In response to high population density, the desert locust, Schistocerca gregaria, becomes gregarious and forms swarms that can cause significant damage to crops and pastures, threatening food security of human populations from western Africa to India. This switch from solitary to gregarious populations is highly dependent on favorable weather conditions. Climate change, which has been hypothesized to shift conditions towards increasing risks of gregarization, is therefore likely to have significant impacts on the spatial distribution and likelihood of outbreak events. However, the desert locust is intensely managed at large scales, which possibly counteracts any increased risk of outbreaks due to a more favorable climate. Consequently, understanding the changes in risks in the future involves teasing out the effects of climate change and management actions. Here we studied the dynamics of gregarization at the very early stages of potential outbreaks, in parallel with trends in climate and management, between 1985 and 2018 in western Africa. We used three different spatial scales, with the goal to have a better understanding of the potential effects of climate change per se while controlling for management. Our first approach was to look at a regional scale, where we observed an overall decrease in gregarization events. However, this scale includes very heterogeneous environments and management efforts. To consider this heterogeneity, we divided the area into a grid of 0.5° cells. For each cell, a climate analysis was performed for rainfall and temperature, with trends obtained by a harmonic decomposition model on monthly data. Analyses of gregarization showed only a few significant trends, both positive and negative, mainly found in western Mauritania where management effort has increased. To improve the statistical power, these cells were then grouped into larger homogeneous climatic clusters, i.e. groups of cells with similar climatic conditions and similar climat