Long-term tree and stand biomass increment patterns derived from tree rings in multiple temperate and sub-boreal forest systems in northeastern Minnesota, USA

Background/Question/Methods

Understanding how woody biomass production changes as trees and forests age is critical for models of forest dynamics and carbon sequestration. The most frequently cited stand-level models of woody biomass change report a rapid increase in biomass accumulation that peaks around canopy closure, and declines to a constant or decreasing rate. Different hypotheses have been tested to explain this pattern, often using chronosequences, fast-growing plantations or even-aged or monospecific forest types as experimental systems. While the hypotheses are widely debated, recent studies of potential production in old-growth forests may call the universality of the underlying pattern into question. The goal of this study was to compare the accepted biomass production model to a wider range of natural forests with varying stand origins and species assemblages. We used increment cores extracted from a census of live trees (DBH > 10cm) in 108 plots (400m²) across 8 forest types in northern Minnesota. Over 3200 tree cores from 20 species were processed, measured and crossdated using standard dendrochronological techniques. Tree diameters were reconstructed from ring-widths and converted to annual biomass increments at tree and stand scales using species-specific allometric equations.

Though temporally rich, tree-ring records from live trees underestimate past biomass increments because growth from trees that died before measurement is unobserved. We simulated missing biomass increments by assuming annual mortality of  0-2% following one of three mortality models. We estimated mean patterns of stand growth by fitting smoothing splines to aggregate biomass increments of the surviving (sampled), and the combined (sampled and simulated) populations, and compared splines and their trends to the general model of woody biomass change.

Results/Conclusions

Most even-aged stands followed the expected pattern of biomass increment when simulated increments were included with measurements from live trees. In younger even-aged stands, biomass increment peaked 20-40 years following stand establishment, with maxima ranging from 5-6 Mg/ha for forests of Pinus bansksiana, to 10-12 Mg/ha for Acer saccarum stands. In contrast, multi-aged Quercus rubra forests on less productive sites showed a pattern of steadily increasing biomass increment over 90 years of stand development. Other multi-cohort forest types showed evidence of multiple peaks in biomass increment; some corresponding to ingrowth of shade tolerant conifer species. These findings of gradual or shifting biomass accumulation among species in multi-aged forests suggest that basic models of forest growth may be inadequate to accurately reflect natural forests with more complex structure.