Do trees make slopes more stable?

31 January 2020

Do trees make slopes more stable?

A common comment on this blog, or on Twitter, in the aftermath of a landslide is that it could have been prevented by the planting of trees. It is widely believed that trees increase slope stability, and thus reduce the likelihood of landslides. This feels logical, and of course there is a long history of landslides in the aftermath of wildfires or in the period after deforestation. However, the quantitative evidence that planting trees decreases landslide susceptibility is surprisingly poor, but some studies have suggested that tree planting can sometimes be detrimental to overall stability.

In that context, a new paper published in the journal Landslides (Lan et al. 2020), which reports a study that seeks to explore this topic, is particularly welcome. To do so, the authors built a model slope, 80 cm long and 20 cm wide, onto which they “planted” artificial trees, with a simulated root network. The slope was then sprayed with simulated rainfall via a nozzle system, and the likelihood of failure was determined. A range of conditions were investigated, with a variety of slope angles and with different patterns of tree planting. In total 12 experiments were conducted, and the behaviour of the slope was both observed and modeled. In each case an experiment was conducted with no trees present as a control, as shown in the figure below.

The development of slope failure in a simulated slope – in this case on a 60 degree slope. From Lan et al. (2020).

The results are interesting and perhaps surprising. For lower gradient slopes (20º and 35º in the model), tree planting increased slope stability. It should be noted that in the real world a 35º slope would be considered to be very steep. However, on steeper slopes (50º and 60º), tree planting reduced stability. The pattern of planting of the trees changed their effectiveness in terms of reducing landslides. The most effective pattern was found to be dense, wide bands of trees with spaces between the bands. This was found to be more effective than an even distribution of trees across the slope.

There are caveats to a study like this, of course. The first is that scaling between model slopes and the real world is notoriously difficult. This is the reason that centrifuges are used to simulate slope behaviour – this study really needs such equipment to be reliable. Second, there is a problem with the grain size of the simulated soil, which does not scale properly. And finally, the model does not really simulate the complex relationships between trees and the soil – for example, to what degree does the presence of trees change the physical characteristics of the soil through time?

But nonetheless, this is a really useful study in that it highlights that tree planting on slopes is a complex issue. It is clear that it can be beneficial in lower gradient slopes, assuming that the slope is sufficiently stable that the trees can become established (this is a big problem in active slopes, like those undergoing coastal erosion). But, effective planting is an expert task, and needs careful design.