Weekly forecasts help prevent birds damaging crops |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Medium and short-term spatio-temporal forecasting of likely breeding areas for the Red-billed Quelea |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Weekly forecasts now warn pest managers in Botswana, Mozambique, Namibia, South Africa, Swaziland and Zimbabwe where and when to expect bird pests. They can then take action to control them and tip off farmers to protect their crops. The Red-billed Quelea devastates millet and sorghum crops throughout southern Africa. The birds migrate long distances to feed on grass seeds, so their migrations follow rains with a predictable time lag. Every week, a map posted on the internet shows where the bird pest season hasn't yet begun, where there's been enough rain to prompt the first migrations, where queleas could breed and where the season is ending. These forecasts have proved so successful that national crop protection teams all over southern Africa now use them. Project Ref: CPP41:
Research Programmes: Crop Protection Programme The activity received funds (1996 onwards) directly from DFID (R6823, R7967, R8314, R8426) and was collaborative with the ICOSAMP project (R7890, R8315). Relevant Research Projects:
Associated projects:
Associated Programme Development / dissemination
Lead Institute:
Partner institutions:
Research Outputs, Problems and Solutions: The Red-billed Quelea Quelea quelea is a major pest of subsistence agriculture in semi-arid areas throughout sub-Saharan Africa. The birds devastate millet and sorghum fields of subsistence farmers and attack wheat and rice produced commercially, causing up to US$70 million worth of damage per annum. The Red-billed Quelea, the most numerous land bird in the world (population 1,500 million), is a long distance (up to 3000 km) migrant. The birds migrate in response to rainfall and the availability of seeds of annual grasses. Queleas breed and roost communally at sites providing targets for control with avicides or destruction with explosives. In South Africa alone there is an annual average of 173 control operations killing c.50 million birds. Hitherto, pest managers did not know where and when to expect queleas to appear and so when crops were threatened they were unprepared. Project technologies forecast (a) in the medium-term (three months) the likelihood of high populations of quelea occurring and (b) in the short-term (one month) the places and times of breeding opportunities. The outputs, developed through projects starting in 1996 [R6823, R7967, R8314 & R8426], can be divided into three processes and two technologies [data-base and short-term forecasting model]: Ø Data collation: An electronic data-base of quelea occurrences in southern Africa from 1834 to 1974 was created in 1999. Data on breeding colonies reported in southern Africa since 1974 were collated. Ø Development of models: The model was designed from knowledge of quelea biology in 1998 and programmed in 2000 using automated systems recording satellite-derived rainfall. Effort was concentrated on a model for southern Africa but a prototype model for eastern Africa was also developed. Ø Model Validation: Outputs using different algorithms were compared with data from the field and the most successful ones adopted. The southern Africa model was validated by analyses of model predictions in relation to the times and places of reported quelea colonies, 2000-2005. Ø Short-term Forecasting model: From Sept 2000 to date a weekly map was provided showing where (a) the season has not yet started; (b) sufficient rain has fallen for the first migrations to be stimulated; (c) where queleas could breed and (d) where the season was ending and new colonies were unlikely but breeding could still be in progress. Means for users to refine the forecasts were made available on the website as maps of known breeding areas, river systems and soil types.
Main commodity: Small-grain cereals (grain sorghum Sorghum bicolor, millet Panicum miliaceum, bullrush millet Pennisetum typhoides, finger millet Eleusine coracana and Italian millet Setaria italica) and rice Oryza sativa. Quelea birds are also major pests of wheat Triticum sp. and attack oats Avena aestiva, barley Hordeum disticum, buckwheat Phagopyrum esculentum, manna Setaria italica, triticale (hybrid between wheat and rye Secale cereale), teff Eragrostis tef and sunflower Helianthus annuus. Therefore the output could be applied to these commodities e.g. rice (Chad, Kenya, Malawi, Mali, Mozambique, Senegal and Tanzania), wheat (Ethiopia, Kenya, Sudan, Tanzania) and teff (Ethiopia).
Value could be added to the output by clustering it with those of other migrant pest projects, in particular those dealing with the forecasting of armyworms and locusts. A possible 'migrant pest' cluster could include Quelea (R8426, R7967, R6823, R8314), Brown Locust Locustana pardalina (R7779), Red Locust Nomadacris septemfasciata (R7818), Desert Locust Schistocerca gregaria (R6809, R6822), Senegalese Grasshopper Oedaleus senegalensis (R6788), Community-based Armyworm Forecasting (CBAF, R8407/R7966/R6762), Novel control of armyworm (R8408), ICOSAMP (R8315, R7890), Armoured Bush Cricket (ABC; R8253, R7428) and Larger Grain Borer Prostephanus truncatus (LGB; R7486,R6684). ABC and LGB are not strictly migrants but their control is often organised by the same organisations. From the perspective of quelea forecasting, the most important link has been with ICOSAMP for information exchange and dissemination, but useful economies were obtained by sharing technologies for satellite-derived rainfall data with the armyworm project. How the outputs were validated: The main validation process for the short-term predictions was to test if the model's forecasts of where and when Red-billed Quelea could breed in southern Africa were accurate. This involved collaboration between the project researchers and the end-user groups (national plant protection organisations) who supplied information on the locations and dates of reports of quelea breeding colonies, either directly or via ICOSAMP. In the three seasons 2002/2003, 2003/2004 and 2004/2005, 95%, 85% and 99% of colonies, respectively, were in 0.5 X 0.5 degree grid squares that the model had predicted to be suitable for breeding. Of those colonies reported from areas where the model had not predicted suitability, most (68%) were located next to squares that were deemed suitable.The value of the model and its potential for eastern Africa was recognized by delegates at a workshop on quelea management held in Kenya during 2005. A prototype model for eastern Africa was developed and was uploaded on the NRI quelea website (see http://www-web.gre.ac.uk/directory/nri/quel/) but it has not been validated. The research on the medium-term forecasts led to the conclusion that quelea abundances in terms of numbers of colonies per season were associated with high rainfall in the December to January period which, in turn, could be predicted in November. This was validated statistically using data on colonies supplied by the end-user groups. Where the Outputs were Validated: The performance of the model was validated in-country by end-user groups (national plant protection organisations) who reported at workshops in South Africa (2005) and Kenya (2005) that they found the model outputs useful for their control operations planning. In 2003, staff from the Botswana plant protection department visited an area denoted as suitable for breeding by the model and this led directly to the finding of a colony in a part of southern Botswana where quelea breeding had not previously been recorded. The conclusions of the medium-term forecasting research were used by the South African quelea control teams to conclude that there would be few quelea breeding during 2003/2004 and many quelea breeding during 2005/2006. The outputs are not targeted directly at particular social groups but to the national plant protection teams who then seek to protect the crops of resource-poor farmers in smallholder rainfed dry/cold and irrigated systems in semi-arid areas. Who are the Users? The model is currently being used by scientists in the national plant protection departments of the Ministries of Agriculture in southern Africa. The model has been handed over to, and is now being run by, the Regional Remote Sensing unit of the Southern African Development Community (SADC), based in Gaborone, Botswana. Each week during the quelea breeding season (September to May), they post a forecasting map of the model's outputs on to a website (http://gisdata.usgs.net/sa_floods/files/region/quel/). This is then available for all to examine and use to target particular zones for quelea surveys and control actions, as appropriate. Where the outputs have been used: The model outputs are currently being used by plant protection personnel in Botswana, Mozambique, Namibia, South Africa, Swaziland and Zimbabwe. Scale of Current Use: Usage was slow until the success of the model's predictions was established. Then, following dissemination of the model via a link on the ICOSAMP website, it is now used on a weekly basis during September-April by plant protection personnel in Botswana, Mozambique, Namibia, South Africa, Swaziland and Zimbabwe. Pre-season rainfall amounts are noted and the results of the medium-term forecast used to predict the likely severity of a season in Botswana and South Africa. Policy and Institutional Structures, and Key Components for Success: The Southern African Development Community (SADC), the Desert Locust Control Organisation for Eastern Africa (DLCO-EA, responsible also for quelea control within its region), The Food and Agriculture Organization of the United Nations (FAO), the International Red Locust Control Organisation for Central and Southern Africa (IRLCO-CSA, responsible for quelea control in some countries within its region e.g. Mozambique, Zambia) and the Information Core for Southern African Migrant Pests (ICOSAMP) have all assisted with the promotion and adoption of the outputs. In addition, national plant protection departments have supported the work. In terms of capacity strengthening, the key facts of success have hinged on the training of SADC staff in how to run the model and put it on a website themselves which they have done very successfully. In addition, three training courses covering aspects of quelea management were supported and attended by representatives of national plant protection departments, DLCO-EA and IRLCO-CSA. At each of these (one held in Botswana in February 2004, one in Kenya in May 2005 and one in South Africa in September 2005) the quelea short-term forecasting model was explained and demonstrated to trainees selected by the national plant protection departments as being those staff responsible for operational control of quelea. During the meeting in South Africa an opportunity arose to explain the project work to Professor Richard Mkandawire the agricultural advisor to the New Partnership for Africa's Development (NEPAD), thus bringing the project's work and the importance of border-crossing migrant pests in general to policy-makers at NEPAD. International policy on quelea control was also raised through meetings with SADC officials instrumental in drawing up their Migratory Pest Control policy. The outputs and national policies on the control of quelea were also discussed at senior levels in the Ministries of Agriculture in Botswana, Mozambique, Namibia and Zimbabwe. Dissemination Direct and Indirect Environmental Benefits: The technologies developed are largely beneficial in their effects on the environment. Early warning of quelea breeding colonies, leads to successful control, which means fewer birds go on to breed elsewhere and thus fewer control interventions are required overall. The fewer control actions that take place the less the environmental damage, be it by contamination with organophosphate pesticides or the result of explosions, and thus the fewer non-target organisms killed or poisoned. Adverse Environmental Impacts: No, except that when control does take place there are negative impacts. The outputs minimise the number of such control actions (see under 24). Coping with the Effects of Climate Change, or Risk from Natural Disasters: The models will predict where any climate-induced changes in rainfall patterns will have effects and thus draw attention to changes in quelea birds' breeding habits. This will in turn be reflected by appropriate responses by the control teams which will lead to benefits for poor people in rural communities. All migrant pests are dependent on rainfall and will thus be likely to increase or decrease in their severity regarding crop damage as the climate changes. The effects of climate change on the ecology of southern Africa are predicted to lead to increased frequencies of drought and greater variability in rainfall patterns. There has already been a 20% decline in summer rainfall over southern Africa between 1950 and 1999, but it is not only the quantity but also the timing and spatial distribution of rainfall that will affect migrant pests. Given that the CPP-funded projects on migrant pests have made considerable progress in establishing the rainfall patterns responsible for outbreaks of armyworm, quelea and locusts, their results will be useful for assisting policy decisions in relation to predictions of different climate change scenarios. These decisions will in turn have impacts on poor people. References Cheke, R. A. & Tucker, M. R. (1995) An evaluation of potential economic returns from the strategic control approach to the management of African armyworm Spodoptera exempta (Lepidoptera: Noctuidae) populations in eastern Africa. Crop Protection 14: 91-103. Elliott, C.C.H. (1989a) The pest status of the quelea. Quelea quelea: Africa's Bird Pest (eds R. Bruggers & C.C.H. Elliott), pp. 17-34. Oxford University Press, Oxford. Elliott, C.C.H. (1989b) The quelea as a major problem in a food-deficient continent. Africa's Feathered Locust (eds P.J. Mundy & M.J.F. Jarvis), pp. 90-99. Baobab Books, Harare, Zimbabwe. Geertsema, L. (1998) Redbilled Quelea: Annual Report. Unpublished report, National Department of Agriculture, South Africa. Publications on the Outputs JONES, P. J., CHEKE, R. A. & DALLIMER, M. (2000) Quelea migration in Africa. Abstract in Astheimer, L. B. & Clarke, M. F. (eds.) Proceedings of the 2nd. Southern Hemisphere Ornithological Congress (SHOC), Griffith University, Brisbane, Australia, 27 June - 2 July 2000. Birds Australia Report Series no. 9: 97. CHEKE, R. A. (2001) DFID CPP Project on forecasting Quelea breeding in southern Africa. Abstract, p. 13. ICOSAMP Workshop, 2-4 May 2001, Pretoria, South Africa. CHEKE, R. A. (2001) Information requirements for the Quelea Project on forecasting Quelea breeding in southern Africa. Abstract, p.15. ICOSAMP Workshop, 2-4 May 2001, Pretoria, South Africa. CHEKE, R. A. (2002) DFID CPP Project: forecasting Quelea breeding in southern Africa. Abstract, p. 27 in M. E. Kieser (ed.) Proceedings of the ICOSAMP Workshop, 2-4 May 2001, Pretoria, South Africa. CHEKE, R. A. (2002) Information requirements for the Quelea Project on "forecasting Quelea breeding in southern Africa". Abstract, p.29 in M. E. Kieser (ed.) Proceedings of the ICOSAMP Workshop, 2-4 May 2001, Pretoria, South Africa. CHEKE, R. A. (2002) DFID CPP Project on forecasting Quelea breeding in southern Africa. Abstract, p. 15-17 ICOSAMP Workshop, 21-23 May 2002, Pretoria, South Africa. CHEKE, R. A. (2003) Environmental impacts of quelea control and a model for forecasting quelea movements and breeding in southern Africa. pp. 58-65 in M. E. Kieser (ed.) Proceedings of the ICOSAMP Workshop, 21-23 May 2002, Pretoria, South Africa. CHEKE, R. A. (2006) Early warnings of threats to food security - the importance of uptake by policymakers. ID21 Insights no 61: 8. (see http://www.id21.org/insights/insights61/art09.html). CHEKE, R. A. VENN, J. F. & JONES, P. J. (2006) Forecasting suitable breeding conditions for the Red-billed Quelea Quelea quelea in southern Africa.Journal of Applied Ecology (accepted subject to revision). CHEKE, R. A., VENN, J. F. & JONES, P. J. (2006) A spatio-temporal model for predicting when and where Red-billed Quelea will breed in southern Africa. Abstract of poster presentation, 24th International Ornithological Congress, Hamburg, Germany, 13-19 August 2006. Journal of Ornithology 147 suppl.: 147. JONES, P. J., CHEKE, R. A., MUNDY, P. J., DALLIMER, M. & VENN , J. F. (2000) Quelea populations and forecasting based in southern Africa. Pp. 139-150 in Cheke, R.A., Rosenberg, L.J. & Kieser, M. (eds.) Proceedings of a Workshop on Research Priorities for Migrant Pests of Agriculture in Southern Africa, Plant Protection Research Institute, Pretoria, South Africa, 24-26 March 1999. Natural Resources Institute, Chatham, UK. TODD, M., CHEKE, R. A., VENN, J. F., KNIVETON, D. & JONES, P. J. (2006?) Variability of annual breeding records of the Red-billed Quelea bird Quelea quelea lathamii in southern Africa and the relationship with climate. Journal of Applied Ecology (submitted). VENN, J., CHEKE, R. A. & JONES, P. J. (1999) Quelea Bird-pest Database: Southern Africa Data. Natural Resources Institute, Chatham, Kent, UK VENN, J., CHEKE, R.A. & JONES, P. J. (2003) Forecasting breeding opportunities for the red-billed quelea in southern Africa. Proceedings of the 2003 EUMETSAT Meteorological Satellite Conference, Weimar, Germany, 29 September - 3 October 2003, pp. 612-617. EUMETSAT, Darmstadt, Germany. VENN, J., CHEKE, R.A. & JONES, P. J. (2004) Forecasting breeding opportunities for the red-billed quelea in southern Africa. Abstract. p. 302 in Proceedings of the 15th International Plant Protection Congress, Beijing, 11-16 May 2004. Foreign Languages Press. Relevant Research Projects,
with links to the
|
For relevant research projects, with links to further information
|
