Purpose: The complex process of post-mining land reclamation poses a great challenge in combatting land degradation and food insecurity. In this study, a field experiment was conducted to determine the effects of different soil amendments on soil organic carbon (SOC) characteristics and plant biomass to reclaim a highly degraded soil after tin-mining. Methods: Quantity and quality of SOC, annual plant biomass, and their relationships were determined over four years in post-tin mining soil that had been treated with six different soil amendments: control, dolomite, compost, charcoal, charcoal with compost and charcoal with sawdust. Results: Charcoal-compost application led to the highest increase in SOC stocks, followed by compost application. It enhanced annual plant biomass to twice that of the control treatment. In contrast to organic soil amendments, dolomite tended to cause SOC stock losses over time. Dissolved organic carbon exhibited a highly dynamic pattern, with a tendency to decrease over time. Dolomite and charcoal application induced SOC chemical composition changes distinct from other treatments. The potential interactions between amendment materials and time as well as crop species were observed for SOC characterization. Conclusions: Charcoal-compost combination with regular compost application was recommended to reclaim the soil for promoting SOC accumulation whilst sustaining crop production.

Natural resource management is a cross-sectoral topic, as reflected by its inclusion in several of the United Nations Sustainable Development Goals (e.g., SDGs 2, 6, 12, 15). In the study area on Bangka Island, Indonesia, agriculture is a pillar of local food security and livelihoods, making restoration of degraded lands also a local issue. The present study aims at restoring degraded land after tin mining and at restoring the natural soil base rendering it more suitable for agriculture. We use co-experimentation with citizens as a tool to develop options for re-habilitation at a pilot scale. The recruitment process in this study was reversed insofar as local citizens were at the origin of the project idea. Consequently, buy-in was high among local stakeholders. This set to increase the probability of successfully scaling up effective and actionable practices that were developed during co-experimentation at both local and regional levels. Co-experimentation provided a platform for exchange between local citizens and scientists. Citizens did not need to learn new skills to be able to participate in the scientific process and could autonomously evaluate results. We see involvement of citizens in this type of scientific projects not only as feasible, but as rewarding for all involved partners and as beneficial for the project outcomes. In light of the call for partnerships to achieve the Sustainable Development Goals, we can only recommend investing in communication and relationship building to work together on better solutions to the challenges we face.

West Java Province is the most landslide risky area in Indonesia owing to extreme geo-morphological conditions, climatic conditions and densely populated settlements with immense completed and ongoing development activities. So, a landslide susceptibility map at regional scale in this province is a fundamental tool for risk management and land-use planning. Logistic regression and Artificial Neural Network (ANN) models are the most frequently used tools for landslide susceptibility assessment, mainly because they are capable of handling the nature of landslide data. The main objective of this study is to apply logistic regression and ANN models and compare their performance for landslide susceptibility mapping in volcanic mountains of West Java Province. In addition, the model application is proposed to identify the most contributing factors to landslide events in the study area. The spatial database built in GIS platform consists of landslide inventory, four topographical parameters (slope, aspect, relief, distance to river), three geological parameters (distance to volcano crater, distance to thrust and fault, geological formation), and two anthropogenic parameters (distance to road, land use). The logistic regression model in this study revealed that slope, geological formations, distance to road and distance to volcano are the most influential factors of landslide events while, the ANN model revealed that distance to volcano crater, geological formation, distance to road, and land-use are the most important causal factors of landslides in the study area. Moreover, an evaluation of the model showed that the ANN model has a higher accuracy than the logistic regression model.