range, standard deviation (s.d.), skewness) using the Means Procedure of Statistical Analysis Systems (SAS Institute 1999). Data distributions were tested for normality. If data were not normally distributed, they were log-transformed. In addition the mean (x) and coefficient of variation (CV) for each property along transects were also calculated, where CV=(s.d./x) x 100. The higher the CV, the more variable is the property (Wilding and Drees 1978; Wilding 1988). The descriptive statistics of the soil data suggested that soil properties at the sites were all normally distributed (skewness of between 1 and -2) except exchangeable Na and Mg (T'sakholo rangeland), exchangeable Na and Ca (Thaba-Putsoa rangeland), and all base cations (T' sakholo stream-bank); these data were log-transformed as the skewness ranged between 1.03 and 6.06 before the calculation of semi-variance (Baxter et al. 2003). Pearson correlation analysis was implemented to determine the relationship between the soil properties. Geostatistical analysis was performed using GS+ Geostatistics for the Environmental Sciences Version 9 software (Gamma Design Software, LLC, Plainwell, MI) to characterise the spatial variability in soil properties. This analysis produces variograms which reveal random and structured aspects of spatial dependence in a dataset of multiple samples collected at increasing distances from each other (the lag interval). A semivariogram (also known as a variogram) is a graphic representation of spatial autocorrelation that is made by plotting the semivariance for several distance intervals (Robertson and Gross 1994). The semivariance is calculated as:
[gamma](h) = [[1/2(h)] [SIGMA][[[Z.sub,i] - [Z.sub.i+h]].sup.2]
where [gamma](h) is semivariance for interval distance class h; [Z.sub.i] is measured sample value at point i; [Z.sub.i+h] is measured sample value at point I + h; N(h) is total number of sample couples for the lag interval h.
The spatial structure of each variable has been defined from semivariogram parameters: nugget, sill (or total semivariance), and range. Nugget is the variance at distance zero and represents the experimental error, and sill is the semivariance value at which the semivariogram reaches the upper bound after its initial increase. The nugget to sill ratio quantifies the importance of the random component and provides a quantitative estimation of the spatial dependence. According to Cambardella et al. (1994), nugget to sill ratios can be grouped into three classes: (i) <25%, which means strong spatial dependence; (ii) 25-75%, moderate spatial dependence; (iii) >75%, spatially independent or pure nugget (i.e. when slopes of semivariograms are close to zero). The ratio of the spatially structured variance to the sample variance ([C.sub.o]/C + [C.sub.0]) reveals the degree of spatial dependence as outlined by Robertson and Gross (1994). The ecologically significant functions of a semivariance analysis of soil resources are to determine whether spatial dependence or patchiness exists for a resource, how distinct the patches are, and at what scale they occur. Contour maps of soil kriging estimates were prepared using Surfer 8.0 (Golden Software Co., Golden, CO) for spatial distribution of silt to clay ratio, organic carbon, and SOC pool. The silt to clay ratio is an indicator of wetland degradation, while the organic carbon and SOC pool are indicators of wetland health. Point kriging with no search radius was used as an unbiased, weighted, linear interpolation method that minimises total parameter variance by incorporating semivariogram functions to create contour maps (Isaaks and Srivastava 1989). The Stepwise Multiple Regression Analysis (SMRA) (Prog Reg procedure; SAS Institute 1999) was used to examine which of the soil properties accounted for most variation in each site.
Soil morphological properties
Description of the sites' morphological properties in terms of land use, colour, horizon boundary, texture, structure, consistency, concretions, roots, and drainage showed that these properties differed greatly between sites (results not presented). In Thaba-Putsoa, soils exhibited mainly black and very dark greyish colours, whereas in T'sakholo, most horizons had light-brown to brown and reddish brown colours. Many of the horizons in Thaba-Putsoa had hues of 10YR and 5Y with low chroma ([less than or equal to]3) colours, which reflects poor drainage or seasonal mottling. Features such as grey or low chroma observed in Thaba-Putsoa indicate soil wetness brought about by oxidation-reduction cycles due to groundwater fluctuations. Wetland soils in a reduced state typically have a dark, grey, mottled appearance with chroma colours [less than or equal to]2. In T'sakholo, soil colour ranged between light-brown to brown and reddish brown with hues 2.5Y, 7.5YR, and 10YR with chroma values 2-8. These colours indicate a relatively high amount of iron-oxide, which may be due to the parent material. Boundaries ranged mostly between wavy and clear on both sites with mainly sandy-loam texture in Thaba-Putsoa and a fairly variable texture (sandy, sandy-clay, silt-loam, and clay) in T'sakholo. Structure was mainly crumb and single-grain at both sites, with more concretions in T'sakholo than in Thaba-Putsoa. Both sites had many roots in the top horizons, decreasing down the profile. Drainage was poor in T'sakholo and moderate in Thaba-Putsoa.
Summary statistics are presented in Table 2 for all land-use types at both sites. Results showed that the standard deviation was higher for soil physical properties than chemical properties. In Thaba-Putsoa, soil physical properties were more negatively skewed than chemical properties. In T'sakholo, only percentage silt and clay and pH, [Na.sup.+], and [K.sup.+] were negatively skewed in rangeland, whereas along the stream-bank, percentage sand and CEC were negatively skewed. The SOC values were 17.20 [+ or -] 5.34 (Thaba-Putsoa), 1.93 [+ or -] 0.85 (T'sakholo rangeland), and 1.12 [+ or -] 1.22 (T'sakholo stream-bank). The SOC in Thaba-Putsoa was less than half that reported by Bai et al. (2010) for alpine wetland in the south-eastern Qinghai-Tibet plateau (35.81 [+ or -] 3.24). The value reported by Bai et al. (2010) was ~19 and ~32 times the SOC values for T'sakholo (rangeland and streambank, respectively). This indicates that drainage and small anthropogenic impacts may have led to an increase in SOC contents in Thaba-Putsoa soils. However, in T'sakholo rangeland and stream-bank, the SOC content was very low, which be attributed to the extensive impact of anthropogenic pressures coupled with climate change (i.e. declining rainfall in the region). This has resulted in the wetlands being colonised by invasive weed species.
Pearson's correlation stepwise multiple regression analyses
Pearson's correlation analysis was conducted to examine the levels of association between the soil properties within each site, and correlations between soil properties are summarised in Table 3. Statistically significant (P < 0.001) negative correlation was observed between percentage silt and percentage sand in Thaba-Putsoa. Soil pH had a strong significant (P<0.01) and negative correlation with percentage clay and it also correlated strongly and positively with [Ca.sup.2+]. The CEC had a significant positive (P < 0.01) correlation with SOC at this site. In T'sakholo, correlation analysis showed more associations between soil properties. Soil pH and available P exhibited significant positive correlation (P<0.01) with silt and clay percentages, and negative correlation (P < 0.01) with percentage sand. SOC had a significant positive (P<0.01) correlation with percentage clay and significant negative (P<0.05) correlation with percentage sand. The correlations between CEC and SOC, and between [K.sup.+] and SOC, were significantly positive (P<0.05) and negative (P<0.01), respectively. Sodium had a significant negative (P<0.05) correlation with percentage clay, and calcium ([Ca.sup.2+]) had a significant positive (P<0.01) correlation with SOC and [K.sup.+]. The results of stepwise regression analysis (Table 4) also revealed that SOC accounted for most of the variation observed at both sites (partial [R.sup.2] at Thaba-Putsoa 44%, and at T'sakholo 94%).
Soil physical properties
Considering variability both within and among the soil pits in terms of CV for textural fractions, sand was the least variable at both sites (Table 5). Clay content varied between the two sites but was more uniform in T'sakholo, with many pits having CV <15%. The CV for a particular property is considered low if it is <15%, moderate if it is 15-30%, and high if it is >35% (Wilding and Drees 1978). Silt content was moderately variable in T'sakholo (CV 15-30%), whereas in Thaba-Putsoa it was moderately variable in transect 1 and highly variable in transect 2.
Soil chemical properties
Of the soil chemical properties, pH across all pits at both sites had low variability (CV <15%) (Table 5). Available P showed extreme variability at both sites, with CV values in the highly variable range (CV >35%). The SOC was moderately variable in Thaba-Putsoa, with CV values mostly 15-33%. However, in T'sakholo SOC was highly variable, especially in transect 2 with CV values mostly >35%.
The geostatistical analysis (Table 6) indicated different spatial distribution models and spatial dependence levels for the soil properties. The coefficient of determination ([R.sup.2]) of all of the variables ranged between 54 and 94% for all properties, indicating a good fit, except for available P (in Thaba-Putsoa rangeland), and CEC and exchangeable K (T'sakholo rangeland), with [R.sup.2] of 22-25%. Soil properties showed only a pure nugget effect fitted by the exponential, spherical, and Gaussian models at different sites. Gaussian model best describes the soil textural properties at both sites and SOC at Tsakholo-Rangeland; Spherical model best describes SOC and SOC-pool (Thaba-Putsoa), Exponential model best describes Silt:Clay ratios (Thaba Putsoa and Tsakholo-Rangeland). It is important to determine the spatial dependence of soil properties because properties with strong spatial dependence are more readily managed, and an accurate, site-specific fertilisation scheme for precision farming can be more easily developed. The geostatistical analysis suggests that all soil properties showed moderate and strong spatial dependence (nugget to sill ratio <50%; Table 6). The kriged estimates for silt to clay ratio, SOC, and SOC pool were contoured and mapped so that their patterns of variation across land-use types are presented in Fig. 1. The patterns of variation in these properties varied slightly differently across sites.
A high degree of soil variability was observed at both sites, especially with regard to soil chemical properties except pH. The observation of pH as the least variable in this study is in line with findings by workers such as Mausbach et al. (1980). However, it has been established that chemical properties of soils appear to be more variable than physical properties (Olaniyan 1998), and this has been attributed to the inherent heterogeneity in the parent material, residual effects of fertiliser application, differences in physiography, soil depth, or horizon, and cultural practices (Ogunkunle and Beckett 1987; Akinbola et al. 2006). The soil chemical properties presented in this study have yielded much higher CV values than the physical properties, and thus indicate their high degree of variability. The CV values for SOC exceeded 20%, indicating considerable variability. In studying soil and vegetation C pools in a mountainous watershed of Nepal, Shrestha and Singh (2007) reported that SOC showed considerable spatial variability, both horizontally according to land use and vertically within the soil profile. Nevertheless, soil texture (though considered uniform as it is a physical property) is reported as exhibiting significant spatial and temporal variability in riparian wetlands (Lyons et al. 1998; Darke and Walbridge 2000; Johnston et al. 2001). Soils in wetlands often exhibit characteristic, complex spatial patterns that indicate heterogeneity in soil resources which affect patterns of soil process rates (Ettema et al. 1998). Bruland and Richardson (2005) show that the combined action of processes such as surface runoff, erosion, overbank flooding, sediment deposition, groundwater inputs, fire, animal burrowing, litter production, and root activity contribute to a high degree of spatial variability in wetlands.
Anthropogenic pressures and climate variability are among other factors that induce spatial variability in wetland soil properties. The most peculiar anthropogenic pressure in the wetlands of Lesotho (including those under study) is overgrazing. Grazing affects the spatial patterns of soil properties through tramping and wallowing, which can increase soil compaction (Knapp et al. 1999) and change nutrient distribution via excreta input (Augustine and Frank 2001). In addition, grazing animals can indirectly influence spatial distribution of soil properties through changing vegetation patterns (Lin et aL 2010). Olofsson et al. (2008) found that the impact of grazers on soil nutrient heterogeneity can be consistent with their influence on vegetation patterns. Soil properties that showed strong spatial dependence indicated that structural factors strongly influence the spatial variability of these properties (Cambardella et al. 1994; Jiang et al. 2010). All of the nugget to sill ratios showed strong spatial dependence except SOC (T'sakholo stream-bank) with no spatial dependence. The texture analysis showed that silt exhibited strong spatial dependence, with the nugget accounting for 23.43%. Sand and clay displayed moderate spatial dependence, with nugget to sill ratios of 32.26% and 49.96%, respectively Liu and Tang (2003) also reported strong spatial dependence for soil texture (i.e. silt content).
Spatial distributions of wetland soil properties showed marked variability across pits at both sites. Soil chemical properties showed more spatial variability than physical properties. Correlation and stepwise analysis showed that low correlation among soil properties on both sites, and SOC accounted for most of the variation observed on both study sites. Thus, further wetland studies in Lesotho should attempt to quantify not only the soil properties or processes under investigation but also their spatial variability, because spatial variability in wetland soils can provide insight into underlying ecosystem processes and may itself indicate wetland condition and subsequent management.
This research was funded by the Regional Universities Forum for Capacity Building in Agriculture (RUFORUM) RU/CGS/GRG/30/06/09. We acknowledge the field and laboratory assistance of Mrs Mashao and Mr Mafona, and the rest of the under-graduate students who were always supportive.
Received 1 June 2012, accepted 10 September 2012, published online 13 November 2012
Adderley WP, Jenkins DA, Sinclair FL, Stevens PA, Verinumbe L (1997) The influence of soil variability in tree establishment at an experimental site in Northeast Nigeria. Soil Use and Management 13, 1-8. doi: 10.1111/j.1475-2743.1997.tb00549.x
Akinbola GE, Ojetade JO, Olaleye AO (2006) Variability of soil properties along two toposequences on basement complex in South Western Nigeria. Discovery and Innovation 18, 44-52. doi:10.4314/dai.v18il. 15725
Augustine DJ, Frank DA (2001) Effects of migratory grazers on spatial heterogeneity of soil nitrogen properties in a grassland ecosystem. Ecology g2, 3149 3162. doi:10.1890/0012-9658(2001)08213149: EOMGOSJ2.0.CO;2
Bai J, Ouyand H, Xiao R, Gao J, Gao H, Cui B, Huang L (2010) Spatial variability of soil carbon, nitrogen, and phosphorus content and storage in an alpine wetland in the Qinghai Tibet Plateau, China. Australian Journal of Soil Research 48, 730-736. doi:10.1071/SR09171
Baxter S J, Oliver MA, Gaunt J (2003) A geostatistical analysis of the spatial variation of soil mineral nitrogen and potentially available nitrogen within an amble field. Precision Agriculture 4, 213-226. doi: 10.1023/ A:1024565507688
Bouyoucus HG (1962) A recalibration of the hydrometer for making mechanical analysis of soils. Agronomy Journal 43, 434-438. doi: 10.2134/agronj1951.00021962004300090005x
Bray RH, Kurtz LT (1945) Determination of total, organic and available forms of phosphorus in soils. Soil Science 59, 39-46. doi: 10.1097/00 010694-194501000-00006
Brouwer J, Fussel LK, Hermann L (1993) Soil and crop growth microvafiability in the West African semi-arid tropics: a possible risk reducing factor for sustainable farming. Agriculture, Ecosystems & Environment 45, 229-238. doi: 10.1016/0167-8809(93)90073-X
Bruland GL, Richardson CJ (2004) A spatially explicit investigation of phosphorus sorption and related soil properties in two riparian wetlands. Journal of Environmental Quality 33, 785-794. doi:10.2134/jeq2004. 0785
Bruland GL, Richardson CJ (2005) Spatial variability of soil properties in created, restored, and paired natural wetlands. Soil Science Society of America Journal 69, 273-284.
Cahn MD, Hummel JW, Brouer BH (1994) Spatial analysis of soil fertility for site-specific crop management. Soil Science Society of America Journal 58, 1240-1248. doi:10.2136/sssaj1994.03615995005800040 035x
Cambardella CA, Karlen DL (1999) Spatial analysis of soil fertility parameters. Precision Agriculture 1, 5-14. doi: 10.1023/A: 1009925919 134
Cambardella CA, Moorman TB, Novak JM, Parkin TB, Karle DL, Turco RF, Konopka AE (1994) Fieldscale variability of soil properties in Central Iowa soils. Soil Science Society of America Journal 58, 1501-1511. doi: 10.2136/sssaj 1994.03615995005800050033x
Cassel DK, Wendroth O, Nielsen DR (2000) Assessing spatial variability in an agricultural experiment station field: Opportunities arising from spatial dependence. Agronomy Journal 92, 706-714. doi: 10.2134/agronj 2000.924706x
Chipps SR, Hubbard DE, Werlin KB, Haurerud N J, Powell KA, Thompson J, Johnson T (2006) Association between wetland disturbance and biological attributes in floodplain wetlands. Wetlands 26, 497-508. doi: 10.1672/0277-5212(2006)261497:ABWDAB]2.0.CO;2
Cohen M J, Dunne JEJ, Bruland GL (2008) Spatial variability of soil properties in Cypress domes surrounded by different land uses. Wetlands" 28, 411-422. doi: 10.1672/06-182.1
Corstanje R, Grunwald S, Reddy KR, Osborne TZ, Newman S (2006) Assessment of the spatial distribution of soil properties in a Northern Everglades marsh. Journal of Environmental Quality 35, 938-949. doi: 10.2134/jeq2005.0255
Darke AK, Walbridge MR (2000) Al and Fe biogeochemistry in a floodplain forest: implications for P retention. Biogeochemistry 51, 1-32. doi: 10.1023/A: 1006302600347
DeBusk WF, Reddy KR, Koch MS, Wang Y (1994) Spatial distribution of soil nutrients in a northern Everglades marsh: Water Conservation Area 2A. Soil Science Society of America Journal 58, 543-552. doi: 10.2136/ sssaj1994.03615995005800020042x
DeBusk WF, Newman S, Reddy KR (2001) Spatiotemporal patterns of soil phosphorus enrichment in Everglades Water Conservation Area 2A. Soil Science Society of America Journal 60, 1273 1277.
Eltaib SM, Soom MAM, Hanafi MM, Shariff ARM, Wayayok A (2002) Spatial variability of N, P and K in rice field in Sawah Sempadan, Malaysia. Songklanakarin Journal of Science and Technology :24, 321-328.
Ettema CH, Coleman DC, VeUidis G, Lowrance R, Rathbun SL (1998) Spatiotemporal distributions of bactetivorous nematodes and soil resources in a restored riparian wetland. Ecology 79, 2721 2734. doi: 10.1890/0012-9658(1998)079[2721 :SDOBNA]2.0.CO;2
Farrish KW (1991) Spatial and temporal fine-root distribution in three Louisiana forest soils. Soil Science Society of America Journal 55, 1752-1757. doi:10.2136/sssaj1991.03615995005500060041x
Gaston L, Nkedi-Kizza P, Sawka G, Ran PSC (1990) Spatial variability of morphological properties at a Florida flatwoods site. Soil Science Society of America Journal 54, 527-533. doi:10.2136/sssaj1990.0361 5995005400020040x
Gaston LA, Locke MA, Zablotowicz RM, Reddy KN (2001) Spatial variability of soil properties and weed populations in the Mississippi delta. Soil Science Society of America Journal 65, 449-459. doi: 10.2136/ sssaj2001.652449x
Grunwald S, Corstanje R, Weinrich BE, Reddy KR (2006) Spatial patterns of labile forms of phosphorus in a subtropical wetland. Journal of Environmental Quality 35, 378-389. doi:10.2134/jeq2005. 0042
Hayati AA, Proctor MCF (1990) Plant distribution in relation to mineral nutrient availability and uptake on a wet-heath site in south-west England. Journal of Ecology 78, 134-151. doi: 10.2307/2261041
He N, Yu Q, Wu L, Wang Y, Han X (2008) Carbon and nitrogen store and storage potential as affected by land-use in a Leymus chinensis grassland of northern China. Soil Biology & Biochemistry 40, 2952-2959. doi: 10.1016/j.soilbio.2008.08.018
Hughes RM (1995) Defining acceptable biological status by comparing with reference conditions. In 'Biological assessment and criteria'. (Eds WS Davis, TP Simon) pp. 31-47. (Lewis Publishers: Baton Rouge, LA)
Hurt GW, Carlisle VW (2005) Using soil morphology for identification, delineation and mitigation of wetlands in coastal zone. In 'Proceedings of the 14th Biennial Coastal Zone Conference'. 17-21 July 2005, New Orleans, Louisiana. (NOAA Coastal Services Center: Charleston)
Iqbal J, Thomasson A J, Jenkins N J, Owens RP, Whisler DF (2005) Spatial variability analysis of soil physical properties of alluvial soils. Soil Science Society of America Journal 69, 1338-1350. doi:10.2136/sssaj 2004.0154
Isaaks EH, Srivastava RM (1989) 'An introduction to applied geostatistics.' (Oxford University Press: New York)
Jiang YM, Chen CR, Xu ZH, Liu YQ (2010) Soil soluble organic carbon and nitrogen pools under mono- and mixed-species forest ecosystems in subtropical China. Journal of Soils and Sediments 10, 1071-.1081. doi: 10.1007/s11368-010-0191-9
Johnson LC, Shaver GR, Cades DH, Rastetter E, Nadelhoffer K, Giblin A, Laundre J, Stanley A (2000) Plant carbon-nutrient interactions control C[O.sub.2] exchange in Alaskan wet sedge tundra ecosystems. Ecology 81, 453-469.
Johnston CA, Lee GB, Madison FW (1984) The stratigraphy and composition of a lakeside wetland. Soil Science Society of America Journal 48, 347-354. doi: 10.2136/sssaj1984.03615995004800020025x
Johnston CA, Bridgham SD, Schubauer-Berigan JP (2001) Nutrient dynamics in relation to geomorphology of riverine wetlands. Soil Science Society of America Journal 65, 557-577.
Knapp AK, Blair JM, Briggs JM, Collins SL, Hartnett DC, Johnson LC, Towne EG (1999) The keystone role of bison in North American tallgrass prairie. Bio-Science 49, 39-50.
Kravchenko AN, Robertson GP, Snap SS, Smucker AJM (2006) Using information about spatial variability to improve estimates of total soil carbon. Agronomy Journal 98, 823-829. doi: 10.2134/agronj2005. 0305
Lin Y, Hong M, Han G, Zhao M, Bai Y, Chang S (2010) Grazing intensity affected spatial patterns of vegetation and soil fertility in a desert steppe. Agriculture, Ecosystems & Environment 138, 282-292. doi: 10.1016/ j.agee.2010.05.013
Liu ZX, Tang LS (2003) Spatial variability estimation of cinnamon soil mechanical composition by rank-order geostatistics. Transactions of the CSAE 19, 27Y32. [In Chinese]
Lyons JB, Gorres JH, Amador JA (1998) Spatial and temporal variability of phosphorus retention in a riparian forest soil. Journal of Environmental Quality 27, 895-903. doi:10.2134/jeq1998.00472425002700040025x
Mausbach M J, Brashner B R, Yeck RD, Nettleton WD (1980) Variability of measured properties in morphologically matched pedons. Soil Science Society of America Journal 44, 358-363. doi:10.2136/sssaj1980.0361 5995004400020030x
Mitsch WJ, Gosselink JG (2000) 'Wetlands.' (Van Nostrand Reinhold: New York)
Ogunkunle AO (1993) Variation of some soil properties along two toposequenee on quartzite schist and banded gneiss in South-western Nigeria. GeoJournal 30, 397402. doi: 10.1007/BF00807220
Ogunkunle AO, Beckett PHT (1987) Comparative influences of soil and management on barley yield in the Vale of Whitehorse, England. The Journal of Agricultural Science 108, 555-560. doi:10.1017/S00218 59600079946
Okae-Anti D (2001) Spatial variability of soil properties--why bother with spatial variation? In 'International Conference on Managing Soil Resources of the Tropics for Sustainable Agricultural Productivity'. 26 February-2 March 2001, Tamale, Ghana.
Olaniyan JO (1998) Variations in profile distribution of soil properties along a catena in Eleja Raji, Kwara State. Centrepoint Scientific Education 9, 52-60.
Olofsson J, de Mazancourt C, Crawley MJ (2008) Spatial heterogeneity and plant species richness at different spatial scales under rabbit grazing. Oecologia 156, 825-834. doi:10.1007/s00442-008-1038-6
Reddy KR (1993) Wetland soils--opportunities and challenges. Soil Science Society of America Journal $7, 1145-1146. doi: 10.2136/sssaj 1993.0361 5995005700040043x
Reese RE, Moorhead KK (1996) Spatial characteristics of soil properties along an elevational gradient in a Carolina Bay. Soil Science Society of America Journal 60, 1273-1277. doi:10.2136/sssaj1996.03615995006 000040045x
Robertson GP, Gross KL (1994) Assessing the heterogeneity of belowground resources: quantifying pattern and scale. In 'Exploitation of environmental heterogeneity by plants'. (Eds MM Caldwell, RW Pearcy) (Academic Press: San Diego, CA) SAS Institute (1999) 'SAS/STAT, Version 8e.' (Statistical Analysis Institute, Inc.: Cary, NC)
Shrestha BM, Singh BR (2007) Soil and vegetation carbon pools in a mountain watershed of Nepal. Nutrient Cycling in Agroecosystems 81, 179-191. doi: 10.1007/s10705-007-9148-9
Sinowski W, Auerswald K (1999) Using relief parameters in a discriminant analysis to stratify geological areas with different spatial variability of soil properties. Geoderma 89, 113-128. doi:10.1016/S0016-7061(98) 00127-X
Soil Survey Staff (1999) 'Soil Taxonomy.' Agriculture Handbook No. 436. USDA-NRCS. (U.S. Government Printing Office: Washington, DC)
Stolt MH, Genthner HM, Daniels WL, Groover VA (2001) Spatial variability in Palustrine wetlands. Soil Science Society of America Journal 65, 527-535. doi: 10.2136/sssaj2001.652527x
Teels BM, Adamus P (2001) Methods for evaluating wetland condition: developing metrics and indexes of biological integrity. U.S. Environmental Protection Agency, Office of Water, Washington, DC, USA. EPA 822-R-01-007f.
Wairiu M, Lal R (2003) Soil organic carbon in relation to cultivation and topsoil removal on sloping lands of Kolombangara, Solomon Islands. Soil & Tillage Research 70, 19-27. doi:10.1016/S0167-1987(02)00 116-2
Walkley A, Black IA (1934) An examination of the degtjareff method for determining soil organic matter and proposed modification of the chromic acid titration method. Soil Science 37, 29-38. doi:10.1097/ 00010694-193401000-00003
Wang GP, Liu JS, Wang JD, Yu JB (2006) Soil phosphorous forms and their variations in depressional and riparian freshwater wetlands (Sanjiang Plain, Northeast China). Geoderma 132, 59-74. doi: 10.1016/ j.geoderma.2005.04.021
Wilding LP (1988) Improving our understanding of the composition of the soil landscape. In 'Proceedings of International Interactive Workshop on Soil Resources: Their Inventory, Analysis, and Interpretation for Use in the 1990s'. (Ed. HR Finney) pp. 13-39. (Extension Service, University of Minnesota: St. Paul, MN)
Wilding LP, Drees LR (1978) Spatial variability; a pedologist's viewpoint. In 'Diversity of soils in the tropics'. Soil Science Society of America Special Publication No. 34, pp. 1-12.
Wilding LP, Drees LR (1983) Spatial variability and pedology. In 'Pedogenesis and Soil Taxonomy'. (Eds LP Wilding, NE Smeek, GF Hall) (Pudoc: Wageningen)
Wuddivira HN, Ogunwole JO, Adeoye KB (2000) Spatial variability of soil physical properties of an Alfisol in Samaru, northern Guinea savanna, Nigeria. Journal of Agriculture and Environment 1, 173-182.
T. Nkheloane (A,B), A. O. Olaleye (A), and R. Mating (A)
(A) Department of Soil Science & Resource Conservation, Faculty of Agriculture, The National University of Lesotho, Roma 180, Lesotho.
(B) Corresponding author. Emails: email@example.com; firstname.lastname@example.org
Table 1. General site descriptions Site Topography Wetland type Altitude (m a.s.l.) Thaba-Putsoa Steep rolling Lacustrine 2638 m T'sakholo Flat to gentle Riverine 1570 m Site Land use Vegetation Thaba-Putsoa Grazing Afro-montane grassland, e.g. Themeda triandra T'sakholo Grazing/livestock Highveld grassland, watering e.g. Eragrostis curvula Site Anthropogenic impact Thaba-Putsoa Little T'sakholo High Table 2. Summary statistics for the study sites Number of samples: T'sakholo 76, Thaba-Putsoa 27 Variable Sand Silt Clay SOC pHw Thaba-Putsoa (rangeland) Mean 65.82 24.39 9.79 17.20 5.00 Range 26.70 26.72 14.00 21.79 1.31 Skewness -0.05 -0.37 -0.56 0.27 -0.003 s.d. 6.37 6.39 3.46 5.34 0.27 T'sakholo (rangeland) Mean 46.24 32.23 21.91 1.93 7.42 Range 18.00 16.02 10.00 3.71 1.71 Skewness 0.39 -0.49 -1.48 0.45 -1.6 s.d. 5.80 4.87 2.02 0.85 0.35 T'sakholo (stream-bank) Mean 56.54 25.23 18.55 1.12 7.17 Range 30.00 21.00 18.00 7.67 1.29 Skewness -1.05 0.54 0.76 3.78 0.09 s.d. 6.21 5.52 2.96 1.22 0.27 Variable Avail. P CEC [Na.sup.+] [Mg.sup.2+] Thaba-Putsoa (rangeland) Mean 1.55 0.59 0.47 0.03 Range 6.63 0.57 0.76 0.59 Skewness 1.80 0.29 0.21 0.06 s.d. 1.15 0.14 0.18 0.14 T'sakholo (rangeland) Mean 2.74 0.50 2.73 0.14 Range 9.68 0.53 1.54 1.09 Skewness 0.92 0.13 -0.01 4.59 s.d. 3.03 0.14 0.45 0.16 T'sakholo (stream-bank) Mean 2.74 0.53 2.74 0.83 Range 8.99 0.87 1.66 1.46 Skewness 0.86 -0.06 0.48 1.98 s.d. 2.59 0.14 0.24 0.24 Variable [Ca.sup.2+] [K.sup.+] Thaba-Putsoa (rangeland) Mean 0.48 0.25 Range 1.88 0.37 Skewness 2.02 0.70 s.d. 0.39 0.10 T'sakholo (rangeland) Mean 0.12 0.97 Range 0.21 0.93 Skewness 0.32 -0.86 s.d. 0.054 0.23 T'sakholo (stream-bank) Mean 1.02 3.50 Range 2.16 4.48 Skewness 0.11 1.69 s.d. 0.34 0.69 Table 3. Correlation matrix of soil properties for each site ([dagger]) P < 0.1, * P < 0.05, ** P < 0.01 Variable Sand Silt Clay pH Avail. P Thaba-Putsoa Sand 1.00 -0.89 ** -0.22 0.16 0.09 Silt 1.00 -0.19 -0.15 -0.16 Clay 1.00 -0.35 ** 0.10 pH 1.00 0.05 Avail. P 1.00 SOC CEC [Ca.sup.2+] T'sakholo Sand 1.00 -0.89 ** -0.64 ** -0.31 ** -0.29 ** Silt 1.00 0.31 ** 0.29 ** 0.22 ** Clay 1.00 0.24 ** 0.22 ** pH 1.00 0.17 * Avail. P 1.00 SOC CEC [Ca.sup.2+] [K.sup.+] [Na.sup.+] Variable SOC CEC [Ca.sup.2+] [K.sup.+] Thaba-Putsoa Sand -0.12 0.14 0.19 0.21 Silt 0.14 -0.11 -0.26 -0.10 Clay 0.04 0.04 0.19 -0.19 pH 0.16 0.12 0.39 ** 0.15 Avail. P -0.16 0.01 0.02 -0.09 SOC 1.00 0.35 ** 0.23 0.19 CEC 1.00 0.24 ([dagger]) 0.18 [Ca.sup.2+] 1.00 1.00 T'sakholo Sand -0.17 * -0.02 0.02 0.02 Silt 0.08 0.05 -0.06 -0.07 Clay 0.27 ** 0.007 0.07 0.07 pH 0.07 -0.04 -0.02 -0.01 Avail. P -0.01 -0.04 -0.02 -0.02 SOC 1.00 -0.17 * 0.72 ** 0.71 ** CEC 1.00 -0.02 -0.02 [Ca.sup.2+] 1.00 0.99 ** [K.sup.+] 1.00 [Na.sup.+] Variable [Na.sup.+] Thaba-Putsoa Sand 0.008 Silt -0.05 Clay 0.16 pH -0.13 Avail. P -0.12 SOC 0.08 CEC 0.09 [Ca.sup.2+] 0.008 T'sakholo Sand 0.05 Silt 0.03 Clay -0.18 * pH -0.01 Avail. P 0.05 SOC -0.03 CEC 0.05 [Ca.sup.2+] -0.05 [K.sup.+] -0.05 [Na.sup.+] 1.00 Table 4. Summary of stepwise multiple regression analysis for physico-chemical properties Variable Thaba-Putsoa entered Partial Model Pr. > F [R.sup.2] [R.sup.2] SOC 0.4424 0.4424 <0.0001 Silt 0.0073 0.9245 0.0213 Sand 0.0229 0.9474 <0.0001 Clay 0.0150 0.9580 <0.0001 Avail. P 0.0035 0.9645 0.0209 CEC -- -- -- [K.sup.+] -- -- -- Variable T'sakholo entered Partial Model Pr. > F [R.sup.2] [R.sup.2] SOC 0.94 0.9412 <0.0001 Silt 0.0008 0.9740 0.0400 Sand 0.03 0.9711 <0.0001 Clay 0.002 0.9660 0.0007 Avail. P -- -- -- CEC 0.001 0.9732 0.0239 [K.sup.+] 0.0012 0.9723 0.0145 Table 5. Comparison of variability of soil properties between two sites Number of samples: T'sakholo 76, Thaba-Putsoa 27. Coefficient of variation (CV): low <15%, moderate 15-30%, high >35% (Wilding and Drees 1978). s.d., Standard deviation Property Thaba-Putsoa T'sakholo Transect 1 Transect 2 Transect l Transect 2 CV s.d. CV s.d. CV s.d. CV s.d. Sand 5.1 63.6 7.9 70.6 11.5 46.4 7.8 56.7 Clay 15.4 11.1 57.4 8.6 8.3 22.1 10.2 18.7 Silt 15.4 25.2 32.2 28.5 15.0 31.8 13.5 25.0 pHw 2.7 4.9 4.2 5.0 2.3 7.4 2.5 7.2 Avail. P 76.6 1.7 70.8 1.5 84.0 2.9 80.4 2.7 SOC 26.1 16.7 27.4 15.0 28.1 1.8 62.7 1.1 CEC 19.8 0.6 17.4 0.6 16.1 0.5 23.4 0.5 [K.sup.+] 30.0 0.2 24.8 0.3 31.1 0.0 66.8 0.0 [Na.sup.+] 34.8 0.4 38.5 0.3 46.4 0.0 54.0 0.0 [Mg.sup.2+] 26.3 1.0 29.5 1.1 51.2 0.2 28.2 0.2 [Ca.sup.2+] 44.2 0.4 53.4 0.5 31.1 0.1 66.8 0.1 Table 6. Sum of geostatistical parameters for silt to clay ratio, soil organic carbon (SOC), and SOC pools x, Mean; s.d., standard deviation; CV, coefficient of variation x s.d. CV Nugget (Co) Thaba-Putsoa rangeland Silt: clay 3.26 2.54 77.9 0.01 SOC 17.20 2.27 30.7 0.01 SOC pools 84.90 38.63 45.5 136.00 Sand 65.96 6.48 9.8 0.10 Silt 24.09 6.06 25.2 0.10 Clay 9.92 3.35 33.7 7.33 Avail. P 1.59 1.15 72.3 1.06 T sakholo rangeland Silt: clay 2.44 2.07 84.8 0.04 SOC 10.16 8.56 84.25 0.21 SOC pools 49.93 48.99 102.21 1.98 Sand 56.50 11.05 19.55 0.10 Silt 27.83 6.85 24.60 0.10 Clay 16.06 6.52 40.59 0.42 CEC 0.55 0.15 27.23 0.01 K 0.13 0.40 307.70 0.011 T sakholo stream-bank Silt: clay 1.39 0.41 29.50 0.02 SOC 1.12 1.21 108.04 1.56 SOC pools 5.25 6.38 121.52 0.10 Sand 54.83 7.26 13.20 0.10 Silt 26.25 5.82 22.17 0.10 Clay 19.22 3.43 17.85 1.31 CEC 0.53 0.15 28.30 0.009 K 0.0003 0.0004 133.30 -- Sill (Co+C) Range Co/(Co+C) [R.sup.2] Thaba-Putsoa rangeland Silt: clay 20.02 11.37 0.001 0.756 SOC 22.59 21.96 0.004 0.108 SOC pools 1422 140.57 0.095 0.03 Sand 81.20 19.02 0.001 0.936 Silt 237.51 37.15 0.0004 0.90 Clay 85.28 40.90 0.085 0.850 Avail. P 4.00 190.00 0.265 0.250 T sakholo rangeland Silt: clay 0.45 11.37 0.09 0.710 SOC 1.37 28.32 0.153 0.154 SOC pools 6.13 169.17 0.323 0.485 Sand 294.64 37.20 0.0003 0.876 Silt 237.51 37.15 0.0004 0.90 Clay 72.05 78.87 0.006 0.84 CEC 0.06 3.07 0.167 0.216 K 0.057 346 0.175 0.232 T sakholo stream-bank Silt: clay 0.199 7.67 0.100 0.393 SOC 1.55 58.1 1.01 0.436 SOC pools 373.67 11.78 0.001 0.964 Sand 142.30 26.05 0.0007 0.89 Silt 133.33 36.39 0.0008 0.836 Clay 96.6 43.70 0.0145 0.536 CEC 0.061 0.670 0.148 0.751 K -- -- -- -- Model Thaba-Putsoa rangeland Silt: clay Exponential SOC Spherical SOC pools Spherical Sand Gaussian Silt Gaussian Clay Gaussian Avail. P Gaussian T sakholo rangeland Silt: clay Spherical SOC Gaussian SOC pools Exponential Sand Gaussian Silt Gaussian Clay Gaussian CEC Gaussian K Spherical T sakholo stream-bank Silt: clay Spherical SOC Spherical SOC pools Gaussian Sand Gaussian Silt Spherical Clay Spherical CEC Spherical K --
|Printer friendly Cite/link Email Feedback|
|Author:||Nkheloane, T.; Olaleye, A.O.; Mating, R.|
|Date:||Oct 1, 2012|
|Previous Article:||Effects of biochars generated from crop residues on chemical properties of acid soils from tropical and subtropical China.|
|Next Article:||Soil selenium in a forested seabird colony: distribution, sources, uptake by plants, and comparison with non-seabird sites.|