Livestock Research for Rural Development 28 (2) 2016 Guide for preparation of papers LRRD Newsletter

Citation of this paper

Evaluation of sampling devices for Glossina species in the Tarangire National Park- Northern Tanzania

M Byamungu and S Matembo

Vector and Vector-Borne Diseases Research Institute
(VVBDI) (Formely TTRI) P.O.BOX 1026, Tanga Tanzania
Tel. 255 – 27-2644572/3


Sampling devices for Glossina species are important in conducting population studies. Based on previous findings on trapping device performances in Tanzania, experiments were initiated to compare the existing devices in capturing the morsitans group of tsetse. Five different trap types including S3, Nzi, Biconical, Pyramidal, and Epsilon were evaluated for their effectiveness in capturing Glossina species in the Tarangire National Park. A total of 25 traps (five per trap type) were compared without odor attractants relative to S3 trap as a standard in a Latin Squares of 5x5x5 (treatment x site x days). The experiments were repeated twice in the rain and dry seasons. The only tsetse species found in the study area were Glossina swynnertoni and G. pallidipes.

Results indicated that there was a significance difference in catches between traps and best traps were S3 and Nzi compared to other trap types. The comparison of catches with seasons showed a significance difference in fly catches  more flies were captured during the dry season than in wet season. There was no statistical significant difference in catches between sex. Observations showed a statistical significant difference between species, G. swynnertoni being captured more compared to G. pallidipes in dry and wet seasons. The Nzi trap was recommended for use in sampling of Glossina species (morsitans group) in the Northern Tanzania due to its design and cost.

Keywords: Glossina, sampling, season, traps


Adult tsetse sampling is done to assess the apparent density of populations of Glossina species in an ecological habitat. The surveys could be longitudinal to assess the population dynamics and seasonal variation or cross-sectional to study their distribution; determine the trypanosome infection rates and the effectiveness of control measures. Tsetse population studies depend on the suitable designs of sampling devices (traps and targets) to capture the representative sample of fly population. The proportion of captured flies depends on the behavioral status of the particular Glossina species (Vale 1982). Moreover, the number of flies caught by any sampling method depends not only on the density of the population but also on the activity of individual flies. Fly activity, hence availability for capture, are conditioned by environmental and physiological factors (Hargrove 1990).

Studies have shown that tsetse flies (Glossina spp) behavior when they approach a trap tends to fly around them and depart before they enter the body of the traps or land on the outside surfaces (Green 1986). Therefore, traps with entrances on all sides might capture these flies and thus increase the effectiveness of the trap. Different designs of traps and targets have been developed for use against tsetse species in particular environments; they differ in cost, maintenance requirements, and are not equally effective for all species. Traps and targets developed in West Africa, Zimbabwe and with their derivatives developed in other parts of Africa such as NGU trap (Brightwell et al 1987), Nzi trap Mihok (2002) and S3 trap Ndegwa Mihok (1999) share a large number of features in common. This reflects the fact that the designs are based upon the principle of exploiting natural patterns of fly behavior (Vale 1982; Flint 1985). G. swynnertoni and G. pallidipes are potential vectors of human (HAT) and animal trypanosomiasis (AAT) in northern Tanzania and affects livestock production and threatens the health and livelihoods of local communities. Ndegwa and Mihok (1999) developed the S3 as an effective trap in capturing G. swynnertoni but studies by Kitwika (IAEA Report 2004) in northern Tanzania indicated that the pyramidal trap was better for trapping G. swynnertoni than S3 trap. Based on these findings, the Tsetse and Trypanosomiasis Research Institute (TTRI) initiated experiments to evaluate some of the traps which are used for riverine and savannah tsetse so as to come up with the simple to make and most cost effective trap for sampling G. swynnertoni and other morsitans group of tsetse.

Material and Methods

Description of the study area

Experiments were conducted in selected habitats for tsetse flies in the Tarangire National Park northern Tanzania (04 13'S and 035 45'E) (Figure 1) with the purpose to identify the best trap for capturing G. swynnertoni and other savannah species. The park is within a discontinuous G. swynnertoni belt that stretches from Tarangire through Manyara National Park to the Serengeti National Park. G. swynnertoni lives in association with G. pallidipes and small populations of G. brevipalpis and G. longipennis. The park experiences bimodal rainfall pattern and has a varied landscape consisting of open grassland with acacia trees and baobab trees, bush steppe, forests and swamps. The temperatures are moderate during the wet season (November to May) the afternoon temperatures are usually around 29C and at night at around 17C. During the dry season (June to October) temperatures are close to 25C and at night close to 14C.

Figure 1. The study area and trapping sites, insert map is Tanzania
Description of the traps

Five traps types made from local cotton material royal blue, black and netting material were compared in the Latin square experiments; the S3 (Ndegwa and Mihok 1999) develop in Kenya for trapping Glossina swynnertoni. The pyramidal trap (Goutex and Lancen 2986) a modification of the biconical trap for riverine species such as Glossina palpalis in West Africa and the blue biconical trap (Challier et al 1977). Others were the Nzi trap (Mihok 2002) developed in Kenya for savanna species of tsetse and effective for stable flies and horseflies. Epsilon trap (Hargrove and Langley 1990) developed in Zimbabwe for capturing savannah species such as Glossina pallidipes and Glossina morsitans.

Experimental design

The VVBD technical staff conducted experiments in February and June 2010. A total of twenty five traps were incorporated in randomized Latin square arrangement of 5 traps x 5 sites x 5days to give a total of 25 daily replicates per season. The traps were rotated everyday for 5 consecutive days to the next randomized position, so as to test each trap design at every site. This also helped to separate the trap positions and day effects from the treatment effect. The experiments were repeated after the rainy and dry seasons and trap positioning was 150-200 metres  apart depending on the vegetation type. The traps operated for 24 hours on each day damaged traps were replaced and catches were collected, sorted according to tsetse species and sex. Identification of tsetse species based on morphological features as outlined in tsetse classification keys (FAO 1982).

Data analysis

Data were analyzed using a SAS program (SAS 2000) where mean catches of different trap types were compared and summarized by Analysis of Variance (ANOVA). Independent variables were traps and seasons and they showed a significance effect on the outputs (tsetse catches) at P>0.05.


Statistical analysis of fly catches

Five trap types were used in the experiments to test their performances in capturing Glossina species in wet and dry seasons. Trap deployment was done in the selected tsetse habitats in five sites within the park.

Table 1 shows the Analysis of Variance (ANOVA) of traps, season, species and sex on tsetse catches. Of all the effects only trap, season and species had statistical significant difference (P< 0.05).

Table 2 shows the least squares mean fly catches of traps, season, sex and specie. Statistical evidence shows that neither sex nor its interaction with season had effect on fly catches (P>0.05). Both interactions between trap and specie, season and specie influenced fly catches (P<0.05).

Table 1. ANOVA showing the effect of trap, season, sex on fly catches in Tarangire National Park
Source DF SS Mean Square F Value P > F


2.19x104 5.5x103 7.17 0.0005**


1.1x104 1.1x104 14.51 0.0008**


3.3x104 3.3x104 42.77 <.0001***


1.56x102 1.56x102 0.2 0.6556NS


4.69x102 4.69x102 0.61 0.4410 NS


1.5x104 3.79x103 4.94 0.0042**


1.09x104 1.09x104 14.34 0.0008**
** Significant, *** highly significant, NS not significant (P>0.05)
Comparison of fly catches

The overall fly mean catches for five trap types in all seasons in this study was 35.72 6.03. For the individual traps the mean trap catches were 75.12 9.78 for S3, 49.8 9.78 for Nzi, 22.8 9.78 for Pyramidal, 15.5 9.78 for Epsilon and 15.25 9.78 for Biconical trap (see table 2). There was a significant difference between traps (P=0.0005) S3 trap had the highest score catching 42.05% of all flies followed by Nzi (23.7%). Considering trap specie specificity in fly catch; Nzi caught 42.25% of the G.pallidipes and S3 caught 93% of the G. swynnertoni. Fly catches for seasons were 19.05 6.19 wet and 52.4 6.19 dry seasons respectively, the latter having the highest fly catches than the previous. G. swynnertonni being the most specie caught in both seasons and the statistical analysis supports the observed difference (P=0.0008). Two specie; G. pallidipes and G. swynnertoni were caught, 90% of the flies caught were G. swynnertoni, the remaining being G. pallidipes. Female flies caught were slightly higher than male flies and there was no statistical evidence to support the difference (P=0.6556).

Generally tsetse catches varied with trap, season, and specie in Tarangire National Park and trap differences had significant (P<0.05) effect on the tsetse catches in the study area. Similarly seasons and specie differences had significant effect (P < 0.05) on the number of caught flies. When interaction between variables were considered; trap *specie and season* trap had significant effect on the fly catches. However, the interaction between season and sex had no significant (P > 0.05) impact on the tsetse catches in the study area.

Table 2. Showing Least Means ( SE) of the trap, season specie and sex on fly catches in Tarangire National Park
Class Level Catches ( SE)
Trap Biconical 15.25 9.78ab
Epsilon 15.5 9.78ab
Nzi 49.8 9.78ab
Pyramidal 22.8 9.78ab
S3 75.12 9.78ab
Season Dry 52.4 6.19ab
Wet 19.05 6.19ab
Specie Glossina pallidipes 7.10 6.19ab
Glossina swynnertoni 64.35 6.19ab
Sex Females 37.7 6.19aa
Males 33.75 6.19aa
Means with the same superscript are not significant different (P>0.05)

The table shows that the S3 trap had the highest means compared to Nzi, biconical, epsilon and pyramidal traps. Dry season had significantly higher mean fly catches (52.4 6.19) compared to the wet season catches (19.05 6.19). Similarly G. pallidipes had significantly lower mean fly catches (7.10 6.19) than Glossina swynnertoni which had higher mean fly catches (64.35 6.19). However, the differences on sex of flies had no significance difference on mean fly catches in the study area.


Trap selection for tsetse sampling should consider the convenience and cost-efficacy of the available traps. The S3 trap was originally designed in Kenya by Ndegwa & Mihok (1999) for capturing Glossina swynnertoni. However, work by Kitwika (2004) in northern Tanzania on the comparison of the tsetse traps revealed that the pyramidal trap was effective than the S3 in capturing G. swynnertoni. These findings together with the need of Institute to study the Glossina populations in Tarangire National Park lead to the initialization of the experiments to evaluate other traps types for capturing the G. swynnertoni and probably other morsitans group of tsetse. In the studies conducted in the Tarangire National Park in two seasons, the S3 trap and Nzi trap performed better compared to other three trap types (epsilon, biconical and pyramidal), the S3 trap being the most superior device for total catches. Similar findings were obtained by Ndegwa & Mihok 2009. The effectiveness of S3 trap probably was due to the three entrances giving more chances for the flies to enter into the trap this was also observed by Goutex & Sinda (1990) while doing trials on various models of the pyramidal trap. Other traps: epsilon biconical and pyramidal had few fly (G. swynnertoni and G. pallidipes) catches in all seasons. Poor performances observed to epsilon trap which was designed in Zimbabwe specifically for savannah tsetse could be due to ecological differences (Kuzoe & Schofield 2005). The poor performances of the biconical and pyramidal in this case could be explained by the fact that these traps were designed for riverine Glossina species. Therefore, such performances could be also due to the ecological differences. The higher mean catches in dry season could be due to the fact that during this period flies tend to concentrate in refuges and they are readily available to traps. Generally as was observed by Vale and Hargrove (1979) there might be natural factors like weather and physiological factors which might have influenced the trap catches.



I acknowledge my institute the Vector and Vector-Borne Disease formally TTRI for funding this study. I also acknowledge Mr. Deusdedit Malulu for data analysis and appreciate the inputs by Mr. H. Nyingilili, Dr. O. Manangwa, Mr. Mashenga and Mr. Kiimbisa. I extend my gratitude to the Tarangire National Park Management and the GIS unit.


Green C H 1986 Effects of colours and synthetic odours on the attraction of Glossina pallidipes and G. morsitans morsitans to traps and screens Physiological Entomology, 11, pp. 411-421

Ndegwa P N and Mihok S 1999 Development of odour-baited traps for Glossina swynnertoni (Diptera: Glossinidae). Bulletin of Entomological Research 89, pp.255-261.

Brightwell R, Drans field R D, Kyorku C, Golder T K, Tarimo, S A, and Mungai D 1987 A new trap for Glossina pallidipes Tropical Pest Management 33, pp. 151-189

Challier A and Laveissiere C 1973 Un nonveau piege la capteur des glossines (Glossina: Diptera Muscidae) Description et essais sur le terrain. Cahiers ORSTOM. Serie Entomologie Medicale et Parasitologie 11, pp. 251-262

Hargrove J W and Langley P A 1990 Sterilizing tsetse in the field - a successful trial Bulletin of Entomological Research 80, pp. 397-403

IAEA Report 2004

Goutex J P and Sinda D 1990 Community Participation in the control of tsetse flies large scale Trials using Pyramidal trap in Congo. Tropical Medicine & Parasitology 41, pp. 49-55

Kuzoe F A S and Schofield C J 2004 Strategic Review of Traps and Targets for tsetse and African Trypanosomiasis control. (UNICEF/UNDP/World Bank/ WHO) Special programme for Research and Training in Tropical Diseases (TDR) pp. 1-20

FAO 1982 Tsetse control training manual, Ecology and behaviour of tsetse. Food and Agriculture Organization of the United Nations, pp.101

SAS 2000 Stastical Analysis System. SAS/STAT users guide. Statistical Analysis Institute, Inc; Cary, Nc. USA.

Vale G A 1982 The improvement of traps for tsetse flies (Diptera: Glossinidae) Bulletin of Entomological Research 72, pp. 95-106

Vale G A and Hargrove J W 1979 A method for studying the efficiency of traps for tsetse flies (Diptera: Glossinidae) and other insects. Bulletin of Entomological Research 69,  pp. 183-193.

Received 2 December 2015; Accepted 6 January 2016; Published 1 February 2016

Go to top