HABITAT PREFERENCES AND SUPPRESSION OF GLOSSINA AUSTENI AND G. BREVIPALPIS IN SOUTH AFRICA.

HABITAT PREFERE ET ELIMINATION DE GLOSSINA AUSTENI ET GLOSSINA BREVIPALPIS EN AFRIQUE DU SUD

J. Esterhuizen¹,  K. Kappmeier¹, E. Nevill¹ & P. Van den Bossche^2,3,*

¹ARC-Onderstepoort Veterinary Institute, Private Bag X05, Onderstepoort 0110, South Africa
²University of Pretoria, Faculty of Veterinary Science, Department of Veterinary Tropical Diseases, Onderstepoort 0110, South Africa.
³Institute of Tropical Medicine, Veterinary Department, Nationalestraat 155, 2000, Antwerpen, Belgium.

            En Afrique du Sud, la trypanosomose animale sévit dans le Zululand, dans la province du Kwa-Zulu Natal. Glossina austeni et G. brevipalpis sont les seules espèces de tsétsé présentes. En vue d’intensifier le contrôle de ces deux espèces, on a étudié leur répartition saisonnière dans les trois pricipaux types d’habitat dans le Zululand en utilisant les pièges-H. L’abondance apparente de G. brevipalpis était très forte dans la forêt indigène et dans les herbages, mais très faible dans les plantations exotiques. C’est la raison pour laquelle G. brevipalpis est beaucoup plus répandue qu’on ne le croyait. Quant à  Glossina austeni, elle était surtout capturée dans la forêt indigène.

            Après l’étude de l’habitat préféré des glossines, on a initié un essai de contrôle des tsétsé avec des écrans à appât olfactif et des écrans traités à l’insecticide. Environ 14 mois après le déploiement des écrans, la densité de G. austeni a diminué de 99% et elle est restée à ce niveau pendant 30 mois. Le contrôle de G. brevipalpis a moins réussi avec une réduction des captures, qui variait entre 60-89%.

            Lorsqu’on a enlevé les écrans, il y a eu une réinvasion  et une reconstitution rapide de la population de G. brevipalpis. En revanche, les captures de G. austeni dans la zone traitée auparavant demeuraient faibles.

            In South Africa, livestock trypanosomiasis occurs in Zululand, Kwa-Zulu Natal Province.  The only tsetse species present are Glossina austeni and G. brevipalpis.  To improve the control of both species, their seasonal distribution was studied in the three main habitat types in Zululand using H-traps. The apparent abundance of G. brevipalpis was highest in indigenous forest and open grassland but was significantly lower in exotic plantations.  Hence, G. brevipalpis was spread more widely than generally thought.  Glossina austeni, on the other hand, was captured mainly in indigenous forest. 

            After the habitat preference study, a tsetse control trial using odour-baited and insecticide-treated targets was initiated.  About 14 months after target deployment, the density of G. austeni was reduced with 99% and maintained at this level for 30 months.   The control of G. brevipalpis was less successful with a reduction in catches of between 60-89%. 

            Removal of the target resulted in a fast reinvasion and recovery of the G. brevipalpis population. In contrast, G. austeni catches in the previously treated area remained low.

Introduction

            In South Africa, trypanosomes are transmitted by the two tsetse species Glossina austeni and G. brevipalpis.  The tsetse-infested area in north-eastern KwaZulu-Natal covers approximately 16,000 km², and is inhabited by about 426 000 humans together with 130 000 small ruminants and about 360 000 cattle mainly in communal farming areas, and about 9000 cattle belonging to commercial farmers.  Due to the agricultural potential of this tsetse-infested area, livestock trypanosomosis constitutes one of the constraints to agricultural development. 

            Although the disease is widespread throughout the Zululand tsetse-infested areas, the incidence of African Animal Trypanosomosis in cattle is relatively low, but is characterised by major disease outbreaks, as was the case in 1990, when Nagana contributed to the deaths of approximately 10 000 cattle. This outbreak was brought under control by dipping of cattle with pyrethroid insecticides, and the therapeutic treatment of infected animals (Kappmeier et al, 1998).  A tsetse control trial with insecticide impregnated targets was also undertaken for localised control of G. brevipalpis, but was unsuccessful due to various factors (Kappmeier et al. 1998).   

             There is evidence that the current prevalence of trypanosomosis has reverted to the high levels of 1990 before these temporary control measures were instigated.  It is envisaged that through area-wide tsetse control interventions much of this constraint in N.E. KZN could be alleviated (Kappmeier et al. 2006). To assist in such area-wide integrated pest management (AW-IPM), the distribution of G. austeni and G. brevipalpis was determined throughentomological field surveys that were conducted from 1993 till 1999 in N.E. KZN (Nevill 1997; Nevill et al. 1995, 1999). The highest densities of flies were found within the game reserves and natural areas (Kappmeier Green 2002). The resulting field data were used to develop a satellite-derived ‘Probability of Presence’ model (AVIA-GIS 2002, Hendrickx et al. 2003). The results of the surveys and the subsequent model clearly indicated that both G. brevipalpis and G. austeni are confined to isolated areas of N.E. KZN.  Both these species are generally considered to be restricted to densely shaded habitats, although G. brevipalpis and to a lesser extent G. austeni, do disperse into open areas adjacent to forested areas (Kappmeier Green 2002).  However, little is known about the relationship between the two species and their immediate habitats.

            Furthermore, small-scale trials were required to evaluate the effectiveness of insecticide-treated targets to control both tsetse species under the conditions prevailing in Zululand. The use of stationary attractive devices, e.g. insecticide-impregnated targets ( Vale et al. 1988) has been promoted in the last two decades as a viable tool to control tsetse, because of the relatively low cost of the materials and the unsophisticated technology.

            The goal is to exert a modest daily mortality of 2-3 % (Hargrove 1988) on the female tsetse fly population by attracting them to a target and inducing a landing response resulting in flies coming into contact with a lethal dose of insecticides applied to the surface of the target.  This principle has been successfully used, amongst other, in Zimbabwe (Vale et al. 1988) and Zambia (Willemse 1991, Van den Bossche 1998).

            The AW-IPM programme against G. austeni on the Unguja Island of Zanzibar highlighted limitations of targets to suppress G. austeni in primary forest habitat (Vreysen et al. 1999), and therefore, an odour-baited insecticide-impregnated target system was developed for the suppression of both G. austeni and G. brevipalpis (Kappmeier and Nevill 1999b) (Figure 3). This black-blue-black target was tested against the two species in a pilot trial, covering an area of 80 km2 in N.E. KZN (Esterhuizen et al. 2001).

Material and Methods.

Study site

            Studies were conducted at the Hellsgate Tsetse Research Station (28°02’ S and 32°25’ E) situated on the Ndlozi peninsula in Lake St Lucia (Figure 1), north-eastern KwaZulu-Natal, where both G. austeni and G. brevipalpis occur. Three major vegetation types occur on the peninsula. Dense indigenous forests occupy a 0.5 to 2.0 km-wide strip along the edge of the peninsula. Open grassland with isolated patches of forest covers the central inland areas. Finally, large plantations of exotic Eucalyptus and Pinus tree species cover the majority of the southern part of the peninsula (Figure 1). A variety of game animals serve as food source for the tsetse flies. Annual rainfall in the research area amounts to a total of about 900mm, with most rainfall recorded in the long, hot and humid summer season from September/October to March/April. The mild and drier winter is from May to August. Mean maximum annual temperature ranges between 24° and 32°C, with mean annual minimum between 12° and 22°C.  Mean relative humidity ranges between 50% and 95%.

Habitat preference studies

            Between May 1999 and May 2000, the tsetse population was monitored along an 11 km transect traversing the three vegetation types.  The transect started at the northern tip of the peninsula in indigenous forest, and proceeded southwards with traps placed every 300 m. A total of 38 H-traps (Kappmeier, 2000, 2001) were deployed, with 17 traps inside indigenous forest, 16 traps in open grassland and 5 traps in exotic plantations (Figure 1). Fly collections were made daily and analysed to produce a monthly Index of Apparent Abundance or IAA (i.e. for each tsetse species the number of tsetse captured per trap per month for each vegetation type). 

Suppression studies

            Between May 2001 and December 2004, a tsetse suppression trial, using insecticide-treated targets, was conducted in 80km² of the Ndlozi peninsula (Figure 2). 

Figure 1.  Location map of Ndlozi peninsula, indicating the research area with the distribution of the three main vegetation types and trap sites. 

Figure 2.  Location map of Ndlozi peninsula, indicating the research area and the three blocks used in the suppression study.

            The trial area was divided into three blocks:  Block A (treatment), Block B (barrier) and Block C (control/untreated).

            Blue and black cloth targets (Figure 3) (Kappmeier & Nevill, 1999) were used, which were immersed into a solution containing 0.8% Deltamethrin (Glossinex 200 S.C.®, Aventis). Targets were deployed using a grid design with gridlines 500m apart. They were placed in the major vegetation types, i.e. indigenous vegetation, open grass areas and exotic plantations, with the majority of targets deployed in indigenous vegetation. Targets were placed at selected densities according to calculations that were made using the estimated population densities of the two species, the target efficiency and the killing percentage (3-4 %) required (Kappmeier Green 2002). At the time of deployment in May 2001, the initial target density was 4 targets per km² in block A and B, with targets placed in  all habitats, including open areas.  In February 2002 target density was increased to 8 targets per km² in block A and B, with targets concentrated in forested and well shaded areas, as well as along the forest edge.  A year later, all the targets from Block B were removed and used to increase the density to 12 targets per km² in Block A in forested areas, the forest edges and isolated forest clumps in open grassland.  Therefore, the target density was increased three times during the study, resulting in three suppression regimes.

Figure 3.  Targets used for suppression of Glossina brevipalpis and G. austeni.

            All the targets were removed in March 2004, and post-suppression monitoring is taking place.

            Nine H-traps were used in each of the three trial Blocks to monitor the tsetse population. Collections from traps were made weekly and catches in each area were expressed as corrected percentage reduction from catches made in the untreated area (Block C).

            In all studies, traps and targets were baited with the SA odour blend (Kappmeier & Nevill 1999), which consisted of acetone (released at ca. 350 mg/h) and a 1:8 ratio of 1-octen-3-ol (ca. 4,4 mg/h) and p-cresol (ca. 15,5 mg/h).

Results and discussion

Habitat preferences

Throughout the observation period of 52 weeks and in all three vegetation types, the IAA of G. brevipalpis was substantially higher than the IAA of G. austeni (Figure 4).

            A total of 4 918 G. brevipalpis were caught during the study, with indigenous forest accounting for 61.6% of the total G. brevipalpis catch, followed by open grassland with 26.6%, and 11.8% in exotic plantations.

            Catches in indigenous forest were significantly higher (P<0.05) than catches in exotic plantations.  Results showed no significant difference in catches between indigenous forest and open grassland (P=0.09), nor between open grassland and exotic plantations (P=0.22). The IAA of G. brevipalpis was highest in the hot months and lowest in the coldest months. This did not differ significantly between vegetation types (P>0.5)

            A total of 1 444 G. austeni were captured during the observation period, with 88.2% captured in indigenous forest.  Open grassland and exotic plantations accounted for 8.4% and 3.4% respectively.  Furthermore, most of the G. austeni captured outside the indigenous forest were captured in traps adjacent to the prime habitat. No significant difference in the monthly average IAA of G. austeni was evident between the vegetation types (P>0.5). 

Figure 4: Monthly mean Index of Abundance of Glossina brevipalpis (A) and G. austeni (B) in indigenous forest (●), exotic plantation (○) and open grassland (▼).  Figure 4C represents the monthly mean maximum (○), minimum (●), average (▼) temperature and rainfall (bars) in the study area.

Suppression study

Glossina brevipalpis

            During the first treatment regime (4 targets per km²) the highest percentage reduction reached for G. brevipalpis in the treatment area (Block A) was 82% (Figure 5A). It increased to 83% after the target density was increased to 8 targets per km² and reached a maximum of 89% after increasing the target density to 12 per km².

     Following removal of all targets from the treatment area in March 2004, there was a rapid increase in numbers of G. brevipalpis in the treatment and barrier areas (Blocks A and B).

Glossina austeni

            Results indicate the efficiency of insecticide treated targets at a density of between 4-12 per km², to reduce populations of G. austeni in their natural habitat.  The percentage reduction for G. austeni in the treatment area (Block A), with a density of targets of 4 per km², reached a peak of 94% (Figure 5B) nine months after target deployment.  When the target density was increased to 8 targets per km², the percentage reduction reached 99.9% within three months.

During the final target density increase to 12 targets per km² in Block A, the percentage reduction remained 99.9%. 

            In the barrier area (Block B), a steady increase in numbers was observed from May 2003 onwards, attributed to the removal of the targets in February 2003 to facilitate the increase in target density in Block A. 

No increase in G. austeni catches occurred in the treatment area, with 99% reduction evident for 9 months after removal of all targets in March 2004.

Discussion and Conclusions

            Studies on habitat preferences showed that G. austeni is restricted in its habitat preference, with very little movement out of indigenous forest.  The limited dispersal rate of G. austeni (Kappmeier Green 2002), and the fact that their peak activity is during the hottest times of day, i.e. early to late afternoon (Kappmeier 2000b), makes movement into unshaded areas undesirable due to heat stress.

            On the other hand, although indigenous forest seems to be the most preferred vegetation type of G. brevipalpis, open grassland and exotic plantations are evidently also suitable for survival.  As the times of peak activity for G. brevipalpis is during dawn and dusk (Kappmeier 2000b), dispersal into the open grass areas will occur mainly during these periods, coinciding with less heat stress and with animal movement from the shade into the grass areas for grazing.  This necessitates new thinking around the distribution and dispersal of G. brevipalpis and potential disease risk habitats, as G. brevipalpis distribution was considered to be confined to well shaded vegetation (Moloo, et al. 1980).

            Promising results were obtained with the suppression of G. austeni using targets optimally deployed at 8 targets per km², achieving 99% population suppression within 13 months.  Less promising results were obtained for G. brevipalpis, with a maximum reduction of 89% achieved only after 32 months after initial target deployment.

            The relatively inefficient and slow suppression rate of G. brevipalpis, may be attributed to several factors, including the removal of all targets from the barrier area (Block B) which resulted in tsetse re-invasion from the south, enabling the mobile G. brevipalpis to move into the treatment area. Added to this is the re-introduction of 200 large herbivores (i.a. buffalo, zebra, wildebeest and rhino) into the treatment area by the conservation authorities.  These large mammals graze mainly in the open grassland, a habitat frequently visited by G. brevipalpis, but not by G. austeni.  This may have reduced the availability of G. brevipalpis for the insecticide-treated targets.  Insufficient target density in the open grassland areas also contributed to lower reduction in G. brevipalpis numbers, due to the ability of these flies to cross open areas (Kappmeier Green 2002).

            A difference in suppression strategy for the two species should be considered when planning tsetse control operations using insecticide-treated targets in areas where both species are present. An area-wide approach has to be adopted especially for G. brevipalpis. Where localized placement of targets in suitable habitat is working for G. austeni, it is clear that targets should also be placed into adjacent areas of open grassland. The required target density of G. brevipalpis in an area-wide approach is, however, not certain and needs to be established.

Figure 5.  Monthly percentage reduction achieved during suppression of Glossina brevipalpis (A) and G. austeni B) in the treatment (▲) and barrier area (■).

             The required high target densities (especially for G. brevipalpis) precludes the deployment of these devices on a large scale i.e. it will be impractical, un-economical and requiring a huge infrastructure. The deployment of these targets could however, be considered for localized use in some areas of N.E. KZN and/or to create barriers between tsetse-free and infested areas to prevent reinvasion.

Acknowledgements

            The authors wish to thank the Director of the ARC-Onderstepoort Veterinary Institute for funding the project.  Drs. Glyn Vale and John Hargrove are thanked for their valuable input and comments.  The field staff at Hellsgate Tsetse Research Station is thanked for their invaluable assistance. 

References

Kappmeier, K., Nevill, E. and Bagnall, R.  (1998).  Review of tsetse flies and trypanosomosis in South Africa.  Onderstepoort Journal of Veterinary Research, 65, 195-203

Kappmeier, K. & Nevill, E.M. (1999). Evaluation of a proposed odour-baited target to control the tsetse flies Glossina brevipalpis and Glossina austeni (Diptera: Glossinidae) in South Africa. Onderstepoort Journal of Veterinary Research,66, 327-332

Kappmeier, K. (2000). A newly developed ‘H-trap’ for the live collection of Glossina brevipalpis and G. austeni (Diptera: Glossinidae) in South Africa.  Onderstepoort Journal of Veterinary Research, 67:15-26

Moloo, S.K., Kutuza, S.B. and Boreham, P.F. (1980).  Studies on Glossina pallidipes, G. fuscipes fuscipes and G. brevipalpis in terms of the epidemiology and epizootiology of trypanosomiases in south-eastern Uganda.  Annals of Tropical medicine and Parasitology, 74, 219-237

Nevill, E.M. (1997). The distribution of Glossina austeni and G. brevipalpis in South Africa. Proceedings of the 24th meeting of International Scientific Council for Trypanosomiasis Research and Control (ISCTRC), Maputo, Mozambique, 1997, 105. [Abstract].

Nevill, E.M., Kappmeier, K. & Venter, G.J. (1995). Recent efforts to determine the distribution of the tsetses Glossina austeni and G. brevipalpis in Zululand. Journal of the South African Veterinary Association, 66, 193. [Abstract].

Nevill, E.M., Kappmeier, K. & Venter, G.J. (1999). Studies on Glossina austeni and G. brevipalpis in South Africa. In: Animal trypanosomosis: Vector and Disease Control using nuclear techniques. Proceedings of the second FAO/IAEA Seminar for Africa, 27 November – 1 December 1995, Zanzibar, United Republic of Tanzania. Backhuys Publishers, Leiden, Netherlands. pp. 35-38.

Vale, G. Lovemore, D.F., Flint, S. & Cockbill, G.F. (1988). Odour-baited targets to control tsetse flies, Glossina spp. (Diptera: Glossinidae), in Zimbabwe.  Bulletin of Entomological Research 78, 31-49

Van den Bossche, P. (1997). The control of Glossina morsitans morsitans (Diptera: Glossinidae) in a settled area in Petuake District (Eastern Province, Zambia), using odour-baited targets. Onderstepoort Journal of Veterinary Research 64, 251-257

Willemse, L. A trial of odour baited targets to control the tsetse fly, Glossina morsitans centralis (Diptera: Glossinidae) in west Zambia. Bulletin of Entomological Research81, 351-357