Proceedings of a Regional Workshop on Identification of a Strategy For

PROCEEDINGS OF A REGIONAL WORKSHOP ON IDENTIFICATION OF A STRATEGY FOR

ALTERNATIVES TO FENTHION FOR QUELEA BIRD CONTROL

4-5 April 2017 Corinthia Hotel, Khartoum, Sudan

Rome/Genève/April 2017

Content 1. Summary ...... 2 2. Country presentations : ...... 4 2.1 Botswana ...... 4 2.2 CHAD: ...... 7 2.3 Eritrea ...... 18 2.4 Ethiopia ...... 21 2.5 Kenya ...... 23 2.6 Mauritania ...... 26 2.7 NIGERIA...... 31 2.8 Senegal ...... 34 2.9 Sudan ...... 38 2.10 TANZANIA ...... 48 2.11 Zambia ...... 52 2.12 Zimbabwe...... 53 3. Presentations of regional organizations on Quelea control ...... 55 3.1 DLCO-EA...... 55 3.2 IRLCO-CSA...... 55 4. outcomes of working groups ...... 56 5. Annex ...... 58 5.1 Annex I : ALTERNATIVES TO FENTHION FOR QUELEA BIRD CONTROL ...... 58 5.2 Annex II : Programme of the meeting ...... 95 5.3 Annex III : PARTICIPANTS ...... 96

1 1. Summary After opening remarks delivered by Mr Ibrahim Adam Ahmed Al-Dekhairi, Minister of Agriculture and Forestry and Mr MukhtarHamid Elwathig, FAO Representative Ad Interim, the first day was devoted to presentations in plenary. Presentations were delivered by the Rotterdam Convention secretariat, the consultant, each country delegate and representative from the Desert Locust Control Organisation for East Africa (DLCO-EA) and the International Red Locust Control Organisation for Central and Southern Africa (IRLCO-CSA). The consultant gave a detailed presentation on all available alternatives to fenthion (see document on Annex I)

Following these presentations there was extensive discussion about the numerous possible methods to control Quelea that could be adopted. However it was recognised that even if fenthion was listed in Annex III of the Rotterdam Convention for Prior Informed Consent (PIC) it was still useable but, due to its risk to human health and the environment, preferably as a control method of last resort. The various alternative control methods discussed could become component parts of an Integrated Pest Management (IPM) approach to be used as and when necessary or possible. It was agreed that without development of radically new control methods or new environmentally friendly pesticides it was not possible to recommend complete replacement of fenthion.

On the second day, the participants were separated into two groups, one consisting of Anglophone representatives and the other of Francophone representatives. Each group was asked to:

(a) prioritise possible alternative methods and

(b) suggest how regional collaboration could be advanced to promote awareness of the problems with fenthion and to strengthen Quelea control.

Alternative methods to be promoted and adopted :

Both groups recommended testing a variety of harvesting methods e.g. nest destruction for chick harvesting and capturing of adults with nets and the use of falcons to scare birds away from crops. The general recommendations were to use:

 fenthion at 600 g a.i. / litre and attempt to reduce concentrations and dosages of fenthion and cyanophos, following trials of the efficacy of lower concentrations and dosages than currently used, but not at the expense of possibly selecting for fenthion resistance.  Harvesting  Scaring by chasing quelea away from crops with falcons (as used in Botswana) or by noise-making machines  Explosions  Repellents such as Anthraquinone, as used in Zimbabwe, at small scales

2 Regional Cooperation on Quelea control :

It was recommended:

o To establish or strengthen regional and sub-regional coordinating focal points dealing with migrant pests, such as DLCO-EA and IRLCO-CSA, to foster information and data exchange, promote monitoring and forecasting methods and conduct research and training.

o To approach potential donors to seek the necessary funding for development and implementation of a regional strategy of Quelea control. At present , DLCO- EA is in charge of Quelea control in eastern Africa and the involvement of FAO/CLCPRO for Western Africa should be considered.

o That a focal point for information exchange in West Africa be established immediately.

o To draft a protocol for spraying procedures for fenthion as a standard to be adopted by all parties, but with options for different circumstances found in differing countries.

o That once a regional organisation was in place it could promote forecasting methods, based on existing models being used for the Desert Locust and by IRLCO-CSA, whose programme was derived from the ICOSAMP programme which is now inactive due to lack of information supply from participating countries.

o That more research on environmental impacts of spraying with fenthion and cyanophos to be conducted

3 2. Country presentations :

2.1 Botswana Introduction Botswana has been controlling quelea for over 30 years. Control operations have been undertaken in order to reduce the quelea problem, which persists and remains very serious. This is due to seasonal migrations from neighbouring countries and because quelea birds have natural habitats locally where they can survive on their natural diet of grass seeds and insects. These suitable conditions lead to successful breeding in areas where control operations are not done.

Migration Quelea is a resident pest in Botswana. The migration within the country is such that there are breeding outbreaks in Central, North East, Kgatleng and Kweneng districts in January to April. Then the birds migrate to Southern district in March / April. In April birds migrate to Chobe district (Fig. 3.1).

Figure 3.1. Map of the different districts of Botswana.

4 Damage

Ecological zones of Botswana support the production of sorghum and millet, as opposed to maize, and so the most vulnerable cereal crops are sorghum and millet. Sorghum is very important to Batswana people as it is the staple food of Botswana.

Control Methods When the country started the official control of quelea, fenthion was used mixed with diesel. It was later noticed that diesel had a devastating effect on the environment. It affected trees and grasses. Later on, fenthion 640g/l, a ULV formulation (trade name Queletox), was used. Thereafter the Government substituted it with cyanophos 520 g/l (trade name Falcolan). Details of quelea control operations from 2007 to 2016 are shown in Table 3.1. The Government has embarked on an IPM strategy to control quelea. This was motivated by the fact that some farmers showed interest in eating the birds. Farmers in the eastern side of the country were against the use of chemicals since they were harvesting nestlings. Therefore the Government included the use of explosive fire bombs. The method is fast and effective. Farmers are also encouraged to use cultural methods at the farm level to reduce crop damage. The methods used include scarecrows, clapping hands, throwing missiles, destruction of nests and harvesting of adults and nestlings. There has been a project on environmental impact assessment on the use of fenthion, cyanophos and explosives. The project was conducted in 2004 to 2005 and 2007 to 2009. The findings of the project indicated that all of these methods are polluting the environment. The project promoted quelea harvesting as an alternative control method. The latest alternative method of control is the use of live Lanner Falcons Falco biarmicus and Peregrine Falcons F. peregrinus kept under captivity and managed by falconers with appropriate equipment. Birds such as quelea are naturally fearful of falcons and quickly flee at the sight of the approaching predator. Prey birds quickly recognize and respect the territory established by a falcon. This natural fear is the main reason why the presence of falcons drastically reduces bird numbers in targeted areas such as the commercial farms in the Pandamatenga region. The method involves flying trained falcons to establish a “no-fly” zone for the local bird life. The falcons are trained

5 to simply chase the birds out of the area and return for a reward on a garnished lure. Local birds immediately flee at the presence of the falcon. The falcon quickly establishes a territory over the target area, which is considered as a dangerous area by the local bird life. The birds will change their daily routine after harassment by the falcons and move to less threatening areas to feed. The falconer flies the falcons several times a day from sunrise to sunset. The falcon patrols the area at a height of 50 to 100 metres and quickly chases any birds out of the area. After clearing the area, the falconer then calls the falcon back to its reward. The falcon's naturally established territory is enforced by the constant presence of falcon and falconer. Effectiveness is further improved by flying more than one species of falcon. For example, flying a female Peregrine Falcon in the morning and flying a male Lanner Falcon in the afternoon. Flying different species puts additional pressure on the birds to avoid the area. One falcon can easily clear a kilometre in a matter of minutes. Falcons are trained to chase rather than catch the birds because once a predator has caught and is eating its prey, it is a signal to surrounding prey that it is safe for the time being, which is why abatement is about the chase and not the catch. Each falcon is flown with a radio- tracking device so in the event of a long distance chase the falcon can quickly be located. Abatement programmes for agriculture are usually started one to two weeks before the vulnerable seeding (dough) stage and continue until harvest. Four falcons can effectively control quelea in 2500 - 3000 ha.

Table 3.1. Details of quelea colonies and control actions against them, 2007-2017. Figures in red refer to the unfinished season of 2016/2017 as control operations are continuing at the time of writing.

Year No. of No. of No. of Colonies Colonies Colonies colonies colonies controlled with controlled controlled migrated Chemicals with Explosives

Ha. Litres No. Ha. No.

07/08 244 136 108

08/09 165 133 32

09/10 42 15 27

6 10/11 61

11/12 57

12/13 140 82 58

13/14 207 120 87 72 224 912 47 160

14/15 23 13 10 0 0 0 13 21.8

15/16 72 54 18 4 24.43 200 50 164.77

16/17 45 5 4 4.25 1 7.7

2.2 CHAD:

Figure 4.1. Filets “Hadjaraye”

Introduction

7 Au Tchad, selon les estimations récentes, plus de 80% de la population vivent des activités agricoles en tirant leurs principales ressources des produits des cultures pour subvenir à leur besoins. Cette agriculture est basée essentiellement sur les cultures vivrières, dont les principales sont le mil, le sorgho pluvial et de décrue (berberé), le maïs, le riz irrigué, le blé, l’arachide, le niébé, le manioc, le sésame, le pois de terre etc. Le secteur agricole occupe une place prépondérante dans l’économie tchadienne par sa contribution de manière constante au Produit Intérieur Brut (PIB) (21 %) et son potentiel important de développement. Cependant l’agriculture tchadienne est confrontée à de nombreux problèmes phytosanitaires parmi lesquels les oiseaux granivores pour qui les récoltes de mil pénicillaire, Au Tchad, la gestion des problèmes phytosanitaires est dévolue à la Direction de la Protection des Végétaux et du Conditionnement (DPVC) qui utilise la lutte chimique et particulièrement le Fenthion 600 UL dans la lutte contre les oiseaux granivores du genre Quelea qui sont détruits. Mais la DPVC encourage aussi les producteurs qui utilisent les méthodes traditionnelles de lutte pour capturer les oiseaux granivores qui sont consommés. Les impacts de ces deux méthodes sur l’homme et l’environnement diffèrent de la méthode chimique à la traditionnelle.

Cultures attaquées par le Quelea, leurs superficies, leur importance dans l’économie du Tchad

Les dégâts causés sur les cultures céréalières sont importants. Ils sont estimés dans l’ordre de 17 à 36% ,soit 10 735 t pour une valeur perdue de 248 319 000 F CFA justifient les deux méthodes de lute (Tableau 4.1).

Tableau 4.1. Cultures attaquées et dégâts.

Cultures Superficie Rendement Production % Estimation Importance attaquées moyenne moyen moyenne Dégâts des dans détruite dégâts l’économie

Mil 25 690 600 kg/ha 15 414 t 36% 5 549 t 83 235 000 penicillaire F CFA

Sorgho 12 564 540 kg/ha 6 784 t 25% 1695 t 25 440 000 (blanc et F CFA rouge)

8 Riz 2 567 800 kg/ha 20 536 t 17% 3 491 t 139 644 000 F CFA

TOTAL 40 821 10 735 t 248 319 000 F CFA

Méthodes de lutte utilisées Face à l’ampleur des dégâts de cette importance économique, le Tchad a adopté deux méthodes de lutte : la lutte chimique (tableau 4.2) et celle traditionnelle.

Tableau 4.2. Méthode chimique. ANNEE Produits et Dose à l’ha Période de Superficies concentration traitement traitées utilisés 2010 Pas de Pas de traitement Pas de traitement Pas de traitement traitement 2011 Fenthion 600 1,9 l/ha Mai à Juin 54,7 ha UL 2012 Fenthion 600 3 l/ha Avril à Juin 53 ha UL 2013,2014, Pas de Pas de traitement Pas de traitement Pas de 2015 et 2016 traitement traitement

C’est au cours de ces interventions chimiques que le Tchad a enregistré l’intoxication, suivi de décès d’un des manipulateurs.

Méthodes traditionnelles A défaut d’une intervention musclée (traitements aériens et terrestres), les méthodes traditionnelles de lutte utilisées par les paysans sont : le gardiennage, le dénichage, ébranchages des arbres reposoirs, et éffarouchages sont parfois menées mais considérées comme un coup d’épée dans l’eau. La capture aux filets de confection locale est assez performante. Il s’agit des filets “Hadjaraï et Massa”

Filet “Hadjaraye”

Il est constitué de deux parties : le filet et les perches.

 Le filet

Il est tissé au fil en nylon n°45 trempé dans une solution de 1,5 kg de sucre pour stabiliser les mailles. Les mailles ont une dimension d’un doigt. Ce filet est de forme triangulaire avec une poche à la base. Les dimensions du filet sont variables (3 à 4 m de haut et 1, 80 m à 2 m de large) en fonction de la taille et de la force du piégeur. Il est bordé d’un fil n°60 en nylon (fil de bordure) pour le montage. Trois fils d’attache sur chacun des sommets permettent de fixer le filet sur les perches. Elles sont taillées à partir de bambous ou des branches d’Eucalyptus. Leurs dimensions sont légèrement supérieures à celles du filet. Les perches portent à leur extrémité des encoches pour la tension des fils d’attache du filet.

Les perches sont rassemblées à leur base par un fil rassembleur. Le filet est ensuite enfilé en quinconce sur les perches par les fils de bordure. Le fil de la poche est attaché sur le fil rassembleur des perches, tandis que les deux autres permettent de fixer le filet sur le sommet des perches au niveau des encoches.

10 Elle se pratique la nuit noire (sans lune). Les positions de capture dépendent de la hauteur du couvert végétal du dortoir: la position au sol et la position sur dos. Elle se fait lorsque la hauteur de l’arbre est petite (4 m).  Position au sol

Assisté de deux aides, le maître piégeur repère la forte concentration d’oiseaux sur l’arbre, lui tourne le dos, déploie son filet en éventail, prend appui des deux perches sur son ventre et d’un geste rapide, le rabat en refermant rapidement les perches pour empêcher les oiseaux de s’échapper. Il secoue son filet pour tasser les oiseaux dans la poche, le vide dans un sac tenu par l’un de ses aides. Il débarrasse le filet des épines qui s’y accrochent, puis continue la même opération sur d’autres arbres. Lorsque l’arbre est assez haut (6m), le piégeur monte en posant un pied sur le dos de deux aides qui, courbés, prennent appui au sol chacun avec un bâton. Il peut aussi monter sur les épaules. L’opération se fait par les mêmes gestes que ceux exécutés au sol.

Sept (7) sacs contenant chacun six mille (6000) oiseaux en moyenne peuvent être capturés en une nuit par l’équipe d’un piégeur appuyé de 2 à 3 aides avec un filet. Les oiseaux capturés sont déplumés , précuits et vendus aux marchés pour la consommation.

. Il ne peut être utilisé que dans les arbres de taille inférieure ou égale à six (6m), ce qui limite son emploi dans les dortoirs arborés de grande taille ;

. La position sur dos s’avère pénible pour les aides durant toute l’opération du maître piégeur pour monter, chercher l’équilibre, prendre l’élan de rabattage du filet et enfin descendre au sol ;

. Le manque de kit adéquat pour les piégeurs qui opèrent en pleine nuit dans les épineux et les souches qui les lacèrent et les blessent, sans compter les serpents à la chasse des oiseaux, parfois des lions.

. Les pesanteurs socio culturelles qui bloquent son développement (la chasse et surtout la capture au filet sont considérées comme activité dégradante, réservée à certaines castes de populations).

13 FILET MASSA C’est un filet de pêche appelé “Epervier” que les pécheurs du Mayo kebbi utilisent pour capturer les oiseaux quand la période n’est pas propice à la pêche. Cette capture est considérée comme une activité de loisir. Elle n’est pas professionnelle.

Il est confectionné avec du fil nylon n°3 avec des mailles d’un doigt. Il est de forme conique. Ce filet mesure quatre (4) mètres de hauteur, prolongé par un fil d’attache de dix (10) mètres de long. La circonférence est de 10m et porte en moyenne cent vingt (120) morceaux de plomb. Il est utilisé aussi bien dans les dortoirs arborés qu’herbacés (roseaux, mimosas). Elle se pratique la nuit par une équipe de deux personnes au minimum : un lanceur de filet et un braqueur de lampe torche. Quand la concentration aviaire est repérée sur l’arbre, le manipulant se met en position de jet avec son filet. Au signal convenu, la deuxième personne placée derrière l’arbre braque la lampe torche. Gênes, les oiseaux fuient la source de la lumière et s’envolent vers le manipulant qui jette le filet et les y prend. Selon les pêcheurs de

14 Kolobo (ethnie Kim), en cas d’une importante prise, le piégeur peut être traîné par les oiseaux. La capture dans les dortoirs herbacés se fait avec ou sans lune dans les mimosas ou dans les roseaux par les pêcheurs de Bongor (ethnie Massa). Elle se pratique par une équipe de deux personnes au minimum. Le manipulant jette son filet sur une concentration d’oiseaux suivi au même moment d’un braquage de lampe torche attachée à son front. Troublés, les oiseaux tentent de fuir par le bas ; puis bloqués, remontent vers la source de lumière, croyant trouver une porte de sortie et se font prendre d’avantage dans le filet. Deux à trois jets de filet sur de fortes concentrations permettraient de capturer 6 000 oiseaux en moyenne.  La pratique de cette capture à titre de loisir ne permet pas de développer cette technique qui pourrait être très prometteuse ;  Le manque d’organisation des piégeurs comme ceux opérant avec le fit Hajaraye ;  Le manque de kit approprié.

Le Tchad a cette opportunité de disposer de ces méthodes de lutte traditionnelle. C’est pourquoi il doit les valoriser en vue de les rendre plus performants Il est à rappeler que les difficultés liées à l’utilisation de filet “Hadjarai” sont :  La capture en monture qui est très pénible;  Le manque de kit de protection approprié;  Les pesanteurs socio culturelles à l’utilisation du filet.

Pistes d’amélioration - utilisation d’une échelle télescopique légère en lieu et place de dos d’hommes ; - utilisation mixte filet Hadjaraye, Massa et qui peuvent éviter la capture sur dos d’homme. - Dotation en kit de protection de couleur noire : bottes, combinaison, gants, chapeaux et paires de lunette. - Mobilisation et sensibilisation de la population des zones à forte concentration aviaire à l‘utilisation des filets.

15 Les difficultés liées à l’utilisation de filet “Massa” sont . La pratique de cette capture à titre de loisir ; . Le manque d’organisation des paysans ; . Le manque de kit de protection approprié.

Pistes d’amélioration - Transformer cette capture de loisir en moyen efficace de lutte contre les oiseaux granivores afin qu’il soit utilisé là où son utilisation permet (mixte avec filet Hadjaraye) ; - Organiser des équipes de plusieurs lanceurs de filets appuyés d’autant des braqueurs de lampes torches, pourrait permettre de réduire la population aviaire de façon significative dans les grands dortoirs comme dans les petits: a) Dans les grands dortoirs, les lanceurs de filet pourront se placer au milieu et un groupe de braqueurs à chaque extrémité du dortoir pour allumer alternativement les lampes torches, chassant les oiseaux vers les piégeurs qui les captureront à volonté. b) Dans les petits dortoirs, les lanceurs de filet et les braqueurs se mettront à chaque bout du dortoir. Les braqueurs allumeront les lampes torches qui chasseront les oiseaux vers les lanceurs.

Dotation en kit de protection de couleur noire : bottes, combinaison, gants, chapeaux et paires de lunette.

16 Tableau 4.4. Impacts des méthodes de lutte sur la santé humaine sur l’environnement.

METHODES DE LUTTE IMPACTS DES METHODES DE LUTTE

Sur la santé humaine Sur l’environnement Fenthion 600 UL - Coût d’intervention élevé; -Effets néfastes avérés sur - Oiseaux tués pas l’environnement (contamination du consommés; sol ,eau, végétation, air) - Source d’intoxication sur l’homme -Pas sélectif Capture au filet Hadjaraye - Economique par son coût -Pas d’effets néfastes sur l’environnement - Sources de protéines aux - Sélective démunis;

- Génératrice de revenu aux femmes revendeuses - -Pas d’intoxication sur l’homme.

17 2.3 Eritrea

Eritrea is divided into six regions (Zobas). These include Gash Barka, Anseba, Debub, Debubawi keih Bahri, Semenawi keih bahri and Maekel (Figure 5.1.). This division is intended to provide each Zoba with sufficient administrative autonomy and agricultural capacities. Agriculture is still the most important sector in Eritrea, employing about 80% of the population. The main crops are sorghum, millet, sesame, vegetables, wheat and beans. Consequently, the Government of Eritrea attaches high priority to rural development to achieve food security.

Figure 5.1. Administrative divisions of Eritrea.

The Government of Eritrea considers migratory pests as the major pest problem of the country, so it gives top priority to control them to safeguard food security. Thus, the country is a fee-paying member of regional organizations namely DLCO-EA (Desert Locust Control Organization for Eastern Africa) and CRC (organization for controlling Desert locust in the central region commission) for sharing common resources and information. The total area of Eritrea is 124,330 km2. Gash-Barka is the most productive and biggest region in the country (Figure 5.2). It covers about 27.5% of the

18 country’s surface area where outbreaks of grain-eating birds mainly Quelea quelea occur frequently. It is well known that quelea outbreaks can cause significant damage to agriculture and crop losses that may affect national and regional food security. These birds can travel 30 kilometres a day looking for food. Each bird can consume 3-5 g of grain and perhaps destroy the same amount each day. A colony composed of a million birds, something very common, is capable of consuming and destroying 70-100 quintals of seeds per day (1 quintal ≈ 100kg), enough to feed 15,000-20,000 people for a day. The main crops attacked by these birds are sorghum and pearl millet. In the rangelands they also feed on grasses such as Urochloa trichopus (Abertata) and Sorghum sudanese (Sudan grass).

Gash Barka region map

Figure 5.2. Map showing position of Gash-Barka.

Appropriate attentiveness is therefore one of the most essential procedures that contribute to the success of quelea preventive control strategies which usually depends upon conducting intensive and frequent surveys, immediate practical actions and developing solid and efficient contingency plans. During outbreaks most control is done by chemical sprays. Seven outbreaks have occurred since 1998 (in the years 1998, 1999, 2004, 2007, 2008, 2009 and 2012). A total area of 8,862 ha was treated

19 and 35,448 litres of 640g a.i fenthion chemical was consumed. Table 5.1 shows the data for some of the sub-regions involved. The total area treated in the last outbreak in 2012 was 189 ha and 756 litres of fenthion a.i 640g ULV was consumed. It was concluded that an estimated 6000 tons of grain were saved, which can feed 100,000 people for four months. In terms of money it estimated to be 90 million Nakfa (15 Nafka ≈ 1 USDollar) saved. This also means saving foreign currency that would have been utilized for importing food in the nation. Table 5.1. Areas treated by Fenthion and the doses used in some sub- Years Amount Treated Sub-regions Doses sprayed (Litres) area (Ha) used (L/Ha)

1998 920 460 Goluj 2

1999 2750 1375 Goluj/Teseney 2

2004 720 360 Goluj/Teseney ?

2007 1,544 772 Goluj 2

2008 3,324 1,108 Teseney /Goluj 3

2009 1,599 533 Teseney/Haykota 3

2012 752 188 Goluj 4

Total 11,609 4,796

regions.

Other control measures practiced Some cultural methods practiced in quelea control included bird scaring using slingshots, shouting, tying plastic on poles, scare-crows and, in rare cases, farmers prune the trees that act as roosting or breeding sites.

Impact of control methods on human health or the environment Following fenthion sprays, wild animals such as non-target birds, foxes, hyenas and snakes have been killed by secondary poisoning. Most control is done immediately after a target breeding colony is established. In such a way crop damage is reduced to the minimum.

20 2.4 Ethiopia

INTRODUCTION

In Ethiopia, the Red-billed Quelea Quelea quelea causes serious damage to cereals in the Awash River basin particularly in Oromiya region, the Omo river basin in South Nations, Nationalities and peoples region and Wabe Shebele basin in Ethiopia Somali region. The crops involved include sorghum (Sorghum bicolor) millet (Pennisetum spp.), wheat (Avena spp.) and teff (Eragrostis teff) (Table 6.1). The quelea population in Ethiopia has been estimated at >30 million and the birds are capable of causing losses of up to US$4million. Lethal control of quelea birds is an important part of the strategy to protect the cereal crop in Ethiopia. Generally the problem of bird damage is reported from June to November along the Rift valley from the Southern to Upper and Middle Awash river basin particularly along the Lake Ziway and Northern Rift valley where the birds prefer to roost. The main cereals that are grown in this area are sorghum, millet and teff. Aerial and ground control is undertaken by spraying the concentration of birds with a lethal avicide at the sites that have been confirmed by the Plant Protection Department (PPD) personnel. The avicide of choice is Queletox® (Fenthion 60%) in ultra-low volume formulation and provided by the MOA while the spray aircraft is provided by the Desert Locust Control Organization for Eastern Africa (DLCO-EA). A generic version of fenthion, Bathion™, has now been registered.

Table 6.1. Areas of crops attacked by quelea birds. No Crop Type Area (ha) Importance

1 Teff 2,739,686 Staple food crop

2 Wheat 1,61,000 Food and cash crop

3 Barley 1,034,615 Food and industrial crop

4 Rice 489,219 Cash and food crop

5 Sorghum 1,620,549 Food crop

6 Millet 904,163 Food and Industrial crop

21 Quelea control Methods

Fenthion is the pesticide of choice for control of quelea populations in breeding colonies or roosts, from which birds are confirmed as either causing serious crop damage or threatening ripening crops. Fenthion is either applied aerially by using spray aircraft or from the ground using applicators, depending on the size and the vegetation of the target site and population size of the birds. Currently Ethiopia is using Bathion 64%ULV at doses of 2 l/ha. Treatments with fenthion are carried out when there are high populations of the pest and when the crop is at the grain-filling or milky stage. Details of areas treated and volumes of pesticide used are given in Table 6.2.

Table 6.2. Details of quelea control operations in Ethiopia, 2010-2016.

Year Bird population Avicide used Area treated controlled (l) (roosts) (ha) (Millions)

2010 15 1150 575

2011 12 750 375

2012 15.5 1250 625

2013 24 1378 689

2014 51 2388 1194

2015 31.4 1683 841

2016 40 2500 125

22 2.5 Kenya

Quelea quelea is considered as perhaps the most numerous and most serious avian crop pest in the world. Enormous flocks of hundreds to thousands of birds descend on maturing crops causing grain losses of up to 100% on individual farm holdings. The birds pick grains from millets, rice, sorghum, pasture and wheat (Table 7.1). The birds are social, feeding in flocks and in the evening congregate to a single roost. 25 to 80 million birds are controlled annually, mostly in roost sites. Over 95% of the control operations use fenthion 640g ULV.

Current quelea bird situation. Quelea outbreaks and control is an annual exercise in Kenya. The extent and level of outbreaks vary from year to year. The high demand for wheat, rice, sorghum and millet for food and industrial use, means more land for production is required. This has resulted in the reclamation of swamps for paddy rice and exploitation of arid and semi- arid areas to meet the ever increasing demand. This has now increased the demand for bird control operations. Quelea Control methods in Kenya The following methods are used: (a) Aerial; (b) Vehicle-mounted ground sprays; (c) blasting by use of explosives; (d) Synchronization in cropping and (e) Scaring. For information on areas controlled from 2010 to 2017 see Table 7.2. Over 98% of the control operations entail the use of fenthion 640g ULV.

Table 7.1. Crops damaged by quelea birds and control methods used in Kenya. Crop Method of Product Product Dose/ha Treatment Control used conc. period Wheat Aerial/Vehicle Fenthion 640 ULV 6 -10 l 6.20 - 7.15pm

Rice 6.30 - 8.00pm Incineration Explosives 50kgs Sorghum /petrol/dies and 400l el Millets Loud noise Pasture scaring, slingshots, whips, scarecrows & tapes

Table 7.2. Areas treated from 2010 to 2017. Source Plant Protection Services Directorate of Kenya.

Year County Roo Population Crop Fenthion Remarks st (millions) Price No (USD/l) 201 Narok 8 11.0 Rice, 0 Kisumu 6 8.0 In -9 Counties were Sorghum, Siaya 7 5.0 dollars$ infested by a Busia 4 1.2 Wheat 28 population of 35.2 Meru 5 4.4 million birds. 201 Uasin 3 1.4 MilletsRice, 1 GishuKirinyaga 64 4.82.5 -839 Counties roost were were Sorghum, HomabayEmbu 52 4.41.2 infestedcontrolled by by a Aerial or KisumuMakueni 63 4.01.1 Wheat 27 populationvehicle mounted of 22.6 SiayaNakuru 3 3.20.8 million- 2 by blast birds. using Millets Busia 6 2.7 -explosives 30 roost were Aerial Narok 4 1.2 or vehicle mounted Nakuru 2 0.9 controlled Nyandaru 0 - 5 roosts blasted a using explosives

201 Kirinyaga 6 4.6 Rice, 2 Narok 5 6.2 -8 Counties were Sorghum, Nakuru 4 0.8 27 infested by a Siaya 5 3.8 Wheat population of Busia 3 3.2 24.3million birds. Millets Machakos 3 2.4 - 27 roost were Aerial Makueni 4 1.5 or vehicle mounted Nyandaru 3 3.3 controlled a - 6 roosts blasted using explosives 201 Narok 13 11.8 Rice, 3 Kisumu 9 12.0 -9 Counties were Sorghum, Siaya 5 6.8 infested by a Busia 2 2.8 Wheat 35 population of Nakuru 5 2.0 45.5million birds. Millets Machakos 3 3.0 -30 roost were Aerial Nyandaru 3 2.2 or vehicle mounted a 9 2.7 controlled Kirinyaga 2 2.2 - 5 roosts blasted Tana using explosives River

24 201 Nyandaru 2 1.6 Rice, -9 Counties were 4 a 7 2.5 infested by a Sorghum, Kirinyaga 17 7.8 population of Makueni 2 1.5 Wheat 39.2million birds. Nakuru 3 4.0 - 49 roost were Aerial Millets Machakos 6 5.0 35 or vehicle mounted 201 KituiKisumu 212 1.216.5 Rice, controlled-9 Counties were 5 Meru 1 0.7 infested by a Narok 4 3.3 Sorghum, - 8 roosts blasted KirinyagaBusia/Siay 1418 12.313 48 usingpopulation explosives of 81.4 Siaya/Busia 3 40 Wheat million birds. Taita 2 3 - 33 roost were Aerial a Millets Taveta 3 2.5 or vehicle mounted Narok 6 5.7 controlled Tana - 12 roosts blasted River using explosives Kitui

201 Siaya/Busi 8 7.5 Rice, 3 Counties were 6 a 3 11.5 48 infested by a Sorghum, Narok 8 5.4 population of 81.4 Kisumu Wheat million birds. Millets

Challenges in Quelea Bird Control 1. It is not viable to carry out an environmental impact assessment on individual roosts. 2. Expertise in explosive use. Method cannot be used where there is civil strife and it requires well trained personnel. 3. Human habitations. Quelea birds are roosting next to human habitations. 4. Water sources. Birds roosting in swamps. 5. Putting more land on sorghum and millets in arid and semi-arid areas. 6. Reclamation of swamps. The areas of Yala Swamp, Ahero, Lake Jipe, Tana delta and Mwea used to be dry season concentration sites of the birds. 7. Dependence on fenthion 640 ULV with no efficacious alternative.

Impact of Control Methods on Human Life. Within the period under review, no fenthion poisoning has been reported in the country. This can more be attributed to the cooperation between the Plant Protection officers, farmers and other arms of the Government. The fenthion based products are only allowed for use by specialized staff in the state Department of Agriculture and by the Desert Locust Control Organization for Eastern Africa (DLCO-EA).

2.6 Mauritania

2.7 NIGERIA

Introduction

Agriculture is next to oil in Nigeria’s economy and contributes about one-third (1/3) of the country’s Gross Domestic Product (GDP). Cereal crops are core raw materials for agro-allied industries which manufacture and process by-products for export and local consumption. Cereals constitute about 70-80% of the meals of an average Northern Nigerian per day. The Red-billed Quelea bird Quelea quelea is one of the major transboundary pests of agricultural importance in Nigeria. These birds are characterized by being highly mobile, voracious, prolific and swarm in colonies with alarmingly large bird populations of between 1-2.5 million birds/colony. Annual bird population increases in Nigeria are due mainly to the cropping system (rain-fed, irrigation) and the influx of migratory populations from the neighbouring countries of Niger, Chad and Cameroon. The Village Weaver Ploceus cucullatus is also an important bird pest requiring control. Quelea feed on the seeds of wild grasses, but also attack cereal crops (wheat, rice, maize, millet, sorghum) and the crops become vulnerable at the milky stage of grain development. Losses may be about 65-70% of crops, if the birds are not immediately contained. Smallholder farmers have often encountered total losses of their expected

31 cereal crop yields, as quelea control is beyond their capabilities. Quelea invasions have also been implicated in the abandonment of well drained fertile farmlands, leading to over-cropping of unfertile lands and undesirable migrations. Quelea quelea are found in the savannah agro-ecological zone of the “Frontline States” (Adamawa, Taraba, Bauchi, Gombe, Borno, Yobe, Kano, Jigawa, Katsina, Zamfara, Sokoto, Kebbi), which lie between latitudes N11∘30’- E12∘30’. These states directly border the Republics of Niger, Chad and Cameroon. Ploceus cucullatus is predominant in the wet southern states of Ekiti, Oyo, Ebony, Imo, Enugu, Rivers, Delta and Ogun, which lie between Latitudes N7∘30’ - E12∘30’.

Chemical control. Chemical control operations against Quelea quelea are carried out aerially and bi- annually due to the prevailing cropping system (rain-fed, irrigation) in the country. It has always proved effective. Ploceus cucullatus however, is selectively sprayed from the ground because it nests and roosts near human habitations. Pre-control surveys using a Global Positioning System (GPS) are conducted to obtain coordinates for precision spray sorties. 20-40 litres of fenthion 60%ULV are aerially sprayed per 2-4 hectares of roost, since control operations are not carried out on crop fields. Chemical spraying is most effective when sprayed at dawn and at dusk. However, aerial spray operations can be successfully carried out at any time of the day provided that there is active hatching/brooding in the roosts. Chemical spray operations may last between 5-7 days, depending on the magnitude and extent of invasion. Table 9.1 gives prices of fenthion in different years. Table 9.1. Prices of fenthion in dfferent years Year Price per litre (Naira) (100 Naira ≈ 0.31 USD)

2010 2, 725.75

2011 2, 725.75

2012 3, 250.15

2013 3, 543.75

2014 3, 725.55

32 2015 5, 525.75

Mechanical control. Quelea often roost in typha grass (Typha latifolia), reeds and thickets in large and deep water bodies, which are not easily accessible. These plants are mechanically cut down and removed in order to reduce the suitability of the roosts for the birds. However, the plants re-generate very fast. This control method is laborious, expensive and ineffective in large scale invasions.

Cultural control. Scarecrows, drum beats, whistling and random shouts on invaded farms are minimally helpful even on small-holder farms, as the birds soon get too familiar with the objects and noises.

Impact on environment and human health i. During aerial/ground chemical operations, the surrounding air, water and soil may be polluted. ii. Phytotoxicity has not been reported, but chemical particles may not be excluded from the food-chain in the immediate or distant environment. iii. Incidences of chemical poisoning by inhalation or physical contacts during field operations have been experienced but not documented due to the lack of poison centres.

Conclusions 1. Nigeria has so far not used any chemical other than fenthion 60% ULV in Quelea control. 2. Nigeria supports the listing of fenthion, subject to the availability of credible alternative(s).

33 2.8 Senegal

37 2.9 Sudan

Introduction Quelea quelea is a small weaver bird known as the most numerous and destructive bird pest of cereals crops in Africa and probably in the World. In Sudan Quelea quelea aethiopica, known by the common local names “Zarzur” or “Gudum Ahmer”, is the major bird pest inflicting serious damage and causing significant loss of cereal crops yearly. Other less injurious weaver species are given in Table 11.1.

Table 11.1. Bird pests other than Quelea quelea in Sudan.

Species English name Common local name

Ploceus sp. Village weaver Chilluk

Passer luteus Golden sparrow Um Awedat

Passer House sparrow Wad Debrag domesticus

Many traditional control methods were used in the past to combat this pest but with little success and were sometimes dangerous, expensive and impractical. Since the 1950s control of Quelea in substantial areas of Sudan under cereal crops was only possible when aerial application was adopted. Fenthion 600 ULV, known by various trade names, is used with success. However, there is a need for better alternatives and more effective management, entailing research and training of personnel. Quelea birds occur in the area between latitudes 8-15 °N and longitudes 23-38 °E with an annual rainfall between 400-800 mm (Beshir 1978, El Bashir & Beshir 1975). The total potential area of infestation is about 800000 square kilometres stretching from the Red Sea Hills in the east to the far western parts of the country (Kaske 1978). Migrations of Quelea follow a relatively simple North - South pattern similar to those of West African flocks and are dependent on seed availability in relation to seasonal movement of the Inter Tropical Convergence Zone (ITCZ). This prolific bird pest has only one nesting season a year in Sudan (Bruggers et al. 1984). In August to October the birds migrate to favoured savannah areas to form large breeding sites in forests, bushes, grasses and weeds. In the dry season roosting concentrations occur near

38 rivers, mainly the Blue and White Niles, Bahr El Ghazal and Sobat (Bruggers et al. 1984). During January to May huge flocks of the birds congregate in night roosts in suitable forests adjacent to water sources and not too far from feeding places (Kaske 1978). The breeding and roosting sites of quelea birds cover 12 states of Sudan, i.e. El Gedarif, Kassala, Sennar, El Gezira, White and Blue Nile, South and West Kordofan, South and East and West and Middle of Darfur States (Figs 11.1-11.3).

Fig. 11.1. The Breeding Sites of quelea in 12 States of Sudan (Gedarif, Kassala, Gezera, Blue Nile, White Nile, South Kordofan, West Kordofan, South Darfur, West Darfur, East Darfur and Middle Darfur).

Fig. 11.2. Quelea early rains migration patterns in East Africa (Ward 1971).

Fig. 11.3. Quelea late rains migration patterns in the breeding season in East Africa (Ward 1971).

Areas of almost 800,000 square kilometres, in the central rain-fed belt of Sudan, constitute the area with the main grain production farms (Fig. 10.1). These farms

40 provide the major bulk of the stable food of the Sudanese people, mainly sorghum, millet and sunflower. Quelea attacks are responsible for the loss of significant amounts of the produce. The introduction of mechanized farming has increased the area under grain production, coupled with intensification and diversification of crops, has created more suitable conditions for quelea attacks. Crop loss estimates vary from 5-10% for sorghum (Bruggers et al. 1984) and 50-100 % was reported by PPD (1990). Lenton (1981) calculated the estimated losses in dollars and believed that loss estimate could be much higher than expected. FAO & DWRC (1978, 1979) and FAO (1993), separately, in a detailed study carried out in El Hawata, concluded that crop losses are huge (Table 11.1). It was calculated that if cultivated areas were left without control above 90% of the crop would have been wiped out, resulting in serious economic consequences and a large food security-gap. Various traditional methods have been used by farmers in Sudan to reduce Quelea depredations to crops (Bruggers et al. 1984), including nest destruction, use of bird lime, poisonous gas, flame-throwers, poisoning ponds, explosives causing fire bombs and the planting of resistant crop varieties. In earlier times, the Colonial Pesticide Unit used modified fishing nets (El Shafie 1957). Currently, with the planting of very large fields, most of the traditional methods have become obsolete, costly, ineffective and impracticable. Since the late 1950s, aerial control of Quelea has been the routine practice. Fenthion 600 ULV is the only avicide used (Tables 10.3-10.5). It is justified due to the cohesive swarming of Quelea and the inaccessibility of many sites. Control operations start after the second week of September. The recommended dose of fenthion for large scale applications is, one litre per hectare. The sites are sprayed 10 minutes before sunset from an aircraft flying 2-3 m above the breeding colony at maximum speed of 150 km/hr. If fenthion is listed in Annex (3) of the Rotterdam Convention, we expect that it will become less easily available for bird control and its price could increase.

41 Table 11.2. Percentage damage caused by quelea birds to cereal production in cultivated areas in different States in Sudan, 1948 – 2016.

Remarks Damage Year State Crop Location & References %

1946,1948, Many subsequent All 1950, South Sorghum, outbreaks cultivated < 65% Darfur Millet (Richards,1955) 1953,1954, areas 1955 *

Eastern

North Sorghum, Kordofan 1953 96% Idris,1957 Kordofan Millet (Sherkiella

) * cultivated 1957 Sennar Sorghum 35% areas

(Beldat) Plantation 100% White at Kenana 1965 Sorghum El Amin,1966 Nile (ARC) Farms at < 65% * Kenana

Huge damage to crops leads to wider food security gap and its result Unity 1975 Sorghum Abyei 100% in starvation, which state supply by surplus of 14000 tons sorghum from Northern States. *

Limited & ineffective All 1989,1990, Sorghum, application and survey Infested All < 60% 1992, 1993 Millet due to muddy & tuff rood States to reach breeding sites. **

Local V. Abu- 1992, 1993 Sennar 44% PPD.Report,1990 Millet Naama

42 Hairy (Seed * 17% V.Millet Production ) Sunflowe 1993 Sennar 50 - 60% r

Rice Introduced as new White EL- early maturing crop in 1993 Rice 15% Nile Dewiem traditional cultivated areas (PPD,1993) *

* No lethal control was applied, but used some of non-lethal methods. ** Applied lethal control method.

Table 11.3. Quelea breeding sites and aerial spraying in Sudan, 2010 - 2016 (September-December). Source. PPD Annual Reports (2010 - 2016).

Flying No. of Areas treated Fenthion consumption times Season Sites (ha.) (Litres) (Hours and minutes) 2010 450 71288 71288 543 : 14 2011 400 60638 60638 328 : 34 2012 480 85323 85323 462 : 32 2013 397 59586 59586 387 : 17 2014 859 111890 111890 719 : 47 2015 644 81035 81035 552 : 31 2016 644 83745 83745 655 : 13 Total 3874 553505 553505 3649:28

Average 553.4 79072.1 79072.1 521 : 21

Table 11.4. Aerial applications of fenthion at Quelea breeding sites in target States for season 2016 (September - December). Source: PPD Annual Report (2016).

43 Flying times No. of Areas sprayed Fenthion State (hours and Sites (ha.) consumption (litres) minutes Gedarif 84 7695 7695 71 : 07 Kassala 31 2450 2450 24 : 08 Gezira 178 17925 17925 129 : 52 Sennar 26 3725 3725 41 : 47 White Nile 163 26900 26900 191 : 47 Blue Nile 15 2450 2450 21 : 10 S. 54 6600 6600 59 : 05 Kordofan S. Darfur 12 3350 3350 19 : 01 W. Darfur 21 2300 2300 06 : 43 E. Darfur 21 4150 4150 16 : 40 C. Darfur 39 6200 6200 66 : 52 Total 644 83745 83745 655 : 13

Table 11.5. Avicide products used in aerial application from 2007 - 2016 (September - December) together with dosages (litres/ha), quantities (litres) and prices (Euros). Source PPD General Pesticide Tender 2007-2016.

Fenthion 600 ULV Trade Names (Price/Litre) Quantity Dos Total Yea Queleto Fencal Fenbe Thion Birdy Vanish (L) e Value r x st (L/h (Euros) a) 329000 200 13.16 15.22 - - - - 50000 1 (380000 7 US$) 200 1 12.2 13.00 - - - - 50000 622000 8 200 1 14.75 13.00 - - - - 80000 1145000 9 201 1 - 12.87 - - - - 60000 772200 0

44 201 1 - 13.25 - - - - 20000 1590000 1 201 1 - 13.15 12.74 - - - 105000 1358200 2 201 1 - 13.15 13.44 12.30 12.29 - 100000 1280950 3 201 1 - - 13.26 12.30 - - 100000 1274000 4 201 1 - - 16.21 14.50 14.35 18.50 170000 2532850 5 201 1 - - 16.21 16.10 15.80 18.50 150000 2479400 6

Impact of Quelea Control Using Fenthion 600 ULV on Human Health and the Environment An environmental study was conducted at Hawata (Gedarif state) where samples of soils and plant materials (grass and acacia) were collected 15-18 hr after spraying with Fenthion 600 ULV for laboratory residue analysis. Residue analyses showed that all soil and grass samples were free from fenthion residues. A level of 0.5 and 0.1 ppm residues were detected in acacia samples treated with Fenthiolina and Quelly, respectively, all below the permissible level. According to this result, the use of Fenthiolina 600 UL and Quelly 600 UL is recommended for control of the Red-billed quelea, Quelea quelea aethiopica as being environmentally safe (Fatima et al. 2016). There have been no observations recorded on effects of fenthion on human health in Sudan.

Suggestions for effective control of Quelea: To minimize the used of Fenthion 600 ULV, we believe that in the future activities should be concentrated in the following general categories; 1. Due to climatic changes, updating studies and training must be conducted to find less susceptible crop varieties, estimate damage and understand the population dynamics of Quelea in Sudan. 2. Capacity building in monitoring, forecasting and use of modern equipment. Development of a computer aided system for monitoring, forecasting and aerial precision application to maximize the efficiency of chemical control. 3. Regional cooperation to establish a "Quelea Information System". 4. Regional monitoring and expert exchanges between Sudan and South Sudan.

45 Acknowledgement. We are grateful to the DG of PPD to his supervision and to the committee for information and their comments on the earlier drafts of this paper and we are also grateful to FAO and the Rotterdam Secretariat's staff for the facilities for the workshop.

References Bruggers, R. L. Bohl, W.H, El Bashir, S.Hamza, M., Ali, B., Besser, J.F, De Grazio, J.W. & Jackson, J.J. (1984). Bird damage to agriculture and crop protection efforts in the Sudan. FAO plant prot.Bull.Vol.32.No.1. Beshir, S.A (1978). The status of vertebrate pest control, PPD Khartoum North, Crop Pest management in the Sudan. Proceedings of a symposium held in Khartoum Feb.1978. Published by MoA & University of Khartoum, Republic of Sudan(1986). El Shafie, B.B. (1957). Note on bird control. Ministry of Agriculture, Research Division, Entomological Section, Locust Bait Stores Khartoum North. No. Ent. /BF/Agric /46-3/2. El Bashir, S. and Bashir, S.A. (1975) The role of aircraft in weaverbird control in the Sudan. Proc. 5th. Int. Agric. Aviat. Congress, Coventry, U.K., pp.92-100. El Amin, E.L.T. Moh (1966) Weaver Bird Control Campaign, Kenana Res. St. Abu Naama KRS-2-D. FAO (1985) Coordination of cooperative action to reduce bird damage to crops in eastern Africa. FAO,UNDP Regional project, Newsletter No.1,January. FAO/DWRC (1978) Vertebrate Pest Control Project, Sudan; in Vertebrate Control Research in Agriculture ,Annual Report, Denver ,USA. FAO/DWRC (1979) Report, Vertebrate Pest Control. Khartoum Base. Fatima ,Amani,Amel,Ismail Sid.Moh.(2016) Evaluation of Fenthiolina 600 UL & Quelly 600 UL (fenthion) for the control of the Red-billed quelea,Quelea quelea aethiopica. Izz Eldin Idris, (1957).Bird Control Report of Agricultural Office Sennar No. DAS/B -P -4. Kaske, R.F. (1978) Logistic improvements & economic advantages of a new weaver bird control method, a comparison between aerial LV & ULV treatment. Crop pest management in the Sudan. Proceedings of a symposium held in Khartoum, Feb1978, published by MoA & University of Khartoum, Republic of Sudan(1986).

46 Lenton, G (1981). Qualitative and quantitative assessment of bird pests in Eastern Africa: Sudan. Proc. 3rd Annu. Tech. Meet, FAO/UNDP Regional Quelea Project RAF/77/042. PPD/ Pesticide Tender in Sudan (2007/2016). Reports of Plant Protection Directorate Head PPD/ Annual Report, (2016). Report of Plant Protection Directorate Head Office, Khartoum North, Sudan PPD/ Vertebrate Section, Quelea Situation in the Sudan 1989/90 a Report DLCO-EA. Head Office Nairobi, Kenya. PPD/ Vertebrate Section, Quelea Situation in the Sudan 1992/93. a Report DLCO-EA. Head Office Nairobi, Kenya. Richards, W.S. (1955) Memo, Bird in Darfur 10 A & F/2.D3: Res. Div. Entom., O/.sec. Kordofan & Darfur. Ward, P. (1971) The migration patterns of Quelea quelea in Africa. Ibis 110: 275 - 297.

47 2.10 TANZANIA

BACKGROUND

Quelea birds are a traditional outbreak and migrant pest in Tanzania, but the magnitude of outbreaks varies in different years. Normally the quelea birds’ outbreak season occurs between January and October/November in irrigated farms. This period coincides with the growing season of cereal crops in most parts of Tanzania. Cereals with small grains are the main preference of the pest. The outbreaks occur in 14 regions of the Tanzanian mainland (Mbeya, Coast, Morogoro, Dodoma, Singida, Tabora, Shinyanga, Mwanza, Geita, Simiyu, Mara, Kilimanjaro, Arusha and Manyara). Quelea bird surveys are conducted by the Central Government in collaboration with Local Government authorities and they are controlled by aerial spraying conducted by the Desert Locust Control Organisation for Eastern Africa (DLCO-EA) using the organophosphate avicide fenthion. The avicides for quelea control are procured by Central Government (Ministry of Agriculture) and then distributed to Plant Health Zonal Centres. On average 6,000 Litres of avicides are procured per season and the average cost is TShs. 210,000,000.00 (US Dollars 121,930.00).

QUELEA CONTROL 2012 - 2016

Details of quelea control operations conducted in Tanzania during the 2012-2016 period are summarised in Table 12.1 and Figures 12.1-12.3. In 2012, if the birds had not been controlled they could have caused crop loss of 569 MT per day, with a similar estimate for 2013 of 486 MT. Two products are registered for quelea control: Queleatox 60% ULV with registration number AV/0001 and Bathion 60 ULV with registration number AV/0003. The average cost of fenthion 60% is 20 USD/Litre. The amounts of fenthion 60% ULV procured were 4000, 6000, 7000, 7000 and 5000 litres in 2012, 2013, 2014, 2015 and 2016, respectively.

48 Table 12. 1. Quelea control in Tanzania 2012 to 2016.

Year REGION AREA Birds Avicide (Ha.) (Millions) (Litres) Dodoma 244.0 12.6 650 Singida 85.0 20.0 250 Morogoro 18.0 4.5 225 2012 Mwanza 75.0 2.0 100 Shinyanga 551.0 11.1 750 Tabora 200.0 3.7 350 Sub Total 1,173.0 53.9 2,325 Dodoma 32.5 2.6 150 Singida 134.0 29.0 450 2013 Shinyanga 425.0 17.6 775 Sub Total 591.5 49.2 1,375 Dodoma 765.0 70.0 1,750 Tabora 470.0 19.7 900 Shinyanga 165.0 7.0 350 2014 Singida 100.0 9.0 400 Mantara 50.0 15.0 175 Morogoro 207.0 21.8 750 Sub Total 1,757.0 142.5 4,325 Mbeya 260.0 11.0 610 Dodoma 81.0 4.5 275 Shinyanga 110.0 2.3 250 Tabora 40.0 0.8 100 2015 Singida 180.0 4.7 470 Manyara 205.0 10.5 570 Morogoro 260.0 7.1 680 Mara 50.0 2.3 150 Sub Total 1,186.0 43.2 3,105 Dodoma 506.0 48.2 860 Singida 380.0 11.9 575 2016 Manyara 145.0 31.0 650 Morogorpo 162.0 6.7 540 mbeya 290.0 19.5 775 Tabora 180.0 7.5 275 Mwanza 60.0 1.5 100 Shinyanga 560.0 26.6 750 Mara 360.0 12.0 475 Sub Total 2.643.0 164.9 5,000

49 Figure 12.1. Quelea control in Tanzania: areas treated in hectares.

Figure 12.2. Number of birds killed (millions).

50 Figure 12.3. Amount of fenthion used (litres)

51 2.11 Zambia

Quelea Control in Zambia While occasional regionalized outbreaks of quelea occur in the southern parts of the country where sorghum and millet are cultivated, the problem of quelea outbreaks is not regarded as a major pest in Zambia and hence no centralised control is undertaken in the country. However, farmers growing sorghum and millet in the Zambezi valley do suffer from occasional damage from quelea. They use traditional scaring methods to protect their crops. These include making noise by shouting, beating tins and throwing stones at invading birds. Harvesting of nestlings for use as a source of protein is also practiced in some communities but this appears to be done more to obtain some good quality protein food, than as a crop protection measure.

Impacts of Control Methods on Human Health and the Environment No systematic follow-ups of non-target kills were made during control operations and there were no reported cases of unintentional poisonings among the pesticide applicators. Unacceptable exposure to the chemicals was inevitable as very little pre- spray training was undertaken. We know that the standard management practice of controlling quelea with organophosphate sprays, leads to more secondary poisoning than was generally supposed until the 1990s (van der Walt et al., 1998).

Conclusion Zambia is one of the countries that previously identified quelea birds as a problematic area that needed regular attention (FAO, 1980). However, due to little sorghum and millet being cultivated in the country now, it is no longer considered a major problem in comparison to the armyworm and locusts. There is no systematic national response to quelea bird control save for localized responses in the Zambezi valley where sorghum and wheat are grown. Concerns regarding the human and environmental negative impacts have remained high as the same group of chemicals, organophosphates, are being used in the control of other migratory pests, although in some cases pyrethroids have been used.

REFERENCES

52 FAO (1980) Cereal Crop Pests in Africa, with Particular Reference to Birds. Unpublished internal report, FAO_UNDP, Rome, Italy.

Van der Walt, E., Meinhardt, H. R., Venter, A. J. & Bouwman, H. (1998) Primary and secondary effects of Red-billed Quelea Quelea quelea control in South Africa. Ostrich 69: 456.

2.12 Zimbabwe

Introduction

The control of quelea in Zimbabwe is conducted by the Problem Bird Control Unit which is within the Zimbabwe National Parks and Wildlife Management Authority. This is provided for by the Quelea Control Act. The chemicals used in its control are registered under the Fertilizers, Farm Feeds and Remedies Act (Chapter 18:12) which is administered by the Ministry of Agriculture, Mechanization and Irrigation Development. Quelea in Zimbabwe attack wheat, sorghum and barley. Wheat is mainly grown in winter from May to September in Mashonaland West, Mashonaland Central and Mashonaland East Provinces (Fig. 1). Sorghum is mainly grown in the drier parts of the country. Zimbabwe is a net importer of wheat. If what is grown is destroyed by quelea, this further complicates the economy as more and more scarce foreign currency will be used to source wheat, thereby making the funds unavailable for other uses in the economy. Sorghum is produced in drier parts of the country where it is the main cereal grown. It is a source of livelihood for many families. If quelea is not controlled, this will negatively impact the livelihoods and the national economy because resources will be channeled towards feeding the people whose crops would have been destroyed by quelea.

Figure 14.1. Provinces of Zimbabwe.

Control Methods Used The main method of control is chemical. This involves the use of avicides fenthion 640 UL and cyanophos 520 ULV. Chemicals are usually applied at application rates ranging from 2 to 15 litres per hectare at the roosting sites either by aeroplanes or by motorized ULV backpacks. This is done at dusk or at dawn. Control is mainly done during the grain filling-stage. A new way of minimizing the negative impact of quelea has recently been adopted, the use of the chemical 9,10 anthraquinone as a repellent. This is sprayed onto the wheat fields and it repels the birds.

Areas treated from 2010 to 2016 Quelea mainly breed in the South-eastern low veld, where treatments are concentrated. Roosting sites scattered in Mashonaland Central and West Provinces were also treated during the last six years. An average of 2000 hectares was sprayed per year.

Prices of fenthion 640 ULV Up until the early 2000s, Zimbabwe used Queletox 640 ULV which used to cost around USD 26/litre. This was then changed to the cheaper generic, fenthion 640 ULV which costs around USD 13/litre CIF Harare from China. From 2011 to 2012, Zimbabwe used cyanophos 520 UL which cost around USD 20 from South Africa. From 2013 to 2016, Zimbabwe used fenthion 640 ULV imported from China, which costs USD 13 CIF Harare.

54 Impact of control methods on human health or the environment Data on this were not readily available but post-application evaluation has shown that non-target species of birds have been killed by fenthion 640 UL application.

3. Presentations of regional organizations on Quelea control

3.1 DLCO-EA

The Desert Locust Control Organisation for Eastern Africa (DLCO-EA) is responsible mainly for control of Desert Locusts Schistocerca gregaria, but it also controls other migrant pests including Red-billed Quelea Quelea quelea. DLCO-EA has nine member countries, Djibouti, Eritrea, Ethiopia, Kenya, Somalia, Sudan, South Sudan, Tanzania and Uganda, which with the exception of Djibouti all have problems with quelea birds. It was pointed out that in some places quelea were changing their habits. For instance in Uganda their populations were increasing and spreading their distributional ranges, with records now near Lake Victoria, and in Ethiopia they had taken to roosting in maize which posed problems for control operations. In addition a habit of roosting near human habitation also posed control risks. It is recommended that no water should be drunk for three days from potentially contaminated sources and that no grazing should take place for seven days in contaminated pastures. Thus pastoralists were often reluctant to report the presence of quelea. It is likely that future climate change will also lead to changes in the birds’ ecology and behaviour.

3.2 IRLCO-CSA

The International Red Locust Control Organisation for Central and Southern Africa (IRLCO-CSA) has its headquarters in Ndola, Zambia. IRLCO-CSA is responsible for migrant pest control, including control of Red-billed Quelea Quelea quelea if requested to do so, in its member countries of Kenya, Malawi, Mozambique, Tanzania, Zambia and Zimbabwe. Fenthion 60% ULV is used and when sprayed aerially the targets need to be within 30 minutes flying time of an airstrip. Dosages of 2-8 l/ha are used against colonies and 5-10 l/ha against roosts. Ground control is also undertaken in roosts or colonies that are less than 5 ha in area and with vegetation less than 5m tall, for which cut-lines 100m apart are recommended.

55 Alternatives to the use of fenthion noted included the harvesting of quelea for food in Malawi, where they can be bought as kebabs at roadsides or cooked in stews. Harvesting is also practised in Mozambique.

4. outcomes of working groups Delegates at the meeting were asked to provide information on their countries’ current quelea control practice, which are summarised in Table 17.1. They were also asked for details of how they made control decisions, if they had regular training programmes for their staff, whether any of their communities ate quelea and, if so, the names of the communities involved or where they lived. Finally, they were requested to provide suggestions for a strategy to develop quelea control methods other than reliance on fenthion. Responses were received from 8 countries. Of these Botswana, Chad, Ethiopia, Kenya and Mauritania reported that they only control the birds when they are threatening a crop. Eritrea apparently attempts control of all breeding colonies. Points not covered in the country reports or in Table 17.1 included that Nigeria used a variety of criteria for control decisions such as the cropping system involved, its economic value, the availability of funds, the general level of preparedness, the size of the bird populations and the crop under threat. Training is conducted in Botswana but was considered to require improvement, partly due to high turnover rates of staff, and there are training programmes in Chad, Eritrea, Kenya, Mauritania and Nigeria. Of non- spraying options, Botswana use falcons managed by a company (African Falcon Services), charging about USD 18 per hectare. The method needs to be evaluated and scaled up. Botswana reported that communities in the Chobe, North East and Eastern areas eat quelea. In Kenya, eating quelea is practised by Luo, Luya, Kamba and Kikuyu people. Communities in northeastern Nigeria also eat the birds, as do communities in the Zambezi valley of Zambia.

56 Table 17.1. Quelea control methods in different countries.

Country Fenthion Fenthion Fenthion Cyanophos Notes g a.i. / l / ha l / ha litre colonies roosts g a.i. / litre Explosions Repellents Harvesting Falcons Database Used fenthion 640g/l in the past. Botswana 640 2 - 4 2 - 4 520 ✔ ✔ ✔ Cyanophos used now at 5-10 l/ha No fenthion since 2012. Used Chad 600 1.9 - 3 1.9 - 3 520 ✔ cyanophos in the past Eritrea 640 2 - 4 2 - 4 Only fenthion used Forecasting study based on breeding sites surveys planned. Cyanophos not registered. Hope Ethiopia 640 2 2 ✔ ✔ ✔ to use explosions in the future. Kenya 640 5 - 7 5 - 7 500 ✔ Have used cyanophos Nest destruction used and this plus use of nets could be Mauritania 640 5 - 7 3 ✔ ✔ developed further Consistent use of fenthion. Nigeria 600 2 - 3 2 - 10 ✔ Cyanophos not registered. Used cyanophos in the Senegal ? ? ? ? past South Only cyanophos and explosions Africa 520 ✔ ✔ used now Sudan 600 1 1 Tanzania 600 ? ? ✔ Zambia ? ✔ Zimbabwe 640 2 - 5 2 - 15 520 ✔ 9,10 Anthraquinone repellent used Concentrations and dosages used dependent on host country DLCO-EA 600 ? ? practice IRLCO- CSA 600 2 - 8 5 - 10

57 5. Annex

5.1 Annex I : ALTERNATIVES TO FENTHION FOR QUELEA BIRD CONTROL

Robert A. Cheke

Natural Resources Institute University of Greenwich at Medway Central Avenue Chatham Maritime Chatham Kent ME4 4TB UK

Email: [email protected]

Report to the Food and Agriculture Organization of the United Nations

EXECUTIVE SUMMARY

1. The reasons for seeking alternatives to fenthion for control of the Red-billed Quelea Quelea quelea are briefly summarised.

2. The only available pesticide that could replace fenthion is cyanophos, but this chemical is also highly toxic to non-target organisms, although less so than fenthion, and may be more expensive. Further research on the environmental impacts of cyanophos is recommended.

3. Apart from chemical avicides, the only technique available for controlling substantial numbers of quelea is the use of explosives combined with fuel to create fire-bombs. Fire-bomb explosions are recommended as the best available alternative to fenthion, although they also have negative effects on the environment, can be

58 dangerous and have associated security issues. The technique is labour intensive and in practice can only be deployed against small (<5 ha) colonies and roosts.

4. An integrated pest management (IPM) approach is the most environmentally benign strategy but, apart from when circumstances permit cultural control measures, most IPM activities only have realistic chances of succeeding in controlling quelea in small (<10 ha) areas.

5. Amongst available IPM means of lethal control, mass-trapping, which also has the advantage of providing a food source, is recommended when quelea roosts and colonies are less than 5 and 10 hectares in area, respectively. Nevertheless with both traps and mist nets, care is needed to minimise non-target casualties. Other IPM measures are reviewed.

6. If fenthion has to be used, means of minimising its use include ensuring that spraying is only conducted when crops are threatened and that the lowest dosages necessary to achieve kills are applied. Regular training of pest control workers in how to use equipment correctly and in what to do in the case of accidental contamination of operators, and training of farmers on IPM principles and quelea biology through farmer field schools are recommended.

59 CONTENTS EXECUTIVE SUMMARY Page

1 INTRODUCTION 3 1.1. Terms of Reference 3 1.2. Background 4

2. ALTERNATIVES TO FENTHION FOR QUELEA CONTROL 5 2.1. Chemical Control 6 2.1.1. Alternative pesticides 6 2.1.2. Bird repellents including narcotics 7 2.2. Mechanical Control 7 2.2.1. Explosions 7 2.2.2. Nest destruction and chick harvesting 9 2.2.3. Trapping 10 2.2.3.1. Chad Traps 10 2.2.3.2. Kondoa basket traps 11 2.2.3.3. Kondoa basket traps made of wire mesh 14 2.2.3.4. Funnel traps 14 2.2.3.5. Miscellaneous indigenous trapping methods 15 2.2.3.6. Mist nets 15 2.2.3.7. Roost traps 17 2.3. Cultural Control 17 2.3.1. Planting and harvest time manipulation 17 2.3.2. Weeding 17 2.3.3. Alternative Crops 17 2.3.4. Quelea resistant crops 18 2.3.5. Protecting crops with netting 18 2.3.6. Trap roosts 18 2.3.7. Scaring 19 2.3.7.1. Scaring by humans 19 2.3.7.2. Scaring with falcons 19 2.3.7.3. Commercial bird-scaring devices 20 2.4. Biological control 20

3. COMPARISON OF RED-BILLED QUELEA CONTROL METHODS WITH THE DESERT LOCUST PREVENTIVE CONTROL APPROACH 21

60 4. RECOMMENDATIONS FOR QUELEA CONTROL MINIMISING THE EXTENT OF FENTHION USE 21 4.1. Forecasting and control planning 21 4.2. Fenthion dosages 22 4.3. Fenthion application methods 23

5. RECOMMENDATIONS FOR QUELEA CONTROL WITHOUT FENTHION: A STRATEGY FOR BIRD CONTROL CONTAINING SELECTED METHODS OF CONTROL 24

Acknowledgements 29 References 29

INTRODUCTION The Red-billed Quelea Quelea quelea is the most important avian pest of small grain crops in Africa, causing damage up to the equivalent of US$79.4 million per annum at 2011 prices throughout semi-arid zones (Elliott 1989a,b). At present control is mostly by aerial and/or ground-spraying of organophosphate avicides, with fenthion (Queletox®) being the pesticide of choice. Regrettably, fenthion is toxic to man and to other non-target organisms so alternatives to its use are urgently sought under the auspices of the Rotterdam Convention. This report, commissioned under the Terms of Reference listed below, aims to review all available alternatives to fenthion for bird control and to draft a bird control strategy based on selected methods of control. The objectives of the Rotterdam Convention (see http://www.pic.int/TheConvention/Overview/tabid/1044/language/en- US/Default.aspx) are:  to promote shared responsibility and cooperative efforts among Parties in the international trade of certain hazardous chemicals in order to protect human health and the environment from potential harm;

 to contribute to the environmentally sound use of those hazardous chemicals, by facilitating information exchange about their characteristics, by providing for a national decision-making process on their import and export and by disseminating these decisions to Parties.

61 1.1. Terms of Reference The Rotterdam Convention is a multilateral environmental agreement, jointly administered by FAO (within AGP Division) and UNEP. It is a full programme aiming at reducing risks from the use of chemicals and in particular hazardous pesticides in agriculture. It is considered a Corporate Technical Activity (CTA) contributing to Strategic Objective 2 (2.3.1 and 2.3.3). It is partly linked to Strategic Objective 3. Under the general supervision of the Senior Technical Officer, and reporting to the Programme Officer of the Rotterdam Convention, the consultant will work on technical assistance activities that move in particular the indicator 2.3.3 and on activities specifically related to indicator 2.1.3. The tasks will be carried out in full compliance with the Programme of Work and Budget 2016-2017 adopted by the Conference of the Parties to the Rotterdam Convention, will address FAO’s Country Programming Frameworks and as far as possible the FAO focus countries. Technical assistance activities under Rotterdam Convention

 Advise and support the Secretariat regarding preparation and implementation of technical assistance in 2016 with a focus on alternatives to Fenthion for Bird Control. o Review of all available alternatives to fenthion for bird control o Draft a strategy on bird control containing selected methods of control o Perform other related duties as required.

1.2. Background The Red-billed Quelea Quelea quelea is restricted to Africa where it occurs in semi- arid areas, feeding principally on native grasses but when these are scarce the birds will attack the seed heads of crops. Principal amongst the latter are millet, sorghum, wheat, rice and teff. There are three subspecies of Red-billed Quelea: the nominate race Q. q. quelea occurs in West Africa from Senegal in the west to Sudan in the east; Q. q. aethiopica ranges in East Africa from Ethiopia to southern Tanzania and Q. q. lathamii is restricted to southern Africa (see Cheke 2014 for a recent summary of the biology and control of quelea). All three of the subspecies are migrant pests which follow rainfall systems. As meteorological conditions vary from year to year, the locations and severity of quelea infestations also vary between seasons. In general, the birds breed 2 or 3 times a year, but up to 5 times per annum in East Africa, during and just after rainy seasons. Quelea coming from huge communal breeding colonies

62 may attack crops. Damage also occurs in dry seasons when the birds continue to flock together and may roost in very high numbers. Both breeding colonies and roosts are the targets of control operations that take place after dark when the birds have settled down for the night. The birds also collect in “day-roosts” or “secondary roosts” during daytime (Ward & Zahavi 1973) when they are susceptible to mass-trapping. As the birds are migrant pests, responsibility for their control rests in some zones with international organisations. Thus, the Desert Locust Control Organization for Eastern Africa (DLCO-EA) uses its aircraft to treat member countries’ infestations and the International Red Locust Control Organisation for Central and Southern Africa (IRLCO-CSA) has a similar role for its areas of responsibility. However, some countries such as Botswana, the Republic of South Africa and affected West African countries now undertake their own control duties. Although fire-bombs are used to destroy quelea breeding colonies and roosts in Botswana, South Africa, Kenya and elsewhere, the principal control agent in all areas is currently the organophosphate avicide fenthion (Queletox®; 640 UL; thiophosphoric acid or O,O-dimethyl-O-[3- methyl-4-(methylthio) phenylphosphorothiote], also known as Baytex, Lebaycid, Tignvon and OMS-2). Fenthion, like other organophosphate compounds, acts by inhibiting acetylcholinesterase, which is essential for normal nerve function. When acetylcholinesterase is inhibited a build-up of acetylcholine results causing prolonged transmission of nerve impulses leading to death from respiratory failure. Fenthion can therefore injure or kill indiscriminately, with consequent adverse effects on non-target organisms (McWilliam & Cheke 2004) including man. Fenthion residues are now known to have a half-life of 45 days, almost twice a previous figure given by Meinzingen et al. (1989), and rainfall after sprays can cause fenthion to leach out of the soil and still be detectable five months later (Cheke et al. 2013). Additionally, persistence of fenthion residues in the air for 64 hours and for 46 days in soil was reported by van der Walt (2000). Recently there was a fenthion-related human fatality in Chad which led to a proposal by that country in 2012 to list fenthion in Annex 3 of the Rotterdam Convention which, if accepted, would lead to the pesticide being subject to the Prior Informed Consent (PIC) procedure (for more details of this procedure and its implications see http://www.pic.int/Procedures/PICProcedure/tabid/1364/language/en- US/Default.aspx). Following the proposal under Article 6 of the Convention, a draft

63 decision guidance document for fenthion (ultra low volume (ULV) formulations at or above 640g active ingredient/l) was prepared and the criteria for including fenthion in Annex III were met according to the Chemical Review Committee. However, the 2015 Conference of the Parties did not accept the proposal. Subsequently, the 7th meeting of the Conference of the Parties mandated intersessional work on “the process of listing chemicals in Annex III to the Rotterdam Convention, including to review cases in which consensus had not been reached, to develop options for improving the effectiveness of the listing process, and to develop proposals for improved information flows” (for further details see http://www.pic.int/Implementation/ProcessforListingChemicals/IntersessionalWorking Group/tabid/5253/language/en-US/Default.aspx). The intersessional working group is expected to report back to the Conference of the Parties at its 8th meeting to be held in Geneva in 2017. Consequently there is a need for information on alternatives to fenthion or, at least, means to reduce the extent of its use, as part of the COP debates.

2. ALTERNATIVES TO FENTHION FOR QUELEA CONTROL To date, measures to control quelea birds without using fenthion have included the use of (a) alternative pesticides; (b) explosives/fire-bombs; (c) a variety of mass trapping methods, sometimes keeping the birds for food; (d) cultural control; (e) quelea resistant crops; (f) protecting vulnerable crops with repellents and netting and (g) scaring the birds, including the use of falcons. These will be described and discussed below under main headings of Chemical Control, Mechanical Control, Cultural Control and Biological Control.

2.1. Chemical Control 2.1.1. Alternative pesticides The most commonly used alternative avicide is cyanophos (Falcolan 520 UL; OMS

226). However, with its chemical formula of C9H10NO3PS (O-(4-cyanophenyl) O,O- dimethyl phosphorothioate) cyanophos is also an organophosphate. It is not registered for use in the United States of America where it is classified as an extremely hazardous substance. Similarly it is not registered for use in the European Union. Nevertheless, it has been used for quelea control in Senegal, Mauritania, Botswana, Ethiopia and elsewhere since it has lower toxicities (e.g. acute oral LD50 for rats 730 mg.kg-1 and

64 3 mg.kg-1 for quelea) than those of fenthion (acute oral LD50 for rats 250 mg.kg-1 and 6-10 mg.kg-1 for quelea). Cyanophos was tested in Tanzania but was not recommended for routine use. In contrast, it is a registered avicide in South Africa where it is used for the majority of quelea control operations since fenthion is no longer available there (E. van der Walt, pers.comm., Nov. 2016). Cyanophos has the disadvantage of a delayed killing effect in comparison with fenthion (Allan 1997), so use of cyanophos could lead to more secondary poisoning of non-target organisms than control with fenthion but few studies of the effects of cyanophos use in the field have been conducted. Mullié et al. (1999) studied non-target organisms after cyanophos spraying against quelea in Senegal and concluded that it seemed to be as damaging as fenthion but that the data were insufficient for adequate comparisons. Cheke et al. (2013) found that cyanophos was still present in soil, at concentrations of from 0.009 to 0.169 g.g−1, 41 days after a spray in Botswana (the maximum residue level for this compound is unknown but UK pesticide authorities recommend a default maximum residue level on food of 0.01 μg.g−1; see https://secure.pesticides.gov.uk/MRLs/). Phoxim, also an organophosphate, has been tested as an alternative to fenthion (Pope & King 1973) and Allan (1997) listed some other alternative chemicals such as mevinphos, another organophosphate. However, mevinphos is even more toxic than fenthion (acute oral LD50 for rats 3-12 mg.kg-1 and 1.43 mg.kg-1 for quelea), so it does not present a suitable alternative.

.1.2. Bird repellents including narcotics Alphachloralose is a narcotic agent added to bait grain or water that has been used in South African trials (see Garanito et al. 2000) which leaves birds so weak that they can be easily picked up or killed, but its potential for affecting non-target organisms renders it unsuitable except, perhaps, in urban areas. Other possible chemicals with repellent abilities that could be used, but which have similar strictures against them, include 4-aminopyradine and aluminium ammonium sulphate, curb (ammonium sulphate) and trimethacarb (predominantly trimethylphenyl methylcarbamate). Use of mesurol, the carbamate methiocarb, a bird repellent, molluscicide and insecticide, also listed by Allan (1997), doubled yields of sorghum in Senegal and in Sudan reduced damage from 85 to 30% in experiments on sorghum and wheat. It is now banned by the EU either for direct use on crops or as a seed dressing. Use of repellents was

65 reviewed by Bruggers (1989). The repellent 9,10 Anthraquinone is in use in Zimbabwe but it is not approved for use in the EU. Attempts have been made in South Africa to spray birds with wetting agents such as dilute molasses to prevent the birds from thermoregulating, but the large volumes required precluded regular use of the technique (E. van der Walt, pers. comm.).

2.2. Mechanical Control 2.2.1. Explosions Explosions/fire-bombs are or were used to control Red-billed Quelea in Botswana (Fig. 1), the Republic of South Africa, Kenya, Zimbabwe and elsewhere. They are often used in or near wetlands, where spraying with organophosphates is contra-indicated. The technique requires highly trained personnel, specialised equipment to transport the explosives safely, and time to deploy the firebombs at the base of vegetation where the birds are either roosting or nesting. This precludes their use except at small (<5 ha) sites. Cheke et al. (2013) described the method used in Botswana as follows: “The technique involves the detonation of 5 L plastic containers, filled with 2.5 L of a mixture of fuels: one-third diesel to two-thirds unleaded petrol was used in 2009 and 2010, but a 50 : 50 mixture of 1 L of diesel and 1 L of petrol was used in 2005; the addition of diesel keeps the flame alight longer than petrol alone, but also gives rise to smoke. Each plastic container (white opaque containers were used in 2005–2008, but green ones in 2009–2010) is placed beneath a bush where quelea birds are either nesting or expected to roost. Each container has an explosive charge placed beneath it. In 2005 this consisted of 150 g of Trojan C150 cast boosters, 38 × 120 mm of pentolite and a mixture of TNT and RDX, encased in yellow plastic [manufactured by Ensign- Bickford, (Pty) Ltd, South Africa]. Each booster had a hole drilled in the middle, through which red detonating cord (plastic cord, 8 g.m−1; Auxim Tech. Ltd, China) was fed. At the ignition site, about 120 cm of yellow safety fuse of slow-burning (8–10 mm.s−1) gunpowder was placed at the beginning of the cord (total length 1050 m for 233 plastic containers at Kotoloname in 2005), giving approximately 2.5 min between ignition and detonation. The fuse was connected to an electric detonator cord containing a white powdered high-explosive core to set off the detonator. This created a shock wave to the detonating cord, along which it travelled at 6400 m.s−1, exploding each booster in turn. In 2009 and 2010 the explosive used was PowergelTM (see www.oricaminingservices.com/download/file id 4292/for information on its toxicology),

66 a commercially available ammonium nitrate product, with a detonation velocity of 1780 m.s−1 [<6400 m.s−1 for TNT (see above), and <8400 m.s−1 for pentaerythritol tetranitrate (PETN), which was also used in years before 2006], mixed with aluminium powder to enhance its performance. These charges were connected by cordex fuse cord, made of powdered PETN, to a central electric detonator that started the reaction with 1 g of metallic-derived explosives or after being activated by a slow-burning safety fuse of gunpowder (black powder). When the explosion takes place, the fuel mixture is first splashed up onto the trees where it forms a mist and then ignites.” In South Africa, the main explosions are preceded a few milliseconds earlier by a small detonation to scare the birds into the air and, in reed beds, the firebombs are raised on poles. In Kenya, small stones are sometimes placed in packages above the explosive apparatus. Allan (1997) also describes the method with particular reference to reed beds and he also discussed the drawbacks of the method including its expense and dangers. There is also a security issue given the involvement of explosives and suitably trained military personnel are often required to oversee the operations. This may explain why explosions are not used in Tanzania, for instance. The method has obvious dangers for the operators and the environment immediately affected by the explosions, but provided that suitable precautions are taken threats to personnel, villagers and livestock can be minimised. It was long thought that explosions were less damaging to the environment than the use of fenthion (e.g. see Meinzingen et al. 1989, Allan 1997) and Jaeger & Elliott (1989) describe people eating the blown-up birds as benefitting from “a much appreciated source of uncontaminated quelea for food”. However, Cheke et al. (2013) showed that explosions are by no means environmentally benign. Many non-target organisms can be killed or maimed (see also reports cited in McWilliam & Cheke 2004), soil is contaminated with concentrations of total petroleum hydrocarbons (TPHs) and phthalates (from the plastic) ranging from 0.05 to 130.81 (mean 18.69) μg.g−1 and from 0 to 1.62 (mean 0.55) μg.g−1, respectively, in the craters formed by the explosions, but the values declined to means of 0.753 and 0.027 μg.g−1 at 10m away. Dead birds will also be contaminated and thus unfit to eat. One year after an explosion, mean TPHs of 0.865 and mean phthalates of 0.609 were still detectable in the soil. In addition, remains of the plastic did not degrade and littered the sites for years after an explosion, bushes and other vegetation were badly burnt, although the bushes tended to recover unless their trunks were broken, and craters were formed at each firebomb location damaging

67 the soil. More than 1% of the area encompassed by the explosion was damaged in this way.

Figure 1. Explosion at a quelea roost, Botswana, 2009. (Photo. R.A.Cheke).

2.2.2. Nest destruction and chick harvesting Bashir (1989) describes how in some communities such as in western Sudan, quelea nests are destroyed by pulling them out of trees with hooks on the ends of long poles or by cutting down the nest trees or with fire including use of flame-throwers. Nest destruction is only useful if conducted after the birds have laid their eggs, otherwise the birds simply repair nests or move elsewhere to breed, and before any fledglings can fly. Amongst various control methods tried against birds in reed beds where chemical control was prohibited, Garanito et al. (2000) concluded that mechanical destruction of breeding and roosting habitat manually or using tractors dragging brushing equipment was the most cost-effective technique.

68 Removal of chicks from nests for later consumption for food is also widely practised. Fig. 2 shows the results of a chick harvesting session in Tanzania. Pelham (1998) reported that up to 3.78 kg of chicks could be harvested per person per hour in Zimbabwe.

Figure 2. Quelea chicks harvested from a colony near Dodoma, Tanzania (Photo: R.A.Cheke).

2.2.3. Trapping It is well known that in many parts of Africa, people eat quelea as the birds provide a nutritious source of protein (Jaeger & Elliott 1989). Indeed quelea colonies are sometimes not reported to pest control authorities when the villagers want to exploit them for food. Various means of trapping the birds are described below. Quelea consumption for food varies regionally and with the preferences of different ethnic groups but it is known to occur in parts of Botswana, Cameroon, Chad, Ethiopia, Kenya, Nigeria, Senegal, Tanzania, Zambia and Zimbabwe. 2.2.3.1. Chad Traps In Chad, farmers adapted nets used for fishing to capture quelea. One of three types of nets used were triangular, suspended on long hand-held poles, and held open in front of roosts and closed to capture birds frightened into them. In this way, on moonless nights at tree roosts about 1,200,000 birds were caught over nine weeks (Mullié 2000). As many as 20,000 could be processed per day by a team of six men for later sale in markets as plucked, fried and dried products. The trapping had a

69 negligible effect on the quelea populations but the revenue from sale of the quelea as food partly compensated the villagers for their crop losses. 2.2.3.2. Kondoa basket traps In Kondoa District, near Dodoma in Tanzania, farmers catch quelea using basket traps woven in star grass Cynodon nlemfuensis (Fig. 3, Cheke 2011). The technique is only used in dry seasons when traps are usually placed with the funnel- shaped opening uppermost near water where the birds come to drink when they are attending “daytime or secondary roosts” (Fig. 4) or else they are placed in fields (Fig. 5). Each basket is baited with grain and heads of millet and, sometimes, with a decoy quelea bird to entice others into the traps. In this way 800 or more birds can be caught per trap per day. The birds are then collected through the hole on the opposite side of the trap after the lid (Fig. 6) has been removed. The birds are killed, plucked, de-gutted and then prepared as food in a variety of ways (Fig. 7, Mtobesya 2012). So successful are the captures that surplus birds are sold at roadsides and elsewhere leading to profits that help the farmers to buy goods and pay for school fees (Mtobesya 2012, Manyama et al. 2014).

70 Figure 3. Kondoa basket trap with funnel-shaped opening uppermost. (Photo:

R.A.Cheke).

Figure 4. A farmer placing a Kondoa basket trap into position near water (Photo:

R.A.Cheke).

Figure 5. Kondoa basket trap deployed in a field, near Dodoma,Tanzania. (Photo: R.A.Cheke).

Figure 6. Kondoa basket trap upside down in comparison with Fig. 3, showing the hole closed with a lid. (Photo: R.A.Cheke).

Figure 7. Adult quelea caught at Kondoa in basket traps prepared for eating (Photo: R.A.Cheke).

Figure 8. A single-hole version of the wire mesh copy of the Kondoa basket trap (Photo: R.A.Cheke).

2.2.3.3. Kondoa basket traps made of wire mesh Given the success of the Kondoa basket traps in catching quelea, experiments were conducted with artificial versions made of wire-mesh (Fig. 8, Mtobesya 2012).

73 The traps were deployed in the same manner as with the traditional model and found to be superior in their catching ability. Also, wire-mesh traps with 3 entrance holes caught more birds than those with 1 or 2 holes and all of the wire mesh versions caught more than traditional grass basket traps, with peak catches of all traps between 0800 and 1000 hours and between 1500 and 1700 hours (Mtobesya 2012).

2.2.3.4. Funnel traps Mitchell (1963) described how very large funnel traps could be used to trap pest birds such as Red-winged Blackbirds Agelaius phoeniceus, Common Grackles Quiscalus quiscala, Common Starlings Sturnus vulgaris and Brown-headed Cowbirds Molothrusater in the United States of America by luring them in with very bright lights (five 1,000 W floodlights). In 101 operations, 672,000 birds were caught and the three best catches yielded 80,000 – 120,000 birds per night. Similarly, such methods are used to capture Common Starlings Sturnus vulgaris in Tunisia in stands of broad- leaved trees such as Eucalyptus, with up to 15,000 caught per night (Elliott et al. 2014). Trials of large funnel traps with attractant strong light have been conducted in Tanzania but with limited success (Figs 9 & 10, Elliott et al. 2014). 2.2.3.5. Miscellaneous indigenous trapping methods Allan (1997) illustrated a variety of basket trap and the means to catch termites to bait it with. Other methods include stick and box traps for which a string is pulled to close the box onto the birds eating bait below, sticky bird lime attached to branches and throwing sticks into the midst of a quelea flock. 2.2.3.6. Mist nets Mist nets are efficient means of catching flying birds but run the risk of catching non-target birds too, which may be killed or injured if not removed quickly from the nets by trained personnel. In a trial of mist-netting operations to catch quelea in Tanzania, nearly 4000 were caught in 5 days using an average of 18 mist-nets (12m long x 3m tall) per day. Although this total is minimal compared with the totals in pest flocks, the villagers nevertheless reported that attacks on their crops did decline during the catching (Elliott et al. 2014). If the method is used repeatedly near colonies with eggs rather than chicks, the birds may desert the colony. Mist nets can also be used to catch birds at roosts (Fig. 11) and are being deployed in Tanzania when possible.

Figure 9. Quelea coming to roost near a funnel trap, Tanzania. (Photo: R.A.Cheke).

Figure 10. A funnel trap erected in front of a quelea colony, near Dodoma, Tanzania. The light source is the black object to the left of the Perspex panel. (Photo: R.A.Cheke).

Figure 11. Quelea trapped in a mist net, Tanzania. (Photo: R.A.Cheke). 2.2.3.7. Roost traps Mtobesya (2012) modified the trap roost concept (see section 2.3.6. below) by devising a “roost trap” consisting of netting that could be drawn over a rigid frame erected over a “trap roost” of Typha grass. Once the birds had settled to roost the net was pulled over the roost trapping the birds which could then be chivvied into a funnel at one end for capture. Approximately 10,000 quelea were caught per night during trials conducted in Tanzania but the catch also included some non-target species although the majority of these were weaver bird species that sometimes also damage crops (Mtobesya 2012).

2.3. Cultural Control 2.3.1. Planting and harvest time manipulation Elliott (1979) and Bullard & Gebrekidan (1989) drew attention to how crop damage by quelea can be minimised if the timings of planting and harvesting can be arranged such that the crop can be harvested when there are few or no quelea present. This method is apt when irrigation facilities are available. Thus in the lower Awash river valley in Ethiopia if irrigated sorghum is planted in September it can be harvested in December when quelea are absent (Bullard & Gebrekidan 1989).

76 Similarly, irrigated rice can be timed for harvesting in mid-May to mid-June in parts of Chad and Cameroon when there are no quelea pests there (Elliott 1979). Harvest time manipulation can also be achieved by growing early-maturing varieties of crops. Even if the latter do not completely escape attack they will be vulnerable for shorter periods than conventional crops (Bullard & Gebrekidan 1989).

2.3.2. Weeding It is important for farmers to keep their fields as weed-free as possible since quelea are attracted to weed seeds and may thus attack crops that they might otherwise ignore (Luder 1985, Rodenburg et al. 2014).

2.3.3. Alternative Crops Crop substitution whereby a crop such as maize, which quelea birds do not attack, is planted instead of vulnerable crops such as millet or sorghum. As maize requires more water to thrive than do millet or sorghum, this measure will only succeed if there is adequate rainfall or irrigation is possible. Other crops that are not attacked by quelea such as groundnuts could also be grown as substitutes.

2.3.4. Quelea resistant crops Bullard & Gebrekidan (1989) described how plant breeders can produce crop cultivars that have morphological or chemical characteristics that are unpalatable to quelea and Tarimo (2000) reported how bird resistance in sorghum is imparted by the cyanogenic glycoside known as dhurrin. However, unless the majority of farmers in an area plant resistant varieties the birds will simply move away from them to seek more palatable cultivars nearby. Furthermore, resistant varieties with high concentrations of tannins are less palatable to people than conventional varieties.

2.3.5. Protecting crops with netting Allan (1997) described a variety of methods whereby nets were used to cover crops and thus prevent birds from attacking them and Elliott & Bright (2007) recommended covering rice fields with nets to reduce quelea damage in Nigeria. This was tested and found to be worthwhile, with damage varying from 0 to 4% with netting

77 compared with 2.7 to 18.8 % with bird scaring. Yields ranged from 565 to 1,448 kg.ha- 1 with netting, but were 296 to 1,250 kg.ha-1 with bird scaring (Ajayi et al. 2007). Allan (1997) illustrated how black cotton threads and metallic tapes can be deployed over crops to deter quelea birds. Such methods may be appropriate for commercial farmers or for small-scale cropping but the expense and labour needed to erect and maintain the systems negates their value for subsistence farmers in general. There is also a tendency for quelea to become habituated to such methods and after a few days they may ignore them.

2.3.6. Trap roosts As it is known that quelea often roost in stands of sugar cane Saccharum officinarum and Napier grass Pennisetum purpurum, these crops have been deliberately sown to act as “trap roosts” (Jarvis & La Grange 1989, Allan 1997). After settling into roosts, the birds then presented a discrete target that could be easily controlled with avicide. Ideally, the roosts should be planted 100m x 100m, with tracks for access on each side. They should also be grown within a few hundred metres of water where the birds can drink before roosting and away from other thickets or similar vegetation that the birds could move to. 2.3.7. Scaring 2.3.7.1. Scaring by humans Bird-scaring methods were reviewed by Bashir (1989). These include visual techniques such as scarecrows, flag-waving and loud noises created by elaborate systems of tins and rattles activated by pulling a connecting string or by cracking whips (Fig. 12). In addition, missiles may be hurled at the birds or shot from catapults or mud is sometimes flicked at the pests from the ends of sticks. All such methods are time- consuming, often conducted by children who are thus absent from schools, and may be effective locally but scared birds will move to other fields where there are no scaring activities. In Ogun State, Nigeria, scaring costs may account for as much as 50% of production costs (Elliott & Bright 2007). 2.3.7.2. Scaring with falcons In Botswana, experiments have been conducted using Lanner Falcons Falco biarmicus to scare quelea away from sorghum crops in the Pandamatenga area (H. Modiakgotla, pers. comm., Gaemengwe 2014). The farmers there reported that the method gave good results and they supported use of the method as it had led to good and high tonnages due to reduced bird damage (H. Modiakgotla, pers.comm., Oct

78 2016). The latter was estimated as 12.1% of sorghum heads damaged on average in 13 fields where falcons were not deployed, but was about half this figure at 6.3% in 6 fields where the falcons were flown (H. Modiakgotla, pers.comm.). However, use of falcons has only been tested in the commercial farms in the Pandamatenga area and has not been applied to protect crops grown by subsistence farmers.

Figure 12. A villager cracking a whip at the edge of her millet field to scare away any quelea, Tanzania. (Photo: R.A.Cheke).

2.3.7.3. Commercial bird-scaring devices In Europe and elsewhere, machines that produce loud bangs at set intervals are available commercially for farmers (e.g. the Bangalore bird-scarer, see http://www.nomorebirds.co.uk/bangalore%2Dbird%2Dscarer%7E230) and it is also possible to purchase varieties that produce species-specific alarm calls or predator calls to scare bird pests (e.g. see http://www.birdstop.co.uk/bio- acoustic_bird_scarers.asp). This approach was tested against quelea using Bird X- Pellers (http://www.bird-x.com/) by Garanito et al. (2000) and trials with similar devices conducted at Pandamatenga in Botswana (H. Modiakgotla, pers. comm.). However, such devices are expensive and the birds are likely to become habituated to them, as they will to other noises used for scaring such as drum beats and tractor horns. An additional possibility is to develop an unmanned aerial vehicle (UAV) or drone that can fly over quelea gatherings and scare them with appropriate noises or with predator- shaped machines. Two such devices have recently been developed for other bird pests (BirdX 2016).

79 2.4. Biological control Barre (1974) reviewed the parasitology of Q. quelea to seek potential biological control agents, but failed to identify any pathogen capable of causing an epizootic that might limit the bird’s populations. Barre recommended a worldwide survey of avian viruses occurring outside Africa and experimental checks to see if any were highly pathogenic to quelea. Q. quelea are hosts to a variety of blood parasites (see Durrant et al. 2007 for surveys of some found in Q. q. lathamii), but none seem to cause morbidity. However, some taxa may be species specific and, if so, there is a possibility that future progress in genetic manipulation might allow the introduction of lethal forms of haematozoa. Quelea are taken by a variety of predators (Thiollay 1989) but these have a negligible effect on the huge numbers of the pest birds, except occasionally when flocks of storks and birds of prey locate breeding colonies. Quelea and their colonies have a sharp and distinct odour, quite unlike that of any other bird’s odour. Since many other species with acute olfactory abilities also have sharp odours, this suggests that odour-based cues serve a communication function in quelea. Likely functions include acting as either a group-specific identification mechanism and/or as a colony- or roost-locating mechanism for either new entrants to colonies, or roosts, and birds returning to their nests or roosts after foraging. In attempt to identify any biologically active odours that might show promise as attractants to traps or as repellents, samples from birds of the nominate race Q. q. quelea were analysed but no compounds of interest were found (R. A. Cheke, D.R. Hall & D. Farman, unpublished data). Samples from other subspecies (Q. q. lathamii and Q. q. intermedia) were also analysed but again no promising compounds were detected. However, none of these samples were from actively breeding birds, which should be the subject of future research on this topic.

3. COMPARISON OF RED-BILLED QUELEA CONTROL METHODS WITH THE DESERT LOCUST PREVENTIVE CONTROL APPROACH

Lethal control of quelea is only advised when the birds are posing a direct threat to a crop (Ward 1972, 1973, 1979). The mere presence of the bird does not justify lethal control as they are often innocuous, especially when their preferred grass food is plentiful. Thus, strategies for dealing with the birds differ markedly from the “preventive control” approach (FAO 2001, van Huis et al. 2007, Magor et al. 2008) applied to other

80 migrant pests such as the Desert Locust Schistocerca gregaria, and the “strategic control” policy for control of the African armyworm Spodoptera exempta (Rose et al. 2000, Cheke & Tucker 1997). For locusts and armyworm, control strategies require “off the crop monitoring” and lethal control as soon as the pest’s populations rise. In this way, if all proceeds according to plan, the pest’s population is prevented from reaching numbers high enough to cause severe damage to crops. Also, locusts and armyworm are remarkable insofar as they change “phase” from the solitarious state to the gregarious condition, a change associated with accelerating population growth that does not occur in birds. One of the aims of the strategic approach to locust and armyworm control is to ensure that the pests do not succeed in changing phase to become gregarious and swarm. Elliott (2000) pointed out that FAO’s approach to quelea control was to adopt Integrated Pest Management (IPM) approaches whenever possible and only to use lethal control as a last resort. The IPM approaches to be tried include many of the options listed above such as “modifying crop husbandry, planting time, weed reduction, crop substitution, bird scaring, exclusion netting etc.”

4. RECOMMENDATIONS FOR QUELEA CONTROL MINIMISING THE EXTENT OF FENTHION USE 4.1. Forecasting and control planning Quelea breeding colonies are often not located in time to control them before the fledgling birds, the juveniles responsible for much of the crop damage at this stage in the birds’ life cycles, have left their nests. Similarly, if birds which are attacking crops can breed successfully then the populations available to attack crops will be augmented. However, if the efficiency of control operations could be improved, then the quantities of fenthion used could be reduced. One way of improving the efficiency of control strategies is to detect the presence of suitable quelea breeding areas by satellite imagery (Wallin et al. 1992) or to forecast where the birds are likely to breed. Given that the birds’ migrations and breeding opportunities are determined by patterns of rainfall (Ward 1971), it is possible to devise forecasting systems to predict where the birds are likely to breed and, thus, to concentrate activities in search of the colonies to areas where the birds are likely to be for control purposes (Cheke et al. 2007). A scheme based on the model described by Cheke et al. (2007) that used satellite- derived rainfall data and knowledge of the threshold amounts of rainfall needed to (a)

81 initiate the migrations at the start of seasons (“early rains migrations”) and (b) to permit the birds to breed was maintained online as a forecasting system from 2001 to 2009, but fell into disuse when the funding for it ceased. That system was for southern Africa only, dealing with populations of Q. q. lathamii, but it is possible to develop a similar system for a pan-African set of forecasts and a prototype model for East Africa was developed (J. Venn & R. A. Cheke, unpublished, Mtobesya 2012). Regrettably, no system of forecasting where quelea roosts will appear in dry seasons has been devised, other than a recommendation to survey sites known to be traditional insofar as they are used regularly year after year.

4.2. Fenthion dosages There is marked variation in the dosages used by different control organisations. For instance at a workshop in Kenya during May 2005, it was reported that Sudan controlled their birds successfully at rates of 1 l.ha-1 (occasionally only 0.5 l.ha-1), DLCO-EA usually used 2-4 l.ha-1 but South Africa reported use of dosages ranging from 7 l.ha-1 up to as high as 14 l.ha-1. Clearly, less fenthion will be used if the dosage is minimised, with 2-4 l.ha-1 recommended. Amounts used by Tanzania during the 2012-2016 period were mostly within this range and were generally lower in the later years (Table 1), a difference which was significant for the roosts (ANOVA, p < 0.0001) so an encouraging trend of gradual dosage reductions was achieved. Interestingly the dosages sprayed on colonies (sample mean 2.66) were consistently and significantly less than those deposited on roosts (sample mean 3.51; Welch two sample t test, t = 8.82, d.f. = 171, p < 0.0001), so if this result is also true elsewhere other control teams could probably minimise potential environmental damage by reducing dosages sprayed on roosts to the levels used on colonies without affecting kill rates.

Table 1. Dosages of fenthion (litres.ha-1) sprayed on quelea roosts and colonies in Tanzania during the 2012-2016 seasons, from analyses of data reported by Mutahiwa (2016). n = sample size; SD = standard deviation.

Year 2012 2013 2014 2015 2016

Roosts

Mean litres.ha- 4.79 4.03 4.20 2.72 2.98 1

82 n, SD 21, 3.95 12, 2.50 10, 2.85 30, 0.56 31, 1.92

Range 1.25 - 15 1.5 - 10 1.67 - 10 2.22 – 4.76 0.55 – 10

Colonies

Mean litres.ha- 1.63 3.02 3.08 2.5 2.47 1 n, SD 11, 0.76 10, 2.92 33, 1.82 1, - 31, 1.92 Range 0.67 - 1.43 – 0.73 - 10 - 0.62 - 10 3.33 11.11

4.3. Fenthion Application methods Studies of environmental effects of fenthion applications revealed that in some cases sprays were conducted incorrectly with regard to speeds and directions of movements of ground-sprayers in relation to wind directions, failing to cease sprays when turning vehicles, incorrect nozzle positioning, equipment maintenance and missed targets during aerial applications (Cheke et al. 2013). Therefore regular training and supervision of pest control workers is recommended, as is correct use of equipment to minimise excessive contamination of the environment and risks to personnel. Furthermore use of the most appropriate equipment may reduce quantities of fenthion needed. For instance, in Tanzania use of ground-based sprays with Micronair AU8000 sprayers required 10% of the volume used by aircraft (B. Mtobesya, pers. comm., October 2016). Training of farmers in quelea biology and IPM principles for damage avoidance and minimisation through farmer field schools is advised. Spray operators also need to know what to do in the case of accidental contamination, which involves washing any affected area immediately with soap and water and discontinuing any further operations until the cause of the contamination has been corrected. The use of unmanned aerial vehicles (drones) for spraying operations would ensure accurate targeting but to date the size of maximum possible payloads has precluded their use. Recent developments have succeeded in increasing payload possibilities up to 80 litres (E. van der Walt, pers.comm., Nov. 2016) so in future this technique may become usable. 5. RECOMMENDATIONS FOR QUELEA CONTROL WITHOUT FENTHION: A STRATEGY FOR BIRD CONTROL USING SELECTED METHODS OF CONTROL

83 The appropriate control measure to be adopted against quelea will vary depending upon the circumstances. Most of the IPM measures described above will seldom be effective on their own, except at scales when infestations are small relative to massive swarms of millions of birds or huge breeding colonies: one colony in March 1998 at Malilangwe in Zimbabwe was 20km long and 1km wide, with nests at densities of 30,000 nests per hectare (Dallimer 2000). However, by combining judicious planning of crop choice and of planting and harvesting times before any expected quelea arrivals with environmentally benign control methods, the cultural control and IPM strategy will succeed under some circumstances. Nevertheless, faced with crop raids by huge quelea flocks, farmers and those in their country responsible for the control will have little choice other than to authorise lethal control with pesticides (cyanophos or other alternatives to fenthion) or explosives. The scheme outlined in Figure 13 is an attempt to provide guidelines on how to minimise chemical use and their unintended effects, together with suggestions on the circumstances under which alternative methods would be appropriate within the context of subsistence agriculture rather than commercial operations. Characteristics of various alternatives to fenthion are summarised in table 2.

84 Table 2. Summary information on alternatives to use of fenthion for quelea control.

Method Application Mode of action Advantages Disadvantages Socio-economic issues

Chemical Methods

Cyanophos Spray Lethal Less toxic than fenthion. High risk of environmental More expensive than organophosphate impacts. fenthion. (Fenthion costs avicide. approx. US$10 per litre Killing action takes longer [www.yufull.com]). Requires than fenthion, so could lead trained personnel and to more secondary poisoning expensive equipment, e.g. than fenthion. applied by Government personnel or international control agencies.

Alphachloralose Narcotic added Immobilises birds. Minimal pollution. Risk to non-target birds. Labour intensive. to bait grain or water Requires birds to be found and killed.

Mesurol Sprayed on Carbamate Risk to non-target birds and Expensive. Approx. US$300 seed heads of pesticide, active mammals. Highly toxic to per litre. Labour intensive. crops or ingredient aquatic fauna. Now not applied as methiocarb. recommended for direct use seed dressing. Repellent. Deters on crops, only as seed birds from crops. dressing. Now banned by the EU.

Method Application Mode of action Advantages Disadvantages Socio-economic issues

Mechanical Methods

Explosions Diesel/petrol- Lethal No organophosphates Risk to non-target birds and Expensive. Requires trained eum firebombs or aircraft involved. mammals. personnel and expensive detonated equipment, e.g. applied by beneath birds. Petroleum product residues Government personnel. pollute soil.

Vegetation damage.

Fire and security risks. Only possible for small sites (< 5ha)

Nest destruction Human Lethal No pollution. Labour intensive. Often Profits possible, if surplus and chick intervention possible only on small scale chicks sold as food or harvesting with sticks on Provides source of but see Pelham (1998). livestock feed. poles or flame protein. throwers

Trapping with Human Lethal No pollution. Labour intensive. Often Profits possible, if surplus Chad or basket intervention possible only on small scale. birds sold as food or traps or other with various Provides source of livestock feed, e.g. annual trapping methods trap designs. protein. value of US$50,000 to 100,000 in Chad.

Trapping with mist Human Lethal No pollution. Labour intensive. Often Profits possible, if surplus nets intervention possible only on small scale. birds sold as food or with various Provides source of livestock feed. trap designs. protein. Needs supervision to avoid non-target mortalities.

86 Locally sourced nets at cost of only US$5 each.

Roost traps Planting of Lethal High risk of environmental Loss of area where crops fodder crops to impacts. could be planted. attract birds to roost, followed by spraying.

Method Application Mode of action Advantages Disadvantages Socio-economic issues

Cultural Methods

Planting and Planting of Avoidance of No pollution. Not always possible. Agronomic advice needed. harvest date fast-maturing quelea attacks on Requires knowledge of likely manipulation crop varieties crops. quelea movements into and or early out of cropped zone. harvesting to minimise risk Alternative crops may not of quelea at flourish in zone, especially in harvest. very arid areas.

Planting of crops that are not susceptible to attacks.

Netting crops Covering Protective No pollution. Only on small scale. May just Expenditure on nets and crops with divert birds to crops with no poles or gantry to rig them on netting netting present.

Scaring by people Farmers and Birds frightened No pollution. Labour intensive. Often Prevents children attending their children away from crops by possible only on small scale. school. Labour intensive. scare birds waving and noise.

87 May just divert birds to crops with no scarers present.

Scaring with Release of Birds frightened No pollution. Requires trained birds and So far only used by large falcons falcons near away from crops. bird handlers. May just move scale commercial farmers not quelea flocks. quelea to fields where by subsistence farmers. falcons not deployed.

Biological Methods

No successful biological control agents identified to date.

Figure 13. Flow diagram of decisions for planning quelea control by responsible authorities for subsistence farmers.

89 Acknowledgements

I am grateful to Hendrick Modiakgotla, Boaz Mtobesya, Etienne van der Walt and Dr Clive Elliott for information and the latter for comments on an earlier draft. I thank Dr Stephen Young for statistical help and I am grateful to the Rotterdam Convention Secretariat in FAO for commissioning this report and Mohammed El Hady Sidatt, Christine Fuell and Elisabetta Tagliati of the Secretariat’s staff for comments on an earlier draft.

References

Ajayi, O., Nwilene, F.E., Gregorio, G., Okoruwa, V., Oladimeji, O.A. & Athanson, B. (2007) Services Agreement No. OFADA 16: Demonstration and financial feasibility of the use of nets to prevent bird damage. Unpublished report to the Pro-poor Opportunities through Commodity and Service Markets (PrOpCom) programme. Africa Rice Center (WARDA), Ibadan, Nigeria.

Allan, R. G. (1997) The grain-eating birds of Sub-Saharan Africa: identification, biology and management. Natural Resources Institute, Chatham, UK.

Barre, N. (1974) La pathologie de Quelea quelea (L.) (Passériformes — Ploceidae). Etude des possibilities de lutte par des agents pathogenes. Institut d'Elevage et de Médecine Vétérinaire des Pays Tropicaux, Maisons Alfort, France, 116 pp.

Bashir, E. A. (1989) Traditional African practices for preventing bird damage. In: Bruggers, R.L. and Elliott, C.C.H. Quelea quelea Africa’s Bird Pest. Oxford Univ. Press pp 248–261.

Bird-X (2016) http://www.bird-x.com/remote-control-drone-pages-64.php. Accessed 18 November 2016.

Bullard, R. W. & Gebrekhidan, B. (1989) Agronomic techniques to reduce quelea damage to cereals. In: Bruggers, R.L. and Elliott, C.C.H. Quelea quelea Africa’s Bird Pest. Oxford Univ. Press pp 281–292. Bruggers, R. L. (1989) Assessment of bird-repellent chemicals in Africa. In: Bruggers, R.L. and Elliott, C.C.H. Quelea quelea Africa’s Bird Pest. Oxford Univ. Press pp 262–280.

Cheke, R. A. (2011) An indigenous trap for mass capture of Red-billed Quelea. In Club announcements. Bulletin of the British Ornithologists’ Club 131: 74.

90 Cheke, R. A. (2014) in CABI, 2014. Red-billed Quelea Quelea quelea [original text by RA Cheke]. In: Invasive Species Compendium. Wallingford, UK: CAB International. http://www.cabi.org/isc/?compid=5&dsid=66441&loadmodule=datasheet&page=481&site=144.

Cheke, R.A., Adranyi, E., Cox, J.R., Farman, D.I., Magoma, R.N., Mbereki, C., Mcwilliam, A.N., Mtobesya, B.N. & Van Der Walt, E. (2013) Soil contamination and persistence of pollutants following organophosphate sprays and explosions to control red-billed quelea (Quelea quelea). Pest Management Science 69: 386–396.

Cheke, R.A., McWilliam, A.N., Mbereki, C., van Der Walt, E., Mtobesya, B.N., Magoma, R.N., Young, S. & Eberly, J.P. (2012) Effects of the organophosphate fenthion for control of the red-billed quelea Quelea quelea on cholinesterase and haemoglobin concentrations in the blood of target and non- target birds. Ecotoxicology 21:1761–1770.

Cheke, R.A. & Tucker, M.R. (1995) An evaluation of the potential economic returns from the strategic control approach to the management of African armyworm Spodoptera exempta populations in East Africa. Crop Protection 12: 91–103.

Cheke, R. A., Venn, J. F. & Jones, P. J. (2007) Forecasting suitable breeding conditions for the Red- billed Quelea Quelea quelea in southern Africa. Journal of Applied Ecology 44: 523–533.

Dallimer,M. (2000) Migration patterns of the red-billed quelea Quelea quelea in southern Africa: genetics, morphology and behaviour. Ph.D. Thesis, University of Edinburgh, UK.

Durrant, K. L., Reed, J. L., Jones, P. J., Dallimer, M., Cheke, R. A., McWilliam, A. N. & Fleischer, R. C. (2007) Variation in haematozoan parasitism at local and landscape levels in the red-billed quelea. Journal of Avian Biology 38: 662–671.

Elliott, C.C.H. (1979) The harvest time method as a means of avoiding quelea damage to irrigated rice in Chad/Cameroon. Journal of Applied Ecology 16: 23–35.

Elliott CCH (1989a) The pest status of the quelea. Pp 17–34 in: Bruggers R, Elliott CCH (eds) Quelea quelea: Africa’s Bird Pest. Oxford University Press, Oxford, UK.

Elliott CCH (1989b) The quelea as a major problem in a food deficient continent. Pp. 90–99 in: Mundy PJ, Jarvis MJF (eds) Africa’s feathered locust. Baobab Books, Harare, Zimbabwe.

Elliott, C.C.H. (2000) FAO’s perspective on migrant pests. Pp. 17–24 in Cheke, R.A., Rosenberg, L.J. & Kieser, M. (eds.) Proceedings of a Workshop on Research Priorities for Migrant Pests of Agriculture

91 in Southern Africa, Plant Protection Research Institute, Pretoria, South Africa, 24–26 March 1999. Natural Resources Institute, Chatham, UK.

Elliott, C. & Bright, E. (2007). Review of the bird pest problem and bird scaring in South west Nigeria. PrOpCom Monographs. Abidjan, Nigeria, Department of International Development.

Elliott, C.C.H., Mtobesya, B.N. & Cheke, R.A. (2014) Alternative approaches to Red-billed Quelea Quelea quelea management: mass-capture for food. Proc.13th Pan African Ornithological Congress, Arusha, Tanzania, 14-21 October 2012. Ostrich 85: 31–37. DOI:10.2989/00306525.2014.900827.

FAO (2001) Desert Locust Guidelines. Second Edition. Food and Agriculture Organization of the United Nations, Rome, Italy.

Gaemengwe, B. (2014) Quelea control alternatives. Agrinews 44: 6–7.

Garanito, M., Botha, M.J. & van der Westhuizen, L. (2000) Alternative strategies for red-billed quelea population management. Pp. 119–123 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.

Jaeger, M.E. & Elliott, C.C.H. 1989. Quelea as a resource. In Bruggers, R.L. and Elliott, C.C.H. (eds.) Quelea quelea Africa’s Bird Pest. Oxford Univ. Press, pp 293–316.

Jarvis, M.J.F. & La Grange, M. (1989) Conservation, quelea control and the trap roost concept. In Mundy, P.J. & Jarvis, M.J.F. (eds.) Africa’s feathered locust. Baobab books, Harare, Zimbabwe, pp. 130–134.

Luder, R. (1985) Weeds influence red-billed quelea damage to ripening wheat in Tanzania. J. Wildl. Manage. 49: 646–647.

Magor, J.I., Lecoq, M. & Hunter, D.M. (2008) Preventive control and desert locust plagues. Crop Protection 27: 1527–1533.

Manyama, F., Nyahongo, J. W. & Røskaft, E. (2014) Factors affecting attitudes of local people toward the red-billed quelea (Quelea quelea) in Kondoa District, Tanzania. International Journal of Biodiversity and Conservation 6: 138–147.

McWilliam, A. N. & Cheke, R. A. (2004) A review of the impacts of control operations against the Red- billed Quelea (Quelea quelea) on non-target organisms. Environmental Conservation 31: 130–137.

Meinzingen, W. W., Bashir, E. A., Parker, J. D., Heckel, J.-U. & Elliott, C. C. H. (1989) Lethal control of quelea. In: Quelea quelea Africa’s Bird Pest, ed. R.L. Bruggers & C.C.H. Elliott, pp. 293–316. Oxford, UK: Oxford University Press.

Mitchell, R.T. (1963) The floodlight trap – a device for capturing large numbers of blackbirds and starlings at roosts. United States Department of the Interior Fish and Wildlife Service Special Scientific Report: Wildlife 77: 1–14.

Mtobesya, B. N. (2012) Non-chemical control of the Red-Billed Quelea (Quelea Quelea) and use of the birds as a food resource. Unpublished M.Phil. thesis, University of Greenwich, U.K. available from http://gala.gre.ac.uk/9814/.

Mullié, W. C. (2000) Traditional capture of red-billed quelea Quelea quelea in the Lake Chad basin and its possible role in reducing damage levels in cereals. Ostrich 71: 15–20.

Mullié W. C., Diallo, A. O., Gadji, B., Ndiaye, M. D. (1999) Environmental hazards of mobile ground spraying with cyanophos and fenthion for quelea control in Senegal. Ecotoxicology and Environmental Safety 43: 1–10.

Mutahiwa, S. (2016) Current quelea outbreak and control in Tanzania. Unpublished report, Plant Health Services, Ministry of Agriculture, Livestock and Fisheries, Dar-es-Salaam, Tanzania.

Pelham, M. (1998) Quelea harvesting in the Malilangwe conservancy. Honeyguide 44: 149–151.

Pope, G. G. & King, W.J. (1973) Spray trials against the red-billed quelea (Quelea quelea) in Tanzania. Misc. Rep. no 12, Centre for Overseas Pest Research, London, U.K.

Rodenburg, J., Demont, M., Sow, A. & Dieng, I. (2014) Bird, weed and interaction effects on yield of irrigated lowland rice. Crop Protection 66: 46–52.

Rose, D., Dewhurst, C. F. & Page, W. W. (eds.) (2000) The African armyworm handbook. Second edition. Natural Resources Institute, Chatham, UK.

Tarimo, T. M. C. (2000) The cyanogenic glycoside dhurrin as a possible cause of bird-resistance in Ark-3048 sorghum. Pp. 103–111 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.

93 Thiollay, J.-M. (1989) Natural predation on quelea. In: Quelea quelea. Africa’s Bird Pest, ed. R.L. Bruggers & C.C.H. Elliott, pp. 216–229. Oxford, UK: Oxford University Press. van der Walt, E. (2000) Research at PPRI on environmental effects of quelea control operations. Pp. 91–95 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. van Huis, A., Cressmann, K. & Magor, J. I. (2007) Preventing desert locust plagues: optimizing management interventions. Entomologia Experimentalis et Applicata 122: 191–214.

Wallin, D.O., Elliott, C.C.H., Shugart, H.H., Tucker, C.J. & Wilhelmi, F. (1992). Satellite remote sensing of breeding habitat for an African weaver-bird. Landscape Ecology 7: 87–99.

Ward, P. (1971) The migration patterns of Quelea quelea in Africa. Ibis 113: 275–297.

Ward, P. (1972) New views on controlling quelea. Span 15: 136–137.

Ward, P. (1973) A new strategy for the control of damage by quelea. Pestic. Abstr. 97–106.

Ward, P. (1979) Rational strategies for the control of queleas and other migrant birds in Africa. Phil. Trans. R. Soc. Lond. B 287: 289–300.

Ward, P. & Zahavi, A. (1973) The importance of certain assemblages of birds as “information-centres” for food-finding. Ibis 115: 517–534.

94 5.2 Annex II : Programme of the meeting Regional workshop on identification of strategy on alternatives to fenthion on Quelea control

AGENDA

Venue: Corinthia Hotel, Khartoum, Sudan Date: 4th – 5thApril, 2017 DATE/TIME ACTIVITY REMARKS Tuesday 4TH April, FACILITATED 2017 BY:

OPENING CEREMONY MASTER OF CEREMONY: 08:00-09:00 Registration of Participants 9:00-09:45 Opening Remarks: Elwathig MukhtarHamid ( FAO representative ad interim)

Ibrahim Adam Ahmed Al- Dekhairi( Minister of Agriculture and Forestry of Sudan)

Group photograph 09:45 – 10:15 COFFEE/TEA BREAK 10:15-10.30 Purpose and structure of the workshop Mohamed El hady Sidatt (RC Secretariat) 10:30-10:45 Introductions of participants All 10:45-11:45 Review of alternatives to Fenthion on Pr Robert Cheke Quelea Control Discussion (Consultant) 11:45-12:30 Video of mass capture of quelea in all Tanzania.

12:30-13:30 LUNCH 13:30-15:30 Participants National presentations on Quelea control : Botswana, Chad, Eritrea, Ethiopia, Kenya, Mauritania,

15:30-16:00 COFFEE/TEA BREAK

95 DATE/TIME ACTIVITY REMARKS 16:00-18:00 National presentations on Quelea Participants control : Nigeria, Sudan, Tanzania, Zambia and Zimbabwe, DLCO-EA, IRLCO-CSA Wednesday 5TH FACILITATED April, 2017 BY:

08:30 – 10:00 Drafting strategy on alternatives to Break out groups fenthion 10:00 – 10:30 COFFEE/TEA BREAK 10:30 – 12:30 Presentation of break out group plenary outcomes Discussion 12:30-13:30 LUNCH 13:30 – 15:00 Preparation of draft meeting report Reporters (Robert Cheke, Djekadom Riabe Samuel & Sergei Mutahiwa) 15:00 – 15:30 COFFEE/TEA BREAK 15:30 – 18:00 Conclusions recommendations Closure of the meeting

5.3 Annex III : PARTICIPANTS

No Name Country Email address Mobile cell No. 1 ABDELMONIEM AHAMID SUDAN [email protected] ABDELGADIR 2 Ms FATIMA MOHAMED ALAMIN SUDAN [email protected] 18256423 3 Ms Fawzia Abass Mokhtar Sudan [email protected] 4 Ms Sania Gobara Ibrahim Sudan [email protected] 5 Hussien Osman Abaker Sudan [email protected] 6 Ms. NELOUMTA MADIBE OKALA CHAD [email protected] 7 Mr. DJEKADOM CHAD [email protected] +23568001222 8 Mr. PETER MUNYAO KMWELLE KENYA [email protected] 9 Mr. ROY NYAMAI KITHAE KENYA [email protected] 10 Mr. JOHN NGONDI KATHERU Zambia [email protected] 11 Mr. ELIKANA ELIONA LEKEI TANZANIA [email protected] 12 Mr. SERGEI JOHN MUTAHIWA TANZANIA [email protected] 13 Mr. KWADZANAI MUSHORE ZIMBABWE [email protected] 14 Mr. BENNY MAVHELA ZEMBABWE [email protected]

96 15 Mr. JOHNSON WAZIRI DZASU NIGERIA [email protected] 16 Mr. OLAWALE AKANDE NIGERIA a_ [email protected] 17 Mr. DOUSSOU SIDI MAHMOUD MAURTANIA [email protected] 18 Mr. MOHAMED ELHADY SIDATT RC [email protected] 699497281 Secretraiat

FAO, Tunisia 19 Mr. LOITSENG SEBETWANE BOTSWANA [email protected] 772465410 20 Mr. KUATE SEBUA BOTSWANA [email protected] 21 Mr. ROBERT ALEXANDER CHEKE BRITAIN [email protected] 7722555057 22 Mr. CLIFF NGWATA ZAMBIA [email protected] 0976009495 23 Mr. ZEBADEWOS SALATO AMBA ETHIOPIA [email protected] 24 Ms. ROMAN KASSAHUN ABERRA ETHIOPIA [email protected] 1794117 25 Mr. JOSEPH MOA OTIENO NDEGE TANZANIA [email protected] 460 26 Mr. EFREM UGABATGI TEILE ERITREA [email protected] 2083