- Hennigar, Chris
New Brunswick Forest Stewardship and Planning Branch
- Weiskittel, Aaron
University of Maine Center for Research on Sustainable Forests
The Acadian Forest has a long history of stand growth research and modeling; however, these efforts have generally been restricted within state or province, rather than across the extent of this forest type. Between 2010 and 2014, approximately 20,000 repeatedly measured plots from Maine, Nova Scotia (NS, New Brunswick (NB), and PEI were compiled and standardized from a combination of available government, industry, and local research datasets. This concerted effort was aimed toward developing statistical tree form, growth, survival, and regeneration models across the Acadian Forest.
To date, two distinct empirical stand model programs have been developed to combine a number of new regional tree-level models together to predict historical forest development with and without silviculture interventions:
·FVS-ACD (Forest Vegetation Simulator – Acadian Variant) - developed by Weiskittel (UM) and Dr. J. Kershaw (UNB) and mostly used in Maine for private land forest management planning.
·OSM-ACD (Open Stand Model – Acadian Variant) - developed by Hennigar and primarily used in New Brunswick for Crown forest management planning.
Both applications are calibrated for use across the region, are open source, and can do batch run simulations using commands; however, the underlying application platforms, equations, and simulation routines have diverged considerably over the past five years. Here, we discuss these differences and compare stand-level net volume prediction performance (mean bias, root mean squared error) of these two models against observed stand development over a 10-50 year period in repeatedly measured plots in Maine, NB, and NS for a range of species, ecological conditions, stand development stages, and management types, including: unmanaged mature-old stands, partially cut hardwoods, spruce (Picea spp.) and pine (Pinus spp.) plantations, thinned, and naturally regenerating clearcuts. Areas of future refinement for each model application will be highlighted and discussed.