Evidence for a recent increase in forest growth

Abstract

Forests and their soils contain the majority of the earth's terrestrial carbon stocks. Changes in patterns of tree growth can have a huge impact on atmospheric cycles, biogeochemical cycles, climate change, and biodiversity. Recent studies have shown increases in biomass across many forest types. This increase has been attributed to climate change. However, without knowing the disturbance history of a forest, growth could also be caused by normal recovery from unknown disturbances. Using a unique dataset of tree biomass collected over the past 22 years from 55 temperate forest plots with known land-use histories and stand ages ranging from 5 to 250 years, we found that recent biomass accumulation greatly exceeded the expected growth caused by natural recovery. We have also collected over 100 years of local weather measurements and 17 years of on-site atmospheric CO(2) measurements that show consistent increases in line with globally observed climate-change patterns. Combined, these observations show that changes in temperature and CO(2) that have been observed worldwide can fundamentally alter the rate of critical natural processes, which is predicted by biogeochemical models. Identifying this rate change is important to research on the current state of carbon stocks and the fluxes that influence how carbon moves between storage and the atmosphere. These results signal a pressing need to better understand the changes in growth rates in forest systems, which influence current and future states of the atmosphere and biosphere.

Fig. 1.

Plot biomass increases with plot age in a predictable pattern. The Monod function (Eq. 1) describes the minimum and asymptotic biomass and the half-saturation point of the system. The red line shows the median curve from parameter estimates, whereas the blue lines indicate curves fit from 50 draws from the posterior distributions of the β and θ parameters. Points indicate site biomass and age at the means of censuses, where multiple measurements were taken (n = 55).

Fig. 2.

Accelerated growth of multiple-censused plots can be seen in A where plot censuses show increased biomass gain. Plot censuses (diamonds) are linked by lines. Plots that had only one census were used to estimate the ensemble curve and are plotted here in gray. The plotted curve is the same as in Fig. 1 but on a log scale with axes adjusted to show site biomass change more clearly. (B) The observed versus expected annual biomass change (Mg ha^-1 yr^-1) with 95% confidence limits (lines) from posterior parameter values of the estimated Monod function. Across all plot ages, consistently higher than expected annual biomass growth can be seen. Even plots that have a below or expected level of biomass increase show other years with higher than average biomass increase. Census intervals with negative growth (14 of 166 intervals) reflect biomass change because of tree death; they were not included in this figure but were included in analyses.

Fig. 3.

Environmental measurements that address hypotheses for accelerated recent growth. (A) Temperature measurements (maximum, mean, and minimum) for Baltimore Washington International Airport are shown. Maximum and mean temperature showed significant increases (P < 0.0001). Minimum temperature showed no trend. (B) Data from Annapolis, MD (15 km north of Edgewater) showed that first frosts arrive later and last frosts arrive sooner, increasing the length of the frost-free growing season (growing season shown as days, first and last frosts as day of the year). (C) Maximum annual CO₂ levels (ppm) at SERC (black dots) and measurements from Mauna Loa (red line; data from ref. 50) are shown.