Perspectives: The wicked problem of defining and inventorying mature and old-growth forests

Mature and old-growth forests are valued for biodiversity, carbon sequestration, habitat, hydrologic function, aesthetics, and spirituality, as well as Tribal and Indigenous histories, cultures, and practices. Over the last 500 years, land use change and industrialization have resulted in global declines in the area of older forests (however defined). The goal of this study was to identify concepts and indicators to define mature and old-growth forests across the vegetation types of the United States in order to quantify their abundance and distribution. Defining old growth has been described as a "wicked problem" that involves values, science, and management; requires multiple disciplines; and can be expressed from many contradictory approaches.The most common approach to defining mature and old-growth forests is to place them in a successional continuum of increases in tree size, biodiversity, habitat niches, and structural diversity with forest age. Time since severe disturbance, including human impact, is often a consideration, although humans have influenced the development of many forests for millennia. The successional framework is less useful in low-productivity or frequently-disturbed forests, or where current structural diversity under fire suppression may not reflect historic or desired future conditions. In order to classify forests into "old" and "not old", existing structure-based ap-proaches apply minima of one or more structural or compositional criteria. Site productivity and/or plant as-sociation is an element of many definitions.Once defined, estimating the area of mature and old-growth forest presents challenges. The only compre-hensive, consistent field data of US forests is the Forest Inventory and Analysis (FIA) network of >140,000 forested plots. While the 0.067 ha sample area of FIA plots limits the number of structural metrics that might be useful and the plot density cannot capture fine-scale spatial heterogeneity, measurements enable a granular application of multiple structural and compositional criteria by vegetation type at broad spatial extents, and the ability to track change consistently over time. Spatial models integrate field and remotely-sensed data to predict the distribution of structural classes at finer spatial grain, but with substantial error in high-resolution estimates. There does not seem to be a readily-available method to map mature and old-growth stands across a landscape with a high degree of accuracy. Identifying mature and old growth forests in a stand management context will likely require additional measurements, adjustments to criteria at local scales, and incorporation of social and traditional knowledge within a consistent definition framework.