Dominant radial and height growth reveal comparable historical variations for common beech in north-eastern France

While forest productivity is usually inferred from height growth indices, retrospective analyses of tree rings have been intensively used to assess long-term trends in forest productivity. However, radial growth is sensitive to the degree of competition between trees and influenced by management practices or local disturbances. Whether radial growth is accurate for diagnosing and quantifying productivity changes remains a debated question.In a previous study (Bontemps et al., 2009), we assessed historical variations in dominant height growth of even-aged stands of common beech in north-eastern France as a proxy for their productivity changes. The analysis was based on a sampling design including 14 pairs of young/old (75/150 yr) neighbour stands growing under the same site conditions. Dominant height was reconstructed from stem analyses and was compared between generations using a statistical modelling procedure.In this analysis, we tested whether radial and height growth of dominant trees may provide compatible indications on long-term trends. We therefore measured and analysed the radial growth of dominant trees at breast height for the same sampled trees. The effects of site, developmental stage, and calendar date were separated by applying a similar modelling approach. Consideration of the developmental stage effect led to the formulation of an original growth equation.Analysis of radial growth revealed: (i) a long-term positive increase; (ii) a magnitude of +50% over the last century; and (iii) growth declines in the 1940s and 1990s. These features were remarkably similar to those reported on dominant height, and indicated that radial growth of dominant trees delivered a sound picture of productivity changes. The radial growth chronology also differed by showing a more acute acceleration phase in the early century, and a recent but significant difference between stand generations.