Modeling the productivity of energy crops in different agro-ecological environments

Abstract

A relatively stable biomass productivity of perennial crop after plantation establishment makes it possible to calculate their total biomass yield through predicting the annual biomass yield. The generic model LINPAC (LINTUL model for Perennial and Annual Crops) is presented to predict annual biomass yield of energy crops on large spatial scales by adding new modules to LINTUL: (1) Leaf Area Index (LAI) is simulated independent of specific leaf area; (2) a species specific daily Light Use Efficiency (LUE, gMJ−1) is modified by temperature and light intensity; (3) crop base temperature is generated by local weather conditions within crop physiological ranges. LINPAC is driven either by site-specific input data or by globally gridded weather and soil data. LINPAC was calibrated on the basis of a model sensitivity analysis of the input parameters and validated against different agro-ecological experimental data sets for two grass species Miscanthus (Miscanthus spp.) and Reed canary grass (Phalaris arundinacea L.), and for two woody species Willow (Salix spp.) and Eucalyptus (Eucalyptus spp.). LINPAC reproduced the biomass yields with a normalized root mean square error (RMSE) of 17%, comparable to the coefficient of variation (CV=12%) of the experimental data. In the model photosynthetic pathways were differentiated by assigning higher LUE values for the C4 crop (Miscanthus) compared with the C3 crops (others), leading to higher simulated biomass yield of Miscanthus (18.8±1.5tha−1) over Reed canary grass (10.5±1.6tha−1) in comparable environments. LINPAC is applicable for local, regional and global estimations of biomass yield of energy crops.