Structure area curves in Eastern Hardwoods: implications for minimum plot sizes to capture spatially explicit structure indices

• Key message: Sampling needs differ by forest type for timber inventory and structural complexity metrics. We demonstrate in a typical mixed Eastern Hardwoods forest that optimal sampling of timber inventory metrics and spatially explicit structure indices may be achieved in one large plot plus a cruise for large diameter trees, but accurately capturing inventory metrics may not be possible with sparse large-scale sampling. • Context: Managing forest stand structures for multiple objectives require accurate and precise estimates of structural features that may be best estimated at different scales. • Aims: We document minimum necessary plot sizes for structural metrics and spatially explicit indices to characterize structure in a mature North American Eastern hardwoods forest. • Methods: Metrics and indices (Index of Aggregation, Diameter Differentiation Index, Dissimilarity Coefficient, Structural Complexity Index) were calculated within 0.05–1.75-ha plots for 1000 iterations of random placement in two 2.0-ha macroplots. Estimation adequacy required (1) precision (varied < 10% among plots) and (2) accuracy (within 10% of the 2.0-ha value at 5th and 95th percentiles). • Results: Minimum single plot sizes to achieve estimation adequacy were 0.25–0.75 ha for spatially explicit indices and 0.5–2 ha for stand metrics. A minimum of five 0.10-ha subplots would be needed for most indices and 6–25 for most metrics, but an untenable 375+ for the density of large diameter trees. • Conclusion: Estimation adequacy for structural complexity requires no greater sampling intensity than for timber metrics, except for density of large trees. A single large plot may be most cost-effective. National inventories in Eastern hardwoods may not estimate structural complexity well due to inadequate sampling intensity.