Browsing by Person "Reinhardt, Nadja"

Now showing 1 - 1 of 1

Gamma-ray spectrometry as auxiliary information for soil mapping and its application in research for development
(2019) Reinhardt, Nadja; Hermann, Ludger
Sustainable yield increase is desperately needed for enhancing global food security, in particular, in Sub-Saharan Africa. There population growth and resulting land degradation accompany with extreme weather events. As a consequence, famines frequently occur. For planning result-oriented agricultural research for development (R4D) like in the Trans-Sec project (www.trans-sec.org), in which this thesis was embedded, local environmental, as well as social realities must be taken into account prior to any cropping experiment. Only this way, cost-efficient and adapted solutions for local subsistence farmers, but also conclusive outcomes for researchers, can be obtained. For this purpose, methods that work quick and cost-efficient are a prerequisite. In this respect, gamma-ray spectrometry as rapid soil survey method is reviewed in the first part of this thesis. Soil or geological exploration are easily accomplishable, in either airborne (with helicopters, airplanes or drones) or proximal (stationary or on-the-go) surveys. Gamma decays of the naturally occurring isotopes 40-potassium (40K), 238-uranium (238U) and 232-thorium (232Th) that appear in sufficient amounts and decay energies for field measurements are counted per time. The counts are then transferred to the respective element contents. Water and soil organic matter attenuate gamma signals, on one hand hampering signal interpretation, on the other hand indirectly enabling soil water content and peat mappings. Gamma-ray signatures of soils depend on (1) mineral composition of the bedrock, as well as (2) weathering intensity and related soil forming processes, that, in turn, influence the environmental fate of 40K, 238U and 232Th. Hence, due to soil formation heterogeneity at the landscape scale, resulting gamma signatures are locally specific and make soils readily distinguishable. In two villages in central Tanzania, participatory soil mapping in combination with gamma-ray spectrometry served as rapid and reliable approach to map local soils for later cropping experiments. Local farmers indicated major soil types on satellite images of the village area, which were the basis for further mapping steps. Fingerprint gamma-ray signatures of reference soil profiles were collected. Subsequent gamma-ray surveys on transect walks accelerated soil unit delineation for the final soil map. Challenges were misunderstandings related to language issues, variable soil knowledge of individual farmers and erosion leading to staggered soil profiles and non-distinctive signatures in some places. The combination of indigenous knowledge and gamma-ray spectrometry, nevertheless, led to a quick overview of the study area and made laboratory soil analyses largely redundant. The gained gamma-ray signal information were further statistically evaluated. For this purpose, distinction of major local soil types via K/Th ratios were graphically and statistically tested. The results showed that gamma-ray spectrometry is a sound method to distinguish certain local clay illuviation soil types by their K/Th ratios. The last part of the thesis covers the Trans-SEC approach of testing innovations for sustainable agricultural yield increase. Pearl millet (Pennisetum glaucum (L.) R.Br.) as the typical staple food in the study region was used as example crop. The process was scientist-led but local farmers selected the innovations that they considered adequate to their needs. Tied ridging for enhancing the water storage and placed fertilizer for increasing fertilizer efficiency was offered for their choice. Transferability of results from on-station experiments and demonstration plots in the village to farmers plots and trans-disciplinary issues are discussed. The number of factors that influence the result, as well as data insecurity increased with every level of spatial aggregation (on-station, demonstration plot and on-farm plots in the village). Soil type, position of the plot in the landscape (lateral water flow, distance to homesteads and, hence, fertility status) were the major influencing factors. In particular, the data insecurity related to on-farm trials due to low control intensity suggests to only conduct such experiments if large numbers of replicates (large N-trials) are feasible in future approaches. In conlusion, the thesis shows, that local knowledge combined with modern science is beneficial for agricultural R4D projects. Shortcomings within the transdisciplinary experimental approaches are pointed out. In particular, with respect to knowledge gained from the linkage of local experience and scientific approaches, there is still high potential. For this purpose, social and applied natural sciences should both strive for more interdisciplinary collaboration.