Classical biological control of waterhyacinth is difficult to evaluate against the backdrop of active herbicide programs. Two experiments evaluated the additive impact of herbivory by two biological control agents with three different rates of 2,4-D on waterhyacinth growth and development in outdoor concrete mesocosms. The herbicide 2,4-D was applied at three rates: (1) control (no herbicide), (2) reduced (2.1 kg ai ha−1), and (3) operational (4.3 kg ai ha−1). Biomass of waterhyacinth populations was reduced by 16.9% by biological control only, 10.5% by the reduced rate of herbicide alone, 44.6% by the operational rate, and 97.3% and 99.9% by the combination of biological control and the reduced and operational rates of herbicides, respectively. These results quantified the relative contributions of both tactics to waterhyacinth management and posit the question of whether further reductions in 2,4-D rates are possible without sacrificing efficacy.

Waterhyacinths grown outdoors were treated with combinations of 0 or 22 kg ai ha-1 of the plant growth retardant paclobutrazol and 0, 25, or 250 waterhyacinth weevils per culture tank under conditions of low (25% coverage) or high (100% coverage) plant densities. The low plant density treatments simulated a regrowth or colonization situation. Weevil feeding activity, canopy height, plant volume, leaf production, and leaf mortality were monitored biweekly for 22 weeks or until all plants had died. In high plant densities, weevil infestations increased leaf mortality rates while decreasing production rates, causing eventual plant death. At low plant densities, waterhyacinth plants partially compensated for weevil feeding by increased leaf production rates, thus enabling them to survive until the end of the study. Paclobutrazol reduced canopy height for about 3 months but had little effect on weevil feeding intensities. Combinations of weevils and paclobutrazol gave complete control of waterhyacinths regardless of plant densities. The combined effects were synergistic, with accelerated leaf mortality rates exceeding production rates leading to early plant death. This study demonstrated the potential for increasing effectiveness of waterhyacinth weevils using integrated strategies.

Common St. Johnswort is widespread in eastern Canada but it seldom constitutes a serious weed problem. A demographic study conducted in 1993 and 1994 at four typical undisturbed sites indicated that 36 to 96% of established St. Johnswort shoots died during the growing season. Mortality was always associated with infection by a host-specific Colletotrichum gloeosporioides. The leaf-feeding beetle Chrysolina hyperici occurred at all sites and caused maximum midsummer defoliation of 27% in 1993 and 51% in 1994. Healthy plants readily recovered from defoliation during pupation of the fourth instar of the insect in June and following adult estivation in August. Although widespread, C. hyperici populations appear transient and alone do not cause sustained feeding pressure resulting in weed control. When Chrysolina larvae and adults were collected at six field sites and placed on healthy seedlings under controlled conditions, up to 36% of the plants became infected with C. gloeosporioides. Scanning electron micrographs commonly showed Colletotrichum conidia among the setae on legs, tarsal pads, and antennae of adults and larvae. In a series of three experiments conducted under controlled conditions in which Chrysolina larvae and adults were placed on healthy plants after feeding on diseased ones, the incidence of infection ranged from 63 to 100%. Hence, under favorable conditions Chrysolina adults may selectively transmit the pathogen in the field. This study demonstrated the potential of enhancing biological control of weeds by insects with the integration of an effective, host-specific pathogen.

In six years of field studies, early-season releases of the native moth Bactra verutana Zeller were evaluated as a biological control for purple nutsedge (Cyperus rotundus L.). Establishment from adult releases required at least 10 pairs of moths per release; establishment from larval release could be achieved with a single release of two or five newly-emerged larvae per shoot. When shoots averaged 4 or 7.5 cm high at the time of such releases, aboveground growth was reduced 32 to 62% after 30 days. Three to eight weekly releases of larvae reduced the topgrowth of nutsedge planted in rows of cotton (Gossypium hirsutum L. ‘Stoneville 213’) 30 to 68% 6 to 8 weeks after the last release. Weekly releases of about five larvae per shoot repeated three, four, or five times in 1976 and 1977 reduced nutsedge growth so that seed cotton yield equalled that of cotton in plots without nutsedge. MSMA (monosodium methanearsonate) applied at the rate of 2.2 kg/ha 5 and 7 weeks after cotton emergence gave greater control of purple nutsedge topgrowth and a yield of seed cotton equal to that resulting from three to five larval releases of about five larvae per shoot beginning when purple nutsedge averaged 7.5 cm high. When MSMA was applied at 7.5 and 8.5 weeks after cotton emergence, the addition of two early-season releases of larvae were also necessary to achieve nutsedge control and seed cotton yield equal to that produced by three to five releases of larvae.

Waterhyacinth [Eichhornia crassipes (Mart.) Solms] grown in screen-enclosed, outdoor aquaria were treated with factorial combinations of 0, 0.75, 1.5, or 3 kg/ha of the experimental growth retardant EL-509 [α-(4-chlorophenyl)-α-(1-methylethyl)-5-pyrimidine-methanol] and zero, one, two, or three pairs of waterhyacinth weevils (Neochetina eichhorniae Warner)/plant, with three replicates of each weevil-number by retardant level combination. When plant growth was evaluated over a time course, significant effects of weevils and retardant were observed. After 127 days, however, waterhyacinth was reduced as weevil numbers increased, but the growth retardant was ineffective without weevils. Weevils appeared to be more effective when used in combination with the retardant, and the combined effects were additive.

In 1985, 1986, and 1987, pendimethalin at 1.7 kg ai/ha plus 2.2 kg ai/ha cyanazine, 2.2 kg ai/ha atrazine, or 1.1 kg/ha atrazine plus 1.1 kg/ha cyanazine was applied delayed preemergence, early postemergence, and postemergence with and without cultivation to evaluate woolly cupgrass control and corn injury. Results varied from year to year. Dry conditions in 1985 resulted in poorer woolly cupgrass control while cold and wet environments in 1987 resulted in corn injury and reduced yields from postemergence treatments containing cyanazine. Good early-season suppression of woolly cupgrass deteriorated to less than 75% control by the late-season evaluation in all experiments. The best woolly cupgrass control and highest corn yields were usually achieved when herbicide applications were followed by row cultivation. Corn yield increases averaged 28, 17, and 11% in 1985, 1986, and 1987, respectively, when a herbicide treatment was followed by row cultivation. Pendimethalin/triazine treatments followed with a row cultivation adequately suppressed woolly cupgrass in field corn, but adverse environmental conditions often reduced herbicide effectiveness or increased corn injury.

Twelve grass genotypes were evaluated for competitiveness with leafy spurge and herbage yield in two sites in North Dakota. ‘Rebound’ smooth brome, ‘Rodan’ western wheatgrass, ‘Bozoisky’ Russian wildrye, and ‘Arthur’ Dahurian wildrye reduced leafy spurge stem density an average of 63% after 3 yr in a silty clay soil at Fargo. ‘Reliant’ intermediate wheatgrass reduced leafy spurge stem density every year for 3 yr, including an 85% reduction the second year after planting, and consistently produced high herbage yields. Rebound smooth brome and Reliant intermediate wheatgrass averaged 72% leafy spurge reduction 3 yr after seeding in a loamy sand soil at Jamestown. Reliant intermediate wheatgrass and ‘Manska’ pubescent wheatgrass provided the most consistent herbage production, which averaged about 2,000 kg/ha annually for 3 yr.

Quackgrass is a serious weed problem in the northern United States and southern Canada. Field experiments were conducted in 1990 and 1991 to examine the effect of herbicide rate, application timing, quackgrass growth stage, and cultivation on quackgrass control with nicosulfuron and primisulfuron in corn. Nicosulfuron (35 g ai/ha) and primisulfuron (40 g ai/ha) each provided 84% or greater control of two-leaf plants and 93% or greater control of four-leaf quackgrass plants. Sequential applications of both nicosulfuron and primisulfuron provided greater season-long control than single applications of the same rate on two-leaf but not four-leaf quackgrass. Cultivation 10 d after nicosulfuron or primisulfuron application often increased early-season control, but by corn harvest few differences were observed. Postemergence (POST) applications of nicosulfuron and primisulfuron provided at least 88% control, compared with 66 to 86% from an early preplant application of glyphosate. Corn yields were similar regardless of the quackgrass control program.

Biological control is one of the most common approaches used to manage invasive weeds of wetlands and other natural areas. Before candidate agents can be released, research is conducted to support biological control, which can be protracted and expensive, leading to a scientific and potentially lengthy regulatory review. To increase biological control safety, efficacy, and transparency, we suggest that during the early phases of a weed project, the feasibility of the invasive plant as a target should be studied explicitly. Our purpose here is to summarize information of an important invasive weed that can serve to judge whether the project is appropriate. Chinese tallowtree, Triadica sebifera, is one of the worst invasive species invading coastal wetlands and other riparian areas of the southeastern United States. Current management practices have not controlled the spread of this weed into these sensitive habitats. Initial surveys in the plant's native Chinese range for potential biological control agents have recovered several herbivore species that could be developed. These potential agents include defoliators, root and foliage feeders, and gall formers, whose biology, apparent host specificity, and impacts on plant fitness suggest that biological control offers great promise against Chinese tallowtree. When conducted during the initial phase of a project, this type of feasibility study can address potential conflicts of interest and risks, ultimately producing projects that are more effective and safer for biological control.

Production of maize in western Kenya is severely constrained by the parasitic weed striga. Although productivity of maize can be improved through adoption of improved varieties, adoption of such varieties remains low in the region, as the majority of smallholder farmers still grow unimproved open-pollinated varieties (landraces). The performance of two improved hybrid varieties was evaluated against six landraces in striga-infested soils in western Kenya. The varieties were planted in plots under natural striga infestation and were supplemented with pot experiments under artificial infestation. Striga emergence was lower in landraces than in the hybrid varieties in both field and pot experiments. Similarly, height of maize plants at harvest and grain yields were higher in the landraces than in the hybrids. After three continuous cropping seasons, in all treatments, striga seedbank density increased two to seven times. Seedbank increase was higher with hybrids and two of the landraces, ‘Rachar' and ‘Endere'. These results provide an insight into the potential role landraces could play in efforts toward an integrated management approach for striga in smallholder cropping systems. They also highlight the need to develop hybrid maize lines with local adaptation to biotic constraints, specifically striga.

Introduction. Du fait des conditionstropicales humides, le contrôle des adventices est un poste importantet permanent dans les dépenses d’un producteur de fruits aux Antilles.La pratique du désherbage chimique est à ce jour la plus économique,mais l’utilisation récurrente d’herbicides dans les vergers de laMartinique conduit à une érosion des sols et à une pollution des eaux.Pour pallier l’usage d’herbicides, nous avons testé l’associationde volailles à un verger productif afin d’en contrôler l’enherbement. Matériel et méthodes. Le dispositif a été installé surune parcelle de 700 m2 subdivisée en sept placettes de 100 m2. Deux placettes ont constitué les traitements témoins ;elles ont été fauchées à la débroussailleuse. Chacune des cinq autresplacettes a été soumise au pâturage de 5 oies pendant une semaine, puisau pâturage de 20 poulets pendant la semaine suivante, puis laisséesans volailles pendant trois autres semaines. Trois cycles de tellesrotations ont été observés. La production de biomasse par groupebotanique (poacées, cypéracées et autres) dans les placettes expérimentales aété mesurée chaque semaine pendant les trois cycles (4 mois). Unefois par mois l’indice de recouvrement par groupe botanique a étérelevé. Résultats. La biomasse totale de la strateherbacée a diminué sur les placettes pâturées, ce qui a permis lecontrôle biologique des adventices pendant quatre mois. La diversitéfloristique a diminué à cause de la sélection opérée par les volailles pourleur alimentation et le couvert s’est homogénéisé par l’augmentationdu pourcentage de la biomasse des cypéracées dans la biomasse totale.Cette famille peu appétible pour les volailles s’est ainsi étenduesur toutes les placettes pâturées. Discussion. Le pâturagedes volailles, des oies en particulier, s’est révélé être une méthodeefficace pour la maîtrise de l’enherbement en verger. Cependant, dansle cas d’une forte proportion d’espèces non appétibles dans la strateherbacée, d’autres méthodes de désherbage devront être combinées.L’installation d’une couverture herbacée composée d’un mélange deplantes de couverture adaptées à l’association verger et volaillesainsi qu’une meilleure gestion de la rotation des animaux devrontêtre envisagées pour optimiser cette technique.

Introduction. In the water-savingand income-generating agroforestry cropping systems developed andpromoted by the ICRISAT in the Sudano-Sahelian zone of West Africa, particularlyin Niger, fruits of domesticated Ziziphus mauritiana (“applesof the Sahel”) are severely damaged by fruit flies (Carpomya incompleta),and chemical pesticide application poses economic, environmentaland human health problems. In the Bio-reclamation of Degraded Lands (BDL)system, apple of the Sahel is the main high-value crop, while inthe Dryland Eco-Farm (DEF), it is grown alongside watermelon. Sclerocaryabirrea (marula plum) is presently being investigated asa dryland tree species for fruit and oil production, either in orchardsor such systems as BDL. Materials and methods. To completepreliminary results of earlier studies published elsewhere, we collectedand incubated in 2010 ripe fruits of marula plum and watermelonin Sadoré, Niger, where the above-mentioned agroforestry systemsare developed, and we recorded emerging fruit flies. We also conducteda spot-spraying experiment (using GF-120) in an apple of the Sahel orchardin Niamey in 2010; we recorded undamaged and damaged fruits and incubatedthe latter. In the same orchard, we set up in 2011 a fruit fly trappingsurvey targeting the invasive species Bactrocera invadens (Chempac®traps using methyl eugenol as the attractant). Results and discussion.The results showed that only Ceratitis cosyra emergedfrom marula plums, and only Dacus spp. from watermelon.The GF-120 spot-spraying experiment showed that C. incompleta was notattracted / intoxicated, contrary to Ceratitis spp.and Dacus spp.; the sprayed trees yielded significantly moremarketable fruits than unsprayed ones; only C. incompleta emergedfrom damaged fruit. Detection trapping revealed for the first timethe presence of B. invadens in the Sahelian zone ofNiger, including at the time of apple of the Sahel fruit production. Conclusion. This, alongside results of earlier studies,suggests a repellent effect of GF-120 on the monophagous C. incompleta species(Trypetinae), while it is attractive to oligophagous/polyphagousDacinae fruit flies; some of them (e.g., Dacus spp.)damage watermelon, which is part of the DEF system, and producefruit at the same time as the jujube tree. The prospects for harnessingthese two opposite regulation pathways are discussed.

In a bid to develop technologies that serve grower needs for economic management of pests and diseases of hot pepper (scotch bonnet) while protecting public health and the environment, a study to assess the potential of cover cropping and/or less-risk pesticide usage in the management of pests and diseases of hot pepper was conducted for two consecutive seasons in Uganda. A split-plot randomized complete block design with three replications was used with cropping system (main plots) and pesticide treatment (sub-plots) as the factors. There were two cropping systems: the hot pepper+cowpea system vs. a hot pepper monocrop and five pesticide treatment options: (i) prophylactic treatment of plots at transplanting with granular carbofuran; (ii) weekly sprays of a neem-based formulation; (iii) combination of the prophylactic carbofuran treatment and neem; (iv) sulphur sprays at 10-day intervals (season 2 only); and (v) the untreated control. Data were collected on population dynamics of aphids, whiteflies, thrips, mites, nematodes and on plant performance parameters of branching and fruit weight. Results indicated that the hot pepper+cowpea system greatly lowered infestations of aphids and nematodes on hot pepper, but did not decrease thrips and whiteflies; and brought about a yield penalty on hot pepper. Prophylactic applications of carbofuran and/or sprays of a neem-based formulation lowered populations of aphids and whiteflies on hot pepper to varying extents; and increased yield of pepper.

Introduction. In-package biofumigation with the volatile-producing fungus Muscodor albus was tested to control fungal decay in organic table grapes stored at a commercial packinghouse. Materials and methods. Sachets containing two different amounts of activated M. albus culture were inserted into shipping boxes containing approximately 4.5 kg of ‘Thompson Seedless’ or ‘Red Seedless’ table grapes. The volatiles were contained inside the boxes either by wrapping pallets of the boxes externally with plastic film after pre-cooling (pallet wrapping) or by using a modified atmosphere liner inside each box. Decay incidence was evaluated after 7 weeks of storage at 0 °C. Results. The M. albus sachets reduced decay incidence among ‘Red Seedless’ table grapes in both wrapped pallets and boxes with liners. In this cultivar, the modified atmosphere liner alone reduced decay incidence by about 70% and the M. albus treatment in the liner further reduced decay incidence, regardless of the amount of M. albus used. The combination of the M. albus sachet and the modified atmosphere liner proved to be the most effective decay control treatment. Decay incidence was lower among ‘Thompson Seedless’ table grapes and a significant decay control was only observed after the grapes had been allowed to warm up after storage with the 50-g rates applied inside the liner. No adverse effects were associated with the treatment or the liners. Discussion. Based on our results, biofumigation with M. albus sachets is compatible with the commercial handling of organic table grapes and could provide significant improvement in their shelf life.

There is growing interest and commitment to the control of schistosomiasis and other so-called neglected tropical diseases (NTDs). Resources for control are inevitably limited, necessitating assessment methods that can rapidly and accurately identify and map high-risk communities so that interventions can be targeted in a spatially-explicit and cost-effective manner. Here, we review progress made with (1) mapping schistosomiasis across Africa using available epidemiological data and, more recently, climate-based risk prediction; (2) the development and use of morbidity questionnaires for rapid identification of high-risk communities of urinary schistosomiasis; and (3) innovative sampling-based approaches for intestinal schistosomiasis, using the lot quality assurance sampling technique. Experiences are also presented for the rapid mapping of other NTDs, including onchocerciasis, loiasis and lymphatic filariasis. Future directions for an integrated rapid mapping approach targeting multiple NTDs simultaneously are outlined, including potential challenges in developing an integrated survey tool. The lessons from the mapping of human helminth infections may also be relevant for the rapid mapping of malaria as its control efforts are intensified.

Introduction. In Mali, mango annual production is estimated at 100 000 t. Because of structural and plant sanitary problems mainly related to the actions of fruit fly, seldom more than 1% of this production is exported. The principal aim of this study was to develop an integrated control of the Tephritidae using spot treatments. One of the preconditions was to identify the species responsible for the damage in the three principal Mali production zones (Sikasso, Bougouni and Bamako). Materials and methods. Mangos belonging to the principal exported varieties were collected from the beginning of April up to the beginning of October 2000 and were observed to identify the Tephritidae species present in the fruits. The installation of sex attractant or of foodstuff attractant traps allowed us to follow the evolution and constitution of the fruit fly population in the orchards of the three selected sites. The effectiveness of a spot treatment to control Ceratitis sp. was tested. Results and discussion. Starting from adult observations and determination, six species of Tephritidae were identified. Among them, Ceratitis cosyra (Walker), C. quinaria (Bezzi) and C. silvestrii Bezzi were earliest and are thus most prejudicial. C. cosyra represented 86% of the total recovered flies. The species C. anonae, C. quinaria, C. fasciventris and C. ditissima were identified for the first time in Mali. The damage due to Tephritidae in the middle of the harvest period bordered on 50% of the crop for the varieties Kent and Keitt and exceeded 60% for that of Brooks. The 180 traps installed on the three studied sites allowed us to capture 128 998 Tephritidae adults belonging to thirteen species; the six species linked to the mango tree represented 99% of this total. The spot treatments carried out on the three experimental sites gave encouraging results, with a reduction in damage of approximately 50% in the treated orchards compared with the control orchards. Conclusions. To prolong the action undertaken, it will be interesting to confirm and look for further results obtained during the years to come. The installation of a regional project for controlling mango tree Tephritidae in West Africa would be fundamental in the medium term to limit such an insect pest which itself prevails on the sub-region scale.

Introduction. Damage of citrus fruits by the Mediterranean fruit fly Ceratitis capitata Weid. results in economic losses in Nigeria. The majority of the citrus producers are small-holders who are yet to adopt standard citrus production practices. A trial was therefore established in a savannah zone of Nigeria with a Southern Guinea agroecology to evaluate simple fruit fly control practices. Materials and methods. The efficacy of combining cultural practices such as removal of dropped fruits, changes of harvest period, and three applications of dimethoate/cypermethrine mixture in controlling fruit fly damage was tested on sweet orange Citrus sinensis (L.) Osbeck var. Valencia Late in 2000 and 2001. Results. Fruits attacked by C. capitata increased with fruit maturity and were positively correlated with the number of dropped fruits. Insecticide applications and removal of dropped fruits significantly reduced the number of fruits attacked by the fruit flies. Similarly, early harvesting at 50% fruit ripening significantly (P < 0.05) reduced the number of fruits attacked by the fruit flies when compared with the late harvesting at 90% ripening. The interactive effects of insecticide applications and the period of harvest significantly (P < 0.05) reduced attacks on fruits. Discussions. Judicious use of insecticides, periods of harvest and removal of fallen fruits reduce fruit fly damage. Late harvest and non-removal of fallen fruits constitute fruit fly reservoirs for the re-infestation of other clean fruits. Since the tested control methods were simple, they can be adopted by small-holders for the control of fruit flies.

Introduction. Au Mali, la production annuelle de mangues est estimée à 100 000 t. Du fait de problèmes structurels et phytosanitaires principalement liés à l'action des mouches des fruits, rarement plus de 1 % de cette production est exporté. L'objectif principal de cette étude a été de mettre au point une lutte raisonnée contre les Tephritidae du manguier au Mali à l'aide de traitements par taches. L'un des préalables a été d'identifier les espèces responsables des dégâts dans les trois zones principales de production (Sikasso, Bougouni, Bamako). Matériel et méthodes. Des mangues appartenant aux principales variétés exportées ont été collectées de début avril à début octobre 2000 et mises en observation afin d'identifier les espèces de Tephritidae présentes dans les fruits. La pose de pièges à attractifs sexuels ou alimentaires a permis de suivre l'évolution et la constitution des populations de mouches des fruits dans les vergers des trois sites choisis. L'efficacité d'une méthode de traitement par taches pour lutter contre les cératites a été testée. Résultats et discussion. À partir des observations et de la détermination des adultes, six espèces de Tephritidae ont été mises en évidence. Parmi elles, Ceratitis cosyra (Walker), C. quinaria (Bezzi) et C. silvestrii Bezzi ont été les plus précoces et sont donc les plus préjudiciables. C. cosyra a représenté 86 % de l'effectif total des mouches. Les espèces C. anonae, C. quinaria, C. fasciventris et C. ditissima ont été identifiées pour la première fois au Mali. Les dégâts dus aux Tephritidae en milieu de campagne ont avoisiné 50 % de la production des variétés Kent et Keitt et dépassé 60 % de celle de Brooks. Les 180 pièges mis en place sur les trois sites suivis ont permis de capturer un total de 128 998 Tephritidae adultes appartenant à treize espèces ; les six espèces inféodées au manguier ont représenté 99 % de cet effectif. Les traitements par taches effectués sur les trois sites expérimentaux ont donné des résultats encourageants, avec une réduction des dégâts d'environ 50 % sur les parcelles traitées par rapport aux parcelles témoins. Conclusions. Pour prolonger l'action entreprise, il sera intéressant de confirmer et approfondir les résultats obtenus au cours des années à venir. La mise en place d'un projet régional de lutte contre les Tephritidae du manguier pour l'Afrique de l'Ouest serait fondamentale à moyen terme pour limiter un tel ravageur qui sévit lui-même à l'échelle de la sous-région.

Introduction. Le ravageur A. aurantii ou pou de Californie, non contrôlé, est capable d'entraîner en deux à trois années le dessèchement complet d'agrumes contaminés. Au Maroc, le nombre d'interventions pesticides, contre cette cochenille, par cycle de produc- tion, n'a cessé de croître depuis les années 1960. Dans ces conditions, une lutte raisonnée s'impose. Le travail entrepris a eu pour objectif de tester conjointement trois techniques d'avertissement agricole. Matériel et méthodes. Les expérimentations effectuées ont étudié la génération printanière du ravageur au cours de deux années pendant lesquelles les dégâts qu'il a entraînés ont été très importants. Onze biotopes répartis dans la zone agrumicole du périmètre irrigué du Tadla (Maroc) ont été étudiés. Le biotope 1, non traité, a servi de réfé- rence. Les trois méthodes d'avertissement utilisées conjointement ont été le piégeage des adultes mâles d'A. aurantii, le suivi de l'évolution de la dynamique des populations larvaires du ravageur et de son principal parasite Aphytis melinus, l'utilisation de la notion de degrés jours. Pour chacune de ces méthodes, un seuil d'intervention a été défini. Résultats et discussion. L'utilisation des trois techniques d'avertissement agricole et la considération des seuils requis par chacune d'elles ont permis de déterminer une période d'intervention précise et d'adapter le traitement préconisé (insecticides, huiles minérales ou aucun traitement, selon les cas) aux risques d'infestation. Par ailleurs, le parasitisme de A. aurantii par A. melinus a réduit significativement les populations du pou de Californie dans les parcelles non traitées. Conclusion. La méthode du calcul de degrés jours combinée au dénombrement des mâles du ravageur par piégeage peut constituer une base d'avertissement agricole efficace et à la portée des producteurs d'agrumes marocains. L'activité naturelle de A. melinus dans les con- ditions du Tadla reste insuffisante pour contrôler l'incidence économique de la cochenille.

Introduction. Pour la mise en place de mesures à promouvoir dans le cadre d'une protection intégrée, celles concernant un aménagement des abords des cultures deviennent un objectif principal. Cependant, jusqu'à présent, cet aménagement s'est fait sans véritable critère. La présente étude montre comment il est possible de raisonner le choix des espèces végétales. Matériel et méthodes. L'entomofaune de 44 espèces ligneuses dont celle du poirier cultivé a été observée, pour chacune de ces espèces, à raison de 30 arbres par mois, pendant 10 mois de l'année 1998. La collecte de l'entomofaune a été réalisée par la méthode du battage de Fauvel. Résultats et discussion. Des recherches fondamentales ont été conduites sur la composition et la structure des communautés animales hébergées par les feuillus et comparées à celles d'un verger de poiriers. Le peuplement de 43 espèces ligneuses a ainsi été étudié. Après avoir montré la spécificité des peuplements, des corrélations de rang de Spearman entre le peuplement d'auxiliaires des 43 essences et celui du poirier ont été mises en évidence et montrent une ressemblance entre l'entomofaune utile du poirier et celle de certaines essences végétales. Conclusion. Les résultats obtenus permettent de raisonner le choix des essences pouvant servir de réservoir à une entomofaune prédatrice dans le cadre de la protection intégrée du verger de poiriers. Cette méthode est applicable à d'autres types de culture.