Recent Trials with Vetiver in Honk Kong

R.D. Hill

Department of Ecology & Biodiversity

The University of Hong Kong


Vetiver has been used in Hong Kong for erosion control for almost 10 years, as has been reported in earlier issues of the Vetiver Newsletter. However, it has not been widely adopted for several reasons. First is the high cost of labour, about US$45 per day for unskilled workers. This has meant that Vetiver has no cost advantage over the well-established methods of hydro-seeding grass on bare slopes or planting trees, usually species of Acacia. Second is that there is a small but active contracting "industry" using hydroseeding, the proprietors of which, having invested in mixers, pumps and hoses, do not wish to use an alternative approach, especially one that is necessarily labour-intensive. Hydroseeding, usually onto cut slopes stabilized with geotextile, is regarded as effective, even though it suffers from some serious defects. The species used have flaccid leaves, especially when young, and while they offer reasonable protection from rain-drop impact, they are not effective where concentrated overland flow develops and, of course, require about six weeks to become effective at all. A further disadvantage is that hydroseeding involves the pumping of a slurry of macerated paper, fertilizer and grass seed. It is thus limited to areas accessible to truck-mounted equipment though these probably account for four-fifths of the slopes to be treated.

The third reason why Vetiver has not been widely adopted is the desire by government agencies, which do most of the tree-planting, for a "one-shot solution", in a word, plant the trees and forget them. It seems easier to do this, using contractors, than to plant Vetiver hedges, which, to obtain optimal performance, require periodical cutting (burning is regarded as too dangerous) and fertilizer application. Contractors also view contour planting as "troublesome".

Fourth, there is the perception that the formation of terracettes behind Vetiver hedges may lead to a localized increase in the static load upon the slope and thus to slope failure. This is despite the fact that in nearly 10 years of trials, both by the author and by Dr Richard Webb under government auspices, there has never been any case of slope failure precipitated by slope overload.

Finally, there is a general reluctance amongst slope stabilization engineers to experiment with a method that relies upon natural processes to revegetate the bare slopes between hedges, one that if it fails may leave them financially liable. This fear to some degree applies even to hydroseeding and many engineers prefer to use "Chunam", a cement/clay mixture applied by pump to slopes as an almost dry plaster which immediately and effectively seals them. Aesthetic considerations do not enter into the equation.

However, there are significant areas distant from road access, where Vetiver can be used though erosion control has a much lower priority than in and near populated areas. In such areas the "traditional" approach has been to plant trees, in the past usually the local Pinus massoniana, now usually Acacias such as A. confusa, A. mangium, A. auriculiformis.

On some sites this has been reasonably successful though it usually takes 3-5 years for a closed canopy to form and even then it is clear that significant erosion continues. The architecture of Acacias is such that they initiate substantial stem-flow, certainly tens and possibly even hundreds of litres a day during rain. This results in local scouring. However, as the Acacia canopy closes, light levels at ground level clearly fall and this appears to reduce the ground cover often leaving only a thin litter of abscissed leaves. On steep slopes this readily washes downslope leaving the soil exposed to drip from the leaves and branches of the canopy. Work by Zhao in south China suggests that at high rainfall intensities drip is even more erosive than direct exposure to rain.

These considerations have led the author to persist with work on Vetiver hedges, especially on highly-eroded nutrient-poor substrates derived from weathered granite. These contain virtually no nitrogen, are indurated when dry though friable when wet. They have low permeability. Whether moist or dry overland flow has been observed to begin within 10-20 minutes of the commencement of the area’s characteristically sharp showers. Most of the slopes have lost between half and one and a half metres of material, more in gullies, as evidenced by vegetated "islands" perched above the general "soil" surface.

The main thrust of recent work has been to improve strike rates following planting, for these have been quite variable from site to site probably in response to weather conditions. Although Hong Kong receives an average of 2,224 mm per year, variability is substantial. However, whatever the total fall may be, the number of rain-days lies in the 90-120 day range annually. Dry spells can occur at any time of the year, the months of September through December sometimes having no rain at all. (However, earlier work has shown that there is probably sufficient residual moisture until October for Vetiver planting to succeed).

Until now bare-root planting, despite the major disadvantage of slow root development, has been the norm for several reasons. First would be the substantial labour cost of preparing slips and "pre-rooting" them in a nursery. Second is the cost and difficulty of obtaining soil for this process. Third would be the additional cost of transporting, say, polybags of slips and soil or mulch, to the actual planting site.


These considerations led to the decision to experiment with the use of a newspaper mulch for pre-rooting and planting out. Clumps of Vetiver were uprooted from the nursery and pulled apart into their constituent tillers, old or very young ones being discarded with stalky and dead material, the tops being cut to 20-25 cm and the roots to 5 cm. Newspaper mulch was prepared with tap-water in a commercial blender using torn-up paper. Three tillers were then bundled with about 100g of wet mulch, wrapped in newspaper and secured with a rubber band. Four treatments, each applied to 40 bundles of slips (i.e. 120 slips), were employed as follows:

A. no treatment (control)

B. 1g granular NPK fertilizer per bundle, applied when wrapped

C. 5m1 stock solution of 1GA rooting hormone (diluted at lml/l from solution as supplied), per bundle, applied when wrapped

D. 1g NPK and 5m1 1GA stock solution per bundle, applied when wrapped


The bundles were placed in separate trays and left out-of-doors in full sun of 21 days, following which the bundles were pulled gently apart and the number of new roots on each tiller counted. The bundles were then reconstituted using new wrappers and planted out on site three days later.

The results are set out in the following Table.


Number of new roots per tiller: Frequency by treatment

Number A B C D
0 10 24 2 19
1 19 21 13 21
2 31 22 25 21
3 26 23 29 27
4 16 7 23 19
5 11 10 16 8
6 3 7 7 3
7 3 3 1 2
8 1 3 2 -
9 - - 1 -
10 - - 1 -
Total 120 120 120 120
Mean* 2.7 2.4 3.3 2.4
Mean SD* 1.0 1.2 1.0 1.5

*Calculated hill-by-hill

 The treatment with rooting hormone (C) is clearly superior with no treatment at all (A) being quite satisfactory. The use of granular NPK, whether alone or with rooting hormone was not satisfactory. The granules were only partly dissolved and it was observed that in a number of cases the roots were chemically burnt. In addition the bundles in treatments B and D smelt strongly of sulphur compounds, presumably formed or released in a reaction with the newsprint as a result of the treatments. (Sulphuric acid is used in the manufacture of cheap papers and a residue remains after the paper is made). The implication of this is that fertilizer should be applied after planting out, not with the bundles.

Analyzed bundle-by-bundle similar results are obtained: A, average 8.1 new roots per bundle, B, 7.5, C, 10.2, D, 7.2. Had the more time-consuming method of measuring the length of the roots been used, it is likely that the differences would have been even more striking for with treatment C many bundles had roots several centimetres long growing through the wrappers. This is important not only supplying nutrient from the soil but in quickly establishing resistance to flowing water which might dislodge the newly-planted slips.

It was intended that the bundles be planted out on site in a regular pattern that would enable their identification as to treatment, A-D, as they subsequently grew. Unfortunately, an error by an assistant at planting out has made such identification difficult. A further trial will therefore be made to monitor performance after planting out. 


A Vetiver nursery is maintained at the University’s Kadooric Agricultural Research Centre which also provided laboratory space and equipment. The planting site is on eroded weathered granite at Mount Butler, Hong Kong Island and is used by permission by the SAR. Government.


Zhao, G., 1996, "Function of forest [with] respect to water and soil conservation 1. Theory of canopy influence on the distribution of rainfall kinetic energy. Acta phytoecol sinica, 20 (4) [in press, in Chinese with English abstract].