Assessing impact of agroecological interventions in Niger through remotely sensed changes in vegetation

Abstract

Water scarcity is a major challenge in the Sahel region of West Africa. Water scarcity in combination with prevalent soil degradation has severely reduced the land productivity in the region. The decrease in resiliency of food security systems of marginalized population has huge societal implications which often leads to mass migrations and conflicts. The U.S. Agency for International Development (USAID) and development organizations have made major investments in the Sahel to improve resilience through land rehabilitation activities in recent years. To help restore degraded lands at the farm level, the World Food Programme (WFP) with assistance from USAID's Bureau for Humanitarian Assistance supported the construction of water and soil retention structures called half-moons. The vegetation growing in the half-moons is vitally important to increase agricultural productivity and feed animals, a critical element of sustainable food security in the region. This paper investigates the effectiveness of interventions at 18 WFP sites in southern Niger using vegetative greenness observations from the Landsat 7 satellite. The pre - and post-intervention analysis shows that vegetation greenness after the half-moon intervention was nearly 50% higher than in the pre-intervention years. The vegetation in the intervened area was more than 25% greener than the nearby control area. Together, the results indicate that the half-moons are effective adaptations to the traditional land management systems to increase agricultural production in arid ecosystems, which is evident through improved vegetation conditions in southern Niger. The analysis shows that the improvement brought by the interventions continue to provide the benefits. Continued application of these adaptation techniques on a larger scale will increase agricultural production and build resilience to drought for subsistence farmers in West Africa. Quantifiable increase in efficacy of local-scale land and water management techniques, and the resulting jump in large-scale investments to scale similar efforts will help farmers enhance their resiliency in a sustainable manner will lead to a reduction in food security shortages.

Figure 1

NDVI map for two locations Danja (a-c) and Elkokia (d-f) for the month of Jun before [year 2013 (a) and (d)] and after [year 2019 (b) and (e)] interventions. The bottom panel (c and f) shows the percent difference in NDVI values from Jun 2013 to Jun 2019.

Figure 2

Mean peak NDVI across all sub-polygons for all sites intervened before and after intervention. The error bars represent the standard error in mean NDVI values from different sub-polygons within a site. [Danja, Koona and Raffa are single polygon sites, hence no error bars].

Figure 3

Scatter plots (a) showing relationship between the mean peak NDVI and total rainfall before (red) and after the intervention (blue) and (b) scatter plot between percent differences in NDVI and rainfall before and after intervention for each of the sites.

Figure 4

Average monthly NDVI values for two sites, Kafat and Danja, from Landsat 7 before and after the intervention. The shaded areas on the left side panels represent one standard deviation computed from multiple polygons at each site.

Figure 5

NDVI values at the control and intervention sites from 2010 to 2020 for Dargue and Karkara. The vertical lines show the beginning and end of control/experiment periods.

Figure 6

Difference in mean NDVI during the baseline and experiment period for control and intervened sites.

Figure 7

WFP intervention sites in southern Niger include interventions intended for agricultural (blue) and pastoral (yellow) uses. The figure was created using ArcGIS Pro 3 (www.arcgis.come).

Figure 8

(a) shows the number of polygons intervened every year between 2013 and 2020. (b) is the histogram of the area in hectares for each of the polygons showing the distribution of polygon sizes. (c) elatively few were intervene represents the number of polygons at each of the sites that were intervened during the study period in the region.

Figure 9

(a) standard rainfall anomaly for southern Niger over the last 40 years (1982–2021). The shaded portion highlights the time period analyzed in this study. (b) shows the percent monthly distribution of long-term rainfall in the region, indicating a clear distinct rainy season and long summer months.