Predictive models of forest dynamics

Computational Ecology and Environmental Science Group, Microsoft Research, Cambridge, UK.
Science (Impact Factor: 31.48). 07/2008; 320(5882):1452-3. DOI: 10.1126/science.1155359
Source: PubMed

ABSTRACT Dynamic global vegetation models (DGVMs) have shown that forest dynamics could dramatically alter the response of the global climate system to increased atmospheric carbon dioxide over the next century. But there is little agreement between different DGVMs, making forest dynamics one of the greatest sources of uncertainty in predicting future climate. DGVM predictions could be strengthened by integrating the ecological realities of biodiversity and height-structured competition for light, facilitated by recent advances in the mathematics of forest modeling, ecological understanding of diverse forest communities, and the availability of forest inventory data.

  • [Show abstract] [Hide abstract]
    ABSTRACT: The Northern Hemisphere's boreal forests and, in particular, the Siberian boreal forest zone, may have a particularly strong effect on the Earth's climate through mechanisms involving changes in the regional surface albedo. Warmer climate has been implicated in the conversion of Russia's Siberian larch forests to evergreen conifer forests implying a potential positive feedback cycle: a warmer climate can accelerate the natural succession from larch to evergreen conifer forest; the resultant albedo change then can promote additional climate warming. Land cover changes in this region can lead to alterations in regional climate through modifications in surface albedo and land/atmosphere energy fluxes, as well as in global climate through changes in C sequestration and release patterns. Utilization of the individual based forest gap model, FAREAST, with historical climate data allowed us to generate baseline biomass and species dynamics across this diverse region. FAREAST has been updated to accommodate daily climate as generated by global climate models to explore detailed impacts of future climate on composition. Biomass and species dynamics resulting from IPCC climate output data for multiple climate change scenarios in comparison to baseline forest structure are used to evaluate detailed changes in forest composition and biomass across Russia for stands of various ages. These results are used to identify the location, age and species composition of forests which are vulnerable to climate change. Assessing the forest vulnerability in congruence with the age and species distribution is a powerful tool in understanding forest response to climate change in addition to the forests role in climate/cover feedback associated with land/atmosphere energy fluxes.
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Many morphological, physiological, and ecological traits of trees scale with diameter, shaping the structure and function of forest ecosystems. Understanding the mechanistic basis for such scaling relationships is key to understanding forests globally and their role in Earth's changing climate system.Here, we evaluate theoretical predictions for the scaling of nine variables in a mixed-age temperate deciduous forest (CTFS-ForestGEO forest dynamics plot at the Smithsonian Conservation Biology Institute, Virginia, USA) and compare observed scaling parameters to those from other forests worldwide. We examine fifteen species and various environmental conditions.Structural, physiological, and ecological traits of trees scaled with stem diameter in a manner that was sometimes consistent with existing theoretical predictions—more commonly with those predicting a range of scaling values than a single universal scaling value.Scaling relationships were variable among species, reflecting substantive ecological differences.Scaling relationships varied considerably with environmental conditions. For instance, the scaling of sap flux density varied with atmospheric moisture demand, and herbivore browsing dramatically influenced stem abundance scaling.Thus, stand-level, time-averaged scaling relationships (e.g., the scaling of diameter growth) are underlain by a diversity of species-level scaling relationships that can vary substantially with fluctuating environmental conditions. In order to use scaling theory to accurately characterize forest ecosystems and predict their responses to global change, it will be critical to develop a more nuanced understanding of both the forces that constrain stand-level scaling and the complexity of scaling variation across species and environmental conditions.This article is protected by copyright. All rights reserved.
    Functional Ecology 05/2015; DOI:10.1111/1365-2435.12470 · 4.86 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: 1. Plant functional traits, in particular specific leaf area (SLA), wood density and seed mass, are often good predictors of individual tree growth rates within communities. Individuals and species with high SLA, low wood density and small seeds tend to have faster growth rates. 2. If community-level relationships between traits and growth have general predictive value, then similar relationships should also be observed in analyses that integrate across taxa, biogeographic regions and environments. Such global consistency would imply that traits could serve as valuable proxies for the complex suite of factors that determine growth rate, and, therefore, could underpin a new generation of robust dynamic vegetation models. Alternatively, growth rates may depend more strongly on the local environment or growth–trait relationships may vary along environmental gradients. 3. We tested these alternative hypotheses using data on 27 352 juvenile trees, representing 278 species from 27 sites on all forested continents, and extensive functional trait data, 38% of which were obtained at the same sites at which growth was assessed. Data on potential evapotranspiration (PET), which summarizes the joint ecological effects of temperature and precipitation, were obtained from a global data base. 4. We estimated size-standardized relative height growth rates (SGR) for all species, then related them to functional traits and PET using mixed-effect models for the fastest growing species and for all species together. 5. Both the mean and 95th percentile SGR were more strongly associated with functional traits than with PET. PET was unrelated to SGR at the global scale. SGR increased with increasing SLA and decreased with increasing wood density and seed mass, but these traits explained only 3.1% of the variation in SGR. SGR–trait relationships were consistently weak across families and biogeographic zones, and over a range of tree statures. Thus, the most widely studied functional traits in plant ecology were poor predictors of tree growth over large scales. 6. Synthesis. We conclude that these functional traits alone may be unsuitable for predicting growth of trees over broad scales. Determining the functional traits that predict vital rates under specific environmental conditions may generate more insight than a monolithic global relationship can offer.
    Journal of Ecology 04/2015; in press. DOI:10.1111/1365-2745.12401 · 5.69 Impact Factor


Available from