Contrasted Cross-Border Agricultural Patterns and Trends in the Irrigated Senegal River Valley since 2016

Kurzfassung

Food production in West Africa is closely linked to two major challenges of the 21st century: population growth and climate change related climate variability. By 2050, West Africa’s population is expected to reach 1.2 billion increasing food demand, changing diets through urbanization, thus posing considerable challenges to food production systems. Key strategies towards sufficient food supply and the Sustainable Development Goals (SDGs) are sustainable intensification of agriculture (increasing yields without additional land consumption and without adverse effects on climate change) and mitigation and adaptation strategies considering local socio-economic conditions. In the Senegal River Valley, irrigation allows for up to three cropping cycles per year. However, environmental factors such as untimely and excessive rains and reservoir releases, as well as heterogeneity and volatility in socioeconomic conditions – e.g., market access and prices – lead to reduced efficiencies. The Senegal River marks the border between Mauritania and Senegal and represents a clear separation between two distinct cropping systems and coping strategies. While the Mauritanian side is characterized by fast-growing large-scale farming systems that are less susceptible to socio-economic fluctuations, smaller farm sizes and a huge subsistence to intensification gradient dominate on the Senegalese side. This work is embedded in the COINS project (Co-developing innovation for sustainable land management in West African small holder farming systems), where we assess which sustainable intensification practices suit for who under which socio-economic conditions with all relevant local stake holders. Therefore, we analyze annual agricultural patterns and trends inside the Senegal River Valley for nearly a decade (2016-today), hence covering the COVID years. We used a random forest classifier to delineate agricultural fields based on Sentinel-1 and Sentinel-2 time series data, trained with samplings from mono temporal high-resolution imagery. Results clearly show contrasted cross-border spatial and temporal dynamics in cropland, especially during the COVID-19 pandemic marked by inter-regional travel restrictions affecting the agricultural labor force. Over the period, a substantial increase of the agricultural area in Mauritania contrasts with a limited expansion in Senegal. There are considerable variations in the annual cropped area in Senegal, with significant areas not put into cultivation notably towards the higher Valley in the East. The study shows the high potential of remote sensing to monitor the agricultural area and to detect abandoned and uncultivated parts at an early stage. In the future, this potential could be leveraged to initiate predict planted area across seasons and guide decision making towards optimized cropland use and sustainable increase in productivity.