Global climate drivers often have strong effects on the carrying capacity of animal populations, but little is known about how effects differ between regional and local scales. In this paper we evaluated how climate variables were correlated with regional and local fluctuations of a small rodent, Necromys lasiurus, in an Amazonian savanna. Between 2000 and 2019, we evaluated the temporal variation in abundance of N. lasiurus in eight 4.0-ha plots separated by 0.8 – 10.6 km. Using generalized linear mixed models, we found that, at a regional scale, the abundance of rodents captured was positively associated with the abundance in the prior year, but had little relationship with the Southern Oscillation Index (SOI), which had been shown to affect rats in a single plot in a previous study. However, variation in densities among years was coordinated among some plots, leading to patchiness in population dynamics. Based on the patterns of density fluctuations, the plots formed three clusters. Analyses based on these clusters indicated that only one was strongly affected by SOI, as in the previous study. The differences in the effects of global climate drivers on populations of a single species in relatively homogeneous habitat indicate that predictions about the effects of climate change should be based on simultaneous studies in a variety of sites or they may lead to spurious relationships.