Liming effect on radial growth depends on time since application and on climate in Norway spruce stands

► Time series of radial increments were analysed from long-term spruce liming trials. ► The growth gain associated to liming (GGL) first increased up to a maximum. ► Then GGL decreased linearly with time while remaining positive. ► GGL inter-annual variations were related to water availability (P–PET). ► Low P–PET values during the current vegetation period reduced GGL. During the 1980s and 1990s, liming was largely used as a tool to protect or restore forest stands located on acid soils from further acidification due to atmospheric deposition. Whereas impacts on soil properties and tree nutrition are well documented, little is known about the long-term temporal impact of liming on radial growth, and on the effect of climate on this response. We therefore analysed time series of radial increment on 128 dominant trees from spruce liming trials located in the Belgian (Gouvy, 1995–2008) and French (La Croix-Scaille, 1981–2008) Ardennes; these stands were limed at ca 40–45years old. The growth gain associated with liming (GGL, %) was defined by the ratio of the difference in Mean Individual Basal Area Increment between treatments (BAINC limed–BAINC control) corrected for initial differences, to BAINC of the control trees. The effect of climate on GGL was assessed by computing the difference between precipitation (P) and evapotranspiration (PET) over three reference periods of the current growing year (March–June, July–September, March–September). Liming resulted in a two stage growth response pattern. In the first stage (4–8years after treatment), GGL increased up to a maximum value of 29% in Gouvy and 54% in La Croix-Scaille. In a second stage, GGL decreased linearly with time from application while remaining >0. In La Croix-Scaille, the residuals around this tendency (Y) were positively and linearly related to the difference (P–PET; X) during the current vegetation period (Y=0.0037X−0.68; R2=0.36, P=0.0029), suggesting tree response to liming depended on water availability. Compared to the control, liming increased the expected growth response that is described by the linear trend, without considering the variation due to climate.