Forests in flux as climate varies

How do changes in climate affect forest ecosystems? A study of temperate forests in the United States has assessed alterations in biomass and tree-species composition across a 20-year period of climate variability.
Sebastiaan Luyssaert is in the Department of Ecological Science, Faculty of Science, VU Amsterdam, 1081 HV Amsterdam, the Netherlands.

Search for this author in:

J. Hans C. Cornelissen is in the Department of Ecological Science, Faculty of Science, VU Amsterdam, 1081 HV Amsterdam, the Netherlands.

Search for this author in:

Documenting and understanding changes induced by climate in the composition and function of vegetation is essential for planning adaptation strategies, because chances to intervene do not arise often for forests that contain trees with long lifespans. Moreover, most of the effects exerted by climate change on the composition and biomass of vegetation probably occur incrementally rather than abruptly1, which makes their detection a challenge. In a paper in Nature, Zhang et al.2 report an analysis of changes in forest biomass in the eastern United States over two decades, during a time when some regions became drier and others became wetter.

Ecological studies have long focused on analyses in which the main groupings for organisms being studied are determined by evolutionary relationships such as belonging to the same genus. However, such kinship can hide functional differences. For example, even among closely related species of oak tree (Quercus spp.), some will thrive under moist conditions, whereas others are more suited to dry climates. In the past decade, the use of functional rather than kinship-driven approaches to grouping has provided many important insights3, and Zhang and colleagues’ work can be added to the list of studies that have successfully done this.

Zhang et al. sought to investigate whether shifts in climate affect forest characteristics such as the prevalence of drought-tolerant species and the total biomass of the tree population. The authors compared temperate-forest inventory data4 gathered during the 1980s and the 2000s. This inventory includes surveys of around 100,000 plots, in which data such as species name and a standardized measurement of tree diameter were recorded for roughly 3 million trees. Diameter measurements allow the amount of biomass that is present above the surface of the ground to be estimated for a particular tree on the basis of previous studies of growth patterns for a given species. Such analyses enabled the authors to assess the aboveground biomass per hectare for tree populations, and also to investigate whether changes occurred in the relative contribution of particular types of species, such as drought-tolerant trees, to the total aboveground biomass.

The authors analysed the data grouped into grids of cells that each covered an area of one degree of latitude by one degree of longitude (about 110 kilometres by 85 kilometres). The plots sampled in the 1980s and the 2000s were not identical; however, the authors were able to check their findings using a subset of plots that had been resampled and found that their conclusions remained the same. The authors determined the tree-age composition of the forests and assigned them into 20-year-interval age brackets. This enabled comparisons to be made between similar types of forest, for example, comparing 20–40-year-old forests in the 1980s and in the 2000s, therefore avoiding possible confounding factors such as changes in drought sensitivity that are linked to the increase in tree height as trees age5.

The authors assigned a numerical score to each recorded species of tree that represented its drought-tolerant characteristics. This score was generated on the basis of information about water availability, derived from measurements of the annual precipitation at grid cells where the species was found, and the minimum water requirements of that species. For each plot, the authors calculated the average drought tolerance of its tree population by using the species’ drought scores, also taking into account the relative abundance of each species. The authors then compared how drought tolerance at the tree-population level for a particular age class in a given grid cell changed in the roughly 20 years between inventories, and assessed whether these changes mirrored any changes in soil moisture (estimated by the Palmer drought severity index) at each location.

Despite the occurrence of unavoidable problems such as logging during the data-collection period, Zhang and colleagues build a strong case that, between the 1980s and the 2000s, the effects of ongoing climate variation in this zone of modest regional climate change have resulted in detectable changes in forest composition and biomass. In areas in which soil moisture increased over time, the authors observed a population-level decrease in drought tolerance and a population-level increase in aboveground biomass per hectare.

The most striking finding made by Zhang and colleagues was that, in areas in which soil moisture had decreased, a decrease in the average aboveground biomass gained per tree was accompanied by an increase in population-level drought tolerance. This increased share of drought-tolerant species occurred because the decrease in the growth rate of moisture-needing (mesic) species during a drought is greater than the decrease in growth rate of drought-tolerant species. Consequently, the forests that reached the particular age range in the 2000s have a lower biomass and a larger proportion of drought-tolerant species than the forests that reached this age range in the 1980s.

In forests containing trees more than 80 years old that experienced drought, the population-level changes in drought tolerance observed by the authors were often driven by an increase in the mortality of mesic species (model a in Fig. 1). The role of preferential tree mortality in driving forest changes that are linked to climate variability has already been reported in a study in Europe6. Yet, for most age ranges, the authors found that the tree population became more tolerant to drought because the growth of the drought-tolerant species was less affected than was that of the mesic species. This means that the drought-tolerant trees increased their proportional contribution to the aboveground biomass per hectare in the study’s two-decade period (model b in Fig. 1).

Figure 1 | Changes in the drought tolerance of tree populations. Zhang et al.2 used an inventory4 of trees growing in the eastern United States in the 1980s and the 2000s to estimate how an intervening period of drought had affected the contribution of drought-tolerant species to total biomass, in forests that had reached a given age. This analysis revealed an increase in the contribution of drought-tolerant trees compared with that of moisture-needing (mesic) trees to the population-level biomass. Three models could account for such an increase. Model a is consistent with the results for forests composed of trees more than 80 years old. Model b can account for the patterns observed by the authors in most other forests. The study’s time span is probably too short to fully capture the sapling emergence of model c. The drought tolerance of trees might partly depend on their entering into symbiotic interactions with root-colonizing fungi that extend the effective root length9. Such interactions could reduce tree biomass. If so, this might need to be considered in studies of this type.

Although the establishment of saplings probably contributes to the response of forests to climate variability (model c in Fig. 1), the short time span of the study, the fact that forest inventories commonly measure only individual trees above a threshold size that it can take a decade or more for a sapling to reach, and the relatively small contribution of saplings to the total aboveground biomass for a given hectare of established forest, make it unlikely that Zhang and colleagues’ approach would capture fully the changes in species composition that are due to sapling emergence.

The authors suggest that their results and observations could have relevance for how climate is affecting other temperate forests or forests in other climate zones. Yet perhaps the priority should be to unravel the mechanisms that underlie the connection between drought tolerance and biomass production at the species level. Zhang and colleagues suggest that the low availability of water should favour tree species that allocate a greater proportion of their biomass to fine roots, thereby promoting drought tolerance at the expense of aboveground biomass production. However, this might be only part of the climate-response phenomenon.

We speculate that, compared with mesic species, drought-tolerant species might have greater investments in symbiotic relationships with soil-dwelling mycorrhizal fungi that can colonize tree roots. Such interactions can extend the effective total root length, thereby extending access to water and nutrients in the soil7. However, such connections would probably come at a substantial cost in terms of tree-biomass reduction because of the need to divert sugars to fungal partners. The type and abundance of mycorrhizal symbioses vary with soil type and climate8,9, so if fungal symbiosis is a major consideration in these scenarios, such factors would need to be considered in future implementations of the approach used by Zhang and colleagues.

There is a pressing need to understand the relationship between water availability and the drought tolerance and biomass of forests. It is necessary, therefore, to ask whether the types of change that the authors observed would be able to keep pace with climate changes that occur on longer timescales. For example, will the drought-tolerance capacity of today’s saplings suffice for the conditions that these trees might encounter when they reach maturity? It’s high time to knock on wood that it will, as well as to continue to investigate the mechanisms that affect forest ecosystems in a changing climate.

Nature 556, 35-37 (2018)

doi: 10.1038/d41586-018-02858-6
Nature Briefing

Sign up for the daily Nature Briefing email newsletter

Stay up to date with what matters in science and why, handpicked from Nature and other publications worldwide.

Sign Up


  1. 1.

    Parmesan, C. & Yohe, G. Nature 421, 37–42 (2003).

  2. 2.

    Zhang, T., Niinemets, Ü., Sheffield, J. & Lichstein, J. W. Nature 556, 99–102 (2018).

  3. 3.

    Kunstler, G. et al. Nature 529, 204–207 (2016).

  4. 4.

    Bechtold, W. A. & Patterson, P. L. The Enhanced Forest Inventory and Analysis Program — National Sampling Design and Estimation Procedures. Gen. Tech. Rep. SRS–80 (US Department of Agriculture Forest Service, 2005).

  5. 5.

    McDowell, N. G. & Allen, C. D. Nature Clim. Change 5, 669–672 (2015)

  6. 6.

    Ruiz-Benito, P. et al. Glob. Change Biol. 23, 4162–4176 (2017).

  7. 7.

    Smith, S. E. & Read, D. J. Mycorrhizal Symbiosis (Elsevier, 1997).

  8. 8.

    Phillips, R. P., Brzostek, E. & Midgley, M. G. New Phytol. 199, 41–51 (2013).

  9. 9.

    Soudzilovskaia, N. A. et al. Glob. Ecol. Biogeogr. 24, 371–382 (2015).

Download references