An index to identify late-successional forest in temperate and boreal zones

Old forests of the temperate and boreal zones are globally critical for forest biodiversity conservation yet are rapidly declining where commercial forestry dominates. Forest certification programs are the only global policy instruments that require private landowners to conserve old forest, but certified landowners lack the necessary practical tools to do so. We developed statistically based, practical late-successional indices that can be used in the field to quickly identify high-conservation value old forest. We measured late-successional forest attributes in economically mature and older stands to be able to distinguish economically mature stands from late-successional stands for two forest types: northern hardwood forest and upland spruce-fir forest. Based on step-wise discriminant function analysis, the best single variable for both forest types was large (≥40 cm dbh) alive and dead tree density. For upland spruce-fir forest, the best linear discriminant function was created by adding one variable: large (≥35 cm dbh) log count. The thresholds levels between economically mature and late-successional stands were: 9 large live and dead trees/0.02 ha plot for northern hardwoods, 8 large live and dead trees/0.02 ha plot for upland spruce-fir forest, and a linear discriminant function score of 2.147 for upland spruce-fir forest. Indices were developed using old-growth data to provide reference levels with scores ranging from 0 to 10. The indices misclassified approximately 15% of economically mature and late-successional stands from an independent dataset. The indices increased across a range of seral classes with the greatest increases occurring between late seral stages. The indices were correlated with other late-successional attributes including: large snags, large logs, abundance of seven LS epiphyte taxa, tree size distribution, and dead wood variables. Our indices have been used to: conserve old forest by certified landowners; train foresters about ecological forestry; verify the implementation of late-successional management goals; and assess trends of late-successional forest in the region. Our approach could be used to develop LS indices for other regions to help conserve old forest.