Tree growth is more sensitive than species distributions to recent changes in climate and acidic deposition in the northeastern United States

1. Tree-growth responses to environmental change could provide early detection of shifts in forest composition and help facilitate ecosystem management and conservation. 2. We studied forest tree responses to recent trends in climate and acidic deposition using analyses of tree rings and long-term climate, deposition and forest plot data along an elevational climatic gradient in the northeastern United States. We analysed how (i) individual growth of dominant species (Picea rubens, Abies balsamea) and (ii) spatial distributions of all species, changed with elevation over time due to changing environment. 3. We observed a mean 220 m upslope shift of temperature envelopes since the 1960s, consistent with regional climate warming, but found no evidence of synchronous upslope shifts in species abundance. Species' ranges were stable although some leaned upslope or downslope, suggesting species-specific migration lags or controls on species' ranges. 4. Compared to species distributions, the growth of dominant species was more responsive to environmental change. Although the basal area of P. rubens declined within its range since the 1960s, its growth has increased recently with increasing precipitation pH and to a lesser extent with warming climate. Abies balsamea has gradually increased in both basal area and density since the 1960s, with its growth responding to precipitation pH but not climate. Historically, P. rubens grew better at lower and A. balsamea at higher elevations, but these elevation effects appeared to be mediated primarily by moisture, and have disappeared over time. 5. Synthesis and applications. Mean tree-growth responses to changing climate (temperature, moisture) and precipitation chemistry were more consistent and more clearly detectable than shifts in tree species' ranges, suggesting that monitoring tree growth across climatically controlled species' ranges (e.g. along elevational or latitudinal gradients) may provide a powerful tool for early detection of potential future changes in forest composition in a changing environment.