A CEES project

Part of our Forest Dynamics research, this project aims to understand how to accurately and usefully scale from the short-term biology of individual trees, to the longer-term dynamics of forest stands, and forested landscapes. Why do this? First, there are good reasons to believe that, in order to make useful predictions about several apects of forest dynamics, it is necessary to represent the fact that forests are built of trees. For example, increased growth causes increased competition for light, which in turn increases mortality rates, which in turn decreases carbon storage. Second, we have lots of short-term measurements of individual trees which, if we only knew how to scale them up, might let us understand and predict the dynamics of forests over much of the globe.

The key to this scaling is the fact that trees interact primarily by competing for shared resources such as light and water. Once the process of competition is understood sufficiently, it can be encoded into a model that translates the biology of individual trees (e.g mortality rates, growth rates, allometry) into the longer-term dynamics of the forest (e.g. carbon storage, canopy height, species composition) at the scale of an individual stand (think 1 acre). From a technical standpoint, the challenge is to find a model that does this properly – i.e. makes reliable predictions that we can trust – whilst being rapid to simulate, easy to parameterize, and easy to understand. Working under Prof Steve Pacala at Princeton University, Drew Purves was fortunate to be involved in developing, parameterizing and validating what appears to be just such a model: the ‘PPA’ or ‘Perfect Plasticity Approximation’ model (e.g. see this publication in PNAS). The key to to the PPA is the assumption that large trees are infinitely flexible in the way they compete for canopy space. This simple assumption makes many other complexities fall out, leaving a minimal and – based on our experience so far – really useful model for addressing dynamics at the scale of individual stands. For this reason, the PPA is likely to form the kernel of most of our forest ecology research for some time. Of course, the PPA model becomes more useful the more realistic the functional forms at its core (determining growth, mortality, reproduction). Our collaborator John Caspersen is currently working on these improved functional forms as part of our work in optimal forest management.

But, there is another half of the scaling: from stands, to forest landscapes. How do the dynamics of a heterogeneous collection of stands, with different combinations of species, subject to different soils and micro-climates, and at different stages of disturbance, add up to the longer-term dynamics of whole forested landscapes? If we could solve this problem, we could get a new level of understanding of (for example) species distributions, and how they might respond to climate change, changes in fire regimes, the loss of particular species, or the arrival of invasive species. For her PhD Emily Lines will be trying, for the first time, to parameterize climate-dependent functions for growth, mortality and reproduction for all of the dominant species of a whole region (peninsular Spain), then enter these functions into simulations of thousands of stands distributed over a wide area (with the stands simulated using the PPA).

Watch a video of Drew Purves talking about the PPA model

 People  

Dr Drew Purves (Microsoft Research Cambridge)

Dr Miguel Zavala (CIFOR-Madrid)

Dr David Coomes (University of Cambridge)

Prof. John Caspersen (University of Toronto)

Emily Lines (PhD student, University of Cambridge)