Environmental changes in biodiversity hotspot ecosystems of
South Ecuador: RESPonse and feedback effECTs (FOR2730)

Abstract:

Soils of Páramo ecosystems regulate the water supply to many Andean populations. In spite of being a necessary input to distributed hydrological models, regionalized soil water retention data from these areas are currently not available. The investigated catchment of the Quinuas River has a size of about 90 km2 and comprises parts of the Cajas National Park in southern Ecuador. It is dominated by soils with high organic carbon contents, which display characteristics of volcanic influence. Besides providing spatial predictions of soil water retention at the catchment scale, the study presents a detailed methodological insight to model setup and validation of the underlying machine learning approach with random forest. The developed models performed well predicting volumetric water contents between 0.55 and 0.9 cm3 cm? 3. Among the predictors derived from a digital elevation model and a Landsat image, altitude and several vegetation indices provided the most information content. The regionalized maps show particularly low water retention values in the lower Quinuas valley, which go along with high prediction uncertainties. Due to the small size of the dataset, mineral soils could not be separated from organic soils, leading to a high prediction uncertainty in the lower part of the valley, where the soils are influenced by anthropogenic land use.

Abstract:

Tropical forests are significant carbon sinks and their soils’ carbon storage potential is immense. However, little is known about the soil organic carbon (SOC) stocks of Tropical mountain areas whose complex soil-landscape and difficult accessibility pose a challenge to spatial analysis. The choice of methodology for spatial prediction is of high importance to improve the expected poor model results in case of low predictor-response correlations. Four aspects were considered to improve model performance in predicting SOC stocks of the organic layer of a tropical mountain forest landscape: Different spatial predictor settings, predictor selection strategies, various machine learning algorithms and model tuning. Five machine learning algorithms: random forests, artificial neural networks, multivariate adaptive regression splines, boosted regression trees and support vector machines were trained and tuned to predict SOC stocks from predictors derived from a digital elevation model and satellite image. Topographical predictors were calculated with a GIS search radius of 45 to 615 m. Finally, three predictor selection strategies were applied to the total set of 236 predictors. All machine learning algorithms—including the model tuning and predictor selection—were compared via five repetitions of a tenfold cross-validation. The boosted regression tree algorithm resulted in the overall best model. SOC stocks ranged between 0.2 to 17.7 kg m-2, displaying a huge variability with diffuse insolation and curvatures of different scale guiding the spatial pattern. Predictor selection and model tuning improved the models’ predictive performance in all five machine learning algorithms. The rather low number of selected predictors favours Forward compared to backward selection procedures. Choosing predictors due to their indiviual performance was vanquished by the two procedures which accounted for predictor interaction.

Abstract:

With respect to sampling for regression-based digital soil mapping (DSM), the above all aim is to ensure that the spatial variability of the soil is well-captured without introducing any bias, while the design remains feasible according to operational constraints such as accessibility, man power and cost. Representativeness of the sample concerning the population to be sampled needs to be guaranteed in any regression-based modelling approach. Four selected sampling designs were adapted to show that basically any design may be optimised to represent the n-dimensional predictor space of a particular area, while selecting points is only permitted from a small accessible sub-area or from outside the area. Sampling efficiency may be evaluated based on the representation of the predictor space. However, not only each predictor’s probability function but also the interaction between predictors may have to be considered, to select a representative sample. Instead of sampling a previously un-sampled area with limited accessibility, the four sampling designs may also be used to subsample an existing dataset and, thereby, optimise a suboptimal dataset based on the predictor space of the area which shall be mapped by DSM.

Abstract:

A digital soil mapping approach is applied to a complex, mountainous terrain in the Ecuadorian Andes. Relief features are derived from a digital elevation model and used as predictors for topsoil texture classes sand, silt, and clay. The performance of three statistical learning methods is compared: linear regression, random forest, and stochastic gradient boosting of regression trees. In linear regression, a stepwise backward variable selection procedure is applied and overfitting is controlled by minimizing Mallow’s Cp. For random forest and boosting, the effect of predictor selection and tuning procedures is assessed. 100-fold repetitions of a 5-fold cross-validation of the selected modelling procedures are employed for validation, uncertainty assessment, and method comparison. Absolute assessment of model performance is achieved by comparing the prediction error of the selected method and the mean. Boosting performs best, providing predictions that are reliably better than the mean. The median reduction of the root mean square error is around 5%. Elevation is the most important predictor. All models clearly distinguish ridges and slopes. The predicted texture patterns are interpreted as result of catena sequences (eluviation of fine particles on slope shoulders) and landslides (mixing up mineral soil horizons on slopes).

Abstract:

The sloping mire landscape of the investigation area, in the southern Andes of Ecuador, is dominated by stagnic soils with thick organic layers. The recursive partitioning algorithm Random Forest was used to predict the spatial water stagnation pattern and the thickness of the organic layer from terrain attributes. Terrain smoothing from 10 to 30m raster resolution was applied in order to obtain the best possible model. For the same purpose, several model tuning parameters were tested and a prepredictor selection with the R-package Boruta was applied. Model versions were evaluated and compared by 100 repetitions of the calculation of the residual mean square error of a five-fold cross-validation. Position specific density functions of the predicted soil parameters were then used to display prediction uncertainty. Prepredictor selection and tuning of the Random Forest algorithm in some cases resulted in an improved model performance.We therefore recommend testing prepredictor selection and tuning to make sure that the best possible model is chosen.This needs particular emphasis in complex tropical mountain soil-landscapes which provide a real challenge to any soil mapping approach but where Random Forest has proven to be successful due to the testing of model tuning and prepredictor selection.

Abstract:

The World Reference Base for Soil Resources (WRB) (FAO, IUSS Working Group WRB, 2007) at present does not acknowledge the spatial soil continuum, but provides a sound basis to do so. Using methods from statistical learning theory to develop digital soil maps is much more efficient and precise while regionalising soil diagnostic properties instead of complex entities such as the soil units assigned by the WRB. Particularly in providing spatial soil information displayed in digital soil maps, any aggregation of this spatial soil information to soil units means a loss of information. The soil landscape can be systematically described in its spatial continuum simply by the vertical order and extent of the WRB diagnostic horizons. The diagnostic horizons are related in their thickness to a standard depth and listed from top to bottom in order of appearance. Typical diagnostic horizon thickness and occurrence probability were predicted from terrain parameters by classification and regression trees (CART), throughout the research area in southern Ecuador. The two disadvantages of CART, abrupt prediction class boundaries and dependence on the dataset, were addressed by hundredfold model runs on different data subsets, leading to a range of possible predictions. Prediction uncertainty was included in the digital soil maps by calculating these predictions' means and standard deviations as well as by horizon occurrence probability prediction. Model performance was evaluated by means of hundredfold external cross validation. Terrain parameters were found to have a strong influence on diagnostic topsoil properties. However, no influence on the vertical profile differentiation was observed. Hence predicting horizon thickness and subsoil diagnostic properties was difficult. The systematic description of the soil continuum of this particular soillandscape resulted in histic and stagnic soil parts dominating the first 100 cm of the soil column for most of the area.

Abstract:

Landslides are a common phenomenon within the Ecuadorian Andes and have an impact on soil-landscape formation. Landslide susceptibility was determined in a steep mountain forest region in Southern Ecuador. Soil mechanical and hydrological properties in addition to terrain steepness were hypothesised to be the major factors in causing soil slides. Hence, the factor of safety (FS) was calculated as the soil shear ratio that is necessary to maintain the critical state equilibrium on a potential sliding surface. Regression tree (RT) and Random Forest (RF) models were compared in their predictive force to regionalise the depth of the failure plane and soil bulk density based on terrain parameters. The depth of the failure plane was assumed at the lower boundary of the stagnic soil layer or soil depth respectively, depending on soils being stagnic or nonstagnic. FS was determined in dependence of soil wetness referring to 0.001, 0.01, 0.1 and 3 mm h−1 net rainfall rates. Sites with FS≥1 at 3 mm h−1 (complete saturation) were classified as unconditionally stable; sites with FSb1 at 0.001 mm h−1 as unconditionally unstable. Bulk density and the depth of the failure plane were regionalised with RF which performed better than RT. Terrain parameters explained the spatial distribution of soil bulk density and the depth of the failure plane only to a relatively small extent which is reasonable due to frequent translocation of soil material by landslides. Nevertheless, their prediction uncertainty still allowed for a reasonable prediction of nconditionally unstable sites.

Abstract:

Soil landscape modelling is based on understanding the spatial distribution patterns of soil characteristics. A model relating the soil?s properties to its position within the landscape is used to predict soil properties in other similar landscape positions. To develop soil landscape models, the interaction of geographic information technology, advanced statistics and soil science is needed. The focus of this work is to predict the distribution of the different soil types in a tropical mountain forest area in southern Ecuador from relief and hydrological parameters using a classification tree model (CART) for soil regionalisation. Soils were sampled along transects from ridges towards side valley creeks using a sampling design with 24 relief units. Major soil types of the research area are Histosols associated with Stagnosols, Cambisols and Regosols. Umbrisols and Leptosols are present to a lesser degree. Stagnosols gain importance with increasing altitude and with decreasing slope angle. Umbrisols are to be found only on slopes <30°. Cambisols occurrence might be related to landslides.The CART model was established by a data set of 315 auger sampling points. Bedrock and relief curvature had no influence on model development. Applying the CART model to the research area Histosols and Stagnosols were identified as dominant soil types. Model prediction left out Cambisols and overestimated Umbrisols, but showed a realistic prediction for Histosols, Stagnosols and Leptosols.