Abstract: Ongoing environmental change threatens the persistence of species in ecological communities. Plant-pollinator mutualistic networks, which support biodiversity and ecosystem functioning, are particularly vulnerable to such disturbances. Although environmental variation is known to modify demographic rates and weaken mutualistic interactions, how its pace and stochasticity jointly affect the stability of mutualistic networks remains largely unknown. Here, we analyse a high-dimensional mutualistic network model with time-varying demographic rates, such as mortality and competition alongside noisy fluctuations. A reduced two-dimensional model reveals that sufficiently fast parameter change can trigger rate-driven pollinator collapse, which is further expedited by noise. These dynamics persist in the high-dimensional mutualistic network. Using 30 empirical plant-pollinator communities, we show that network structure is associated with vulnerability to rapid demographic change, as measured by critical rates: higher connectance and nestedness increase tolerance to rapid changes across noise intensities, whereas higher modularity reduces stability, leading to collapse at lower rates of change. Overall, our results highlight that the pace and variability of demographic changes interact with network structure to shape the stability and persistence of mutualistic communities. *The article is published in "Proceedings of the Royal Society B" under the DOI: https://doi.org/10.1098/rspb.2026.0296.