Does acclimation of plant traits improve dynamic vegetation modelling of a tree species?

Dynamic vegetation modelling is intensively used with plant functional types which limits the range of interest of obtained outputs for other fields of knowledge like conservation science. An alternative approach is to simulate plant species. This however requires additional data, i.e. morphological and physiological traits values characterizing the species and determining their functional properties. However, not only many traits vary among the species belonging to the same plant functional type but also the traits vary broadly according to climate factors.

Since most of the traits are functional, their values may be critical for dynamic vegetation model outputs. We measured several traits (specific leaf area, leaf and sapwood C:N) of Cedrus atlantica in its native range, the Rif and Middle Atlas Mountains of Morocco, as well as in some plantations in western Europe. Trait values exhibit significant variations between the sampled sites. It is possible to predict these trait values using multiple regression with climate factors as explanatory variables. Using regression equations, we produced spatial- and time-varying traits over the study area. We implemented these equations in the CARAIB dynamic vegetation model and tested whether they improve the simulation of C. atlantica in the Rif and Middle Atlas Mountains, by comparing the net primary productivities and biomasses computed with and without trait variation, with those retrieved from measurements on the sampled sites. We then performed simulations of the future using climate projections of the regional climate model RCA4 nested in HadGEM2 general circulation model under the RCP8.5 scenario, in order to test the influence of trait acclimation on the predicted future changes in the range and productivity of the species.