By artificially regulating key drivers, trade-off and synergy relationships of ecosystem services can transform into each other. Under the background of continuously promoting the construction of ecological civilization in China, scientific regulation of ecosystem services trade-off and synergy in "the areas with extremely important ecosystem service functions" is of great significance for maintaining regional ecological security and promoting the sustainable supply of ecosystem services. Recognizing the long-time patterns of change in ecosystem service trade-off and synergy and their drivers is important for the scientific regulation of ecosystem services. However, there is lack of understanding of the drivers of ecosystem service trade-off and synergy and their non-linear impacts based on long-time sequence analysis. In this paper, we took "the areas with extremely important ecosystem service functions" in Wuhu City as a case study, considering three key ecosystem services, including habitat quality, carbon sequestration, and soil conservation, using multi-source data and pixel-by-pixel trend superposition analysis method to evaluate the change trend of ecosystem services and the spatial distribution of trade-offs in the region from 1990 to 2020. This paper also used the random forest model to explore the key drivers of ecosystem service trade-off synergy and its nonlinear impact effect. The results showed that: (1) In the past 31 years, the area where the habitat quality in Wuhu City had significantly declined was about twice the area that has increased, and there was a risk of further decline. At the same time, urban expansion led to a significant decline in carbon sequestration and soil conservation in the suburban fringes. (2) The supply capacity of ecosystem services in the study area was significantly affected by the trade-off and synergy, and the affected area accounted for 64.48% of the total area. Significantly spatial differences existed between trade-off affected areas and synergistically affected areas. The area affected by synergy was higher than the area affected by trade-off, and the area affected by synergy was concentrated in hilly areas with low negative impacts of human activities and good vegetation and hydrothermal conditions. (3) Land use intensity played a dominant role in transformation of the managed ecosystem service trade-offs to synergies. The impact of land use intensity on trade-off and synergy did not change linearly with the increase of land use intensity value, but showed nonlinear characteristics and had an impact threshold. When the intensity level was between weak and medium and tended to be weak, it had a positive effect on ecosystem services synergy.