Using a Process-Based Model (3-PG) to Predict and Map Hybrid Poplar Biomass Productivity in Minnesota and Wisconsin, USA

Abstract

Hybrid poplars have demonstrated high biomass productivity in the North Central USA as short rotation woody crops (SRWCs). However, our ability to quantitatively predict productivity for sites that are not currently in SRWCs is limited. As a result, stakeholders are also limited in their ability to evaluate different areas within the region as potential supply sheds for wood-based bioenergy facilities. A reliable method for predicting productivity across the region is needed; preferably, such a method will also lend itself to generating yield maps that stakeholders can use to inform their decision making. In this study, the Physiological Processes Predicting Growth model was (1) assigned parameters for hybrid poplars using species-specific physiological data and allometric relationships from previously-published studies, (2) calibrated for the North Central region using previously-published biomass data from eight plantations along with site-specific climate and soils data, (3) validated against previously published biomass data from four other plantations using linear regression of actual versus predicted total aboveground dry biomass (R2 = 0.89, RMSE = 8.1 Mg ha−1, mean bias = 5.3 Mg ha−1), (4) evaluated for sensitivity of the model to manipulation of the parameter for age at full canopy cover (fullCanAge) and the fertility rating growth modifier, and (5) combined with soil and climate data layers to produce a map of predicted biomass productivity for the states of Minnesota and Wisconsin. Mean annual biomass productivity (total aboveground dry biomass divided by age) ranged from 4.4 to 13.0 Mg ha−1 year−1 across the states, with the highest productivity mainly concentrated in the area stretching from south-central Minnesota across southern Wisconsin.