Using heterogeneity and representation of ecosite criteria to select forest reserves in an intensively managed industrial forest.

摘要

The 189 000 ha privately owned, industrial Black Brook District in New Brunswick, Canada, contains 27 forest reserve areas, totaling 9500 ha (5%), with a primary management strategy of conservation protection. We conducted a gap analysis of the current reserve network, to evaluate representation by ecosite, and tested two approaches to identify potential additional reserves to increase to target areas of 6%, 7%, 10% and 15%: (1) selecting areas with maximum diversity of forest species composition, soil types, and elevation classes ('heterogeneity method', to incorporate the greatest amount of forest features in the smallest area); and (2) selecting areas to cover a total of 10% of the landbase, with 10% coverage of each ecosite area, if possible ('representation method'). With a 10% target, the heterogeneity and representation methods resulted in 29 and 42 reserves, ranging from 490-2000 to 50-1500 ha, respectively. The heterogeneity method resulted in 13 fewer reserves, 10% more core area, 30% larger mean patch size, and 30% less edge. Representation of ecosites ranged from 0.9% to 23.3% in the current reserve network, 6.7-26.5% for the heterogeneity method, and 8.0-26.4% for the representation method. In simulations of management scenarios with 39-64% plantations and 10% unharvested reserves, projected softwood harvest was similar with both methods, at 2.3-2.5 million m3 per 5-year period, while hardwood harvest was 0.02-0.09 million m3 higher, at 0.79-0.82 million m3 per 5-year period, with the representation than the heterogeneity method. Projected areas of old hardwood, softwood, and mixedwood habitats, defined based upon stand structure characteristics, were similar between the two methods, with up to 1000 ha more with the representation method in some projection periods. We conclude that the heterogeneity method was more effective in that it selected fewer reserves, with larger areas, more core area, and less edge..