Biology, images, analysis, design...
Clegs etc Ticks Mossies Tsetse
"It has long been an axiom of mine that the little things are infinitely the most important" (Sherlock Holmes)



By simple search and observation

Horseflies - biting flies which feed on horses, cattle, and sometimes man, are an important part of our biodiversity, affect the distribution of grazing animals, and may transmit disease.

Whether your aim is to reduce the number of horsefly bites suffered by people or livestock, or to find out what species are present as part of a biodiversity study, you need to be able to catch them. The simplest way to catch tabanids is to walk through their habitat, preferably on a hot sultry day, and see what turns up. Tabanids can (occasionally) be observed just resting on vegetation, such as the male common cleg Haematopota pluvialis  shown below.

Both males and females of many species of tabanids can also be found at flowers since they regularly feed on nectar (Hayakawa (1986),  Kniepert (1980) ). Some species may also take moisture and/or salts from damp soil (see Raymond (1982) ). The picture below is of a band-eyed brown horsefly Tabanus bromius  doing this - one of several along a woodland track.

Nectar and other sources of sugars provide the only source of energy for male tabanids, but females employ an additional food source - blood.


By attraction to a live host

In summer your most likely encounter with a horsefly will be a female fly trying to bite you and take a bloodmeal. This is especially true for Chrysops (deer flies) and Haematopota (clegs).

They can often be caught while feeding with a simple swat - which is the way non-entomologists tend to catch horseflies!

The red colour in the picture above is the author's blood, generously donated to obtain interesting photos of a common cleg (Haematopota pluvialis) feeding. A simple hand-net makes this sampling method more efficient, and also provides specimens in somewhat better condition for subsequent identification...

Contrary to what you might feel on such occasions, most tabanid species are not strongly attracted to man. The most frequent hosts are cattle, horses and deer, but most tabanids are opportunistic and will also feed on a wide range of hosts including dogs and even large reptiles. The first picture below shows the common Tabanus bromius feeding on a cow.


The second picture above shows the author hand-netting for tabanids around horses in West Sussex, UK.

Hand-netting around the host, whilst effective, has some important limitations:

  1. Catches depend very much on the skill of the person using the hand net.
  2. The host may react strongly to being bitten, or to your catching / photographic activities.
  3. They can be problematic for surveying or regular monitoring...

Some of the limitations of hand-netting can be overcome by a rather more high-tech way to catch biting flies around livestock - that is the use electric nets around a live host. This method was pioneered in Zimbabwe and was then used in several parts of Africa to investigate the behaviour of another group of biting flies tsetse  around both hosts and traps.

The picture above shows an experimental set-up in Kenya where we were estimating the number of biting flies, mainly Glossina (tsetse) species, attracted to a single cow. Below we can see where electric nets were used to surround a small group of cattle in Galana Ranch, Kenya. It is assumed the nets knock-down every insect they intercept, and none fly above them, and that they do not otherwise affect the insects' behaviour.

Catching tabanids around the host (whether with a hand net, sticky trap or with electric nets) is the only way to be certain about which species are approaching the animal. - More sophisticated methods are required to find what proportion of those insects actually feed.

Unfortunately electric nets are not very portable, or rain-proof, or robust, or otherwise easy-to-use - even for experts.


Visually attractive devices and traps

A popular alternative is to use a visual attractant, or a trap. The earliest designs attempted to mimic animals (for example see Morris (1963) ). More recent devices tend to make use of the fact that many biting flies are attracted to large single-coloured areas. There are innumerable such designs, such as carrying a blanket over a pole, through sticky-coated panels, to 3-dimensional cloth traps of varying complexity utility and cost. Some methods also incorporate movement which is strongly attractive to some tabanids - see the trolling trap of Mizell et al. (2002). 

  • Black blanket with hand net

    Many such devices were originally developed for tsetse flies (Glossina) owing to their economic and medical importance (sleeping sickness). The picture below shows people hand-netting tsetse flies as the flies are attracted to a black blanket. Fortunately these devices are often very effective for tabanids, a point first noted by Morris (1963). 

    Other devices, detailed below, do away with the need for the hand net by capturing the flies in a variety of ways.


  • Coloured panel coated with a non-setting adhesive

    Sticky panels are cheap, simple to erect and simple to replicate, and have been used to trap both tsetse and tabanids (see Hall et al. (1998) ).

    This white sticky panel was used to sample coastal tsetse (Glossina austeni) in Kenya. Black panels are sometimes used for tabanids.

    Sticky panels are not much loved by entomologists - partly because they damage / kill the specimens, and may render them impossible to identify, but also because nonsetting adhesives are very unpleasant to handle. In contrast, "sugaring" for butterflies and moths (by painting a mixture of molasses, amyl acetate, and beer on trees) uses the resulting odour to attract the insects, but they are not intended to get stuck to the bait.


  • Water-trap

    Water-traps are popular among agricultural entomologists for trapping many different kinds of flying insect. They usually consist of a simple enamelled metal tray (approx 30 by 40 cm, and 5 cm deep) filled with water, plus a little detergent and set about 45 cm above ground. Instead of detergent the surface can be covered with paraffin oil, as was done in the trap shown below. Trays of larger area catch more flies, but are difficult to handle. Whilst water traps are moderately labour intensive and inefficient (when used without odour bait) they are comparatively inexpensive and easy to erect. They enable you to trap in many locations simultaneously, in all manner of sites, and can be operated 24 hours a day for extended periods.

    One might expect that black water traps would be optimal for biting flies, and indeed that is the case for the riverine tsetse, Glossina tachinoides. But for the savannah tsetse species Glossina morsitans submorsitans significantly more were attracted to white water traps (Dransfield et al., (1982). ) In Nigeria in 1980 we used arrays of 3 black and 3 white water traps to trap several species of tsetse and tabanids, as shown in the second picture below. It seems likely that as more research is carried out, more differences will be found in the responsiveness of different tabanid species to different visual and odour attractants.


    The humble water trap has recently taken on a new lease of life under the name horizontal polarizing liquid trap. Egri et al. (2013)  describe a circular black plastic tray (with a diameter of 50cm) plus overflow tube, filled with water, and then with vegetable oil poured onto the water. They suggest that male and female tabanids are attracted to horizontally polarized light because these insects find water by the horizontal polarization of light reflected from the water surface (Horváth et al. (2008) ). In addition host-seeking female tabanids are attracted to the linearly polarized light reflected from the coats of hosts. The higher the degree of polarization, the more attractive is the host, independently of the direction of polarization of coat-reflected light (Egri et al. (2012) ).

    Our first report of the efficacy of simple horizontal water traps for catching both tsetse and tabanids (Dransfield & Brightwell, (1983) ) was met with some scepticism, since tsetse flies do not drink water (nor, unlike tabanids, do they oviposit in marshy places). At the time we explained it on the basis of responses to colour (as with coloured panels) with the water+detergent simply acting as the trapping medium, but the resulting shiny surface with a high degree of polarization may have contributed to the high catches we observed. Interestingly Egri et al. (2013)  found that no flies were caught if the traps were raised on packing crates or mounds. We found that traps mounted on 30 cm high wooden stands whose tops were smaller than the water trap being supported, were just as effective as traps on the ground, and were much better protected from grazing animals and wind blown debris. More curiously in Kenya we found very large (2 × 1 m) brown trays at ground-level caught effectively no tsetse, nor indeed anything much else (which is why we used them to collect flies intercepted by electric nets ).


  • Malaise trap

    The Malaise trap is an open-sided tent-like structure with a central vertical panel which reaches down to the ground. Fast-flying insects such as tabanids hit the central panel and, attempting to escape, fly upwards towards a sloping roof that directs them towards a collecting chamber at the high end of the trap. The collecting chamber can be filled with preservative if required. Malaise traps are commonly described as "flight-interception" traps.

    Guest image, Copyright Alexandra Shaw, all rights reserved.

    The malaise trap pictured above is made from black netting, but they are also made from white netting. As such they are intended to be non-selective amongst flying insects, but like all traps, they are biased towards certain insect types. The effectiveness of Malaise traps can be increased by using odour baits,  such as those described below, but then of course the trap will have a different bias.


  • Canopy traps

    Several traps have been designed specifically for tabanids, employing largish black attractive surfaces or objects, combined with some sort of cone and a cage to collect and hold the insects. Probably the most widely used trap is some form of the "canopy trap" (Adkins et al. (1972) ). This consists of a large canopy with a black skirt and a netting apex set about 50-70 cm off the ground. A shiny black spherical visual target is often suspended underneath the canopy. Figures showing the canopy trap are available from University of Missouri Extension.  Other, similar, widely-used traps are the Manitoba trap, and the box trap.


  • NGU traps

    A number of other very effective traps used for tabanids are modifications of the triangular NG2B trap, developed specifically for controlling tsetse flies(Brightwell et al. (1987) ) which also catch large numbers of tabanids. This family of traps include the NG2F trap (Brightwell et al. (1991) ) figured here, plus two variants: the epsilon trap and the NZI trap. We currently use odour-baited NG2F traps for tabanid surveys in UK.

    This design is particularly useful for catching tabanids because, being designed for African farmers to control tsetse, it is efficient and comparatively easy to construct. It is also quite portable, can yield live specimens in good condition, and is reasonably specific to biting insects. In this picture the trap is fitted with a small sampling-cage, but a much larger polythene bag is required if the trap is left operating for long periods of time. In UK we have always used the trap baited with cow-urine and acetone to maximise catches.



Many attempts have been made to improve the effectiveness of visually attractive insect traps - of which only odour-baiting has proven notably successful.

  • Carbon dioxide (CO2 a component of cow breath) whether dispensed from a cylinder or as 'dry ice' can give large increases of all biting flies, but too much is needed to be worthwhile under field conditions.

  • Acetone (as in nail-varnish remover, but a component of cow breath) can yield small-to-moderate increases for some species at comparatively small doses, and being cheap, easy to dispense and readily available it has proven useful. However, acetone does not seem to be effective for all tabanid species.

  • Octenol (1-octen-3-ol, a component of cow sweat) also gives moderate increases for some species at small doses (French & Kline (1989) ). Unfortunately octenol is not especially cheap, nor available, nor easy to dispense.

  • Ammonia (NH3 vapour) gives moderate increases for some species (Hribar et al. (1992) ) but has so far been insufficiently tested.

  • Cow urine (aged for at least 1 week) was first found to be a highly effective attractant for biting flies by Dransfield et al. (1986).  Being freely-available and quite easily-dispensed, it is a popular attractant especially among cattle farmers, albeit very high doses repel. Several active phenolic-compounds have been isolated from cow urine, namely 4-methyl phenol and 3-n-propyl phenol. They can be obtained in plastic dispenser-sachets, sometimes mixed with octenol, but these are neither readily-available nor especially cheap.

Research is still ongoing on odour attractants for tabanids, with a need for more information on the dose response curves for the different attractants and on their effects (or otherwise) in combination with other attractants.


Trap efficiency

Trap efficiency is usually considered to describe the percentage of flies that enter the collecting device of the trap of the total that approach close to it. Unfortunately many of the flies that enter the vicinity of traps do not enter and go on to bite another day.

A conventional way to evaluate trap efficiency is to use electric nets  (Vale (1974) ) to form an incomplete ring around the trap. By counting the number of flies arriving (on the outer sides of the nets) and departing (on the inner sides of the nets), and comparing this to the number actually caught by the trap, one can estimate trap efficiency. The picture below shows an incomplete ring of electric nets around a NG2G trap: brown water-filled trays collect insects stunned by the nets.

For this method to give an accurate estimate one has to make several rather unrealistic assumptions about fly behaviour, in particular that they fly straight to and away from the trap and do not circumnavigate it. We therefore used an alternative approach to estimating efficiency - that is to compare the catch of a trap operating with no electric nets with the number of flies arriving on the outer sides of nets forming a complete ring around the trap (Dransfield & Brightwell (2001) ). A complete ring of electric nets around a biconical trap is shown in the picture below.

Unlike the previous approach, there is no 'internal' control, so a crossover design was used with multiple replicates of each of the two treatments - with and without electric nets.


Tabanid biodiversity survey

We have been using NG2F traps baited with acetone and cow-urine to survey tabanid flies in UK in summer 2014. As well as catching vast numbers of the commoner non-man-biting species (in particular Tabanus bromius), we have found host-seeking females of the very rare and local Atylotus rusticus both within and somewhat outside its usual haunts.

We also caught several female Hybomitra ciureai at Rye Harbour. This species is found in south east coastal areas, and is very seldom reported in Britain. Until now only one (a male) had been recorded at that reserve.


Our sincere thanks go to the following people:

Useful weblinks 


  •  Adkins, T.R. et al. (1972). A modified canopy trap for collecting Tabanidae (Diptera). Journal of Medical Entomology 19, 183-185. Abstract 

  •  Böse et al. (1987). Transmission of Trypanosoma theileri to cattle by Tabanidae. Journal of Parasitology Research 73(5), 421-424. Abstract 

  •  Brightwell, R. et al. (1987). A new trap for Glossina pallidipes. Tropical Pest Management 33(2), 151-159. Abstract 

  •  Brightwell, R. et al. (1991). Development of a low-cost tsetse trap and odour baits for Glossina pallidipes and G. longipennis in Kenya. Medical and Veterinary Entomology 5, 153-164. Abstract 

  •  Dransfield, R.D.. et al. (1982). Population dynamics of Glossina morsitans submorsitans Newstead and G. tachinoides< Westwood (Diptera: Glossinidae) in sub-Sudan savanna in northern Nigeria. I. Sampling methodology for adults and seasonal changes in numbers caught in different vegetation types. Bulletin of Entomological Research 72, 75-192.Abstract 

  •  Dransfield, R.D. & Brightwell, R. (1983). A new sampling technique for studying population dynamics of tsetse flies. pp 17-24 In: Ilemobade, A.A. (ed.) Proc 1st Nat. Conf. Tsetse Tryp. Res. Nigeria, Kaduna, 10-12 August 1981. Abstract 

  •  Dransfield, R.D., Brightwell, R., Chaudhury, M.F., Golder, T.K. & Tarimo, S.A.R. (1986). The use of odour attractants for sampling Glossina pallidipes Austen (Diptera: Glossinidae) at Nguruman, south-western Kenya. Bulletin of entomological Research. 76, 607-619. Abstract 

  •  Dransfield R.D. & Brightwell R., (2001). Trap efficiency for Glossina pallidipes (Diptera: Glossinidae) at Nguruman, south-west Kenya. Bulletin of entomological Research. 91(6), 429-444. Abstract 

  •  Egri, A. et al (2012). New kind of polarotaxis governed by degree of polarization:attraction of tabanid flies to differently polarizing host animals and water surfaces. Naturwissenschaften (2012) 99 407-416. Full text 

  •  Egri, A. et al. (2013). A horizontally polarizing liquid trap enhances the tabanid-capturing efficiency of the classic canopy trap. Bulletin of Entomological Research 103(6), 665-674.Full text 

  •  French, F.E, & Kline, D.L. (1989). l-Octen-3-ol, an effective attractant for Tabanidae (Diptera). Journal of Medical Entomology 26(5) 459-461. Abstract 

  •  Hall, M. J. R. et al. (1998). Use of odour-baited sticky boards to trap tabanid flies and investigate repellents. Medical and Veterinary Entomology 12, 241-245. Abstract 

  •  Hayakawa, H. (1986). Studies on flower-visiting and nectar-sucking of tabanid flies 1. The rate of flies with pollen in each species examined on specimens (in Japanese). Japanese Journal of Sanitary Zoology 37(4): 389-393.

  •  Horváth et al. (2010). An unexpected advantage of whiteness in horses: the most horsefly-proof horse has a depolarizing white coat. Full text 

  •  Hribar, L.J. et al. (1992). Ammonia as an attractant for adult Hybomitra lasiophthalma (Diptera: Tabanidae). Journal of Medical Entomology 29(20), 346-348. Full text 

  •  Kniepert, F. W. (1980). Blood-feeding and nectar-feeding in adult Tabanidae (Diptera). Oecologia 46, 125-129. Full text 

  •  Magnarelli, L.A. & Anderson, J.F. (1980). Blood and sugars are needed to nourish deer flies and horse flies. Frontiers of Plant Science 32(2): 7-8. Abstract 

  •  Mizell, R.F. et al. (2002). Trolling: a novel trapping method for Chrysops spp. (Diptera: Tabanidae). Florida Entomologist 85(20), 356-366. Abstract 

  •  Morris, K.R.S. (1963). A study of African Tabanids made by trapping. Acta Tropica 20 Full text 

  •  Raymond, H. L. (1982). Preliminary notes on the location and behaviour of the males of Tabanidae (Diptera) of alpine pastures of Combes (Hautes-Alpes). Entomologiste 38(1): 21-25. Abstract 

  •  University of Missouri Extension (1996). Protecting cattle from horse flies. Full text