Recently there has been much discussion in the media, parliament and wider society of ways in which we can reduce the impacts of flooding. One method that has been suggested and is currently being seriously considered by government is “to plant trees in the uplands to help slow the flow of rain water” The Times 25/01/16.
Many commentators have cited as evidence for tree planting the results of a study we carried out at Pontbren in mid-Wales between 2005 and 2012.
In the study we planted trees on previously grazed pasture and measured the subsequent effects on soil hydraulic properties and runoff processes. We found that soil infiltration rates were 67x times faster and surface runoff volumes were reduced by 78% under trees compared with grassland.
Given the high level of attention given to our results and the significance to them attached by many I thought I should draw attention to some of the caveats presented in the full research paper. These suggest that a degree of caution is required when extrapolating the significance of our findings, particularly if one is planning major changes to existing countryside management policies.
What we found
Prior to our work several studies had shown that high stocking densities can cause soil surface compaction leading to increased surface runoff. Our study at Pontbren clearly indicated that excluding sheep and planting trees resulted in reduced near surface soil compaction, increased infiltration rates and reduced surface runoff volumes. However, what we didn’t measure was whether there was any change in the ability of the soil to store water and whether the water was able to penetrate deeper into the soil profile as a result of the presence of tree roots. These are important parameters that we need to know if want to try and estimate what size of storm this type of land use intervention may be effective against.
Tree age (may) matter
Our study and other work undertaken at Pontbren shows that the age of the trees is important and it has been predicted that further improvement in soil hydraulic properties could be achieved under mature trees. We also know that different tree species have different root architecture which will have an impact on the way that water is able to penetrate into the ground. Further work is needed to understand the full impact of trees as they reach maturity and whether the ability of soil below trees to store water could be further improved through tree species selection.
Soil types are variable
Across the UK, the landscape is highly variable with many soils types, often with very different characteristics. Many studies have shown a high degree of variability in hydrological function even for a given soil type depending on how intensely that soil is managed. What we found at Pontbren was that despite choosing four replicate sites with broadly similar soil characteristics, measured infiltration rates and runoff volumes were highly variable. Therefore we cannot say with any certainty what impacts planting trees would have on different types of soil. This indicates a strong need to measure these soil physical properties across a range of soil types to quantify the relative impact.
The results reported in our paper were from a field study undertaken at relatively small scale (each of the plot being 12m x 12m). The UK’s landscape has a very complex structure and one cannot simply upscale the results measured in plot-scale studies like ours to the catchment scale in order to predict the impacts that planting trees might have on flooding. In order to do this we need to employ hydrological models which take account of land use and land use change in their predictions.
Developing the evidence base
To gain a better understanding of whether tree planting would have a positive impact in reducing flooding we need to develop the evidence base. Our work is one of very few studies which provide any empirical information. In future we need to measure what effects planting trees has on soil hydraulic properties, such as the water storage capacity in the soil, under a range of soil types and conditions, as well as looking at different tree species and ages. An example of work towards this end is the Multi-land project, currently under way and funded by the National Research Network for Low Carbon Energy and Environment. We can then use revised estimates of soil parameters to improve flood model predictions to get a better understanding of where tree planting may be effective, what area they may be effective over and what size of storm they might be effective against.
A cautionary tale – more research needed
Caution must be given to the expectation that tree planting is the panacea to all flooding. When soils are already saturated as has often been the case during the current winter, the positive contribution that trees may have in terms of providing additional water storage space in the soil below will be greatly diminished. Nevertheless, the added surface roughness provided by the trees and their understory could aid in flood mitigation by delaying runoff. It is questionable however whether any land use intervention would be effective against the extreme events that the UK has experienced in recent months, we just need more research to work out the magnitude of these effects.
Finally, we need to think of the catchment as a whole when thinking of ways to reduce flooding. Planting trees is only one option amongst a suite of measures that we should consider.
Dr Miles Marshall works in the CEH “Catchment management and soils systems” research group. He is based at our site in Bangor, North Wales. His research interests focus on biogeochemical and hydrological processes and how they vary spatially and temporally, and the linkage to other ecosystem processes and functioning.
The publication discussed in this blog post is: Marshall, M.R.; Ballard, C.E.; Frogbrook, Z.L.; Solloway, I.; McIntyre, N.; Reynolds, B.; Wheater, H.S., 2014, The impact of rural land management changes on soil hydraulic properties and runoff processes: results from experimental plots. Hydrological Processes, 28, 2617-2629