Use of a spatial process-based model to quantify forest plantation productivity and water use efficiency under climate change scenarios

Abstract

The area of commercial Eucalyptus plantations has expanded dramatically in many countries during the last three decades, and continues to do so. This is causing concerns about the potential impacts of these plantations on water resources and uncertainty about productivity under different environmental conditions. We present a study where the effects of climate change on the spatial variation in climate are taken into account when predicting potential productivity in terms of dry mass (DM) of wood or mean annual volume of wood increment and water use efficiency (WUE) of hybrids of Eucalyptus grandis and Eucalyptus urophylla plantations across more than 32 million ha located near the Atlantic coast of Brazil. Our main objective was to estimate the effects of future climate and increasing CO2 concentration on planted forests in this region. Predictions of mean annual increment in wood production and of water use efficiency were generated with an updated spatial version of the process-based growth model 3-PG (Landsberg and Waring, 1997). The model has been modified to include the direct effects of increasing levels of atmospheric CO2 on the vegetation. We assume that light saturated assimilation rate and light use efficiency increase as atmospheric CO2 concentration increases, while maximum stomatal conductance declines. The study considered three climatic periods with different CO2 concentrations: the historical scenario has 350 ppm, while the 2030 and 2050 scenarios correspond to 450 ppm and 520 ppm, respectively. Sensitivity analyses quantified the effects of the parameters used in the model to account for the effects of atmospheric CO2 on the predicted forest productivity. Stem mass is strongly sensitive to changes in canopy quantum efficiency, and hence to the effect of CO2 on light use efficiency, but less so to changes in stomatal conductance, and hence to the effect of CO2 on conductance. WUE, defined here as DM of wood per mass of water evapotranspired, is also sensitive to these parameters. Analysis of the climatic data for the 2030 and 2050 scenarios in the study area suggests a reduction of 2% and 3% in annual precipitation and an increase of 8% and 15% in vapour pressure deficit in 2030 and 2050, respectively, compared with the period 1971 to 2000. Application of 3-PG with the 2030 and 2050 climates suggests that, averaged over the study area, forest productivity may increase by of the order of 6 m 3 ha -1 year - 1 by 2030, and 10 m 3 ha -1 year -1 by 2050, corresponding to 17% and 26% increments compared with the historical period. WUE increases by an average of 1.0g DM kg -1 H2O in 2030 and 1.7 g DM kg -1 H2O in 2050 compared with the historical scenario, which is equivalent to increases of 29% and 51% in WUE, respectively. This shows that with increasing CO2 the trees are more efficient in using water. If these changes do occur it will increase the amount of land with higher potential productivity for Eucalyptus plantations in the study area. From the total area of 32 million hectares (Mha), 8.5 Mha currently have potential productivity above 40 m 2 ha -1 year -1 . With increasing CO2 this area increases to 20.5 Mha in 2030 and 26.0 Mha in 2050.