How fast do ecosystems recover from disturbance? It’s complicated.

ResearchBlogging.orgIn the 21st century, human activity promises to impact the natural world on an unprecedented scale. In order to decide where to focus conservation effort, one thing we need to know is how permanent the damage from a forest clear-cut or a collapsed fishery actually is. A paper in this week’s PLoS ONE looks at natural systems’ ability to recover after human and natural disturbances, and the authors say the results are hopeful. I’m not so sure.

A clearcut.
Photo by : Damien.

The authors, Jones and Schmitz, assemble a meta-dataset of ecological studies published from 1910 to 2008, all examining the recovery of either ecosytem functions (like total nutrient cycling rates) or plant or animal diversity following disturbances as diverse as hurricanes and oil spills. They then calculated the proportion of measured variables that recovered, or failed to, within the period studied by each paper in the dataset, how much the measured variables had been altered by the disturbance, and how long it took before they returned to their pre-disturbance state.

The results are complicated, to say the least. For example, here’s Figure 2, which charts the times to recovery for variables measuring animal community recovery (black bars), ecosystem function (white bars) or plant community (gray bars), broken down by ecosystem type in the top panel, and by disturbance type in the bottom panel:

The authors’ conclusion? There is “no discernable pattern.” Which I can’t really dispute — recovery times look highly idiosyncratic. An ANOVA performed on the data finds significant effects of ecosystem type and disturbance type, but what does that tell us? Different ecosystems recover differently. Forests take the longest to recover, which makes sense given that trees grow slowly, and succession from clearcut to mature forest can take centuries. Similarly, ecosystems experiencing multiple types of disturbance took the longest to recover.

Of the ecosystems that do recover, the authors point out, recovery occurs comparatively rapidly:

Among studies reporting recovery for any variable, the average recovery time was at most 42 years (for forest ecosystems) and typically much less (on the order of 10 years) when recovery was examined by ecosystem. When examined by perturbation type, the average recovery time was no more than 56 years (for systems undergoing multiple interacting perturbations) and typically was 20 years or less …

The authors then perform a regression of the strength of disturbance (i.e., how much the measured variables changed due to disturbance) against the time needed for recovery. The data set is necessarily small, because not many studies follow an ecosystem all the way from disturbance to complete recovery, and they find a significant effect of disturbance strength on recovery time mostly because of a single data point.

Jones and Schmitz conclude from this dataset that ecosystem recovery from human disturbance is frequently possible within human lifetimes, especially if we put in the effort for restoration. I’ll buy that; but I think the more important lesson to draw from this paper is that, after a century of watching the natural world respond to human activity, we still can’t predict what the results of our actions will be. It shouldn’t need saying, but when we fiddle with our life-support systems, we must proceed cautiously.


Jones, H., & Schmitz, O. (2009). Rapid recovery of damaged ecosystems PLoS ONE, 4 (5) DOI: 10.1371/journal.pone.0005653