This respotory gathers the scripts used for mapping genosoils and phenosoils and assessing changes in soil conditionas part of the SELVANS project.
Soils are essential for facing humanity’s existential challenges of food security, water security, climate change mitigation, and ecosystem functioning and biodiversity. Whilst the increasing demand for food, clean water and energy by the growing world population intensify the pressures on soils and threaten their functioning, the importance of maintaining the soil resources has gained recognition in the society and policy in the last decade. Soil functions are in the core of the updated EU Soil Thematic Strategy , and improving soil conservation is intrinsic to the EU Green Deal , Biodiversity strategy 2030 , Soil Mission Health , and EU Forest Strategy 2030. Improving the soil capacity for delivering ecosystem services is needed for achieving the SDGs. Soil-based ecosystem services (SES) in forests and agroforestry systems are linked to multiple SDGs, notably combat climate change (SDG13), sustaining the life on land (SDG15), achieving food security (SDG2), ensuring health and promoting well-being (SDG3), ensuring clean water and sanitation (SDG6) and affordable and clean energy (SDG7). The capacity of forests in the Iberian Peninsula for providing SES is expected to decrease due to the increase in aridity and climate change related disturbances (e.g., higher frequency and severity of wildfires, pathogens outbreaks). Drought events are related to crown defoliation and tree mortality with the consequent decrease in the capacity to sequester C and supply biomass, but also with the decline in soil regulating functions like nutrient cycling due to, among other reasons, the sensitivity of soil microbial communities to water availability . Tree species selection in forest plantations modify soil properties (pH, C:N ratio) through their mycorrhizal association and litter chemistry, which in turn influence the soil microbial community composition and functioning . At the same time, forestry practices inadequate for the site conditions (e.g., clearcutting in steep slopes) or that foster a single ecosystem function (e.g., monospecific commercial plantations) can lead to soil erosion, alter soil functioning and increase the vulnerability to disturbances. Hence, to guarantee the long-term provision of SES, forest soils should be managed according to their capability, with focus on regulating functions mediated by the soil microbial community. Soil capability and condition are two dimensions of soil security, which connects the biophysical, economic, and social soil attributes with SES, and can be used as roadmap towards the SDGs. Soil condition and capability are linked to the concepts of soil quality and soil health. The EU aims to “ensure that 75% of soils are healthy by 2030 and are able to provide essential ecosystem services”. But for evaluating whether soil management optimizes the supply of SES according to its potential, it is necessary to set soil-class specific baselines for local management. The first studies mapping soil security for Tasmania, China, and Italy, assigned thresholds for indicators of soil condition by soil order (e.g., Soil Taxonomy, Australian Soil Classification) and land use type. However, the definition of baselines specific to local conditions has not been addressed sufficiently. The concepts of genoform and phenoform connect detailed soil maps, management, and soil functions, and have been applied for investigating soil change. Recently, I have developed a digital soil mapping (DSM) approach for identifying soil-class references and assessing changes in soil condition over large areas based on a new concept, the pedogenon. This soil taxon is not associated to any established soil classification system, but it integrates a conceptual model of soil change. Kuzyakov and Zamanian proposed that fostering a single ecosystem function (crop production) leads to convergence of soil properties after long-term intensive agricultural use. The rates of change would differ for different soil properties, as a function of climate, biophysical conditions, and land use intensity. Similarly, the patterns of soil organic carbon (SOC), SOC chemistry and stability, and microbial functioning after agricultural abandonment vary across soil types, climate and vegetation, and hence control the direction of pedogenesis. Empirical relationships between key soil properties have been proposed for identifying critical values (e.g., SOC) after which changes in related soil properties affect soil functioning. Other studies have investigated the effects of management on soil health combining the concept of genoforms and phenoforms with clustering analysis. For forest soils, the effects of forest management practices and harvesting methods on SOC storage, soil nutrients and subsequent tree growth have been widely assessed with reviews and meta-analyses, long-term experiments and observational studies, and process-based modelling. The response of the soil property and recovery time after management practices often varies by soil type. Yet, we need a better understanding of the abiotic and biotic controls (climate, soil type, tree species) on microbial functioning and their response to management. SELVANS will expand the knowledge on soil change for forest soils from the Iberian Peninsula and investigate the links between pedogenesis and soil microbial functional diversity, and its effects on carbon and nutrient cycling. Multiple studies have quantified or mapped the ability for performing soil functions and SES, investigated how their trade-offs vary spatially depending on climate, land use and management, advancing in the spatial optimization of soil multifunctionality. The models and decision support tools for assessing the potential for supplying SES include multicriteria decision modelling, functions using static or dynamic soil properties, empirical and process-based models. However, only in few cases SES are evaluated with respect to the individual soil intrinsic potential. SELVANS will apply integrative modeling using reference soils to compare how land use history has modified the potential for supplying different SES in soils with similar pedogenesis. The overall goal of SELVANS is to implement an innovative framework for assessing the effects of forest management on soil condition with respect to reference soils. The sensitivity of forest ecosystems to global change requires to identify thresholds for key soil properties that alert of the potential loss of soil functioning. This knowledge will inform scientifically sound management strategies for maintaining forest soils multifunctionality, counteract the risk of soil degradation, and upscale estimates on soil condition and capability.
The specific objectives of SELVANS are to:
- Define pedogenon classes for the Iberian Peninsula at regional and local scale and adapt the pedogenon mapping approach to ecosystems with long history of intensive human use and pressure. The underlying hypothesis is that pedogenon classes result from similar historic anthropedogenetic processes, have developed similar inherent soil properties, and hence would have the same capability or potential for providing SES.
- Identify a reference state for each pedogenon class (i.e., soils with least change since the time chosen as benchmark). This objective can be reached by developing indices of human pressures on soil across the landscape combining remote sensing and management history data (plot and stand scale).
- Investigate the trajectories of soil condition indicators in response to contemporary land use and forest management history for different pedogenon classes. I hypothesize that the response (magnitude, direction) of dynamic soil properties to management practices will differ among pedogenon classes, as well as within a pedogenon class depending on management intensity (e.g., rotation length, harvesting method, tree species).
- Investigate the linkages between anthropedogenesis, soil inherent properties, soil microbial community, microbial functions and biogeochemical processes. In the case of forest plantations, I will test whether the effect of the forest species on dynamic soil properties is stronger that the pedogenon effect (i.e., inherent soil properties, climate, and environmental site conditions).
- Quantify the capacity of soils for delivering SES across pedogenon classes, specifically on climate regulation and nutrient cycling. The comparison between reference soils (e.g., forest stands with native species) and human-affected forest soils will indicate whereas the soil is being managed according to its capability and whether the ability to deliver SES has been reduced due to changes in condition.
- Develop a tool for guiding the sustainable management of forest soils. Detailed information -quantitative and spatially explicit- on soil condition and capability generated to address the previous objectives will be integrated into a data-driven knowledge tool for guiding the decision-making process.
SELVANS will be implemented through three phases that will permit me to achieve both the specific and overall goals. The first and second objectives of SELVANS will be achieved by the development of an improved DSM framework, which will permit pedogenon and pedophenon mapping of forest soils at regional and local scales (WP1). For this purpose, I will adapt the current DSM framework for assessing soil change that I designed in collaboration with Prof. Alex McBratney and Prof. Budiman Minasny at the University of Sydney. The two-stage DSM framework has been implemented at local, state (New South Wales), and continental scale (Australia). Pedogenon classes are modelled with data mining algorithms (unsupervised classification, e.g., k-means clustering) from a set of environmental covariates that represent the soil-forming factors for a given reference time, and can include information on relatively stable soil properties (e.g., soil texture, mineralogy). The assumption is that pedogenon classes result from similar multimillennial natural pedogenesis and historic anthropedogenesis (i.e., legacy of human activities on soil-forming processes over centuries) up to a reference time (WP1.1). Subsequently, each pedogenon is divided into subclasses by incorporating information on human forcings (drivers of contemporary soil change), ranging from the least disturbed pedogenon (i.e., reference state) to several pedophenons (i.e., variants resulting from contemporary management) depending on the type, intensity, and duration of the human activity. This novel framework produces soil spatial information at a relevant level of detail for both local and regional management. In the context of the Iberian Peninsula, a question arises with respect to the choice of a reference time and characterization of human forcings: How can we identify reference soils when these ecosystems are the result of long-term intensive human influence? For example, if the reference is the beginning of the XXth century, cadastral maps, historic aerial photographs, and maps of estimated natural vegetation can be included as proxies for organisms. In addition, human-shaped ecosystems (e.g., high-elevation managed pastures, dehesas) are assumed to define the soil system in quasi-steady state for that reference time. This methodology will be implemented using metropolitan France as case study, and developed under the supervision of Dr. Sophie Cornu (CEREGE, INRAE, Marseille) and in collaboration with Dr. Eva Rabot and Dr. Nicolas Saby (INRAE, Orléans). This scripts are gathered in the directory "FRANCE". Next, I will conduct a literature review and select the methods for integrating information available at stand scale (e.g., forest management) with spatially exhaustive layers to develop an index (or multiple indices) of anthropogenic pressure on forest soils (WP1.2). Time since disturbance or land use change is essential for assessing the effects of human activities on soils and the departure from its reference state. Long-term satellite imagery datasets will allow to identify forest disturbances and land use conversion (e.g., abandonment of agricultural areas) at high spatial and temporal resolution, using powerful software platforms like Google Earth Engine (GEE). Additional regional and local datasets on historic forestry practices (Spanish national forest inventory (NFI), ICP Forests) and maps of historic forest management will characterize the management intensity (e.g., rotation length, harvesting method). The last stage of pedogenon and pedophenon mapping consists in evaluating the maps (WP1.3). Legacy soil data on relatively stable (e.g., soil texture) and dynamic soil properties (e.g., SOC) from the CARBOSOL dataset will be used for map evaluation at regional scale. CARBOSOL, which gathers data of physical and chemical properties from 22,100 horizons and 6,609 georeferenced soil profiles, has a wide representation of forest ecosystems in Spain (1943 profiles in coniferous, 294 mixed, 949 in Quercus and 569 in broadleaf forests) with dates ranging between 1960-2011. In addition, soil data from previous and ongoing projects directed by Dr. Jorge Curiel Yuste (supervisor at BC3) in Mediterranean and Atlantic forests (ATLANTIS, HoliSoils) will be also used. The second phase of SELVANS will assess soil condition. The third objective will be achieved by applying different data visualization and multivariate data analyses on a comprehensive set of physical (e.g., available water capacity, bulk density), chemical (pH, SOC, nutrients) and biological (e.g., microbial community composition, microbial and fungi biomass, microbial diversity and functioning) indicators of soil condition. Pedogenon class and degree of human pressure will be assigned to soil data from ATLANTIS, IBERYCA and HoliSoils, and used to assess the change in soil condition (magnitude, direction) by soil class. Phase diagrams with key soil properties will be used to differentiate phases in soil change, identify thresholds of soil degradation, and for investigating the interdependence between key soil properties along the gradient of anthropogenic pressure (WP2.1). This research objective specially meets the research lines of the Terrestrial Ecology lab at BC3 (host group) and incorporates knowledge and methods of data analysis from the fields of soil ecology and biogeochemistry. This objective builds on previous and ongoing projects led by Dr. Curiel. The study sites from ATLANTIS include mixed stands of native species and exotic species plantations (Pinus radiata D. Don) in the Atlantic basin in the Basque Country along different conditions of management intensity and topography (12 plots). Given the richness of these soil datasets I anticipate that the gradient of anthropogenic pressure will be represented within several pedogenon classes. The same datasets will be used to determine the relationships between soil classes, management history, inherent soil properties, and biogeochemical processes (Objective 4, WP2.2). Causal relationships will be investigated with structural equation modelling and the hierarchy of controlling factors with linear mixed models. The third phase of SELVANS will model soil capability. A spatially explicit assessment of a subset of SES (climate regulation, nutrient cycling and supply) will be applied to quantify soil capability applying the Soil Security Assessment Framework developed at the University of Sydney (Objective 5, WP3.1). Soil data will come from all available datasets. The comparison between the capacity to supply SES by reference soils and soils managed/altered by human pressures within each soil class allows to evaluate the legacy of past management on SES due to differences in soil condition. Finally, the results generated to fulfil the previous objectives will be integrated into a tool for sustainable management of forest soils, available with an interactive web app created with Shiny for RStudio (Objective 6, WP3.2). The end user will be able to delineate an area of interest in the regional and local pedogenon maps, visualize phase diagrams on soil condition indicators for pedogenon classes of interest, enter soil test data from their forest stands to gain insight on the soil condition and functioning in their forests, and SES supply.