RT Dissertation/Thesis
T1 Impact of environmental and socio-economic factors on soil fertility variability and microbial carbon use efficiency in tropical smallholder farming systems
A1 Agumas Endalew,Birhanu
WP 2022/04/19
AB The main drivers of soil fertility variability across Sub-Saharan Africa (SAA) must be understood to develop tailor-made integrated soil fertility management (ISFM) strategies, considering agro-ecological zones, smallholder farmers’ resource endowment and their indigenous knowledge of soil fertility. Moreover, most soil fertility indicators including, but not limited to total soil organic carbon (SOC) content, lack in sensitivity and accuracy. The insensitivity and inaccuracy of these indicators impedes their application for soil fertility surveys in smallholder farming systems across larger spatial scales. Hence, the verification of novel soil fertility indicators, such as SOC functional groups and microbial carbon use efficiency (CUE) as influenced by environmental factors (e.g. soil pH, organic input quality), become paramount important to overcome this constraint. The implementation of such methodological innovation would help to better understand the extent of regional soil fertility variability and subsequently design niche-based ISFM strategies for smallholder farming systems in SSA.
Therefore, the first aim of this study was to explore the interrelated effects of biophysical and socio-economic factors on soil fertility variability, as reflected by soil nutrient contents as well as SOC content and quality parameters (i.e., SOC functional groups). The second aim was to evaluate soil microbial CUE as an additional proxy to assess soil fertility considering the influence of environmental and methodological variations on CUE calculation.
The specific objectives of this PhD study were to: 
• verify that soil fertility variability across two model regions in Central and Western Ethiopia with four distinct agro-ecological zones could be determined by the inter-related effects of agro-ecology and farmers’ resource endowment (“wealthy” versus “poor” farmers). 
• confirm this approach of local soil fertility assessment in Ethiopia by including “market distance” as an additional factor for soil fertility variability, as exemplified in the Democratic Republic of Congo (DRC). 
• test whether farmers’ indigenous knowledge on soil fertility status is driven by inter-related effects of agro-ecology, market distance and farm typology, considering the continuous knowledge transfer among farmers within and across agro-ecological zones.
• evaluate the potential of SOC functional groups and soil microbial CUE as promising indicators of soil fertility status influenced by physico-chemical soil properties and organic input management.
• to modify the exsiting single C-cycling enzymatic stoichiometry (SCE-STM) through proposing novel “multi”-C-cycling enzymatic stoichiometry (MCE-STM) methods for soil microbial CUE estimation.
To tackle objectives 1-3 of the presented PhD study, two local field-based soil fertility surveys were conducted in Ethiopia and DRC. A lab-based incubation study was implemented for objectives 4 and 5. For the soil fertility surveys, mid-infrared spectroscopy coupled to partial least squares regression (midDRIFTS-PLSR) and wet-lab analyses were used to assess the soil fertility (i.e., soil pH, total soil carbon (TC), total soil nitrogen (TN), plant-available phosphorous (Pav) and potassium (Kav), exchangeable calcium (Caex) and magnesium (Mgex)) across four agro-ecological zones in Ethiopia. MidDRIFTS peak area analysis of spectral frequencies (2930 (aliphatic C-H), 1620 (aromatic C=C), 1159 (C-O poly-alcoholic and ether groups) cm-1) were applied to characterize SOC quality and to calculate the SOC stability index (1620:2930, 1530:2930). While in DRC, both techniques were employed to assess soil fertility proxies across market distances (defined as walking time) in distinct regions. For the lab-based incubation study (60 days), two soils differing mainly in acidity level mixed with two specimens of plant residues differing mainly in lignin (L) and polyphenol (PP) content were used. For estimating soil microbial CUE during plant residue decomposition in the different soils, single C-cycling enzymatic stoichiometry (SCE-STM) and the newly proposed “multi”-C-cycling enzymatic stoichiometry (MCE-STM) methods were validated against the conventional C-balance method.
MidDRIFTS-PLSR and peak area analysis results of the Ethiopian case study showed that the inter-related effects of agroecology and farmers’ resource endowment determined the observed soil fertility variability across four agro-ecological zones. Resource endowment dependent soil fertility management options revealed higher TC in the high altitude agro-ecological zone, while higher TN and Kav was found in the lower agro-ecological zones in the fields of wealthy farmers. Similarly, SOC of higher quality was found in soils of wealthy than poor farms in higher altitude zones. Thus, agro-ecological zone distinctions contributed to these differences in soil fertility variability. It was dedeuced that this difference in soil fertility status between wealthy and poor farmers’ fields across agro-ecological zones has been due to the high variability in landholding size per capita, livestock population and amount of fertilizer used per unit area. 
Complementary, the results of the DRC case study revealed that “market distance” and “farm typology” were key determinants of soil fertility variability, both with contrasting trends in the study sites. Decreasing soil fertility status was noted across all farm typologies with increasing market distance. A significant influence of “farm typology” was found for Caex and Mgex, while factor “site” resulted in a significant difference of Pav. For SOC quality indices (i.e., ratio 1530:2930), factor “site” was decisive, as reflected in its interaction with “market distance”. However, the effect of market distance became obvious in the medium wealthy and poor farmers fields, where an increasing SOC quality index of 1530:2930 with increasing market distance implied a lower SOC quality in remote farms.
Soil depth and soil color were the most frequently used soil fertility indicators by farmers across agro-ecologies, market distances, and farm typologies. Concerning farmers’ indigenous knowledge across the study regions in Ethiopia and DRC, fertile and less fertile fields were distinguished visually by soil color. Higher pH and Pav were found in fertile (brown/black) than less fertile (red) soils in most agro-ecological zones of the Ethiopian case study. Furthermore, higher peak areas of 1159 cm-1 and SOC stability indices were observed in less fertile compared to fertile soils in Ethiopia. In close agreement with farmers’ indigenous knowledge in the DRC study region, soil fertility levels were higher in deep than shallow soils, which was reflected in higher nutrient stocks in deep soils receiving organic amendments. Accordingly, site-specific soil management strategies with the integration of farmers’ indigenous knowledge will be a feasible option to overcome the low adoption of ISFM.
This PhD study suggested the use of more sensitive indicators, such as soil microbial CUE, to accurately assess soil fertility status for the design of niche-based soil fertility management decisions. Furthermore, the PhD study showed that higher CUE was recorded in more fertile and less acidic (pH 5.1) soils amended with residues of higher quality than the other three combinations. It was deduced that microorganisms invested more energy to support growth in more acidic (pH 4.3) soil to tolerate soil acidity, which, in turn, suppressed N-acquiring enzymatic activities and further decreased CUE. Lower CUE values were recorded from multi-C cycling enzymatic stoichiometry modeling (MCE-STM) as compared to the CUE values obtained from C-balance and single-C cycling enzymatic stoichiometry modeling (SCE-STM) methods. The modification of the MCE-STM method for CUE determination proposed in this dissertation work was capable to quantify the combined effect of soil pH and plant residue quality on the efficiency of microbial metabolism. As a result, it improved the original stoichiometric modeling approach (SCE-STM), which relied only on the concept of nutrient availability. 
In conclusion, for regional soil fertility assessment, midDRIFTS-PLSR predictions along with midDRIFTS peaks representing SOC functional groups proved to be a sensitive as well as more efficient and robust approach as compared to the existing aproaches relying on classical soil properties (e.g., SOC content) assessed by wet lab analyses. Based on the generated data using midDRIFTS, the main drivers of soil fertility variability have been revealed, considering specifically the interrelated effects of agro-ecology, resource endowment, market distance and farmers’ indigenous knowledge. Furthermore, integration of soil microbial CUE (e.g. MCE-STM) in soil fertility assessments does not only provide a clearer picture of soil fertility statuts. It also serves for better understanding of ecologcical processes in soils in general. With his, this PhD study fostered the knowledge of soil fertility drivers across spatial scales and laid the scientific basis for the furthering of novel soil fertility indicators based on soil microbial CUE. This outcome will benefit the design of niche-based soil fertility management strategies, which are of paramount importance to secure the livelihoods of SSA smallholder farming systems.

K1 Bodenfruchtbarkeit
K1 SOC
K1 Stabilitätsindex
K1 Agrarökologische Zonen
K1  Extrazelluläre Enzyme Qualität von Pflanzenresten
K1 Marktdistanz
PP Hohenheim
PB Kommunikations-, Informations- und Medienzentrum der Universität Hohenheim
UL http://opus.uni-hohenheim.de/volltexte/2022/2023