The formulations of site-scale processes affect landscape-scale forest change predictions: a comparison between LANDIS PRO and LANDIS-II forest landscape models

Forest landscape models (FLMs) are important tools for simulating forest changes over broad spatial and temporal scales. The ability of FLMs to accurately predict forest changes may be significantly influenced by the formulations of site-scale processes including seedling establishment, tree growth, competition, and mortality. The objectives of this study were to investigate the effects of site-scale processes and interaction effects of site-scale processes and harvest on landscape-scale forest change predictions. We compared the differences in species' distribution (quantified by species' percent area), total aboveground biomass, and species' biomass derived from two FLMs: (1) a model that explicitly incorporates stand density and size for each species age cohort (LANDIS PRO), and (2) a model that explicitly tracks biomass for each species age cohort (LANDIS-II with biomass succession extension), which are variants from the LANDIS FLM family with different formulations of site-scale processes. For early successional species, the differences in simulated distribution and biomass were small (mostly less than 5 %). For mid- to late-successional species, the differences in simulated distribution and biomass were relatively large (10-30 %). The differences in species' biomass predictions were generally larger than those for species' distribution predictions. Harvest mediated the differences on landscape-scale predictions. The effects of site-scale processes on landscape-scale forest change predictions are dependent on species' ecological traits such as shade tolerance, seed dispersal, and growth rates.