4. Conclusions
Information on the variation in soil element concentrations at different
spatial scales is needed for, e.g., planning land use and environmental
management. Knowledge of the concentration and spatial
distribution of naturally occurring elements in the soils of Sub-
Saharan Africa (SSA) is limited and there is a need for a better understanding
of the factors that might regulate the variations. TXRF provides
chemical fingerprinting for inferring soil chemical and physical
functional properties which is of interest in the African soil contexts
for agricultural and environmental management at large scales. The
AfSIS baseline dataset of SSA provides an excellent base for studying
chemical variations in soil composition at the continental scale related
to factors such as mineralogy, climate, topography, vegetation
and land use. Thus, this study helped to establish the baseline
concentrations of 17 elements for soils occurring within 34 sentinel
sites across SSA and documented variability in total element concentrations
within and between sites, which appeared to relate to differences
in mineralogical ‘functional groups’.Weobserved strongwithin site and
between site heterogeneity in many element compositions which were
related to soil forming factors. The exploratory analyses of the relationships
between element composition data and other site factors using
Random Forests regressions have demonstrated that all site or soilforming
factors (e.g., mineralogy, climate, topography, vegetation and
land use) have an important influence on total elemental concentrations
in the soil. The fact that the soil-forming factors are related to
the concentration of the naturally occurring elements in the soil gives
rise to the notion that they might be predicted from the soils' element
composition. The results also implied that N70% of variation in soil
element composition patterns can be predicted using information in
existing databases or readily observable features. Thus, future studies
should investigate the feasibility of quantitatively predicting soil
functional properties from concentrations of elements and the existing
databases or readily observable features e.g., for digital soil mapping.
4. ConclusionsInformation on the variation in soil element concentrations at differentspatial scales is needed for, e.g., planning land use and environmentalmanagement. Knowledge of the concentration and spatialdistribution of naturally occurring elements in the soils of Sub-Saharan Africa (SSA) is limited and there is a need for a better understandingof the factors that might regulate the variations. TXRF provideschemical fingerprinting for inferring soil chemical and physicalfunctional properties which is of interest in the African soil contextsfor agricultural and environmental management at large scales. TheAfSIS baseline dataset of SSA provides an excellent base for studyingchemical variations in soil composition at the continental scale relatedto factors such as mineralogy, climate, topography, vegetationand land use. Thus, this study helped to establish the baselineconcentrations of 17 elements for soils occurring within 34 sentinelsites across SSA and documented variability in total element concentrationswithin and between sites, which appeared to relate to differencesin mineralogical ‘functional groups’.Weobserved strongwithin site andbetween site heterogeneity in many element compositions which wererelated to soil forming factors. The exploratory analyses of the relationshipsbetween element composition data and other site factors usingRandom Forests regressions have demonstrated that all site or soilformingfactors (e.g., mineralogy, climate, topography, vegetation and
land use) have an important influence on total elemental concentrations
in the soil. The fact that the soil-forming factors are related to
the concentration of the naturally occurring elements in the soil gives
rise to the notion that they might be predicted from the soils' element
composition. The results also implied that N70% of variation in soil
element composition patterns can be predicted using information in
existing databases or readily observable features. Thus, future studies
should investigate the feasibility of quantitatively predicting soil
functional properties from concentrations of elements and the existing
databases or readily observable features e.g., for digital soil mapping.
การแปล กรุณารอสักครู่..
