Systems

Systems for Collectors

There are many practical choices for collecting flies caught by traps (netting cages, wire mesh frames, plastic bags and/or bottles, and various combinations), but there are few experimental data to support the choice of one system over another. Many researchers do not report these details, so it is difficult to provide advice on a best system. In general, people appear to simply adapt convenient materials to achieve a smooth exit from the cone, and to minimize escape from the collecting chamber. Below, I provide a summary of common systems of collection, with insights from a few experiments - published and unpublished.

Tabanid Traps in Temperate Environments

Manitoba Trap Collector Traps set for tabanid surveys often use a wide-mouthed funnel inserted into a large plastic jar as a collector. The system illustrated here at the top of a Manitoba trap was made from a 2-litre plastic juice bottle (35 mm diametre exit hole) and a 4.8 litre Snapware plastic jar. This brand has a convenient double-lid for insertion into the trivet at the top of the trap.

Insecticide is sometimes added (e.g. a small piece of Vapona® containing dichlorvos, an organophosphate insecticide). Although insecticide helps to prevent escape, it may also act as a repellent. This possibility is not well-documented (Hargrove, 1977).

The jar generates high temperatures in bright sunlight, and hence flies die quickly and accumulate at the bottom. This is not the best system for identifying specimens. When insecticide is not used, special care is also required to remove the jar quickly to not lose any flies.

In drawings or photographs of traps in publications, exit holes appear to be a few cm in diametre, but exact sizes are rarely provided. Below are some examples:

	Box Trap for Greenhead Control - a wire cone with an 89 mm base and a 13 mm exit hole, offset 5-7 cm from the edge of one corner of each trap tier, and inserted into a round plastic container, 10 cm deep and 25 cm wide (Hansens & Race, web document).
	2-Tier Box Trap for Chrysops - a funnel with a 30 mm exit hole running into a 30-cm long tube, inserted into the top of a 2 litre bottle, suspended vertically at the side of the trap (French & Hagan, 1995)
	Canopy Trap for Tabanids - a funnel with a 196 mm base with an exit of about 25 mm (approximate size provided by Hribar), inserted into a large vertical jar at the apex of the trap cone (Hribar et al., 1991). This "cone in a large jar" collector is typical of many tabanid traps from pictures in numerous papers.
	Manitoba Trap for Tabanids - a "cone in a large jar" collector with a 30 mm exit hole in a cone that is 60 mm high, with a 86 mm base (Thompson, 1969)
	Malaise Trap for Tabanids - a wire cone with a 15 mm exit hole leading into a 1.4 litre container (Anderson et al., 1974)

Unpublished Experiments in Canada

These experiments are from Russell, an area with small horse flies (~ 11-17 mm).
These results may not apply for much larger species.

There may be a benefit in adding an extension to the exit to prevent escape, especially with wide-mouth funnels. I tested this with a "chimney" inserted into the funnel. It was made from the inverted top of a plastic water bottle; this also reduced the limiting exit size to 25 mm. For 28 days, I monitored the catch in a Manitoba trap, alternating use of the chimney. At a mean of ~ 5 tabanids per day (max ~ 30), catches of Tabanus and Chrysops spp. were almost the same with or without the chimney.

Funnel Tops In two experiments (one during the tabanid season, one during the stable fly season), I compared catches in Nzi traps for exit sizes of 35 mm (2-L juice bottle) or 21 mm (2-L soda bottle). This experiment used a chamber + netting sleeve collector as shown in Collecting Systems for the North. Catches did not differ significantly by exit size with N=20 per test for both tabanids and stable flies (ANOVA). The smaller hole caught slightly more flies of all types, so I arbitrarily adopted a 21 mm exit hole as a standard for trapping at my home in Russell.

For stable flies, I have also tested the configuration at left. The objective was to compare a large exit (35 mm) with a long neck (additional 50 mm extension) to a small exit (21 mm) without an extension. Catch differences were just significantly different at a mean catch of ~5 stable flies per day (max 18, N=24 per treatment, ANOVA). The use of this system resulted in a 36% reduction in the catch of stable flies relative to the use of a plain soda bottle top. As before, there appears to be a benefit in keeping the exit size small to prevent any escape back down into the trap body.

Tabanus quinquevittatus 11k

Tabanus quinquevittatus (11-13 mm) crawling up through the 21 mm - wide neck of a bottle-top funnel.

With the use of an exit just larger than the size of most flies, tabanids typically enter the collecting chamber by crawling through, rather than flying through. Horse flies spend only a minute or two flying about before finding the top exit. They often bang repeatedly against the funnel before crawling up and out. Deer flies behave the same way but spend much more time resting inside the cone. They also tend to focus their activity on corners and edges instead of just focusing on the top of the cone. Stable flies may also take several minutes to exit as they only gradually work their way to the top of cone. Mosquitoes spend the most time of all flies resting on surfaces inside the cone; I suspect that many may simply die and fall onto the grass without being counted when traps are checked in the evening. Black flies spend most of their time crawling inside the cone.

Once in the top chamber + sleeve collector, flies constantly fly around and crawl about on all surfaces. The use of a long, transparent netting sleeve or some other diversion system is important as it concentrates activity away from the point of potential exit.

In areas with large flies, the cone and the collecting system can become clogged when the exit hole is small, or when the trap is not well-stretched, e.g. when the cone is creased or floppy. For these situations, a wide exit (e.g. 35 mm), and the use of bottle/bag collectors (Tropics) are more practical than the small funnel - chamber - sleeve collecting system illustrated here, and in more detail in the North.

Tsetse Traps in Africa

The many systems for tsetse are illustrated in the book of Cuisance (1989), and in the Electronic Training Resources section of the PAAT Information System (Program Against African Trypanosomiasis) at FAO in Rome (Volume 4, Chapter 3 of the Training Manual for tsetse control). Practical information is also available in French at La Maladie du Sommeil. The Natural Resources Institute in the UK has posted a particularly well-illustrated guide in their overview of Bait Technologies.

In West Africa, researchers use relatively small exit holes for riverine tsetse. Collectors are typically made from a large wire cone supporting a rectangular cage, with exit holes between 10 - 20 mm in diametre. For savannah tsetse such as Glossina pallidipes in East Africa, the FAO training manual recommends an exit size of 13 mm, fashioned by punching a hole in a bottle cap. I used a flat washer with an inner diametre of 15 mm in early work on biting flies, e.g. for catching stable flies with Vavoua traps (Mihok et al., 1996). Currently, many researchers use plastic water bottles to make exit funnels and collectors for traps. For example, I used a 1.5 litre water bottle with a 21 mm exit hole and a 27 mm neck during development of the Nzi trap. In the late 1990's, the Epsilon trap from Zimbabwe was supplied with a bottle-top that had a 24 mm exit hole and a 17 mm neck. This funnel was inserted into a bottle to form a vertical exit "chimney". The chimney led into a final small angled bottle with an attached plastic bag.

Experiments in Africa

The use of mostly small exit holes for tsetse in Africa, and mostly large exit holes for tabanids elsewhere, has evolved with little documentation as to why people have chosen one option over another. Most of the earliest traps for tsetse actually used large and complicated collectors; these were shown to be unnecessary by Hargrove (1977). Practical traps were developed in the 1980's by various groups, but no parallel work seems to have been conducted on optimizing collectors. Altogether, there appears to be only one modest test of various collectors by Vale (1982) in Zimbabwe, with a focus on Glossina pallidipes and Glossina morsitans.

Vale (1982) simultaneously tested exit options with an experimental trap with four exits. The smallest exit hole was large by today's standards (50 x 50 mm). Flies left a "cone" in the experimental trap and entered into a final collector that was quite large (500 x 250 x 250 mm).

Size (mm) Tsetse Stable Flies Non-biting Muscids Tabanids

50 x 50 226 459 917 28

71 x 71 208 536 989 23

100 x 100 136 379 1024 21

141 x 141 124 476 1117 25

Based on these results, and a follow-up experiment on the shape of the exit hole, he suggested an optimum size would be 50 x 50 mm or 50 x 100 mm (the larger size was suggested to avoid clogging of the exit by butterflies).

Vale (1982) also tested the utility of adding a small extension chimney to a 50 x 100 mm exit hole (40-120 mm long), but found no change in tsetse catch with this added feature. Lastly, he manually released tsetse into a cone-collector system. He documented that only about 85% of the flies were retained after one hour with a 50 x 100 mm exit hole.

I conducted a similar test of exit hole sizes at Nguruman in Kenya (unpublished) with a Nzi trap over a period of eight days. The trap was equipped with an exit at the top of the netting cone made from the middle portion of a 1.5 litre plastic water bottle with the top/bottom removed. This large plastic "chimney" was inserted into a large tetrahedral plastic bag collector. On alternate days, I inserted the original plastic bottle top into the chimney to make a small exit funnel. Hence, on alternate days, flies exited the trap using either a large (73 mm) or a small hole (21 mm). The catches were:

Exit Glossina pallidipes Glossina longipennis Stable Flies Non-biting Muscids

Small 1712 144 783 140

Large 803 191 227 95

Ratio S:L 2.13 0.75 3.45 1.47

Stable flies were mainly Stomoxys niger niger and Stomoxys n. bilineatus. Very few tabanids were caught. Trends were similar for males and females.

Given the large differences in catch, there is considerable potential for escape back into the trap when a very large exit is used (73 mm). This appears to be important for stable flies and for the savannah tsetse G. pallidipes. In contrast, the catch may improve for very large flies, such as the fusca tsetse G. longipennis. More work clearly needs to be done to appreciate the subtleties of fly behaviour inside the confined space of the trap cone and the collecting chamber. A simple option for monitoring fly behaviour would be to use insecticide in the collector to retain flies (so long as the insecticide does not have a repellent effect), or alternatively, to coat the inside of the collector with a sticky material.

Amsler & Filledier (1984) examined catches of tsetse and tabanids in Burkina Faso using collectors of roughly similar size, but of different form, in a well-replicated series of Latin Square experiments with two traps (biconical and monoconical). They compared a plastic funnel + bottle collector to a traditional wire cone + frame covered with netting. The bottle was attached to the top of the funnel horizontally. It was a 1.5 litre bottle with an exit of about 25 mm (Amsler, personal communication). The frame was rested vertically on the wire cone. Statistical analyses revealed some differences in catch, but these were not large. Some total catches are presented below.

Trap Collector Glossina tachinoides Glossina morsitans Tabanids

Biconical Bottle 1177 426 163

Cage/Net 1635 415 102

Ratio B:C 0.72 1.03 1.60

Monoconical Bottle 980 319 139

Cage/Net 958 256 90

Ratio B:C 1.02 1.25 1.54

Conclusions

From the few data available, one should obviously not take for granted that all collectors are equivalent. It is also clear that systems are not 100% efficient in retaining every fly that enters. Common sense should be used to choose practical collectors, keeping in mind the potential for both entrance and exit. If there is a general rule, it might be simply to keep the exit hole small rather than large.

References

Amsler, S. & Filledier, J. (1994) Comparaison de différents systèmes de collecte avec deux types de pièges pour la capture des glossines et des Tabanidés. Revue d'Élevage et de Médecine Vétérinaire des pays tropicaux 47, 387-396.

Anderson, J.R., Olkowski, W. & Hoy, J.B. (1974) The response of Tabanid species to CO₂-baited insect flight traps in northern California (Diptera : Tabanidae). Pan Pacific Entomologist 50, 255-268.

Cuisance, D. (1989) Le piégeage des tsé-tsé. Etudes et Synthèses de l'I.E.M.V.T. 32, 1-172.

French, F.E. & Hagan, D.V. (1995) Two-tier box trap catches Chrysops atlanticus and C. fuliginosus (Diptera: Tabanidae) near a Georgia salt marsh. Journal of Medical Entomology 32, 197-200.

Hargrove, J.W. (1977) Some advances in the trapping of tsetse (Glossina spp.) and other flies. Ecological Entomology 2, 123-137.

Hribar, L.J., LePrince, D.J. & Foil, L.D. (1991) Design for a canopy trap for collecting horse flies (Diptera: Tabanidae). Journal of the American Mosquito Control Association 7, 657-659.

Mihok, S., Maramba, O., Munyoki, E. & Saleh, K. (1996) Phenology of Stomoxyinae in a Kenyan forest. Medical and Veterinary Entomology 10, 305-316.

Thompson, P.H. (1969) Collecting methods for Tabanidae (Diptera). Annals of the Entomological Society of America 62, 50-57.

Vale, G.A. (1982) The trap-oriented behaviour of tsetse flies (Glossinidae) and other Diptera. Bulletin of Entomological Research 72, 71-93.