Spatial and vertical analysis of soil properties in a Mediterranean vineyard soil

Soil physical properties influence vineyard behaviour; therefore, the knowledge of their spatial variability is essential for making vineyard management decisions. Little work has been conducted at high spatial resolution on soil properties at depths lower than 0.30 m which is of special relevance to perennial crops. The objectives of this work were to (i) analyse the spatial and vertical variability of soil depth, particle size fractions and water‐holding capacity (WHC) by geostatistical techniques; (ii) study the causes of the variability, with additional information from classical soil sampling; and (iii) assess the significance of WHC through its relationship with vine vigour. The work was carried out in a vineyard of eight hectares within the D.O.Ca. Rioja (northern Spain). Soil variability was determined via grid sampling at three depth ranges (0–0.30, 0.30–0.60 and 0.60–0.90 m). A conventional soil survey provided additional information on soil variability. Clay, sand and silt fractions, soil organic matter content, WHC and pruning weight were determined. Most soil properties had strong or moderate spatial dependence, with the exceptions of sand at 0.30–0.60 m and silt in the topsoil. Topography and soil erosion caused the spatial variability of soil depth and contributed to the spatial distribution of particle size fractions in the topsoil, while the heterogeneity of parent material influenced the spatial pattern of soil properties at 0.60–0.90 m. The WHC and soil depth spatial distributions related well to that of vine vigour, demonstrating the importance of knowing the spatial variability of these soil properties.     

Key soil and topographic properties to delineate potential management classes for precision agriculture in the European loess area

Recent advances in on-the-go soil sensing, terrain modelling and yield mapping have made available large quantities of information about the within-field variability of soil and crop properties. But the selection of the key variables for an identification of management zones, required for precision agriculture, is not straightforward. To investigate a procedure for this selection, an 8 ha agricultural field in the Loess belt of Belgium was considered for this study. The available information consisted of: (i) top- and subsoil samples taken at 110 locations, on which soil properties: textural fractions, organic carbon (OC), CaCO₃ and pH were analysed, (ii) soil apparent electrical conductivity (EC_a) obtained through an electromagnetic induction based sensor, and (iii) wetness index, stream power index and steepest slope angle derived from a detailed digital elevation model (DEM). A principal component analysis, involving 12 soil and topographic properties and two EC_a variables, identified three components explaining 67.4% of the total variability. These three components were best represented by pH, EC_a that strongly associated with texture and OC. However, OC was closely related to some more readily obtainable topographic properties, and therefore elevation was preferred. A fuzzy k-means classification of these three variables produced four potential management classes. Three-year average standardized yield maps of grain and straw showed productivity differences across these classes, but mainly linked to their landscape position. In the loess area with complex soil-landscape interactions pH, EC_a and elevation can be considered as key properties to delineate potential management classes.     

中国东南地区两土链土壤中土壤性质和水稻产量的空间变异研究

In the light of an increasing demand for staple food, especially rice, in southeast China, investigations on the specific site potential expressed as the relationship between soil and crop yield parameters gain increasing importance. Soil texture and several soil chemical parameters as well as plant properties such as crop height, biomass and grain yield were investigated along two terraced catenas with contrasting soil textures cropped with wet rice. We were aiming at identifying correlative relationships between soil and crop properties. Data were analyzed both statistically and geostatistically on the basis of semivariograms. Statistical analysis indicated a significant influence of the relief position on the spatial distribution of soil texture, total carbon and total nitrogen contents. Significant correlations were found for the catena located in a sandstone area （Catena A） between rice yield and silt as well as total nitrogen content. Corresponding relationships were not detectable for paddy fields that developed from Quaternary clays （Catena B）. As suggested by the nugget to sill ratio, spatial variability of soil texture, total carbon and nitrogen was mainly controlled by intrinsic factors, which might be attributed to the erosional transport of fine soil constituents, indicating the importance of the relief position and slope in soil development even in landscapes that are terraced. The crop parameters exhibited short ranges of influence and about one third of their variability was unexplained. Comparable ranges Of selected crop and soil parameters, found only for Catena A, are indicative of close spatial interactions between rice yield and soil features. Our findings show that especially in sandstone-dominated areas, a site-specific management can contribute to an environmentally safe rice production increase.     

Assessment of gully erosion susceptibility using different DEM-derived topographic factors in the black soil region of Northeast China

As a primary sediment source, gully erosion leads to severe land degradation and poses a threat to food and ecological security. Therefore, identification of susceptible areas is critical to the prevention and control of gully erosion. This study aimed to identify areas prone to gully erosion using four machine learning methods with derived topographic attributes. Eight topographic attributes (elevation, slope aspect, slope degree, catchment area, plan curvature, profile curvature, stream power index, and topographic wetness index) were derived as feature variables controlling gully occurrence from digital elevation models with four different pixel sizes (5.0 m, 12.5 m, 20.0 m, and 30.0 m). A gully inventory map of a small agricultural catchment in Heilongjiang, China, was prepared through a combination of field surveys and satellite imagery. Each topographic attribute dataset was randomly divided into two portions of 70% and 30% for calibrating and validating four machine learning methods, namely random forest (RF), support vector machines (SVM), artificial neural network (ANN), and generalized linear models (GLM). Accuracy (ACC), area under the receiver operating characteristic curve (AUC), root mean square error (RMSE), and mean absolute error (MAE) were calculated to assess the performance of the four machine learning methods in predicting spatial distribution of gully erosion susceptibility (GES). The results suggested that the selected topographic attributes were capable of predicting GES in the study catchment area. A pixel size of 20.0 m was optimal for all four machine learning methods. The RF method described the spatial relationship between the feature variables and gully occurrence with the greatest accuracy, as it returned the highest values of ACC (0.917) and AUC (0.905) at a 20.0 m resolution. The RF was also the least sensitive to resolutions, followed by SVM (ACC = 0.781–0.891, AUC = 0.724–0.861) and ANN (ACC = 0.744–0.808, AUC = 0.649–0.847). GLM performed poorly in this study (ACC = 0.693–0.757, AUC = 0.608–0.703). Based on the spatial distribution of GES determined using the optimal method (RF + pixel size of 20.0 m), 16% of the study area has very high level susceptibility classes, whereas areas with high, moderate, and low levels of susceptibility make up approximately 24%, 30%, and 31% of the study area, respectively. Our results demonstrate that GES assessment with machine learning methods can successfully identify areas prone to gully erosion, providing reference information for future soil conservation plans and land management. In addition, pixel size (resolution) is the key consideration when preparing suitable datasets of feature variables for GES assessment.     

加拿大西部起伏地貌的地形指数与产量变异性

Understanding the relationships between topographic indices and crop yield variability is important for soil management and crop production in rolling landscape. Two agricultural fields at Alvena and Hepburn, Saskatchewan, Canada were selected to examine how topographic indices were related to wheat yield under two topographic and weather conditions in the Canadian prairies. The landscapes of the two sites are classified as hummocky and the dominant soil type is an Aridic Ustoll. The relationships among yield, topography, soil, and weather were analyzed using wheat (Triticum aestivumL.) grain yield from Alvena in 2001 (dry year) and 2004 (wet year) and from Hepburn in 1998 (dry year). Topographic/soil indices included relative elevation, wetness index, upslope length, curvature, soil organic matter, and soil moisture storage before seeding. The results indicated that, in the dry years, the correlation coefficients between upslope length and grain yield were 0.79 for the typical rolling landscape (Alvena) in 2001 and 0.73 for shallow gentle rolling landscape (Hepburn) in 1998. In the wet year (2004), the relationships between yield and topographic/soil attributes were not as strong as in dry years. Therefore, upslope length was the best yield indicator for the two landscapes in dry years, whereas no topographic indices were highly correlated to crop yield in wet years. Those topographic indices seemed useful in identifying the yield variability and delineating the proper management zone.     

Characterization of Spatial Variability of Soil Parameters in Apple Orchards of Himalayan Region Using Geostatistical Analysis

ABSTRACT

Knowledge of spatial variability in soil properties is critical for effective nutrient management plan in crop production. Poor productivity of apple orchards in Indian Himalayan Region (IHR) is due to lack of information on the variability of soil properties important for enhancing productivity in the region. The study was conducted in apple orchards with the hypothesis that spatial variability of soil properties is high due to adoption of varied management and passive soil factors. The major objectives of the study were to assess the spatial variability of soil parameters, viz. soil pH, electrical conductivity, soil organic carbon along with plant available soil nitrogen (N), phosphorus (P), potassium (K), exchangeable calcium (Ca), and exchangeable magnesium (Mg) at a regional scale through geostatistical methods. Coefficient of variation (CV) indicated that the fertility parameters varied from medium to high in heterogeneity (CV > 20%). Available N was found critical/medium in 69.6% of soil samples and might be one of the limiting nutrients for crop growth, P and K were in high, and OC in very high range. Significant correlation was found between OC with N; pH with K, Ca and Mg and EC with OC, P and K. The semivariogram parameters indicated that the spatial distribution of soil fertility parameters were inconsistent and showed strong to weak degree of spatial dependence for all parameters. The study highlighted the importance of delineation of soil fertility management zones in the apple growing region as a guide for precise and site-specific nutrient.     

基于GIS的亚热带典型地区土壤有机碳空间分布预测

Spatial distribution of organic carbon in soils is difficult to estimate because of inherent spatial variability and insufficient data. A soil-landscape model for a region, based on 151 samples for parent material and topographic factors, was established using a GIS spatial analysis technique and a digital elevation model (DEM) to reveal spatial distribution characteristics of soil organic carbon (SOC). Correlations between organic carbon and topographic factors were analyzed and a regression model was established to predict SOC content. Results for surface soils (0-20 cm) showed that the average SOC content was 12.8 g kg^-1, with the SOC content between 6 and 12 g kg^-1 occupying the largest area and SOC over 24 g kg^-1 the smallest. Also, soils derived from phyllite were the highest in the SOC content and area, while soils developed on purple shale the lowest. Although parent material, elevation, and slope exposure were all significant topographic variables (P < 0.01), slope exposure had the highest correlation to SOC content (r = 0.66). Using a multiple regression model (R² = 0.611) and DEM (with a 30 m × 30 m grid), spatial distribution of SOC could be forecasted.     

结合统计和数字地形数据的可视化方法预测土壤深度

Information about the spatial distribution of soil attributes is indispensable for many land resource management applications; however, the ability of soil maps to supply such information for modern modeling tools is questionable. The objectives of this study were to investigate the possibility of predicting soil depth using some terrain attributes derived from digital elevation models (DEMs) with geographic information systems (GIS) and to suggest an approach to predict other soil attributes. Soil depth was determined at 652 field observations over the Al-Muwaqqar Watershed (70 km²) in Jordan. Terrain attributes derived from 30-m resolution DEMs were utilized to predict soil depth. The results indicated that the use of multiple linear regression models within small watershed subdivisions enabled the prediction of soil depth with a difference of 50 cm for 77% of the field observations. The spatial distribution of the predicted soil depth was visually coincided and had good correlations with the spatial distribution of the classes amalgamating three terrain attributes, slope steepness, slope shape, and compound topographic index. These suggested that the modeling of soil-landscape relationships within small watershed subdivisions using the three terrain attributes was a promising approach to predict other soil attributes.     

An application of PCA and fuzzy C-means to delineate management zones and variability analysis of soil

Within-field variability is a well-known phenomenon and its study is at the centre of precision agriculture (PA). In this paper, site-specific spatial variability (SSSV) of apparent Electrical Conductivity (EC_a) and crop yield apart from pH, moisture, temperature and di-electric constant information was analyzed to construct spatial distribution maps. Principal component analysis (PCA) and fuzzy c-means (FCM) clustering algorithm were then performed to delineate management zones (MZs). Various performance indices such as Normalized Classification Entropy (NCE) and Fuzzy Performance Index (FPI) were calculated to determine the clustering performance. The geo-referenced sensor data was analyzed for within-field classification. Results revealed that the variables could be aggregated into MZs that characterize spatial variability in soil chemical properties and crop productivity. The resulting classified MZs showed favorable agreement between EC_a and crop yield variability pattern. This enables reduction in number of soil analysis needed to create application maps for certain cultivation operations.     

Temporal changes in the spatial distributions of some soil properties on a temperate grassland site

Abstract. Precision Agriculture seeks to match resource application and agronomic practice with soil and crop requirements as they vary in both space and time. Therefore, an understanding of both the temporal and the spatial components of variability is essential before decisions can be made about the feasibility of site-specific management. In the present study, the spatial and the temporal components of variability in certain key soil properties of a grassland field were evaluated to assess the likely feasibility of adopting a site-specific approach to grassland management. A 7.9 ha grassland field was selected for the study and soil samples were taken three times at regular 25 m intervals across the field over a two year period, and chemically analysed. Classical and geostatistical procedures were used to evaluate the spatial variability and the temporal stability of soil property distributions. Soil extractable P and K had the greatest within-field variability and soil pH the least. Soil K distributions were also highly unstable over time and it was concluded that the optimal risk aversion strategy would be to apply uniform dressings of this nutrient to the entire field. In contrast, soil pH, P, Mg and sulphate distributions were not only temporally stable, but were also spatially correlated over reasonably large ranges. It was concluded that these properties might be managed in a site-specific way based on the results of periodic soil testing in three clearly defined management sub-units within the field. Over the two year period, C and N accumulated in the soil at surprisingly high rates on certain parts of the field but not in others.     

Cotton lint yield variability in a heterogeneous soil at a landscape scale

Landscape variability associated with topographic features affects the spatial pattern of soil water and N redistribution, and thus N uptake and crop yield. A landscape-scale study was conducted in a center pivot irrigated field on the southern High Plains of Texas in 1999 to assess soil water, soil NO₃-N, cotton (Gossypium hirsutum L.) lint yield, and N uptake variability in the landscape, and to determine the spatial correlation between these landscape variables using a state-space approach. The treatments were irrigation at 50 and 75% cotton potential evapotranspiration (ET). Neutron access tubes were placed at a 15-m interval along a 710 m (50% ET) and 820 m (75% ET) transect across the field. Soil NO₃-N in early spring was autocorrelated at a distance varying between 60 and 80 m. Measured soil volumetric water content (WC), total N uptake, and lint yield were generally higher on lower landscape positions. Cotton lint yield was significantly correlated to soil WC (r=0.76), soil NO₃-N (r=0.35), and site elevation (r=−0.54). Differences of site elevation between local neighboring points explained the soil water, NO₃-N and lint yield variability at the micro-scale level in the landscape. Soil WC, cotton lint yield, N uptake, and clay content were crosscorrelated with site elevation across a lag distance of ±30–40 m. The state-space analysis showed that cotton lint yield was positively weighted on soil WC availability and negatively weighted on site elevation. Cotton lint yield state-space models give insights on the association of soil physical and chemical properties, lint yield, and landscape processes, and have the potential to improve water and N management at the landscape-scale.     

Relationship between soil depth and terrain attributes in karst region in Southwest China

Purpose

Soil depth generally varies in peak-cluster depression regions in rather complex ways. Because conventional soil survey methods in these regions require a considerable amount of time, effort, and consequently relatively large budget, new methods are required in karst regions.

Materials and methods

This study explored the relationship between soil depth and terrain attributes abstracted from digital elevation models (DEMs) at different spatial resolutions in the Guohua Karst Ecological Experimental Area, a representative region of peak-cluster depression in Southwest China. A uniform 140 m?×?140 m grid combined with representative hillslope methodology was used to select 171 sampling points where soil depth was measured. Nine primary and secondary terrain attributes, such as elevation, slope, aspect, especial catchment area, wetness index, length-slope factor, stream power index, relief degree of land surface, and distance from ridge of mountains, were computed from DEMs at different spatial resolutions. The optimal DEM spatial resolution was determined by Grey relational analysis (GRA) to reflect the correlations between soil depth and terrain attributes.

Results and discussion

GRA revealed that the 10-m spatial resolution DEM can best reflect the relationship between soil depth and terrain attributes; therefore, the terrain attributes at this resolution were used for multiple linear stepwise regression (MLSR) analysis. The result of MLSR indicated that slope, TWI, and elevation could explain about 61.4 % of the total variability in soil depth in the study area.

Conclusions

The terrain attributes of slope, WTI and elevation can be used to evaluate soil depth in this region very well. This proposed approach may be applicable to other peak-cluster depression regions in the karst areas at a larger scale.     

基于Markov链地统计模型的区域土壤性质研究进展

土壤水力学性质的空间变异对于区域土壤水分溶质循环模拟研究至关重要。基于Markov链的条件模拟是一种能融合多源信息技术的地统计学模拟方法,与传统插值法和基于变异函数的条件模拟相比有诸多优势。目前,该方法在土壤水力学性质空间变异性领域的研究并未全面展开。实现区域土壤水力学模型参数的随机模拟,对于实现区域土壤水分运动和溶质运移的随机模拟,分析土壤水力学性质空间变异性对土壤水分运动和溶质运移模拟结果的影响,特别是参数采样点变化对土壤水分运动和溶质运移结果影响的不确定性等研究都有重要意义。本文旨在综述基于Markov链的地统计学模拟在土壤学相关领域的研究进展,以期为区域模拟中面临的参数获取难题提供帮助,为区域农业生产管理,水分高效利用,农田生态环境保护提供科学依据。     

基于地形因子和随机森林的丘陵区农田土壤有效铁空间分布预测

为了掌握丘陵地区农田土壤有效铁含量及其空间分布,本文以重庆市江津区永兴镇内同源成土母质的典型丘陵(2 km2)为研究区,采集309个土壤样点,利用普通克里格(Ordinary Kriging,OK)、多元线性回归(Multiple Linear Regression,MLR)、随机森林(Random Forest,RF)模型,结合高程、坡度、坡向、谷深、平面曲率、剖面曲率、汇聚指数、相对坡位指数、地形湿度指数等地形因子对土壤有效铁进行空间分布预测,并通过85个验证点评价、筛选预测模型。结果表明:1)土壤有效铁与谷深、地形湿度指数存在极显著水平正相关关系,与坡度、平面曲率、剖面曲率、汇聚指数、相对坡位指数存在极显著水平负相关关系。2)随机森林模型的预测精度明显高于多元线性回归和普通克里格插值,其平均绝对误差为22.33 mg·kg-1、均方根误差为27.98 mg·kg-1、决定系数为0.76,是研究区土壤有效铁含量空间分布的最适预测模型。3)地形湿度指数和坡度是影响该区域土壤有效铁含量空间分布的主要地形因子。土壤有效铁与坡度、谷深、平面曲率、剖面曲率、汇聚指数、相对坡位指数、地形湿度指数均达到极显著水平相关关系。4)研究区土壤有效铁含量范围为3.00~276.97 mg?kg-1,水田有效铁含量大于旱地;土壤有效铁具有较强的空间相关性,土壤有效铁含量空间变异主要受到结构性因素的影响。可见,基于地形因子的随机森林预测模型可以较好地解释丘陵区农田土壤有效铁含量的空间变异,研究结果为丘陵区土壤中、微量元素含量及空间分布预测提供方法借鉴和理论依据。     

Tillage effects on soil hydraulic properties in space and time: State of the science

Soil tillage practices can affect soil hydraulic properties and processes dynamically in space and time with consequent and coupled effects on chemical movement and plant growth. This literature review addresses the quantitative effects of soil tillage and associated management (e.g., crop residues) on the temporal and spatial variability of soil hydraulic properties. Our review includes incidental management effects, such as soil compaction, and natural sources of variability, such as topography. Despite limited research on space–time predictions, many studies have addressed management effects on soil hydraulic properties and processes relevant to improved understanding of the sources of variability and their interactions in space and time. Whether examined explicitly or implicitly, the literature includes studies of interactions between treatments, such as tillage and residue management. No-tillage (NT) treatments have been compared with various tillage practices under a range of conditions with mixed results. The trend, if any, is for NT to increase macropore connectivity while generating inconsistent responses in total porosity and soil bulk density compared with conventional tillage practices. This corresponds to a general increase in ponded or near-zero tension infiltration rates and saturated hydraulic conductivities. Similarly, controlled equipment traffic may have significant effects on soil compaction and related hydraulic properties on some soils, but on others, landscape and temporal variability overwhelm wheel-track effects. Spatial and temporal variability often overshadows specific management effects, and several authors have recognized this in their analyses and interpretations. Differences in temporal variability depend on spatial locations between rows, within fields at different landscape positions, and between sites with different climates and dominant soil types. Most tillage practices have pronounced effects on soil hydraulic properties immediately following tillage application, but these effects can diminish rapidly. Long-term effects on the order of a decade or more can appear less pronounced and are sometimes impossible to distinguish from natural and unaccounted management-induced variability. New standards for experimental classification are essential for isolating and subsequently generalizing space–time responses. Accordingly, enhanced methods of field measurement and data collection combined with explicit spatio-temporal modeling and parameter estimation should provide quantitative predictions of soil hydraulic behavior due to tillage and related agricultural management.     

Spatial variability of Southeastern U.S. Coastal Plain soil physical properties: Implications for site-specific management

Our objectives were to describe the field-scale horizontal and vertical spatial variability of soil physical properties and their relations to soil map units in typical southeastern USA coastal plain soils, and to identify the soil properties, or clusters of properties, that defined most of the variability within the field. The study was conducted on a 12-ha field in Kinston, NC. A 1:2400 scale soil survey had delineated three soil map units in the field: Norfolk loamy sand, Goldsboro loamy sand, and Lynchburg sandy loam. These are representative of millions of hectares of farmland in the Coastal Plain of the southeastern USA. Sixty soil cores were taken to ∼ 1-m depth, sectioned into five depth increments, and analyzed for: soil texture as percentage sand, silt, and clay; soil water content (SWC) at − 33 and − 1500 kPa; plant available water (PAW); saturated hydraulic conductivity (Ksat); bulk density (BD); and total porosity. A penetrometer was used to measure cone index (CI) at each sample location. Variography, two mixed-model analyses, and principal components analysis were conducted. Results indicated that soil physical properties could be divided into two categories. The first category described the majority of the within-field variability and included particle size distribution (soil texture), SWC, PAW, and CI. These characteristics showed horizontal spatial structure that was captured by soil map units and especially by the division between sandy loams and finer loam soils. The second class of variables included BD, total porosity, and Ksat. These properties were not spatially correlated in the field and were unrelated to soil map unit. These findings support the hypothesis that coastal plain soil map units that delineate boundaries between sandy loams versus finer loam soils may be useful for developing management zones for site-specific crop management.     

Spatial variability of soil organic carbon stock in an olive orchard at catchment scale in Southern Spain

Orchards have a high potential for carbon sequestration. However, little research is available on the spatial variability at catchment scale and on the difference between the tree area and the lanes. We analyzed theik spatial variability of soil organic carbon stock, SOC_stock at 90 cm depth in an 8-ha catchment in Southern Spain with olives on a vertic soil. Results showed higher soil organic carbon concentration, SOC, in the tree area as compared to the lane up to 60 cm depth, but its impact on SOC_stock was negligible since it was compensated by the higher soil bulk density in the lane. SOC at different depths was correlated with that in the top 0–5 cm. The overall SOC_stock of the orchard was 4.14 kg m⁻², ranging between 1.8 and 6.0 kg m⁻². This SOC_stock is in the mid-lower range of values reported for olive orchards, measured at smaller scale, and similar to those other intensive field crops and agroforestry under comparable rainfall conditions. The spatial variability in SOC_stock was correlated to several geomorphological variables: elevation, cumulative upstream area, topographic wetness index, sediment transport index, and tillage erosion. Differences in SOC and SOC_stock are driven by the sediment redistribution downslope, mainly by tillage erosion, and higher soil water availability in lower areas allowing higher biomass production. These topographic indexes and the correlation between SOC in the topsoil and SOC_stock up to 90 cm should be further explored in other typology of olive orchards for facilitating the mapping of SOC_stock.