logo 2
   
09 Nov2008

Overstory #212 - Forests on sites with high landslip risk

Written by Lawrence S. Hamilton.

Forests on sites with high landslip risk

With their undergrowth, leaf litter, forest debris and uncompacted soils, forests are almost without question the best and safest land cover for minimizing surface erosion of all kinds (Wiersum, 1985; Kellman, 1969). On sloping lands within any climatic regime, the earlier FAO land suitability classes for lands that might safely be cleared and used for crops or grazing were predicated on this function of minimizing erosion under various uses (FAO, 1976). With their stronger, deeper root systems, forests are also the best land cover for minimizing the hazard of shallow landslips (Rapp, 1997; O'Loughlin, 1974; Ziemer, 1981). Such landslips are often catastrophic, and this section presents the case for keeping slip-prone lands in forest cover. Short-duration, intense storm events generally create shallow landslips, while prolonged, low-intensity events produce the deeper, larger landslides for which forests may be ineffectual - witness the estimated 20 000 landslips and landslides that occurred in a single day in the Sikkim-Darjeeling area in a 1968 event (Ives, 1970).

What was possibly, in terms of human life, the Western Hemisphere's worst natural disaster in 500 years occurred on 16 and 17 December 1999. (The December 2004 tsunami disaster in South and Southeast Asia dwarfs this.) Hundreds of landslips and landslides killed 50 000 of the 500 000 people living on the coastal slopes and at the base of Venezuela's Cordillera de la Costa range (Myers, 2000); 40 000 homes were destroyed, and most roads. The previous year, Hurricane Mitch had triggered thousands of landslips and landslides in Central America - the worst natural disaster to strike the region in 200 years. Shallow landslips, deep landslides and flooding affected 6.4 million people, killing 9 976 and leaving 11 140 unaccounted for, 13 143 injured and 500 000 homeless (International Federation of the Red Cross, 2000).

A large number of internationally noteworthy mass erosion events occurred at the end of the 1980s: southern Thailand in 1988 after torrential rainfall; Cyclone Bola in New Zealand in the same year; Puerto Rico during Hurricane Hugo in 1989; and the Philippines in 1990. Most damage was due to slope failures in hundreds or thousands of shallow landslips, with some large landslides. It is interesting to note that devastation in southern Thailand and the Philippines was blamed on logging, although most of the failures were on land cleared for crops (Rao, 1988; Hamilton, 1992). It was the same in the 1999 Venezuelan catastrophe - almost all the landslips were on cleared land. Slope failures and high damage occur periodically, generating much discussion and research on the role of forest cover in reducing the incidence or severity of these mass movements. Case study 8 gives a post-disaster analysis of the true causes of southern Thailand's floods in 1988.

It is difficult to distinguish human-caused slope failures from natural ones (Bruijnzeel and Bremmer, 1989). There is a common lay misperception that slope failures do not occur in undisturbed forest lands, and popular confusion about the types of mass movements that can be influenced by vegetation.

The following classification was developed from studies in the United Republic of Tanzania and seems to be applicable in many locations (Rapp, Berry and Temple, 1972):

Class 1: numerous, small, 1 to 2 m-deep, 5 to 20 m-wide earthslides and mudflows, triggered by heavy rainstorms, occurring at intervals of about ten to 20 years.

Class 2: occasional, large earthslides, cutting many metres in depth into the weathered bedrock, below the anchoring effect of tree roots, and occurring at much longer intervals.

Much scientific attention has focused on the role of forest cover in reducing the incidence and severity of shallow landslips (Class 1). It is accepted that these can both be reduced or even eliminated by good forest cover. Large, deep-seated mass movements (Class 2) seem to be beyond the control of vegetation, but most research has been done in temperate countries, especially Japan (protection forests), New Zealand (pastureland and logging), Taiwan Province of China (protection forests) and the western United States (associated with logging).

In a study of land degradation in the Uluguru Mountains of the United Republic of Tanzania, Rapp (1997) examined numerous, small (1 to 2 m-deep, 5 to 20 m-wide) earthslides and mudflows in a 75 km2 area, which occurred after more than 100 mm of rain had fallen in less than three hours, He found that of 840 landslides only three started on slopes under forest cover - the remainder were in cultivated or grazed areas on similarly steep slopes. The importance of tree roots in providing shear resistance in slip-prone soils has been demonstrated in studies by O'Loughlin (1974) and Ziemer (1981). This work and its subsequent confirmation provide the basis for guidelines in this area.

All very steep lands benefit from being in forest cover, especially those in seismically active areas; pressure to clear land occurs where slopes are intermediate but still slip-prone - and this is where the red flag needs to be raised. Prudent land use could avoid many of the catastrophic results that were seen in Thailand in the 1988 storm, when former forest land that had been cleared to establish rubber plantations failed in thousands of landslips. The subsequent ban on logging did not apply to land clearing (Hamilton, 1991). In several regions of New Zealand that were largely cleared of native forest for grazing in the 1870s, slip-prone land has experienced serious erosion from major storm events. The erosion rate measured in the Wairarapa region was 2.8 percent per decade in the period 1938 to 1977, and resulted in 56 percent eroded land in hill-slope scars in 1984 (Trustrum, Thomas and Douglas, 1984). This material appears as excessive sedimentation in watercourses.

In September 2004, Hurricane Jeanne passed over the Dominican Republic, Puerto Rico and Haiti, resulting in landslips and flooding. Aide and Grau (2005) report that although these countries received similar levels of precipitation, there were seven flood-related deaths in Puerto Rico, 24 in the Dominican Republic, and more than 3 000 in Haiti. There are complicating factors, but these authors attribute the diverse landslip and flood damage to differences in vegetative cover, with forested areas and abandoned, reverting shrublands being less damaged by erosive agents.

The challenge is to identify these red flag areas in advance and to maintain them in forest or woodland. The general factors that influence mass wasting are known; presence of water, type of rock or mineral and state of weathering, number and density of natural fracture planes, and structure and inclination of slope. Practical field guides are needed for identifying slip-prone areas where forest retention is desirable. Megahan and King's 1985 guidelines are particularly good for this. They point out that in highly erosion-prone areas, the major hazards are found on slopes greater than 45 to 55 percent, with maximum frequency of about 70 percent; concave slopes, which concentrate water; and soils that are low in cohesion. Shallow soils over bedrock or with pronounced discontinuity of texture or structure may become saturated, buoyant and slip-prone. Megahan and King discuss rainfall erosivity, and the difficulty of obtaining good data in the tropics.

Based on reported research worldwide, Blaschke, Trustrum and Hicks (2000) produced a map showing the approximate extent of areas where mass-movement erosion affects land productivity, as well as a table showing land use, rainfall, land form, area affected, duration of event and soil loss rate, by country or region. Good data are few and scattered. Studies such as Humphreys and Brookfield (1991) in Papua New Guinea report that shallow slope failures are by far the most common erosion forms in cultivated steep lands. The result is not only reduced productivity, but also sediment loads in watercourses, which impair water quality, have an impact on aquatic life and promote flooding.

Well-managed agroforestry or silvopastoral systems with high percentages of deep-rooted trees would give more safety than crop or pasture alone, because of root shear strength. No research indicates the necessary tree density on slip-prone sites, but the more the trees, the greater the safety margin.

Roads associated with logging or other uses are often a triggering agent, because the cut and fill on side slopes further destabilizes these troublesome sites.

In summary, trees are the safest land cover where high-intensity or prolonged rainfall is characteristic of slopes of about 70 percent - but also as low as 45 percent - and that have spoon-shaped concavity or shallow, planar surfaces. When there are no tree roots, there is high risk of slope failures.

Clearing is inadvisable and should only be sanctioned if there is the certainty of:

  • quick re-establishment of tree crops such as rubber or tree plantations (although there is a long vulnerable period until new tree roots become effective);
  • immediate terracing for crop production plus trees in an agroforestry system, assurance of adequate terrace maintenance and the ability to repair minor landslip damage quickly; or
  • a silvopastoral system with abundant deep-rooted trees and conservative grazing management.

Forest harvesting reduces the stabilizing ability of roots when cut trees die, and there is a vulnerable period of several years before new roots are effective. If a serious storm event occurs during this period, landslips may ensue. Logging roads represent a further destabilizing force on these unstable sites.

Established in 2002 with its headquarters in Kyoto, Japan, the International Consortium on Landslides (ICL) promotes landslide research for the benefit of society and the environment. The interface between forests and water and the role of forests in mitigating landslide risks and hazards are important components in ICL's activities.

Guidelines

Slip-prone areas are probably the most serious form of erosion that can be minimized by sound land-use policies and management. Tree roots impart a margin of safety by improving soil shear resistance. To reduce the occurrence and severity of shallow landslips, slip-prone areas need to be kept in forest cover, woodland or agroforestry/silvopastoral systems with fairly dense tree cover. Such areas can be identified in advance of land-use decisions, based on erosivity, slope, slope shape, and soil shallowness and cohesiveness. Roads are a particular problem.

Start case study: Floods and landslips in southern Thailand

In November 1988, following an unprecedented downpour and flash floods, mud slides descended from the hills surrounding Nakhon Si Thammarat province in southern Thailand. Hundreds of landslides littered the hill slopes almost overnight, killing 200 people, burying 300 houses under sand and knocking down hundreds of fruit trees.

Floods uprooted trees along the path of their flow, and debris caused blockages and dammed up water in some locations. As the rainfall continued, these blockades broke up and the released waters coalesced and surged forward, carrying sand, uprooted trees and other debris. This unprecedented flow scoured the stream banks, flooded houses and fields and changed the course of rivers. The people affected included small farmers cultivating rubber, fruit growers and the landless.

This area was originally covered by a typical tropical, moist forest with a complex of vegetation consisting of the predominant trees, the dominant canopy, other layers of trees, shrubs, climbers and undergrowth. Some areas were logged until some years ago, when they were cleared and converted into rubber plantations, which were established even on steep slopes. In most cases, there was no cover crop and very little vegetation to protect the ground - just the rubber trees, many of which were only recently planted.

Several news reports attributed the flood damage to logging operations. The outcry resulted in a government decree banning logging. A less sensational but more realistic explanation for the flood damage is the combination of several factors that together proved to be disastrous:

  • The slopes where the slides occurred were steep, often with angles of more than 25 degrees.
  • The basic geological underpinning for the topsoil was deeply weathered and extremely fractured granite formations, which are easily eroded.
  • The soil blanket was not sufficiently anchored by the young rubber trees' roots - the most common vegetation in the area of the slides.
  • The vegetation was sparse, with no vegetative cover between the rows of rubber trees.
  • Records show that from 20 to 23 November 1988, hill areas of the province received 1 022 mm of rainfall, which saturated the soil.
  • The intense rainfall could not be absorbed, particularly on the steep upper reaches, and the resulting sheet of water flowed downwards.
  • Runoff and slope failures created landslides on the generally steep hill slopes.

The numerous scars and deep gullies that today mark the landscape of Nakhon Si Thammarat province present an unrivalled opportunity to demonstrate how landslips can be recovered and brought into productive use. In response to the Government of Thailand's request, FAO has approved a Technical Cooperation Programme project that will work with foresters and agronomists in Thailand to attempt rehabilitation of some landslips.

Source: Adapted from Rao, 1988.

End case study: Floods and landslips in southern Thailand

References

Aide, T.M. & Grau, H.R. 2005. Will rural-urban migration reduce floods? Arborvitae: IUCN/WWF Forest Conservation Newsletter, 27: 11.

Blaschke, P.N., Trustrum, N.A. & Hicks, D.L. 2000. Impact of mass movement erosion on land productivity: a review. Progress in Physical Geography, 24(1): 21-52.

Bruijnzeel, L.A. & Bremmer, C.N. 1989. Highland-lowland interactions in the Ganges Brahmaputra River Basin: a review of published literature. Occasional Paper No. 11. Kathmandu, International Centre for Integrated Mountain Development (ICIMOD).

FAO. 1976. A framework for land evaluation. Soils Bulletin No. 32. Rome. Hamilton, L.S. 1991. Tropical forests: identifying and clarifying the issues. Unasylva, 166(42): 19-27.

Hamilton, L.S. 1992. Storm disasters - has logging been unfairly blamed? Further note on Philippine storm disaster. IUCN Forest Conservation Programme Newsletter, 12(5) and 13(3).

Humphreys, G.S. & Brookfield, H. 1991. The use of unstable steeplands in the mountains of Papua New Guinea. Mountain Research and Development, 11: 295-318.

International Federation of the Red Cross. 2000. Central America: Hurricane Mitch emergency relief. Situational Report No. 4 (final). Geneva.

Ives, J.D. 1970. Himalayan highway. Canadian Geographical Journal, 80(1): 26-31. Kellman, M.C. 1969. Some environmental components of shifting cultivation in upland Mindanao. Tropical Geography, 28: 40-56.

Megahan, W.F. & King, P.N. 1985. Identification of critical areas on forest land for control of nonpoint sources of pollution. Environmental Management, 9(1): 7-18.

Myers, L. 2000. Students raise funds for disaster relief in Venezuela. Cornell Chronicle, 31(21): 1, 3.

O'Loughlin, C.L. 1974. The effect of timber removal on the stability of forest soils. Hydrology, 13: 121-134.

Rao, Y.S. 1988. Flash floods in southern Thailand. Tiger Paper, 15(4): 1-2.

Rapp, A. 1997. Erosion and land degradation in drylands and mountains. In D. Brune, D.V. Chapman, M.D. Gwynne and J.M. Payne, eds. The global environment: science, technology and management, Vol, 1, pp. 207-224. Weinheim, Germany, Scandinavian Science Publisher.

Rapp, A., Berry, L. & Temple, P.H., eds. 1972. Studies of soil erosion and sedimentation in Tanzania. Geografiska Annaler, 54A: 105-379.

Trustrum, N.A., Thomas, V.J. & Douglas, G.B. 1984. The impact of forest removal and subsequent mass-wasting on hill land pasture productivity. In C.L. O'Loughlin and A.J. Pearce, eds. Symposium on effects of forest land use on erosion and slope stability, p. 308. Honolulu, Hawaii, USA, East-West Center.

Wiersum, K.F. 1985. Effect of various vegetation layers in an Acacia auriculiformis forest plantation on surface erosion in Java, Indonesia. In S.A. El-Swaify, W.C. Moldenhauer and A. Lo, eds. Soil erosion and conservation, pp. 79-89. Ankeny, Iowa, USA, Soil Conservation Society of America.

Ziemer, R.R. 1981. Roots and the stability of forest slopes. In T.R.H. Davies and A.J. Pearce, eds. Erosion and sediment transport in Pacific Rim steeplands, pp. 343-359. Publication No. 132. Washington, DC, International Association of Hydrological Sciences.

Original Source

This article was excerpted with the kind permission of FAO from:

Lawrence, L.S. 2008. Forests and Water. FAO Forestry Paper 155. Food and Agriculture Organization of the United Nations. Rome.

About the Author

Lawrence S. Hamilton is Emeritus Professor of Natural Resources of Cornell University, having taught and researched there for 29 years, 1951-1980. He has carried out consultancies in Australia, Costa Rica, Venezuela, Trinidad and Bhutan for IUCN, USAID, Sierra Club, The World Bank and UNESCO, and has had writing consultancies for GEF and FAO. More than 290 publications have been authored, co-authored or edited during this lengthy career, including several books. The two most recent hardcover books were Ethics, Religion and Biodiversity (1993), and Tropical Montane Cloud Forests (1995). He is partner with his wife Linda Hamilton in ISLANDS AND HIGHLANDS, Environmental Consultancy based in rural Vermont, USA, and can be reached at:

Lawrence S. Hamilton Islands and Highlands, Environmental Consultancy 342 Bittersweet Lane, Charlotte, VT 05445, USA Email: hamiltonx2@mindspring.com

Related editions to The Overstory