Effectiveness of a soil mapping geomatic approach to predict the spatial distribution of soil types and their properties
Introduction
There is an increasing demand for soil information, since the knowledge and understanding of soil and how it is distributed across the landscape is considered essential for its effective use, management and conservation (Grealish et al., 2015). Consequently, soil information is also essential to help decision makers in land planning and in drafting environmental management policy (van Delden et al., 2011, Brungard et al., 2015), and may even be required by law (Vacca et al., 2014).
Soil information can be provided by soil maps, which are graphic representations for transmitting information about the spatial distribution of soil attributes (Yaalon, 1989). In general terms, soil maps can be produced in a conventional or in a digital way. Conceptually, conventional soil maps and digital soil maps (DSM) are very similar (Kempen et al., 2012), since both approaches use a soil-landscape model to predict soil at unobserved locations (Hudson, 1992). The main difference is that in conventional soil maps the soil-landscape model is a qualitative model based on soil surveyors’ expert knowledge, while in DSM the soil-landscape model is quantitative. Comprehensive overviews of DSM were provided by McBratney et al., 2003, Grunwald, 2006, Minasny and McBratney, 2016, and Arrouays et al. (2020). Among the first conceptualizations of DSM, McBratney et al. (2003) formalized the so-called scorpan model as “empirical quantitative descriptions of relationships between soil and other spatially referenced factors with a view to using these as soil spatial prediction functions”. The possibility of producing DSM strongly depends on the availability of ancillary data (Zeraatpisheh et al., 2017), including existing soil data (e.g. polygon-based soil maps and soil profile databases), which can serve as both training and validation datasets (Zhang et al., 2017). Consequently, in areas with limited existing soil data, producing an accurate DSM can be challenging (Stoorvogel et al., 2009), so this method has rarely been used for routine production mapping or addressing land management questions (Grealish et al., 2015). In these areas, pragmatic and easy-to-apply relationships for predicting soil properties under different environmental conditions, and assist in soil data collection, are needed to provide answers for the current issues that require a fast delivery of information (Gray et al., 2009, Grealish et al., 2015). Several different statistical approaches have been tested to generate quantitative predictions of categorical soil variables from limited samples with the general aim of producing soil maps for unsampled or sparsely sampled areas at different scales, from national to sub-regional (Minasny and McBratney, 2016). Grunwald (2009) provided a multi-criteria characterization of digital soil mapping and modelling approaches, classifying DSM techniques in three wide categories based upon predictor variables and modelling approaches, which can then integrate data driven statistical approaches with pedotransfer functions and dynamic mechanistic modelling of soil properties.
In Sardinia (Italy), the use of soil information and maps in land use planning is specifically required by law (RAS, 2006, RAS, 2008). Because the scale of the three available soil maps covering the island (Arangino et al., 1986, Aru et al., 1990, Madrau et al., 2006) was considered not adequate for local land planning strategies, a new project was recently initiated for the production of a new soil map, at a scale of 1:50,000. The general structure of the project and the methodology used were described in Vacca et al. (2014). The existing soil dataset (point data and maps) was considered insufficient and inappropriate to produce a DSM without resorting to the support of ancillary variables. Adopting a cost-effective approach, existing digital environmental data, along with soil data, were therefore used to delineate homogeneous spatial areas in terms of soil, geological substrate, landform, and land cover in a GIS environment.
This paper aimed to verify, in two of the pilot areas and by means of statistical analysis, the effectiveness of the adopted methodology (Vacca et al., 2014) in representing and predicting the spatial distribution of soil types and soil properties. This is considered crucial, as it affects the reproducibility of the model. There appears to be a need for clarification of the quantitative relationships between soil properties and environmental covariates in order to reduce the uncertainty of the model and allow better prediction.
The specific objectives of this paper were to (1) evaluate the influence of landforms and parent material on soil types; (2) evaluate their influence on soil properties; and (3) evaluate if the adopted methodology is suitable for calibrating a model to predict land units composition in terms of principal soil types.
Section snippets
Study area
This study was conducted in two pilot areas of Pula and Muravera, located in southern Sardinia (Italy), as shown in Fig. 1. The two areas have similar geology, topography, climate and land use. There are two distinct physiographic regions within each area: a hilly part and a coastal plain. Geology of the hilly sectors consists mainly of Paleozoic metamorphic rocks, which were deformed and affected by low-grade metamorphism during the Hercynian orogenesis, and granitoids of the Carboniferous (
Relationships between morphometric parameters and parent material
Considering only the most common parent materials among those described in Table 2 (AL, DP, DC, DV, M, Y; 1358 observations), some relationships with morphometric parameters were found (Fig. 2). Metamorphic rocks, M, were significantly (p < 0.05) associated to >15% sloping convex slopes (35% on landform unit 3, 28% on landform unit 2). The same trend was observed on granite even if a relatively higher (and not statistically significant) percentage of observations fell in areas with slope
Conclusions
In the two pilot areas of Pula and Muravera, the distribution of soil types varied with landform and parent material. The relationships between soil types and landforms reflected the influence of slope gradient and curvature. On steeper slopes, due to morphodynamic processes, mainly very weakly (Leptosols and Regosols) and weakly (Cambisols) developed soils were present. Leptosols (shallow soils) were more frequent where a predominant erosional character prevailed (convex slopes), whilst
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
We are grateful to three anonymous reviewers, whose critical comments and suggestions greatly helped to improve the quality of the manuscript. We are also grateful to Agris Sardegna for the availability of soil data collected in the Muravera pilot area. We also thank Dr. Alison Garside for her revision and editing of the English language. This work was supported by the RAS under the decree DGR n. 56/36 dated 29.12.2009.
References (52)
- et al.
Impressions of digital soil maps: The good, the not so good, and making them ever better
Geoderma Regional
(2020) - et al.
Machine learning for predicting soil classes in three semi-arid landscapes
Geoderma
(2015) - et al.
Paleosols provide sedimentation, relative age, and climatic information about the alluvial fan of the River Tirso (Central-Western Sardinia, Italy)
Quat. Int.
(2006) - et al.
Digital soil mapping of a red clay subsoil covered by loess
Geoderma
(2014) - et al.
Relationships in soil distribution as revealed by a global soil database
Geoderma
(2009) - et al.
Extrapolating regional soil landscapes from an existing soil map: sampling intensity, validation procedures, and integration of spatial context
Geoderma
(2008) Multi-criteria characterization of recent digital soil mapping and modeling approaches
Geoderma
(2009)- et al.
Methods to interpolate soil categorical variables from profile observations: lessons from Iran
Geoderma
(2007) - et al.
On digital soil mapping
Geoderma
(2003) - et al.
Effect of granite crystal grain size on soil properties and pedogenic processes along a lithosequence
Geoderma
(2015)
Digital soil mapping: A brief history and some lessons
Geoderma
Soil genesis, morphodynamic processes and chronological implications in two soil transects of SE Sardinia, Italy: Traditional pedological study coupled with laser ablation ICP-MS and radionuclide analyses
Geoderma
Modelling soil thickness in the critical zone for Southern British Columbia
Geoderma
Fundamental quantitative methods of land surface analysis
Geoderma
Correlations between pedological parameters in relation to lithology and soil type in Almería (SE Spain)
J. Arid Environ.
Implementation and evaluation of existing knowledge for digital soil mapping in Senegal
Geoderma
A GIS based method for soil mapping in Sardinia, Italy: A geomatic approach
J. Environ. Manag.
Comparison of scale and scaling issues in integrated land-use models for policy support
Agric. Ecosyst. Environ.
Recent progress and future prospect of digital soil mapping: a review
J. Integr. Agric.
Comparing the efficiency of digital and conventional soil mapping to predict soil types in a semi-arid region in Iran
Geomorphology
Soils and Geomorphology
Soil Genesis and Classification
Cited by (9)
Saprolite: A bibliometric study from 1990 to 2020
2022, Journal of South American Earth SciencesCitation Excerpt :Therefore, saprolite is the non-transported material that has no significant change in its volume as compared to the unweathered material, despite some loss of mass (Buol, 1994). The saprolite is strongly influenced by the properties of the original rock (Calzolari et al., 2021; Castillo et al., 2021). There is growing evidence that the saprolite performs fundamental functions in the environment (Le Pera et al., 2001; Frey et al., 2010; Dumig et al., 2011), such as water filtration (Taylor et al., 2010), pollutant retention (Drahota et al., 2018), phosphorus solubilization (Spohn et al., 2020), plant nutrient availability (Nicolitch et. Al, 2021) and gas flux (Smith et al., 2003; Nan et al., 2020).
Measurement of Soil Chemical Properties for Mapping Soil Fertility Status
2023, International Journal of Design and Nature and Ecodynamics