FAO Home Page French version Spanish version Waicent Free Text Search

Use of attractive devices for tsetse survey and control

Chapter 2: Odour Attractants

General Principles

When tsetse flies seek a host, they use a variety of sensory perceptions. For a long time it was thought that the visual sense was the most important, including colour, contrast with the environment, and movement. Despite the fact that early work indicated that odour was also important, this has only recently been utilised for control purposes.

The odour of hosts can act on the flies in several ways. Some odours stimulate tsetse to fly and start searching for the host, for example carbon dioxide, but not acetone or octenol; others will induce flight toward the host, whilst others promote entry of flies into traps, this may be the equivalent of a landing response when natural hosts are involved.

There has been much recent research to determine how flies use the odour cues (signals). Tsetse flies generally fly upwind towards the source when they detect an attractive odour. If they loose the scent in flight, they turn sharply, and may settle and reassess the wind direction before flying upwind again. They may also, however, be able to assess the wind direction whilst in flight. Once close to the host, they mainly use sight to make the final approach.

Flying upwind to find the odour source may be difficult under field conditions, especially if the vegetation is dense. As the wind is constantly changing direction, flying upwind may take the fly in the wrong direction. However, it is now thought that even if the odour results in a small bias in flight in the correct direction, it will be enough to give the tsetse a good chance of finding the host. There is also evidence that tsetse may reduce their flight speed and turn more frequently when they detect high odour concentrations.

It is often thought that odours attract flies from a great distance. In fact evidence now shows that the range of attraction of ox odour is only about 90 m downwind of the odour source, and may be less for combinations of synthetic baits. The range of visual attraction of a stationary trap is only about 15 to 20 m, but is greater for a moving host.

It seems that all tsetse, including members of the palpalis group, such as G. fuscipes, are attracted by the carbon dioxide given off in animal breath. This can produce a substantial increase in catch if dispensed near a trap. However, it is not suitable for practical field use both because of cost, and the need for either pressurised gas cylinders or dry ice.

We will, therefore, concentrate on the currently known chemicals that can be used for sampling and/or control. These consist of three main components: ketones; octenol; and urine or phenol mixtures. Information on the degree to which these chemicals increase the catch of different species (section 2.3) is taken from the literature. However, it should be remembered that results have not always been analyzed correctly in the past (see section 4.5), and so some of the indices of increase presented may only be approximate.


Odour dispensers

Requirements for a good dispenser
A good dispenser should ensure a steady release of the odour at the correct rate for as long as there is odour in the dispenser. It will also be sufficiently robust to operate under field conditions for long periods.

Ideally the release rate should be independent of environmental conditions such as temperature and wind strength. The only way to do this is to use a valve release mechanism, as is done for carbon dioxide, but this is far too expensive for routine use, and so with the techniques used, release rates usually vary, being primarily dependent on temperature and humidity. This variation can be reduced by partly burying the odour dispenser or by protecting it with a cover.


Types of dispenser

(a) Glass jars and bottles
These are most frequently used for the ketones and for urine of various host animals. The volume of the bottle does not affect the release rate; this is controlled by the size of the aperture and may be set by using a lid drilled with a hole of the correct diameter (Fig. 1A).

Acetone is normally used at dose rates varying from 150-2500 mg/h. At temperatures varying from about 20C at night to about 35C in the afternoon, a hole of 2 mm diameter will give an average release rate of about 150 mg/h; one of 6 mm will give about 500 mg/h whilst one of 22 mm will give 2500 mg/h.

For cow urine a release rate of about 1000 mg/h is required. This can be obtained using an aperture of about 45 mm diameter. Alternatively a plastic container, or empty tin covered with polythene, can be used, with a aperture cut in the side,near the top, measuring 2 x 4 cm (Fig. 1B).

The advantage of the latter is that a rain cover is not required. If the aperture is exposed to the rain, some form of cover (Fig. 1A) must be provided to prevent the dispenser from filling with water.

Dispensing through an aperture will generally give too high a release rate for the phenols and octenol. These are best dispensed from a small glass bottle fitted with a rubber septum, which is kept constantly damp with the chemical by using a wick, often a pipe cleaner, or by laying the dispenser on its side. The chemical then diffuses gradually through the rubber septum. A loose fitting cloth cover will prevent the rubber hardening and cracking from exposure to sunlight. Sealed polythene sachets, through which the odours slowly penetrate, is the preferred method for dispensing octenol and the phenols (see below).

Figure 1. Odour dispensers A. 500 ml glass bottle for acetone with rain cover; B. 2 l plastic jar for cow urine. A suitably cut plastic insert below the cap can be used to support an octenol sachet. Dimensions are in cm.


(b) Plastic tubing
Tubes of low density polythene with a wall thickness of about 1500 um, surface area about 45 cm2, have been tested for dispensing mixtures of phenols and octenol. A small plastic plug is inserted in each end of the tubing and the chemicals diffuse through the walls of the tubing. Unfortunately the plastic tends to harden with age and exposure thus reducing the release rate.

(c) Polythene sachets
Polythene film of about 125-175 um thickness can be used to make sachets. These have the advantage that the release rate remains reasonably constant with age and they are cheap and easy to make. Flexible lay-flat polythene tubing can be obtained as a roll, and the sachets are made by hand using a heat sealer. These are commercially available in a variety of models.

Usually two seals are made across each end. The more expensive sealers enable the sachets to be filled first and then sealed. Alternatively the completed sachet can be injected with the chemical by syringe and the resulting small hole sealed. Presence of chemical in the area of the seal will increase the sealing time required. Care should be taken when handling phenols and protective clothing worn.

Sachets can either be flat and rectangular (Fig. 2A), or pyramidal in shape (Fig. 2B). A sachet made of 125 um polythene film will need to be about 5 x 4 cm to give an octenol release rate of 0.5 mg/hour, depending on temperature.

Figure 2. Odour dispensers for octenol/phenols made from lay-flat polythene tubing A. flat; B. pyramidal. Both give the same dose-rate but the latter holds more. Dimensions are in cm.


Attractiveness of odours to species

Morsitans species

(a) G. pallidipes
Several effective odours are known for this species. There is some evidence, however, that they vary in effectiveness for geographically separated populations.

In Kenya, acetone, when compared to a trap without odour, will give increases in catch of 1.5-2.5 times depending on the release rate (150-2500 mg/h). In Zimbabwe, the increases range from 1.5-4.0 times using release rates of 50-50,000 mg/h. In Mozambique, increases of 1.7-2.3 times can be expected with release rates of about 1200 mg/h. Whilst, in Somalia, acetone released between 5-50,000 mg/h does not seem to produce any increase when using an F3 trap.

Whereas a high release rate of acetone may be justified for survey purposes this may be expensive for control and an average rate of 100-150 mg/h is normally used. Butanone, or methyl ethyl ketone will double the catch in Kenya and Zimbabwe at a lower release rate than acetone (10-50 mg/h), but not in Somalia.

Octenol alone at 0.5 mg/h increases trap catches by about 1.5times in Kenya, Ethiopia and Zimbabwe, but has no effect at higher or lower doses. It can double the catch again if used with acetone or butanone, but has less effect, 1.5 times, if urine or phenols are also present. Octenol, either alone, or with ketones, seems to be ineffective in Somalia.

Bovid urine (cow or buffalo) can give substantial increases in catch (2-5 times) if dispensed at about 1000 mg/h. If a much higher release rate of urine is used, it becomes repellent. Acetone at 150-500 mg/h in combination with cow urine, in separate dispensers, gives increases in catches of 9-14 times in Kenya. In Ethiopia increases of 5-7.5 times were recorded. Pig urine is less effective than bovid urine.

Cattle urine increases in effectiveness if allowed to 'age' for about ten days in a sealed bottle before use. This may be because phenol concentrations in the urine increase as a result of microbial action. Subsequently urine can remain effective for several months if topped up with 50 percent cow urine/water at 1-2 monthly intervals. If urine is poured on the soil, it looses its attractiveness to tsetse within a few days.

The most active components in urine are 4-methylphenol, also known as para-cresol, and 3-n-propylphenol. The two phenols are synergistic, in other words they have little effect on catch when used alone, but put together give increases of up to 4 times. If the two phenols are mixed together with octenol in a sachet in an 8:4:1 ratio by weight of 4-methylphenol, octenol and 3-n-propylphenol respectively, release rates should be about 1.5, 0.2 and 0.5 mg/h respectively. Used together with acetone, this can increase catches by 10-20 times, in Zimbabwe.

A 9:1 mixture of 4-methylphenol and 3-n-propylphenol increases catches in Somalia by 1.6 times, and this mixture is further synergised by acetone (500 mg/h) and octenol (0.5 mg/h) to give an overall increase of 4 times.

Urine and phenols tend to increase the catch more for traps than for targets. This is because they also improve the entry response to traps, i.e. the trap efficiency.

In Zimbabwe, hessian sacks on which bushpigs have lain increase trap catches of G. pallidipes by 1.5-3 times in the presence of ketones and octenol. Ox sebum, skin secretions, gives slight increases for both traps and targets (1.3 times). Both these animal host residues may yield useful new attractive components.


(b) G. morsitans
Acetone is also effective for this species, and in Zimbabwe release rates of 50-50,000 mg/h give increases in trap catches of 1.7-7 times. In Burkina Faso release rates of 400-1200 mg/h increase catches of G.m. submorsitans in biconical traps by 1.8-2.3 times, depending on the season.

Octenol alone dispensed at up to 0.5 mg/h has little effect on trap catches of G.m. morsitans in Zimbabwe and is repellent at higher dose rates. Used with acetone it increases the catch by 1.5-2 times. Octenol alone at 0.5 mg/h has little effect on the catch of G.m. submorsitans in Burkina Faso but used with acetone it may more than double the catch, giving overall increases of 2.8-6.7 times, again depending on the season. Increases with acetone and octenol for this species in the Gambia range from 1.3-5 times, with seasonal variation.

Cattle urine usually does not increase the catch of G. morsitans. This may be because it contains both attractive and repellent constituents, 3-n-propylphenol increases the catch slightly by about 1.5 times, but 4-methylphenol decreases it. If both G. morsitans and G. pallidipes are present in the same area, a different synergist for 3-n-propylphenol can be used, eg. 4-ethylphenol, which although rather less effective for G. pallidipes, will not reduce the catch of G. morsitans.

Hessian sacks on which bushpigs have lain have also been shown to increase trap catches of G. morsitans in Zimbabwe in the presence of ketones and octenol, but increases were not as great as for G. pallidipes. Ox sebum increases catches at traps by 1.4 times and at targets by 1.8 times. 3-methyl indole is being tested as an attractant for G.m. submorsitans.


(c) G. longipalpis
Cow urine gives a slight increase in catch for this species, 1.1-1.3times. Bushbuck, domestic pig, warthog and hartebeest urines all give higher indices of increase than cow urine, with bushbuck urine the best, 2.5 times. A mixture of the seven main phenolic components of cow urine gives a 1.5 times increase. Of these, the most important is 3methylphenol, also known as meta-cresol, a major constituent of bushbuck urine, followed by 4-methylphenol.

Both acetone, at 500 mg/h, and octenol, at 0.5 mg/h, used in conjunction with the phenolic compounds increase the catch further, with the most effective combination being the phenols and acetone, 3.4 times increase.

For control purposes a 1:1 mixture of 3-methylphenol and 4methylphenol in a polythene sachet, together with acetone, or a 1:1:2mixture of these phenols with octenol is recommended.

d) G. austeni and G. swynnertoni
Relatively little work has been done on odour attractants for these species, although phenols and butanone were found ineffective for G. austeni on Zanzibar. Under laboratory conditions a positive response to ox breath odours has been noted, the active ingredients of which would include carbon dioxide.


Palpalis species

(a) G. palpalis
Despite much effort, there is generally little evidence of any effective odours for G. palpalis. In two trials in Liberia, however, acetone, at 100 and 500 mg/h, apparently increased the catch by 2 and 5.8 times respectively. A mixture of the phenols found in cattle urine had no effect, either alone or in combination with acetone and octenol.

(b) G. tachinoides
This is the one species of the palpalis group for which attractants have been found, although the level of increase is not as great as that for G. pallidipes and G. morsitans. The odour of live hosts, men, pigs and cows, has been shown to be attractive and the use of charcoal filters confirmed that carbon dioxide was not the only attractant.

In Burkina Faso, cow urine will increase the biconical trap catch by 1.7-3 times, and urine from Baoule cattle seems more effective than that of Zebu cattle. Bushbuck urine is even more effective, 4-5 times. The main active constituent in urine for this species is 3-methylphenol. The activity of 3-methylphenol is increased by the presence of octenol, at 0.5 mg/h, and the two together will give increases of 1.4-2.5 times. Octenol alone did not increase catches in Burkina Faso.

In Cte d'Ivoire octenol alone at 0.6 mg/h increases catches slightly, 1.3 times. Ketones are generally ineffective, as are phenols and indoles dispensed individually. However, an indole mixture, indole and 3-methylindole in a 5:1 ratio, increases the catch by 1.5 times at unspecified low dose rates, whereas, a phenol mixture, 4-methylphenol, 4-ethylphenol, 4-propylphenol and 2-methoxyphenol in a 100:5:5:1 ratio, increases the catch by 1.4 times at a higher doser rate.

For control purposes, a full agreement has not been reached amongst field workers on whether any odours can be recommended. The most effective combinations are probably either a mixture of 3methylphenol and octenol in a 3:1 ratio or a mixture of 3-methylphenol, 4-methylphenol and octenol in a 1:1:2 ratio. Cattle urine with octenol has been recommended in Ethiopia.


(c) G. fuscipes
The search for odour attractants for this species has so far been unsuccessful, with the exception of carbon dioxide. Various ketones, octenol, urine and phenols have all been tested but results have been inconsistent. In Kenya, acetone has sometimes increased the catch, and at other times appeared repellent; in one trial in Uganda, acetone at 500 mg/h doubled the catch. However, no odours can yet be recommended for control purposes.

fusca species

(a) G. medicorum
An artificial solution containing eight of the phenolic constituents of urine at 10-fold natural concentration, release rates unspecified, gives a 2.9 fold increase in catch of this species in biconical traps. 3methylphenol dispensed from a polythene sachet will increase catches by about 2.8 times.

(b) G. longipennis
On their own, acetone and cow urine have relatively little effect on the catch. Used together at release rates of 150-500 mg/h and 1000mg/h respectively, they increase catches by 2.5-4 times.

Octenol is a very effective attractant for this species, and when used in combination with acetone, gives a further increase of 2-4 times. Phenols, 4-methylphenol with 3-n-propylphenol, are as effective as the cow urine, possibly more so when the octenol is not present.

(c) G. brevipalpis
The only odour attractant so far known to be effective is acetone, dispensed at fairly high release rates of 500 mg/h and above. Neither cow urine nor octenol increase the catch, either alone or in combination with acetone. Further research is needed before odours can be recommended for this species.


Factors affecting efficacy

Rate and method of dispensing
It is clearly important to use the optimal release rates of the different chemicals. These may vary geographically and seasonally. Moreover it cannot be assumed that a dispenser that gives the correct release rate in one locality, will do so in all situations, especially in the case of acetone. Hence the need to test and compare individual and combinations of odours in order to identify the most effective for local conditions and species.

Some odour baits become repellent at high release rates, e.g. cow urine and octenol for G. pallidipes. Butanone is effective at low release rates but less effective at higher ones.

There is very little information available on the optimum site for the release of odours. Within 30 cm of the trap is probably best. Putting them inside the trap may reduce effectiveness.

The optimal position for the odour may vary in relation to wind direction, but other than for research purposes, this information cannot be used practically since wind direction will change frequently over a period of time.


Male and female flies do not always respond to odours in the same way. For G. pallidipes in both Kenya and Ethiopia, the acetone/cow urine combination increases the catch more for females than for males. Acetone and octenol, on the other hand, increase the male catches more.

Physiological state
The efficacy of some odours, e.g. acetone and cow urine for G. pallidipes, may change substantially through the day. This is probably because the activity of flies of different hunger stages also similarly changes.

For G. pallidipes in Kenya, indices of increase are much higher in the middle of the day, when mainly high-fat flies are active, than in the morning or late afternoon when low-fat flies also become active. Similar variations have been noted in Ethiopia, where indices of increase for acetone and cow urine are highest between 15.00 and 18.00 h.

Environmental factors
Since temperature affects both the release rate of the chemicals, and the categories of the fly population active at any particular time of day, this must be taken into account before assuming any direct effect of temperature on the flies' responses. Differences of indices through the day in relation to temperature, with higher indices of increase with increased temperatures, probably result from the temperature influencing the categories of flies which are active.

Wind speed has been suggested as one of the reasons why odours proved less effective for G. pallidipes in Somalia than in Kenya or Zimbabwe. If the wind speed and consequent turbulence is high, it may prevent the establishment of an odour plume.

Trap visibility may also be a factor. In Burkina Faso, acetone and octenol increased catches of G.m. submorsitans by only 2.8 times in the dry season but by 6.7 times during the rains. The situation is similar in The Gambia. This difference may result from the lower visibility of traps during the rains when the grass is very high; in such conditions, odour may become relatively more important in enabling the fly to find the trap.


Table of Contents

Other Chapters


Comments: AGA-Webmaster