Comments on: The hydrological hazards of tree planting

By: More forest mythology

More forest mythology — Mon, 30 Jun 2008 04:37:00 +0000

[...] is a nice example of the hydrological mythology that often informs discussions of forest in southeast Asia, and elsewhere. This is from a [...]

By: Josh

Josh — Mon, 16 Jun 2008 03:22:13 +0000

Sorry. I did say that base flows would increase. That was a mistake. They could depending on the type of land cover present before reforestation, but they won’t for sure. My apologies.

By: Josh

Josh — Mon, 16 Jun 2008 03:15:17 +0000

Hi Andrew,

I don’t want to divert from the initial argument quite yet. Lets remind ourselves what you stated: you think reforesting a catchment can be hazardous. I think you’re overstating the reduction in average stream flows involved and disingenuously equating this reduction with true hydrologic hazards. It should be plainly obvious that a forested landscape is less prone to hazardous hydrological events than a crop or urbanized area. And for the record I never once stated that base flows would increase with reforestation. Please re-read what I’ve said.

Again, I am arguing that increasing forested area within a catchment would stabilize base flows, reducing annual base flow coefficients of variation, and that the question of which landscape will loose more water to the atmosphere depends on the proportion of time the canopy is wet. In total water balance terms, a crop system will generally output more water than a forested system and store less water over a given time period. This is because crop systems have lower infiltration rates and higher runoff rates during wet periods, as well as higher transpiration rates during dry periods. Consider the following closed system water balance when comparing the two systems: dS/dt = P – E – Q, where S is system storage, t is time, P is precipitation input, E is evaporation and transpiration output, and Q is discharge output including deep percolation and surface runoff. The change in average base flow, on an annual basis, brought on by reforestation would depend on catchment characteristics; topography, soil texture, site history etc., i.e. it’s inherent ability to store water. But generally, average stream flows would decrease due to reforestation (remember that base flows and stream flows are different metrics and different things). But more importantly forested catchments are less likely to be subject to life damaging no flow, or zero flow occurrences because of their storage effect (or the sponge analogy you seem to like so much). Furthermore, forested systems are less likely to be subject to flood events because they store more water due to their higher infiltration rates, and reduce both instantaneous and short term runoff rates due to their above ground woody biomass. These qualities make forested catchments less prone to commonly identified hydrological hazards, namely no flow and flood events. Don’t forget what I am arguing against – namely your assertion that reforestation is somehow a hydrological hazard. Reforestation also provides a number of ancillary climate, ecological and social benefits as well. No hazards there!

In answer to your question: this would depend on the species planted and the management practices used. I’ve seen white pine plantations in Ontario and Massachusetts root to 3 m, grow to 8 m and achieve an LAI of 12 in less than 15 years. I know eucalyptus, bamboo and other common plantation species in your part of the world grow even quicker. Regardless, the probability of hazardous events occurring would begin to decrease shortly after planting. Weighing the associated hazards is the real question. Do you still think you can somehow unduly create hazards by planting trees and increasing biomass on a landscape?

In the likely event that you’ve skimmed my rebuttal, the bottom line is that reforestation will reduce the frequency of hazardous hydrological events when compared with crops or urban areas. Not cause “hydrological hazards”.

By: Andrew Walker

Andrew Walker — Sun, 15 Jun 2008 21:29:59 +0000

Thanks again Josh for your detailed comment. We have discussed many of these issues in our book (Forest Guardians, Forest Destroyers). You can get a copy from Amazon – not so expensive!

I think we agree that forests take more water out of the system than other land-covers.

I think we also agree that deforestation is likely to increase annual stream flow and that reforestation is likely to reduce annual stream flow.

Where we disagree (I think) is on the effect of reforestation on dry-season flow. You are arguing that greatly improved infiltration will lead to increased base flow (even if the total annual flow is reduced). In other words you are arguing that, in relation to dry season flow, the sponge effect of reforestation will outweigh the pump effect.

There are many aspects of this we could debate. But, for now, can I ask just one question – how long do you think it will take for reforested land to recover the sponge effect that you suggest is lost when deforestation occurs?

By: jeplang

jeplang — Sun, 15 Jun 2008 13:59:10 +0000

Thanks Roy for the additional information,and Josh for the detailed mini-article.
I ,also ,believe that Andrew has over-simplified the issue but then I’m not a forest hydrologist or forest meteorologist.
Is Andrew’s cautionary tale out of concern in part for the abuse directed at the highland agriculturalists by the lowland agriculturalist?
Shouldn’t one major concern be the efficient use of water during the dry season ?How efficient is dry season water use in Thailand?

There is no doubt that the supply of freshwater for agriculture is going to diminish over time ,something that geographers have been warning the world at large for at least 15 years or more.
Agricultural planners should by now be thinking of crop species and varieties that require less water .Are they?
Is IRRI investigating ways and means of reducing the water consumption of rice,which of course is one of the world’s thirstiest
crops?

By: Roy

Roy — Sun, 15 Jun 2008 06:04:30 +0000

Further to Jeplang’s question. The link that Andrew calls report is a short VoA news blurb that does not more than mention a report from the Lao Ministry of Agriculture and Forestry that states the intentions of covering an area of 25,000 hectare with trees. The news clipping further states that this is part of Lao government ambitions to increase forest cover up to 53% in Laos and to restore an area of 2,5 million hectare affected by heavy deforestations.

As Jeplang pointed out, this of course leaves lots of questions unaddressed. So, some information about the content of the actual report would be helpful indeed.

A Vientiane Times article (28th March 2008) entitled ‘Govt orders more trees to fight climate change’ adds the following. It explains that following the Asean Inter-Parliamentary Assembly meeting in KL (August 2007) Asean members countries declared (not binding) to start planting trees in urban areas as a means to ‘fight climate change and to assist the environment’.

In the Lao case this means according to the Vientiane Times that one million ornamental trees will be planted in urban areas, starting with 500,000 in Vientiane this year. The trees include the ‘champa’ (frangipani), the Lao national flower, and will be planted along main roads, public places and at the new national stadium where the 2009 SEA Games will be held. The Vientiane Times article further states that the 1 million ornamental trees are separate from the 25,000 hectare to be planted on Arbor day.

Although this Vientiane Times article does not present any new information about large scale reforestation ambitions it illustrates that the planting of trees is about much more than plain environmental concerns.

By: Josh

Josh — Sun, 15 Jun 2008 04:35:14 +0000

P.s. I am trying to refine my debating skills and using the blog-o-sphere as a virtual firing range to practice with. Please don’t be offended by the tone. I do however genuinely disagree with you on some of the finer points laid out on your blog. They are my targets, not you. I found you through google alerts. Thanks for playing.

By: Josh

Josh — Sun, 15 Jun 2008 04:16:51 +0000

I am taking issue to your position that reforestation is a hydrological hazard. I feel it isn’t an accurate statement. I don’t feel I’ve oversimplified anything. To the contrary, I take issue to your statement because I think you’ve oversimplified forest water use. I’ll grant that I’ve only studied hydrometeorology for a few years, but I’m at my third university now and I do have something to say on the matter.
The thing you’re oversimplifying relates to the source of forest evaporation and why forest canopies loose more water than crops do over the course of a year. Forests don’t loose more on balance because they transpire greater amounts of water (in fact crops generally transpire more), they loose more because they intercept precipitation and evaporate it. Forest canopies effectively trap precipitated rain. Sometimes this can amount to a large proportion of local precipitation, depending on the characteristics of the canopy and intensity of the precipitation event (Arnell, 2002). The surfaces of crops and grasslands are not able to effectively intercept as much precipitation as forest canopies simply because they don’t have comparable amounts of foliage. This fact coupled with the fact that the aerodynamic conductance of a forest is at least an order of magnitude greater than that of crops (Arnell, 2002) or grasslands (Kelliher, 1993), make forests evaporate greater amounts of interception and evaporate it more efficiently as well. This makes it essential for any assessment of land use change effects on water resources, to incorporate knowledge of the proportion of time that the evaporating surfaces are wet. Without a forest canopy the local precipitation regime has control over water availability and deforestation is a real threat to precipitation regimes. In fact it has been shown that significant changes to cloud cover in the Amazon basin occurred when a threshold value of 20% deforestation in the region was reached due to a significant reduction in evapotranspiration (Nosetto et al., 2005). Without forests what little rain falls in a drought stricken area will leave the system at a much greater rate. I guess this may not be considered hazardous to someone concerned with the operation of a hydroelectric dam, but it certainly is a hazard. Too little water downstream to turn turbines is less of a hazard in my eyes.
Since transpiration is either halted or significantly reduced when leaf surfaces are wet, the partitioning of ecosystem evaporation between transpiration and interception is greatly simplified. Though forests have the ability to evaporate much larger amounts of water on a seasonal or annual basis than areas where forest trees have been removed and replaced with grasses or crops due to their greater rates of interception evaporation, this doesn’t mean that forests transpire more water. Typically forests have lower canopy conductances and higher aerodynamic conductances than do crops (Oke, 1987). One reason for the difference between forests and crops is energy related. The transpiration rates of crops have been shown to be closely tied to variations in net radiation, where as rates in forests are generally tied to variations in vapour pressure deficit, VPD (Oke, 1987). The lower surface roughness of crops is thought to reduce turbulence within crop canopies allowing canopy temperatures to rise above air temperatures and thus allowing greater vapour pressure gradients to develop between short, even height crop species and the atmosphere (Oke, 1987). This reduces resistance to water vapour flow and consequently, the transpiration of agricultural land is usually higher than forested landscapes (Oke, 1987). This explains why there is a tendency for the transpiration from crops to be closely tied to variations in Rn rather than VPD, and also explains the reverse for forests. In general, it can be said that forests are therefore more closely ‘coupled’ with the state of the atmosphere than are crops during dry periods. Here’s a summarization from Tim Oke’s Boundary Layer Climates (ISBN 0-415-04319-0), a modern classic used in Universities all over the world: Et crop > Et forest but Ei crop < Ei forest (p156 where E is evaporation, i is interception and t is transpiration). The water that isn’t intercepted is runoff and this is a real hydrological hazard. Flooding is a genuine hazard in my mind. Forests reduce the occurrence of flash flooding.
Crops categorically do loose much greater amounts of water to runoff than forests and do have significantly lower infiltration rates over any time span. As for the soil moisture of a crop compared to a forest, it depends more on soil texture than on land cover and I have never read a review article that quantifies agricultural landscapes as generally having higher soil moistures than forests. I think you’re completely mistaken about that fact in particular.

Arnell N (2002) Hydrology and Global Environmental Change. Pearson Education Ltd., Edinburgh Gate, Harlow, Essex, UK, 346 pp.

Barbour MM, Hunt JE, Walcroft AS, Rogers GND, McSeveny TM, Whitehead D (2005) Components of ecosystem evaporation in a temperate coniferous rainforest, with canopy transpiration scaled using sapwood density. New Phytologist, 165, 549-558.

Calder IR (1998) Water use by forests, limits and controls. Tree physiology, 18, 625-631.

Farley KA, Jobbagy EG, Jackson RB (2005) Effects of afforestation on water yield: a global synthesis with implications for policy. Global Change Biology, 11, 1565-1576.

Kelliher FM, Leuning R, Raupach MR, Schulze ED (1995) Maximum conductances for evaporation from global vegetation types. Agricultural and Forest Meteorology, 73, 1-16.

Kelliher FM, Leuning R, Schulze ED (1993) Evaporation and Canopy Characteristics of Coniferous Forests and Grasslands. Oecologia, 95, 153-163.

Kurpius MR, Panek JA, Nikolov NT, McKay M, Goldstein AH (2003) Partitioning of water flux in a Sierra Nevada ponderosa pine plantation. Agricultural and Forest Meteorology, 117, 173-192.

Nosetto MD, Jobbagy EG, Paruelo JM (2005) Land-use change and water losses: the case of grassland afforestation across a soil textural gradient in central Argentina. Global Change Biology, 11, 1101-1117.

Roberts J (2000) The influence of physical and physiological characteristics of vegetation on their hydrological response. Hydrological Processes, 14, 2885-2901.

Roberts J (1983) Forest Transpiration – a Conservative Hydrological Process. Journal of Hydrology, 66, 133-141.

By: Andrew Walker

Andrew Walker — Sat, 14 Jun 2008 23:47:28 +0000

Thanks for your comment Josh. In our book, Tim Forsyth and I deal with the impact of forest cover on the seasonal distribution of flow in some detail (108-113). I don’t think the situation is as simple as you suggest. Reforestation may increase infiltration (but improvements in infiltration may take a very long time and be quite modest). This is the “sponge” effect. But reforestation, as you know, also increases evapotranspiration – taking a lot more water out of the soil than most other land covers. This is the “pump” effect. The effect on base flow of reforestation will depend on the trade-off between the “sponge” effect and the “pump” effect. This is the trade-off we were referring to when we wrote, as quoted above:

There is also a good chance that extensive tree planting will reduce dry-season flow, because the medium- to long-term benefit in terms of enhanced infiltration on reforested soil may well be limited and strongly outweighed by short- to medium-term increases in the level of water “lost” due to the increased evapotranspiration.”

We can’t assume that sub-surface soil under forest cover will contain more moisture than under other land covers. Here is another quote from our book:

The work of Takahashi and colleagues (1983) provides one illustration of this trade-off between infiltration and evapotranspiration. … [T]his study found that infiltration on cultivated upland fields was significantly lower than infiltration in forest. However, when researchers examined the soil itself they found that the cultivated areas had higher levels of soil moisture, which is the basis of dry-season flow. In the forested area, “the soil was drier in deeper horizons and always in the condition of low soil moisture, compared to the other plots” (Takahashi et al. 1983:97). The lower soil moisture under forest resulted from the trees using more water than the crops cultivated on the cleared land.

I would be very interested to hear about other studies from the region that compare soil moisture under forest and other land covers.

By: Josh

Josh — Sat, 14 Jun 2008 12:29:09 +0000

Reforestation will increase annual evapotranspiration and decrease annual surface runoff when compared to most crops or bare soil (as you say), but it will serve to stabilize base streamflows during dry periods, not significantly reducing them as you have stated. Here’s an analogy. A crop, urban area or cut over site will act like a plate, with water from a tap (i.e. rain) running off its surface quickly. A forest will act like a sponge that’s dry at the top and saturated on the inside, with water from a tap running from it’s volume at a slow steady pace, while at the same time allowing more water to evaporate. Reforestation will increase infiltration, as you have said, but this increases base flows and decreases high flows following precipitation events (like flash floods for example). The distinction, in terms of runoff response, between a crop or bare soil and a forest is especially stark during drought, because of the large decrease in infiltration rates as soils dry (like baking a clay pot, soils also bake during drought and harden). So when looking long term you must ask yourself whether it is better to have more stable streamflows in a narrower range (i.e. a forested landscape), or more erratic streamflows with more frequent extreme events ranging from no-flow to flood stage (i.e. a crop or urban area). As an added bonus the evaporation from forests also reduce local temperatures by consuming heat and storing it in the form of latent heat rather than converting it to sensible heat. The fact is that adding biomass to a watershed stabilizes its hydrology, moderates temperatures, and moderates local atmospheric water vapour contents. I don’t consider this a hydrological hazard.