September 2000, Volume 21 No. 3

• General News
• IPM Systems
• Training News
• Internet Round-Up
• Announcements
• Conference Reports
• Proceedings

IPM Systems

This new section replaces the 'Biorational' section, but will essentially cover the same topics - integrated pest management (IPM) including biological control, and techniques that are compatible with the use of biological control or have little impact on natural enemies.

Banana Flavouring

The articles in this section have a common ingredient in this issue: bananas and plantains. This is not a comprehensive summary, but indicates the sheer effort going into overcoming the very serious constraints to production of this key crop around the world. Bananas and plantains can be divided into those which are sweet and eaten as a dessert fruit and those that are starchy and only eaten after cooking. In many parts of the world cooking bananas and plantains are staple foods, and 90% of the total world production (some 85 million tonnes) is grown for domestic consumption. On the other hand, in the developed world dessert bananas are a favourite fruit, and this makes them the fourth most valuable food crop and the most important fruit, worth US$2.5 billion annually in export trade.

One of the most serious pests of bananas is a boring weevil, and the many promising avenues being explored for controlling it are described. There is also an array of nematode pests that infest the crop in different regions and climatic conditions, and progress against these is also reported. Above all, though, bananas are threatened by diseases. There are a host of these, and new ones are still emerging. The most serious of them, black Sigatoka was first seen (in Fiji) in 1963 but has now found its way to all the banana-growing regions, although some countries are still free of the disease. Research on management of this disease is in the early stages, and we hope to report on it in a later issue. Here we report on initiatives and progress against other diseases, both `old' (such as yellow Sigatoka) and `new' (such as banana leaf streak disease in Uganda). Lastly, we look at control of post-harvest diseases, which are particularly important in export bananas because they affect quality and storage.

Although wild species are seed-bearing, modern cultivars are sterile. Propagation is by vegetative means, and this has implications for the incidence of pests and diseases and how they may be managed. Vegetative propagation means that both have been passed on in planting material between and within countries, and the provision of clean planting material has been an area of intense research that is yielding dividends. To add to the problems, bananas and plantains are perennial crops, in which new plants emerge as suckers from a shared rhizome. (The rhizome and bananas/plantains attached to it are collectively called a mat.) Pest and disease problems tend to get worse with successive ratoon crops.

Two wild species (Musa acuminata and M. balbisiana) gave rise to modern commercial bananas and plantains. These diploid species have produced diploid, triploid and tetraploid offspring. The conventional way of indicating the (probable) parentage of current clones is by using `A' and `B' for acuminata and balbisiana, respectively (e.g. Cavendish AAA). However, the narrow genetic base of modern export, dessert bananas makes them susceptible to pest and disease attack, and a key element in banana and plantain IPM programmes is improving host plant resistance.

IPM and INIBAP

INIBAP (the International Network for the Improvement of Banana and Plantain) was established in 1985 as an international organization with a mission to sustainably increase the productivity of banana and plantain grown on small-holdings for domestic consumption and for local and export markets. INIBAP is a programme of the International Plant Genetic Resources Institute (IPGRI). It has a small headquarters staff in Montpellier, France and regional offices in the four major banana-growing areas of the world.

IPM is a very important area of work for INIBAP, and a number of different activities are ongoing. These include supporting and carrying out research on various component parts of IPM (host plant resistance, biological control, disease management, etc.), organizing meetings and workshops, and distributing relevant information. This report provides an overview of INIBAP's different IPM-related activities. It should be noted that INIBAP is a networking organization. It does not conduct research in-house. Instead it works through outsourcing, thus making use of and strengthening existing facilities and expertise. INIBAP also plays an important role co-ordinating and catalysing research carried out by its partners worldwide.

Host plant resistance is one of the most important components of any IPM programme. In the case of bananas, despite the fact that breeding started as long ago as the 1920s, new pest and disease resistant varieties developed by breeding programmes have only become available for distribution in the last few years. INIBAP is playing a number of different roles in helping to ensure the continual availability of improved varieties for use in IPM programmes worldwide.

Support to Breeding Research

The absence, until very recently, of improved, pest and disease resistant varieties has been attributed to the difficulties of breeding bananas at the genetic and practical levels (most important varieties are highly sterile and can only be used in conventional breeding programmes with difficulty), together with the very low levels of investment in banana breeding. In an effort to accelerate the development and release of new varieties, INIBAP works in a number of ways:

• Providing direct support to research on Musa improvement, with a focus on the use of biotechnology and molecular approaches to increase breeding efficiency.
• Providing the secretariat and co-ordination for PROMUSA, the Global Programme for Musa Improvement. PROMUSA brings together all the major players in Musa improvement worldwide, allowing researchers to develop collaborative projects and forge partnerships, thus creating synergies and avoiding unnecessary duplication of research.
• Organizing a worldwide, multi-site evaluation programme (International Musa Testing Programme) which allows breeders to test their materials over a wide range of environments and disease pressures, while at the same time providing national programmes early access to improved varieties.

Distribution of Germplasm

INIBAP maintains the world's largest collection of Musa germplasm. This material is held in trust for the global community and is freely available for distribution to bona fide users. The collection contains wild species of the genus Musa, as well as a wide diversity of cultivated varieties and improved varieties from breeding programmes. All material is introduced and distributed under the terms and conditions of Material Transfer Agreements, to ensure that the germplasm and information related to it remain in the public domain.

In the case of improved varieties, INIBAP has introduced a specific Germplasm Acquisition Agreement and a Material Transfer Agreement. These agreements provide a framework within which improved germplasm can be distributed by INIBAP, while at the same time providing breeders the opportunity to benefit from the commercial use of their material. Through this innovative approach, INIBAP has been able to acquire improved varieties from all the major banana-breeding programmes and is working to ensure that such varieties will continue to be available free-of-charge to those who need them most - small-scale farmers in developing countries.

All material is virus indexed by INIBAP before distribution. Only accessions in which no viruses have been identified are available for dissemination. As plants are distributed in the form of in vitro plantlets, a clean source of planting material is ensured for the end users.

Improved pest and disease resistant varieties have been distributed to over 50 countries by INIBAP, and in several countries, most notably in Cuba, Nicaragua and Tanzania, smallholder banana producers are gaining significant benefits from the use of these varieties.

Clean Planting Material against Viruses

A number of viruses affect bananas and these include Banana Bunchy Top Virus (BBTV), Cucumber Mosaic Virus (CMV), Banana Bract Mosaic Virus (BBrMV) and Banana Streak Virus (BSV). Banana viruses, especially BBTV and BBrMV (this one was only identified for the first time in 1989), have had a significant effect on the production of non-export, traditional dessert bananas in several Asian countries, including the Philippines, Sri Lanka, Bangladesh, Vietnam and Pakistan. Control measures for these viruses are based on quarantine, sanitation (use of clean planting material) and roguing. INIBAP's regional office for Asia and the Pacific has been active in helping to develop virus control programmes in several countries in Asia. Workshops and training courses on virus control, virus indexing and the use of clean planting material have been held in the Philippines (for the region) and Sri Lanka. As a result of these workshops, national banana rehabilitation programmes based on IPM, with clean planting material as the key component, have been put in place in both the Philippines and Sri Lanka.

Controlling Bacterial Diseases in Indonesia

`Blood disease', a destructive banana bacterial disease, is serious in the Sulawesi area of Indonesia and has devastated the cooking banana Kepok (Saba) in this region. In recent years severe outbreaks of this disease have also been seen on the southern tip of Sumatra, in the region of Lampung, where it is now seriously threatening the smallholder banana crop. INIBAP's Regional Office for Asia and the Pacific is working with the Research Institute for Fruits in Indonesia, to conduct demonstration trials on the management of this disease. Disease management technologies, effective in combating a similar disease known locally as `tibaglon' or `Bugtok', have been developed in the Philippines. A study previously carried out in Negros Oriental (Philippines) showed that an integrated disease management programme including sanitation, early debudding, disinfection of tools and fruit bagging can reduce disease incidence by up to 100% over a 12-month period. It was discovered however that farmers found bagging impractical because of the height of the local varieties of cooking bananas being cultivated. Nevertheless, the practice of sanitation and early debudding alone was shown to reduce infection from an initial incidence of 88% to 6% after 12 months. These methods were considered practical and have been well accepted by farmers. The disease management strategies developed for Bugtok in the Philippines are being tested for their effectiveness against Blood disease. It is hoped that the simple techniques, either alone or in combination with other techniques, such as the use of clean planting material, will help to solve the Blood disease problem in Indonesia.

Mobilizing IPM for Sustainable
Production in Africa

An Africa-wide workshop on banana IPM was organized by INIBAP, together with the International Institute of Tropical Agriculture (IITA) and the Institute of Tropical and Subtropical Crops, South Africa (ITSC) in the framework of the Banana Research Network for Eastern and Southern Africa (BARNESA) in South Africa in November 1998. This attracted IPM researchers from across Africa and beyond. The proceedings of this meeting were published by INIBAP in 1999 and provide state-of-the-art information on banana IPM. The meeting identified technologies ready for testing in farmers' fields. Such technologies are: the use of clean planting material, host plant resistance and a range of cultural management practices. The meeting called for greater efforts in on-farm testing of banana IPM strategies. The main research gaps were identified as:

• Biological control: Beauveria bassiana holds promise for the control of weevil borer (Cosmopolites sordidus), but other predators and pathogens should also be investigated.
• Biologically-enhanced planting material: endophytes, arbuscular mycorrhizal fungi, etc. The screening and testing of organisms and formulations is required.
• Biological pesticides: the efficiency of neem needs to be tested under farmers' conditions.
• Enhanced trapping technologies for weevils should be further investigated.
• Use of semio-chemicals (kairomones and pheromones): improved efficiency and delivery mechanisms are required.
• Resistance breeding (including genetic engineering): new hybrids must be evaluated across years and environments. Decentralized breeding strategies and the use of varietal mixtures could also be investigated.
• Various aspects of cultural management (soil fertility, water management, mulches, intercrops, etc.) and their effects on pathogen levels need further study.

On-Farm Testing of IPM Strategies in East Africa

Pest and diseases have been identified as the major constraint to banana production throughout the East African region [see below]. However, options for IPM are known to be available and following the recommendations of the IPM workshop organized in South Africa in 1998, these should now be tested on-farm. A project proposal was therefore developed by INIBAP in collaboration with the national agricultural research services of Uganda, Kenya and Tanzania. The project was accepted for funding by DFID (the Department for International Development, UK) and commenced in May 2000.

The project will allow combinations of IPM options to be tested on-farm in a range of settings in benchmark sites in Kenya, Tanzania and Uganda. Such settings include agro-ecological variation in terms of varieties grown, disease/pest pressure, soil fertility levels, climatic factors, etc., in combination with the differing socio-economic situations of the participating farmers. One of the most important aspects of the IPM project will be the testing of new varieties.

One other INIBAP initiative is described later in this news section. An INIBAP Associate Expert, located at the Centre de Recherches Régionales sur Bananiers et Plantains (CRBP) in Cameroon, is working on the development of an IPM strategy against the banana weevil borer, C. sordidus.

By: Suzanne Sharrock, INIBAP,
Parc Scientifique Agropolis 2,
34397 Montpellier Cedex 5, France
Email:
Fax: +33 4 67 61 03 34

Pearls of Banana Research

Uganda is the world's largest banana producer, with a 15% share of total global yield; in 1996, it produced about 9 million tonnes of the fruit. Highland cooking banana is the most important staple in the Great Lakes region. Uganda is also a secondary centre of banana diversity and there are many locally evolved cultivars; an average of 12 different cultivars were found growing on farms surveyed by IITA. However, the many different clones of the East African highland bananas are all Musa (AAA) and have therefore evolved (by somatic mutations) from a relatively narrow genetic base. The ancestral cultivars that would have been introduced are unknown. There is little difference in the susceptibility of these cultivars to the more important pests and diseases that have been introduced in recent times. The Uganda highland banana cultivars have recently been taxonomically characterized by Deborah Karamura at the National Agricultural Research Station (NARO).

Since the 1970s, banana has undergone a decline in the traditional growing areas in central Uganda. During the same period, it has expanded into the southwest. The increasing importance of banana in this region has been related to (a) improved food security and (b) increasing urban market demand following rapid population growth in Kampala and Jinja. Cooking banana is the preferred urban staple, in contrast to many urban areas of the world which rely on grains. It remains, also, the preferred rural staple food throughout southern Uganda.

Partly associated with this geographical shift, there has been a major decline in production in the last 25 years. Yields have fallen to as low as 6 tonnes/ha and the longevity of banana plantations has fallen from about 50 years to only 5-10 years in some areas. The decline in productivity and plantation longevity is attributed to social, economic and biological factors. The (often quoted) decline in soil fertility can be traced to labour costs and availability associated with the traditional system of culture with organic manures and regular mulching. These practices are now less common. Land ownership is an issue as farms are divided by inheritance. Replanting bananas will be less acceptable on the smaller farms because there is a 12-15 month wait for bananas to come into production, the reason to change to short duration crops is thus compelling.

As if these problems are not great enough the arrival of alien diseases and pests like black Sigatoka (Mycosphaerella fijiensis), yellow Sigatoka (M. musicola) and the burrowing nematode (Radopholus similis) has accentuated the decline. The most significant pest recognized by farmers is the weevil Cosmopolites sordidus. When this pest was introduced is unknown but it is likely to have arrived with the newer `exotic' cultivars in the late 19th or early 20th century. The first cited record for Uganda is 1918. The combined effects of nematodes and weevils include destruction of the root system, reducing anchorage and water and nutrient uptake, which leads to reduced bunch weight or plant toppling. Diseases reduce photosynthetic capacity and therefore fruit production and often quality, or even cause the death of the plant.

Devastating banana weevil and nematode outbreaks in the mid 1980s led to widespread crop failure in Masaka and Rakai districts. Although these outbreaks persisted for only a few years, early stages of yield decline are already in evidence in certain areas in the southwestern region. This has caused apprehension about the future of the crop in Uganda. Moreover, commercial banana production has been shifting further and further away from Kampala. Currently, much of the banana serving Kampala comes from Mbarara and Bushenyi districts (240-300 km from the capital). As a result, revitalization of banana production in the central zone and stabilizing production in the southwest became high priorities within the Uganda Ministry of Agriculture and NARO. Banana weevil is considered a priority concern. It is an important pest of highland cooking banana in East Africa, and has been implicated in the decline and disappearance of cooking banana from its traditional growing areas in central Uganda and western Tanzania. Since the 1960s, accelerated yield decline in this region has led to the replacement of cooking banana with exotic beer bananas (types AB and ABB) and annual crops.

However, Uganda is also home to exciting and innovative research into banana pest and disease control. There is an active national  programme, and they have involved many international organizations, so the prospects for bananas are becoming rosier.

In 1990, the International Institute of Tropical Agriculture (IITA) and the newly formed Uganda National Banana Research Programme (UNBRP) developed a collaborative programme to address the problems of banana weevil, nematodes, diseases and other production constraints. The first activity was a rapid rural appraisal at 25 banana-producing villages across southern Uganda. Pests and soil fertility decline were ranked as key constraints in nearly all sites. Farmers ranked the banana weevil as the leading pest/disease problem at 18 sites. Within the central region, many farmers were replacing cooking banana in favour of exotic brewing bananas such as  Pisang awak (ABB) and Ney poovan (AB) which are resistant to the weevil and require less management attention. In many areas, farmers have lost confidence in a crop that had long been their mainstay.

Later in this section, progress made by this  programme in developing strategies to manage the constraints to banana production and to ensure that Uganda retains its place at the top of the world's banana producers is reported.

From: Cliff Gold and Simon Gowen.

Contact: Wilberforce Tushemereirwe, UNBRP, Kawanda Agricultural Research Institute, Box 7065, Kampala,
Uganda
Email:

A Cosmopolitan Weevil

The banana weevil, Cosmopolites sordidus, was first recorded in connection with bananas in Guadeloupe in 1889, but has been distributed throughout the tropics on banana planting material. It has no other host and all types of banana are susceptible to varying degrees. In some of the banana-growing regions it has serious impact (for example, in East Africa) but elsewhere its importance is not as great. The reasons for this are not clear but environmental factors and differences in crop growth and management could be influential. There is yet no evidence to suggest that the weevil populations vary in aggressiveness.

The first attempt at biological control of this pest was in Fiji by Frank Jepson, the Government Entomologist. In 1913 he made collections of the  histerid beetle Plaesius javanus and other natural enemies in Java. Subsequently, this predator was released in several countries, including Uganda by C. C. Gowdey in 1918-20, but it never became established.

Biological control was forgotten during the early insecticide era. Aldrin was recommended for weevil control in Uganda during the 1950s until resistance developed. Post-colonial funding from the UK Government was for chemical control of the banana weevil which was done by Graham Mitchell of the Centre for Overseas Pest Research (COPR) in the Caribbean in 1974-78 when alternatives to the  organochlorine insecticides were being sought. At this time much research was done on damage assessment, population estimation and chemical control throughout the commercial export-producing countries.

The opportunities for developing biological control have been revived over the last 10 years and the most promising biological control agent, Beauveria bassiana, has been isolated from weevils in many countries. This fungus has been used in banana plantations in Brazil. In Uganda, isolates of B bassiana have been collected by Caroline Nankinga (National Agricultural Research Organization, NARO) and significant reductions in weevil populations have been achieved with formulations of this fungus under field conditions. Another promising line of attack is with entomopathogenic nematodes (epns). These nematodes (Steinernema spp. and Heterorhabditis spp.) have also been tested successfully under field conditions in Australia and Brazil. Recently, epns have been recovered from bananas in Kenya by Charles Waturu (Kenya Agricultural Research Institute, KARI) and there is a survey planned for Uganda later this year as part of a weevil management project.

The chemical nature of the male-produced aggregation pheromones has been identified and both lures and traps have been developed commercially in Costa Rica from where there have been reports of successful reductions in weevil damage in commercial Cavendish plantations.

By: Simon Gowen

Many Hands Tackle Weevil Problem

The IITA/UNBRP rapid rural appraisal described in `Pearls of Research', above, was followed by a study on the geographic shifts of banana production, which highlighted the importance of banana weevil, Cosmopolites sordidus, in the decline of cooking banana in central Uganda. The study focused on villages in central Uganda in which banana had largely disappeared; such villages presented a truer picture of crop dynamics in central Uganda than those used in the initial IITA/UNBRP study (i.e. sites still considered banana-producing villages) described above. Farmers attributed increasing weevil pressure to reductions in available labour and resulting changes in management practices. For example, crop sanitation, trapping and other cultural controls have been largely abandoned. Field verification confirmed very high weevil levels in survey study sites.

Diagnostic surveys showed that the banana weevil is widespread within Uganda, but unimportant above 1600 m above sea level. Weevil pest status varied considerably across sites and among farms within sites. For example, weevil population estimates (based on mark and recapture studies) within one watershed in Ntungamo district found density to range from 1600 to 149,000 weevils/ha. These studies found only a weak relationship between adult density and damage.

Salient features of the weevil's biology include a cryptic life style, long life span, limited mobility, low fecundity and slow population build-up. The adult weevil commonly lives more than one year, while some adults have been reported to live more than 4 years. The weevil is very sensitive to desiccation but can live for several months without feeding. These factors mean that weevils are often favoured by mulches (which contribute to soil moisture retention during dry periods) and that extensive plant loss may occur when banana is planted in a previously infested field without an adequate break between the crops.

The adults are most often found in the leaf sheaths, in the soil around the base of the mat or associated with cut residues. They are nocturnally active and not casually observed. Adults are often sedentary for extended periods and most do not move more than a few metres at a time. They rarely fly. The weevils are attracted by plant volatiles (especially those emanating from cut rhizomes) and also produce a male aggregation pheromone. Under field conditions, mean oviposition may be 0.5-2 eggs per week. The eggs are inserted singly into the leaf sheaths or rhizome. The larvae feed within the rhizome cortex, central cylinder and, occasionally, the (pseudo)stem. The insect passes from egg to adult in 6-8 weeks.

With low fecundity, population build-up is slow and, following crop establishment, weevil problems are most pronounced in ratoon crops. In one trial, yield loss increased from 5% in the plant crop to 44% in the third ratoon. This loss reflects reduced bunch size, snapping, toppling and mat die off.

Current research results suggest that no single control strategy will be likely to provide complete control for banana weevil. Therefore, a broad IPM approach might provide the best chance for success. The components of such a programme include habitat management (cultural control), biological control, host plant resistance and (in some cases) chemical control. The poor relationship between weevil adult density and damage suggest that factors which target the adult stage (e.g. trapping) may be less effective than those directed at immatures. However, the immature stages are hidden within the banana plant and largely immune to many control strategies.

Cultural Control

Dispersal of the banana weevil is primarily through infested planting material, so clean planting material could protect new stands from banana weevils for at least several crop cycles, if this is planted some distance from other infested stands. However, replanting previously infested fields will only be possible after an interval of many months to allow the existing population to die out. Otherwise, these weevils will readily attack suckers (that have been detached from mother plants and have an exposed cut rhizome surface). In one field trial, more than 40% of planted suckers were killed by weevils, as were an additional 40% of the replants.

The use of clean planting material, disinfected of weevils through paring and/or hot water treatment, has been recommended as a cultural control strategy to reduce initial infestation levels and retard pest build-up. Paring of the rhizome surface exposes damaged suckers, which can then be rejected. Paring also removes most eggs and many first instar larvae. Hot-water treatment of suckers has also been recommended for both weevil and nematode control. However, hot water regimes commonly used for nematode control (i.e. 54ºC for 20 minutes) killed only 32% of weevil larvae. Therefore, paring alone is probably adequate for reducing initial weevil infestations. [The IITA/UNBRP programme initiatives against nematodes are described below.]

In field studies, weevil populations were lower in cleaned planting material plots than in controls for up to 27 months after planting. Weevil damage levels in controls were 70-200% higher than in cleaned planting material treatments for the plant crop. Plant loss to weevils and nematodes was 21-34% in controls, compared to 2-6% in treated plots. Bunch size was similar among treatments, but clean planting material plots provided higher yields resulting from the greater number of harvested bunches.

Pseudostem trapping has also been widely recommended in Uganda for weevil control. Farmer participatory research in Ntungamo district, Uganda showed that intensive pseudostem trapping (1 trap/mat/month) could substantially reduce weevil populations, while more moderate trapping (0.3-0.6 trap/mat/month) could also lower weevil numbers. However, most farmers indicated that the labour and material required for intensive trapping was beyond their resources.

IITA and the National Agricultural Research Organization (NARO) are currently investigating enhanced trapping using pheromones. The presence of a male aggregation pheromone (i.e. produced by males and attractive to both sexes) was first recognized at ICIPE (the International Centre for Insect Physiology and Ecology, Nairobi). The chemical structure of the pheromone (called sordidin) has since been identified. ChemTica in Costa Rica has synthesized and is distributing commercially a mixture of the pheromone and weevil-attractive plant volatiles in a formulation called Cosmolure+. Preliminary studies in Uganda show that Cosmolure+ can catch up to 30 times as many weevils as conventional pseudostem traps. Further work is being undertaken on field efficacy under farmer conditions in Uganda.

Crop sanitation (i.e. destruction of crop residues) is also widely recommended to control banana weevils. The supposition is that crop residues left in the ground after harvest serve as shelters and breeding grounds/oviposition sites for banana weevils, leading to population build-up and increased damage on plants. Farmers implement a range of sanitation practices including cutting and chopping of spent stems and digging out or burying old rhizomes. The intensity of sanitation practices is often greater on commercial farms. In Ntungamo district, dominated by subsistence banana production, nearly 80% of the farmers practised little or no sanitation. It is also possible that crop residues serve as `trap crops' drawing gravid females away from growing plants. Current studies are investigating weevil oviposition preferences, larval development and the effects of sanitation on weevil populations and damage.

Prospects for Biological Control

Biological control efforts against banana weevil have included the use of exotic natural enemies (classical biological control), endemic natural enemies, secondary host associations and microbial control (e.g. entomopathogens, endophytes, entomophagous nematodes). Microbial control agents that require repeated applications may be considered as biopesticides, although they lack the toxic side effects of chemical insecticides. As such, they may entail repeated application costs on the part of the farmer.

Classical biological control of banana weevil may be possible. The banana weevil originated in South-east Asia, coincident with the centre of origin of bananas. The banana weevil is not believed to be a serious pest in most areas within its area of origin. However, searches for arthropod natural enemies of banana weevil in Asia, conducted in the first half of the 20th century, provided only generalist predators (e.g. histerids and hydrophilids) which attack the immature stages. Releases of these natural enemies in Africa and elsewhere met with little success. In collaboration with the Research Institute for Fruits in Solok, Indonesia, IITA will be undertaking further exploration for natural enemies later this year. It is hoped that such searches may reveal the presence of egg parasites.

ICIPE has conducted studies on endemic natural enemies of banana weevil in Western Kenya. These included adults of three staphylinids, three histerids, one hydrophilid, one carabid, one tenebrionid, two labiids and one carcinophorid earwig. In laboratory studies, these predators variously searched the rhizomes of living plants and crop residues. Most of the predators attacked the egg and early larval instars, although four also attacked older larvae. Some of the predators were able to reduce weevil levels in pot trials, but predator densities under field conditions suggest limited potential.

In contrast, the Cubans have employed the myrmicine ants, Pheidole megacephala and Tetramorium guineense, which they report as providing partial control. The ants can be encouraged to nest in pseudostem pieces which can then be transferred to other banana stands. IITA and NARO have surveyed ants at several sites in Uganda and found P. megacephala, four other species of Pheidole and one unidentified Tetramorium species. A graduate student will begin her Ph.D. research in late 2000 on the potential of controlling banana weevils with myrmicine ants.

In West Africa, IITA explored the possibility of secondary host association, by evaluating the potential of two strains of a carrot weevil parasitoid, Anaphes victus, for parasitizing banana weevil. These parasitoids readily oviposited in banana weevil eggs but failed to successfully complete their development or to be able to emerge from the weevil eggs due to their relatively larger size. Larvae of A. victus failed to consume all of the banana weevil egg contents with decomposition of unconsumed material contributing to pupal failure. Most of the few parasitoids which successfully reached the adult stage then failed to emerge through the relatively thicker chorion of banana weevil eggs.

IITA and NARO have also been collaborating on microbial control of banana weevil, using endophytes and B. bassiana. Endophytes are mutualistic, non-pathogenic fungi that exist within a host plant. The objective of this study was to identify endophytes within banana plants which (a) could kill weevils (and nematodes) and (b) be introduced into tissue culture plants to provide extended protection against pests. Protocols include (a) isolation of endophytes from banana plants, (b) screening against weevil immatures, (c) determination of modes of action, (d) identification and markers of candidate strains (especially important if what is subsequently isolated is the same as what had been introduced), (e) studies on distribution and persistence within the host plant, (f) pathogenicity testing, (g) efficacy in pot trials and (h) efficacy in field trials under different environmental conditions.

Surveys of highland cooking banana and Pisang awak revealed an abundance of endophytes in the central cylinder and cortex of the rhizome. In laboratory experiments, 12 of 200 strains killed a high percentage of weevil eggs. Some of these strains also produced moderate levels of larval mortality. Weevil mortality was caused by direct colonization and by fungal exudates. The most promising strains are within the genus Fusarium. A number of these strains have been successfully inoculated into tissue culture plants. Preliminary studies on the efficacy of the strains in controlling weevils in pot experiments have not yet produced clear trends. Currently, further work is being undertaken on endophyte persistence and efficacy against weevils.

NARO has also screened strains of B. bassiana for efficacy against banana weevil adults. These strains were isolated from banana weevils, Galleria baits placed in banana stands and from other insects. A number of isolates effected mortality of more than 95% in the laboratory. Tests on potential delivery systems revealed that maize-based substrates and oil formulations were more effective than water suspensions. In field experiments, trap captures and weevil damage were lower in plots where maize-based formulations were applied to soil at the base of banana mats than in controls. However, the quantity of substrate required was cost-ineffective. Therefore, future work will continue to address the development of more cost-effective delivery systems, as well as pathogen performance under different ecological conditions.

Host Plant Resistance

In a screening trial in Uganda, plantains and highland bananas appeared more susceptible than other groups. Cluster analysis suggested 19 clones were highly susceptible, 17 clones were intermediate in susceptibility and 9 clones were resistant. Within the highland group, 15 clones were susceptible, while 11 were intermediate in susceptibility. Resistance did not appear to be related to host plant location or acceptance (antixenosis). There was no relationship between weevil trap captures, oviposition and damage. Instead, antibiosis appears to be the predominant mechanism affording resistance to the weevil, especially as it affected larval developmental times and survivorship. For example, in one study, survivorship in susceptible highland banana clones was 16-23 times as high as in resistant Pisang awak. Current work is attempting to explore compounds conferring resistance against banana weevil.

By: Cliff Gold

Work described in this article was conducted by: C. S. Gold, S. Okech and P. R. Speijer (IITA-ESARC), E. B. Karamura (INIBAP), A. Kiggundu, C. M. Nankinga and W. K. Tushemereirwe (NARO), D. Rukazambuga (National Banana Programme, Minagri, Tanzania) and M. Griesbach (University of Bonn, Germany).

Contacts: Cliff Gold [weevil biology, population dynamics, pest status, cultural controls; endophytes]; Deborah Karamura [weevil IPM]; A. Kiggundu [host plant resistance]; Caroline Nankinga [microbial control]; Suleman Okech [on-farm testing]; Wilberforce Tushemereirwe [banana programme strategy including benchmark sites].
c/o IITA-ESARC, P.O. Box 7878,
Kampala, Uganda
Fax: +256 41 223459
Email:
or c/o UNBRP,
Kawanda Agricultural Research Institute, Box 7065, Kampala, Uganda
Email:

Beauveria Bait for Banana Borer

Brazil has also been at the forefront of research into  biocontrol of banana weevil, Cosmopolites sordidus. Amongst the crop protection problems faced by Brazilian banana growers, the banana weevil aptly known here as the `banana rogue', stands out in importance. The damage caused by the insect ranges from a drop in plant productivity to plant death. The tunnels the weevil makes in the rhizome allow the entry of pathogenic microorganisms, which accelerate the decline of the plant. Research on control of this pest increased from the 1950s onwards. There was increasing use of chemical control, initially organochlorine insecticides and more recently other types of active ingredient.

Continuous and prolonged use of  organochlorine insecticides led to the development of resistance in the weevil populations. Following on from this development and the withdrawal of organochlorine insecticides, banana growers resorted to using systemic compounds, but residues of these could make their way into the fruit. This fact, combined with ecological disturbances resulting from the misuse of these products, stimulated work on alternative methods for banana weevil control. The most promising avenue is the use of entomopathogenic fungi, and research began by evaluating the efficiency and potential of the fungi Beauveria bassiana and Metarhizium anisopliae as biocontrol agents for the weevil. The Instituto Biológico in São Paulo State has carried out laboratory and field studies, the latter in the Vale do Ribeira region.

The project was initiated back in 1984, when the fungi were evaluated under laboratory conditions. The pathogens were grown on two different substrates (moist autoclaved rice and sterile ground beans). Inoculations were carried out in two ways: by making the insects crawl over the culture medium containing the fungi and then transferring them to pieces of banana pseudostem, and by placing the  inoculum directly onto the pseudostem so that the insects contaminated themselves as they entered. Pest mortality of more than 85% was recorded in all treatments except for M. anisopliae grown on bean substrate (56% mortality). Under field conditions, B. bassiana was found to be the most promising agent for controlling the weevil, particularly when grown on rice medium, on which it developed best and to which the pest insect was attracted as a bait.

The results of a study on the virulence of different isolates of B. bassiana identified the isolate, CB-66 (originally obtained from the coffee berry borer, Hypothenemus hampei), as most effective for banana weevil control. In field trials, isolate CB-66 was prepared as a paste which was applied to the cut surface of longitudinal sections of banana pseudostems, and these (telha baits) were placed cut-side down on the soil surface next to the banana plant. Evaluations over three seasons showed that B.  bassiana could reduce the adult weevil population by up to 61% over an extended period.

Under laboratory conditions, it was found that mineral oil at 3-5% EC added to strain CB-66 in paste form reduced C. sordidus populations by 77.5-100%, while the fungus alone gave only 37.5% control. Tests on the compatibility of the pathogen-oil association showed that although spore germination was reduced, virulence was increased. In the field, B. bassiana with 3% mineral oil applied to telha baits reduced pest populations below economic damage levels. The timing of application of the fungus was determined by the population level of the insect. It was applied when this averaged five or more adults per bait (recorded monthly.) Other authors have found this to be the level at which action is required to control the pest.

Over the 12-month study period, four applications and 20 evaluations were undertaken. After the first application, there was a clear reduction in the population of adult weevils in the treated area, from 9.0 to 3.4 insects per bait (equivalent to a control efficiency of 46.7%). The last application reduced the pest population to a level below that necessary for control. The control population was 45.8% higher, with 7.0 adults per bait.

These results indicate that B. bassiana is capable of reducing weevil infestations below damaging levels and, therefore, that it should be possible to establish a programme for management of C. sordidus based on the use of the fungus. Currently we are looking at whether we can use pheromones in association with the fungus to further increase disease incidence and improve weevil control.

By: Antonio Batista Filho, Centro Experimental do Instituto Biológico,
Caixa Postal 70, 13001-970 - Campinas, São Paulo, Brazil
Email:

Weevil Control a Piece of Cake!

A study conducted in West Africa complemented the work conducted by Caroline Nankinga in Uganda [described above], and may provide the basis for developing affordable biological control measures using indigenous pathogens for the banana weevil Cosmopolites sordidus in Africa.

Strains of Beauveria bassiana were isolated from a range of hosts (C. sordidus and Hypothenemus hampei) in East and West Africa by staff from IITA, IMI (now part of CABI Bioscience) and ARSEF (Agricultural Research Service Collection of Entomopathogenic Fungi), and the most virulent selected on the basis of pathogenicity tests. Mass-production techniques for them were adapted from a technique developed for Metarhizium production at IITA under the LUBILOSA programme.

Key to the success of a fungal biocontrol agent is ease of mass production and the development of an effective means for its application. Researchers in the University of Ghana and IITA chose to work with a virulent B. bassiana isolate, IMI 330194, that was also robust and able to resist invasion of three common contaminating competitor species (Aspergillus niger, Fusarium moniliforme and Penicillium hirsutum). They also developed a method for formulating B. bassiana on oil-palm kernel cake (OPKC), which it has since been shown to persist for several months in the field, and to control weevils during the first few critical months after planting. This work may be the basis for an effective management tool for banana weevil.

Plantain (Musa AAB) is the preferred staple food of Ghana. The principal constraints to production are declining soil fertility, black Sigatoka disease, nematodes, and banana weevils. For smallholder farmers, biological control using Beauveria bassiana was considered the most promising management option for the weevil.

Researchers in Ghana began by looking at the efficacy of a water-based formulation of the fungus. In laboratory studies they found that B. bassiana applied to corm and pseudostem pieces had the potential to control all stages of C. sordidus, with up to 25%, 46% and 59% of eggs, larvae and adults, respectively, showing disease symptoms. In pot experiments they began to look at other formulations and application methods. Adult mortality in pot experiments was in the range 26.4-62.0%, and best results were obtained using dry conidia in a mixture of kerosene + groundnut oil (70:30 v/v) applied to the soil surface. These encouraging results translated to the field. In preliminary trials, weevil mortality of 53-81% was recorded on suckers dusted with B. bassiana, compared with 7-8% in an untreated control.

However, the researchers were looking for a method that would not only control the weevil, but would do so over the critical establishment phase of young plantains. In trials to investigate how well the OPKC formulation performed in the field, they lost no plantain suckers in the OPKC-treated plot, while 17% were lost in a plot treated with dry spores of B. bassiana and 19% in an untreated control.

In further field trials, they looked at whether conidia or conidial powder in OPKC gave best results. Both treatments gave equivalent and high levels of weevil mortality (75.5%) in artificial infestations, compared with only 1% in an untreated control. However, with natural infestations, the conidial treatment led to 41.7% mortality, compared to only 5.7% for the conidial powder treatment (and 3.3% for the control). No suckers were lost during the 2-month study in the conidia-OPKC treatment, but 17.7% and 19.4% were lost in the conidial powder-OPKC and control treatments, respectively. A study of the spread of fungal conidia using artificially infected and uninfected adult weevils indicated that B. bassiana conidia might be able to spread up to 18 m from the release point.

Contact: Ignace Godonou, CABI Africa Regional Centre, ICRAF Complex,
PO Box 633, Village Market,
Nairobi, Kenya
Email:
Fax: +254 2 522150

Pathogens for a Pair of Weevils in Peru

Laboratory and field studies have been conducted in Peru to assess the prospects for using local strains of Beauveria bassiana for control of not only Cosmopolites sordidus (known locally as black weevil) but also the streaky weevil, Metamasius hemipterus. The black weevil is an important economic pest of bananas in Peru, as elsewhere in the world. The larvae cause damage by eating developing stems, which can cause wilting, stunted development and reduced production in attacked plants. The streaky weevil is a secondary pest. Its larvae do not have the capacity to cause damage to healthy stems, and its presence is always associated with damage from other pests, including the black weevil.

Laboratory assays were carried out at the National Agricultural University to assess pathogenicity. A fungal strain was isolated from a natural infection of coffee berry borer (Hypothenemus hampei), and this was mass produced on rice grains. Weevils were inoculated by either applying 10 g rice-fungus substrate directly, or by spraying conidia at 2 x 10⁹ conidia/ml. Inoculation method did not affect efficacy (judged by mortality and percentage sporulation). However, streaky weevil appeared to be the more susceptible, with 100% mortality reached 10 days after inoculation, compared to 45-50% for black weevil.

Subsequently, field trials were conducted in two commercial plantations in Tingo Maria using two entomopathogenic strains (GM1 and GG1) mass produced on rice substrate. These strains were isolated from black weevil adults from banana plantations around Tingo Maria. Traps used to attract the weevils were made from two discs of plantain pseudo-steams with 10 g rice-fungus substrate sandwiched between. These were laid out at a density of 30-40/ha at the stem-base of plantain trees and were changed every 7 days. Trapped weevils were recovered every 2-3 days and subsequent mortality and sporulation were evaluated in the laboratory. Strain GM1 induced 55.3% and 86.7% mortality in black and streaky weevils, respectively; sporulation occurred in 92.4% and 98.4% of these, respectively. Strain GG1 gave lower mortality (38.6% and 78.4%, respectively) and sporulation (74.2% and 82.1%, respectively).

Work is continuing on evaluating the pathogenicity of B. bassiana to the larvae, and also on formulations and mass production.

By: Enrique Arévalo, Oscar Cabezas, Luis Zúñiga and Nilda Albornoz

Contact: Enrique Arévalo,
Instituto de Cultivos Tropicales (ICT),
Jr. Tarapoto No. 247, Banda de Shilcayo - Tarapoto, San Martín, Peru
Fax: +51 94 522361
Email:

Effective and Economic Neem

Studies conducted since 1996 at the International Centre of Insect Physiology and Ecology (ICIPE) Mbita Point Field Station (MPFS) and in farmer's fields in western Kenya have been directed at finding ecologically sound and affordable technology for control of banana pests. Results indicate that neem products can give control that is not only as effective as that provided by pesticide treatment, but also far cheaper. The pests targeted were the banana weevil (Cosmopolites sordidus) and the parasitic nematodes Pratylenchus goodeyi and Meloidogyne spp. These are the predominant pests in the major banana-growing areas in East Africa, which are normally over 1200 m above sea level. The pests occur together in the same plant and attack both root and rhizome tissues causing severe fruit yield losses. Although effective synthetic pesticides exist, they are expensive and hazardous to use. In recent years, neem (Azadirachta indica) has come under close scientific scrutiny worldwide as a rich source of natural pesticides.

Neem seed powder (NSP), neem kernel powder (NKP), neem cake (NC) and neem oil (NO) containing 4000, 5500, 5800 and 850 p.p.m. of azadirachtin, respectively, were tested as such or in aqueous form. In laboratory choice tests, less than 12% of weevils had settled under neem-treated banana 48 h after release, while more than 50% had settled under untreated corms. In a feeding test, weevil larvae caused little damage to neem-treated corms, indicating a strong antifeedant effect. In addition, females laid 3-10 times fewer eggs in neem-treated than untreated corms, and only 25% of eggs laid in neem-treated corms hatched. Forty to sixty per cent of 2nd-instar larvae died within 14 days when confined to neem-treated banana pseudostems; the survivors were small in body size and weighed 4-6 times less than those in the control. The higher the concentration, the greater was the effect of neem materials.

Effective methods, frequency and rates of application of neem materials were determined at MPFS and in farmers' fields, under different levels of soil fertility and pest infestations using a highly susceptible banana cultivar (Musa AAA-East Africa). Results indicated that application of 100 g NSP, NKP or NC at planting around the base of pared or unpared banana suckers, planted in drums and inoculated with weevils and mixed nematodes, significantly reduced the nematode population and weevil damage, and results were on a par with Furadan treatment at 40 g/plant. In addition, 10 months after inoculation, NSP- and NC-treated unpared suckers supported eight times fewer nematodes than pared suckers treated with the same neem product. Paring is generally recommended because it reduces initial infestations, but it is labour intensive. The results of these studies indicate that NSP and NC treatment obviates the need for paring. The NKP and NO applications were phytotoxic, however.

Soil application of powdered NSP or NC was more effective than their application in aqueous forms. Application of powdered NSP or NC at planting time and then at 1- to 4-month intervals to plants grown with controlled pest infestations in drums significantly reduced nematode density and weevil damage. Their application at 5-month intervals and above was ineffective. Their application in the farmers' fields at 60-100 g/mat at planting and then at 4-month intervals significantly reduced nematode density, and weevil and nematode damage, and increased fruit yields by 30-60% in the second crop. Yields obtained with Furadan at 60 g/plant at 6-month intervals were equal or less than that in control plants. Even with low soil fertility and high pest infestation levels, the neem treatments controlled the pests and markedly increased the yield 7-10 times more than the control. However, application of NSP or NC at more than 200 g/mat at 6-month intervals was phytotoxic.

Depending on the soil fertility and doses of application, the net gain in the NSP or NC treatments was in the range US$70-800/ha while Furadan use led to a loss of $700/ha.

By: Thaddée Musabyimana
Current address: Agriculture and Agrifood Canada, Horticultural Research and
Development Centre 430,
Boul. Gouin, Saint-Jean-sur-Richelieu, Quebec, Canada J3B 3E6
Email:
Fax: +1 450 346 7740

Integrated Approach for Weevil in Cameroon

An integrated approach to weevil control is being pioneered in an INIBAP-CRBP (Centre de Recherches Régionales sur Bananiers et Plantains) project in Cameroon. Four species of weevil are present in smallholders' fields in southwest Cameroon: Cosmopolites sordidus, Metamasius hemipterus, M. sericeus and Pollytus melleborgi, although the most severe infestations are caused by C. sordidus. Plantains with the genotype AAB appear to be most susceptible. Weevil larvae bore into the corm and weaken the plant, resulting in high plant mortality, reduced bunch weight and reduced hand number.

Development of IPM strategies for C. sordidus for plantain production in smallholder conditions in Cameroon and West Africa involves control with botanical insecticides, biological control, genetic control and chemical control. Studies on population dynamics are on-going in order to assess the mobility of weevils in relation to the natural population build-up in plantain fields. Yield loss of plantain over different crop cycles is being assessed. Pheromone-baited traps have been tested on a small scale. Technology validation through farmer participatory trials with NGOs and extension agents is planned for 2000.

In a field trial in southwest Cameroon, neem (Azadirachta indica) seeds were efficient in protecting young suckers of plantain against weevil attack for 3 months when they were dipped in a 20% solution of crushed neem seeds before planting. Weevil damage was tempered and sucker mortality due to weevil attack was reduced by 25-30%. The efficiency of neem, when applied closely to the sucker using this dipping treatment, can be explained by the multiple mode of action of neem on the life cycle of C. sordidus, which was assessed under laboratory conditions for neem and three other substances (household ashes, coffee husk and hot pepper). Neem had a repellent effect on adult C. sordidus and slowed down oviposition. Fecundity of females in contact with suckers rolled in neem seed powder was affected. Hatching of eggs in contact with a 10% solution of neem seeds was blocked. The toxicity of neem seeds to adult C. sordidus was low compared to a classical contact insecticide, which explains why a crown application (encircling the mat) with crushed neem seeds at 40-100 g/mat did not reduce weevil damage or weevil populations. The toxicity of neem seeds varied according to their origin and storage conditions. Neem seeds did not have any nematicidal effect on R. similis in the field, which was abundant and caused severe damage and toppling in our field trial.

Wood ash, which is being used by more than 30% of smallholder farmers in Cameroon, because they believe it controls weevils, had a moderate repellent effect, but did not have any effect on oviposition, hatching of the eggs or adult survival. Coffee husk (Coffea spp.), which is reported to have an insecticidal effect, did not have any effect on weevils. Hot pepper (Capsicum spp.) blocked the hatching of eggs under laboratory conditions and had a moderate repellent effect on adults.

Three strains of entomopathogenic fungi have been isolated from infected adult weevils in southwest and west Cameroon. Pathogenicity and viability tests were done over various infection cycles under laboratory conditions. Germination tests have indicated that viability of the strains had diminished after multiplication and sporulation on artificial (Vegetal) media. Nevertheless, in vitro, one strain caused 94% mortality of weevils after 9 days in the first and second infection cycle. Strains will be re-isolated from infected adults in order to regenerate their viability and mass produce them on an adequate delivery system. Entomopathogenic nematodes have been isolated from dead C. sordidus larvae.

CRBP has one of the largest Musa collections in the world. Preliminary screening identified various genotypes with promising levels of resistance, including plantain hybrids and parents used in the CRBP breeding programme, which are being evaluated in the field. Weevil resistance will be incorporated in the CRBP plantain-breeding programme.

Several new insecticides are being tested in order to broaden the possibilities for chemical control. In Cameroon only one efficient insecticide (fipronil, as Regent) is available on the market for controlling weevils.

In a field trial in southwest Cameroon population dynamics are being assessed using mark-recapture techniques. In this trial, which had very low weevil populations at time of planting (April 1999), the population started to build up in July, with a peak in October 1999, at the end of the rainy season, after which the population dropped and remained constant for several months. Migration of weevils between plots was observed in July and August, but not later. Trapping results indicated that C. sordidus migrates by flight. Daily trapping and removal of weevils reduced weevil populations, but did not control them.

The trapping efficiency of a ramp trap baited with sordidin pheromone lures (commercialized by ChemTica in Costa Rica) was compared to split pseudostem or sandwich traps in a small plantain plot in terms of number of weevils per traps per week. These ramp traps were only two to three times more effective then the pseudostem traps, with the number of weevils caught being highly variable. Pheromone lures remained attractive for 3-4 weeks.

By: Stijn Messiaen, INIBAP Regional Office for West and Central Africa,
BP 12438, Douala, Cameroon
Email:
Fax: +237 42 91 56

UK Support for Bananas

The UK government has recognized the significance of bananas and plantains to the rural population of East Africa and, in a cluster of programmes funded by the Department for International Development (DFID) through their Crop Protection Programme (CPP), is providing financial support for research into sustainable solutions to pest and disease problems. Collectively these projects address the problems of banana weevil, foliar diseases, wilts, banana streak virus and nematodes. In one of these, new technologies will be incorporated into an IPM system for management of the banana weevil in Uganda. There will be close linkages in this work with the regional project concerning on-farm testing of technologies also funded by the UK government and with a project funded by the Gatsby Foundation which has similar objectives but emphasizes the large-scale introduction of new disease-free varieties to many farmers in the central region of Uganda.

The DFID Crop Protection Programme has recognized that the Uganda National Banana Research Programme (UNBRP) has the trained staff available to address the different constraints to banana productivity. With several promising techniques and practices available together with some new varieties developed by the International Institute of Tropical Agriculture (IITA) and FHIA (Fundación Hondureña de Investigación Agrícola) banana improvement programmes, work in Uganda should make significant advances in alleviating poverty and improving rural livelihoods: the major purpose of the current DFID policy.

Contact: Simon Gowen, Department of Agriculture, The University of Reading, Earley Gate, Reading, RG6 6AT UK
Email: Fax +44 1189352421

Managing Diseases in Uganda

Banana diseases, including wilts, leaf spots and parasitic nematodes, have been found to be major constraints to the production of both indigenous and exotic bananas in Uganda and a key contributor to a recent and significant decline in production. Management of these will be addressed in a programme for the integrated management of banana diseases funded by the DFID Crop Protection Programme. It began in January 2000, and is being led by CABI Bioscience in collaboration with the University of Reading, UK, the Uganda National Banana Research Programme (UNBRP), the International Institute of Tropical Agriculture (IITA), Uganda and the Natural Resources Institute, UK (NRI).

UNBRP have already identified or developed a number of cultural farming technologies with potential for alleviating disease constraints and, consequently, for increasing yields and reversing the downward trend in production. As part of its intervention or management phase, UNBRP is now placing emphasis on the evaluation and transfer, to farmers, of such technologies as part of an IPM approach. The DFID-funded programme will enable new technologies developed for banana disease management to be evaluated under farmer field conditions at benchmark sites in Uganda. These include banana varieties with different yield/growth characteristics and disease and nematode resistance, improved use of organic and mineral fertilizers, and related cultural treatments to improve plant vigour, clean planting material and break crops.

The project will identify technologies that are most effective in minimizing losses due to banana diseases under different agroecological and farm management conditions. Those technologies found to be most appropriate for widespread adoption in Ugandan farming systems will then be selected with a view to subsequent widespread adoption by farmers as part of an IPM approach.

A new wilt syndrome, referred to locally as `matoke wilt', has recently been observed on highland banana types. Its development appears to be related to particular farm management practices, including the use of household refuse and manure. The project will investigate these and other factors. Options for more effective management of this newly emerged problem by farmers may then be developed.

The project aims to provide stakeholders, including farmers, with information to allow them to judge and decide on the suitability of the various disease management technologies being assessed. It will provide a better understanding of the effects of particular farm management practices on plant health and on the prevalence and significance of major banana diseases. Direct participation by farmers, extension services and national agricultural research programme staff will greatly enhance their ability to recognise banana pest and disease constraints and increase their awareness of the beneficial effects of cultural farming practices. By identifying and facilitating the uptake of practices that most effectively improve banana plant health and reduce losses due to diseases, the project will enable farmers to increase yields and produce a crop of higher quality, thereby helping to reverse the general decline in banana production in Uganda seen in recent years. This, in turn, will enhance the livelihood of smallholder banana farmers in Uganda, by providing a more reliable staple food source and by increasing farm income generated through the sale of bananas at local markets. The demonstrative and training elements of farmer participation will contribute significantly to the rapid uptake of suitable technologies to achieve early impact for improved and sustainable banana production.

Contact: Mike A. Rutherford,
CABI Bioscience UK Centre (Egham), Bakeham Lane, Egham,
Surrey TW20 9TY, UK
Email:
Fax: +44 1491 829100

Managing Banana Streak Virus

Another of the DFID cluster of projects in Uganda is concerned with epidemiology, vector studies and control of banana streak virus in highland bananas, which is perceived as one of the major threats to this crop in the country.

Banana streak virus disease (BSV) is now a major constraint to increasing banana production in many regions of the tropics. In Uganda, and other countries of East Africa where bananas are a major component of the diet, declining yields and the abandonment of banana plantations by some growers is attributed to the effects of the virus. BSV is widely distributed in most common cultivars of banana across the country, with high incidence near the borders. The disease appears to be causing significant yield loss in certain locations, although reliable information on the relationship between disease severity and yield loss is lacking.

Because many aspects of the dynamics of BSV are poorly understood, disease management strategies are poorly developed and banana farmers in Uganda have no clear control options available. This project, which began in November 1999, is led by the Natural Resources Institute (NRI) in collaboration with the National Agricultural Research Organization in Uganda (NARO), the University of Reading and the International Institute for Tropical Agriculture (IITA) in Ibadan. It aims to gain a better understanding of the epidemiology and ecology of BSV, and its importance and effect on banana production in Uganda. The information generated will be used to provide recommendations on optimum crop and pest management practices to limit the spread of BSV and reduce its effect on crop productivity.

Many of the activities in the project will be conducted in selected `benchmark' sites where related studies on constraints to banana production are being carried out with groups of participating farmers. One such activity is an examination of the main factors influencing the activation of the virus, the expression of disease symptoms and the damage caused to the plant. The interaction of the virus with the host plant is complex; symptoms can be very variable, often being similar to symptoms of nutrient or water stress, and there can be periods of symptom remission. Most, if not all, Musa species and varieties tested to date have BSV-like sequences integrated into their genomes, which cause no symptoms in the host. However, it is suspected that some of the integrated forms can be activated under certain conditions. This may be influenced by climatic factors, plant nutrient status, and crop management. For example, water stress and cool temperatures are suspected to be the cause of localized outbreaks of BSV.

Spread of BSV will be examined through field experiments at the benchmark sites to monitor the natural spread of BSV in blocks of `trap plants' of virus-indexed Cavendish `Williams'. The experimental plots will be monitored visually for symptom expression and by enzyme-linked immunosorbent assay for the presence of BSV. In a related study, researchers at the John Innes Institute, UK are investigating the possibility that BSV is present in different strains in Uganda.

The role of insect vectors in the spread of the virus in the field is being assessed. BSV is a badnavirus, most of which occur in clonally-propagated tropical crops. In natural systems, the most important means of spread for badnaviruses is probably by vegetative propagation, but transmission by mealybug vectors has been demonstrated. Indeed BSV has been transmitted to banana under experimental conditions by three species , although not by African mealybugs nor in the field. An identification key to banana mealybugs is being developed by Jerome Kubiriba (NARO) and Gillian Watson (CABI Bioscience).

Contact: Tim Chancellor, Pest Management Department, Natural Resources Institute, University of Greenwich,
Central Avenue, Chatham Maritime,
Kent ME4 4TB, UK
Email:
Fax: +44 1634883379

Forecasting Less Fungicide for Sigatoka Disease in Guadeloupe

Yellow Sigatoka disease caused by Mycosphaerella musicola (anamorph Pseudocercospora musae) is one of the most important threats the banana industry has to face in Guadeloupe. The disease causes leaf spot, and heavy attacks can reduce considerably the number of leaves and, ultimately, bunch weight. Nevertheless the most important effect of this disease is indirect. Heavy spotting results in a reduction in the greenlife of the fruit, which considerably depreciates its export marketability. At least ten viable leaves at harvest are necessary for good fruit marketability and therefore excellent control of this disease is essential.

From 1937, when yellow Sigatoka disease was first reported in Guadeloupe, 40-50 fungicide treatments per year have been used to control the disease, and treatments have been applied systematically according to a pre-established `calendar' programme. The main objective of CIRAD-FLHOR (Centre de Coopération Internationale en Recherche Agronomique pour le Développement - Département Productions Fruitières et Horticoles) (ex-IFAC, ex-IRFA) research has been to reduce the number of treatments to the minimum necessary for ensuring good fruit quality, so decreasing the cost of control, as well as minimizing risks of fungicide resistance and environmental effects. Key to this was gaining a better knowledge of the disease, so that a forecasting system could be developed for optimizing fungicide applications.

Disease Epidemiology and Disease Control

Germination and stomatal penetration by the fungal agent of Sigatoka disease are impeded by fungal antagonism in old leaves. Stomatal penetration only occurs on the unfurled leaf (cigar) or on the first full leaf (leaf 1). Bananas produce new leaves at a rate of about one per week. Under climatic conditions most favourable to disease development, first symptoms are observed after 12 days on leaf 2. In less favourable conditions, the disease first appears on leaf 3 or older leaves, or not at all. There is a gradient of evolution of the disease from the top to the bottom of the banana tree, and fungicide applications should be aimed at the top of banana trees to control new infections.

The incubation period (from stomatal penetration to first symptom of disease) and the transition period (from stage 1 (minute yellow point) to stage 5 (necrotic grey spot), according to Brun's scale) vary widely with climatic conditions, and can be up to 100 days. Several steps of the disease cycle are highly dependent on the water status at the leaf surface. With such variability in the development of the disease, a forecasting system has real potential to reduce fungicide applications.

In the field, it is essential to control the disease before necrotic formation, as sporulation occurs in stages 4 (waterborne conidia) and 5 (airborne ascospores). Where leaves are heavily spotted they should be removed since they can produce ascospores for many months. Contact fungicides are not curative and are effective only before fungal penetration, so they are useless in forecasting strategies that rely on symptom observations. Systemic fungicides have a curative effect on streaks (stages 1, 2, 3), but not on necrotic lesions (stages 4, 5), although sporulation is temporarily decreased.

Area-wide control is important, as wind-transported ascospores can disseminate the disease over long distances, and failure of control in one area can affect neighbouring areas.

Four Keys to Successful Forecasting

The forecasting system relies on the timing of decisions and applications, treatment efficacy, and organization of control.

1. Decision Making. In Guadeloupe, both biological and climatic data recorded every week are used to decide on the application timing. The information they give is complementary. Climatic information is predictive and is useful in preventing spread of the disease. However, more importance is attributed to biological data, since they represent the real status of the disease. Comparison of theoretical and observed data is also essential to detect any disruption in the control strategy.

The biological forecasting system is based on early detection of new attacks by continuous monitoring. The stage of evolution of the disease (SED) is calculated as the product of the rate of new leaf production (foliar emission rate) and the speed of development of the disease. The speed of disease development is monitored continuously by inspecting the youngest five leaves on ten plants in a plot each week. The most advanced stage of the disease (according to Brun's scale) is scored for each leaf. A coefficient for each leaf number/disease stage association (which increases with disease severity and decreases with leaf age) has been calculated, and these are summed to provide an estimate of the speed of disease development. The foliar emission rate is an indicator of plant vigour, and the more vigorous the growth of the banana trees, the faster the disease develops.

SED is an indicator of the potential of development of the disease and graphic representation of its weekly value is used for decision timing. There is a threshold value for spraying, but attention is also paid to the slope of the graph of SED against time. Experience has also shown the level of SED up to which fungicide efficacy is maintained.

Climatic information identifies periods when conditions are not favourable to disease development. As thermal conditions are always favourable to disease development in Guadeloupe, temperature is not used for forecasting. On the other hand, Piche evaporation, assessed under an open-air station, represents well the water status at the leaf surface, taking into account global radiation, air saturation, wind and temperature. A relationship found between Piche evaporation and duration of treatment efficacy is used in decision timing.

2. Spray Timing. The time between decision and execution of one application should not exceed 7 days and the whole spraying area should be sprayed on the same day. Treatments are made by aeroplane, which facilitates a swift operation, but good logistics are essential. The climatic conditions required for aerial spraying are limiting, and the `windows' are small: only in the early morning and late afternoon do thermal inversion and air turbulence not interfere with spray deposition. Aerial application is not possible at all on rainy and windy days. Miss a `window' and control may break down.

3. Effective Treatment. This is dependent on the quality of the foliar application and good coverage is essential. Bad weather conditions on the day of application, irregular topography in the spray zone or the presence of obstacles can alter its uniformity. The use of mineral oil carriers has considerably improved the quality of coverage through aerial spraying with low volumes (at 12-15 litres/ha).

Efficacy relies also on a strong curative effect and systemic fungicides are thus preferred to contact fungicides. The systemic fungicides used for yellow Sigatoka control have an antimitotic mode of action or are ergosterol biosynthesis inhibitors of group 1 (DMI group) and 2. Oil carriers strengthen the curative effect because mineral oils have a fungistatic effect.

It is important to manage development of fungicide resistance, and alternation of groups of fungicides with different sites of action is essential. Regular monitoring of resistant strains using a methodology based on a germination test of conidia determines any changes in sensitivity.

Keeping the sources of inoculum at a very low level is also important. Chemical sprays do not eliminate the disease from spotted leaves, so where extensive spotting is present, new infections will develop quickly and the only solution is to remove leaves mechanically from the banana tree.

4. Organization. Since ascospores are transported by wind over long distances, the control strategy should be the same in all banana plantations. Organization is more efficient if centralized under a single technical service operating according to rational guidelines, rather than each grower implementing his own strategy, often with short-term objectives. Banana growers are grouped in an association that performs the control strategy. The cost of the phytosanitary campaign is covered by a tax on exported bananas.

Progress and Prospects

In Guadeloupe, the forecasting system has been operating for 25 years and consequently the control has been centralized over 6000-7000 ha of bananas. Disease assessment is done and meteorological records are recorded by a technical team from the banana growers association, and treatments are applied by the banana growers association or by a private company. An equilibrium of six treatments/year has been achieved since 1973 through a control strategy including timing of decisions and the use of a systemic fungicide in pure oil, compared with 10-20 treatments/year in other countries (Ecuador, Surinam, Dominican Republic, Jamaica, Windward Islands) where fungicides are applied on a calendar basis.

However, forecasting did not eliminate all problems. The exclusive use of benomyl from 1973-1982 led to the build-up of fungicide resistance. Fortunately, fungicides with novel modes of action were available at that time and their introduction in an alternation strategy enabled us to return to equilibrium.

Today, yellow Sigatoka disease is under effective control and is not affecting the quality of export fruit. The cost of control (0.08FF/kg) represents less than 3% of the production costs. The number of treatments and the quantity of pesticides discharged in the environment have been reduced 8-10 fold by the forecasting system. Nevertheless, we should not be complacent, for new fungicide resistance may develop. There is a need for more new fungicides with more novel modes of action, especially since antimitotic fungicide resistance is widespread in banana plantations. Products belonging to the strobilurin family are still under evaluation.

The problems faced by the banana industry are quite different to those for other crops, because only one group of cultivars (the Cavendish group) with low genetic variability is grown for export. A new approach to yellow Sigatoka disease control should combine genetic resistance or tolerance with a rational use of fungicides. It is vital to begin to look at these options because another important curse of bananas, black Sigatoka (or black leaf streak disease) (Mycosphaerella fijiensis), which is similar to yellow Sigatoka disease but more difficult to control, is now a serious threat for the Caribbean banana industry.

By: L. de Lapeyre de Bellaire,
CIRAD-FLHOR, Station de Neufchâteau,
Sainte Marie, 97130 Capesterre belle eau, Guadeloupe, French West Indies
and J. Ganry and X. Mourichon,
CIRAD-AMIS, BP 5035, 3
4032 Montpellier, France

Contact: Luc de Lapeyre de Bellaire
Email: /
Fax: +590 86 80 77

Integrated Action against Nematodes in Uganda

Plant parasitic nematodes are a major constraint to sustainable Musa production. In Uganda, which is the world's largest producer of East African highland bananas (Musa spp., AAA group), nematodes have been identified as a major factor contributing to declining production. The major nematode species affecting banana in Uganda are Radopholus similis and Helicotylenchus multicinctus at an elevation of between 1000-1300 m above sea level. At higher elevations, the most common nematode species is Pratylenchus goodeyi. At Sendusu, near Kampala (1120 m), production losses in the commonly grown cultivar, Mbwazirume, from R. similis and H. multicinctus were 30-38% under a variety of management regimes. Damage is characterized by reduced flower production and bunch weight, and an increase in plant toppling because of poor root development.

Nematodes can be controlled with chemicals, but these may have adverse environmental effects and the use of nematicides is too expensive and the products too dangerous for subsistence farmers. An integrated strategy (cultural, biological, cultivar selection) may be the best solution for nematode control, and this is being developed by staff at IITA-ESARC (the International Institute of Tropical Agriculture - Eastern and Southern Africa Regional Centre).

Diagnostic survey activities in Uganda were carried out in representative banana-growing villages. Data were collected on pest and disease incidence and severity and current pest management methods. In collaboration with the Uganda National Banana Research Programme (UNBRP), four benchmark sites have been selected for in-depth research and farmer participatory research. Currently work is focusing on three control strategies: host plant resistance, biological control by the use of endophytes and clean planting material. At the moment, most of our work is concentrated on R. similis, as this is the most damaging nematode species.

Host Plant Resistance

Research is directed at identification of durable nematode resistance sources and genetic analysis of nematode resistance. Screening for resistance to nematodes using field trials is a very time and labour-consuming effort. To screen the large available Musa germplasm, including landraces, commonly used exotic bananas and the IITA hybrids, a fast and reliable screening protocol is needed.

The method of single root inoculation was initiated in 1998, and improvements since have meant that the protocol is now fully standardized and can be used as a routine screening method. Four weeks after planting, three roots are selected from each plant and a cup is placed around them. Fifty females of R. similis in suspension are placed directly on the selected root. Evaluation 8 weeks later includes root health assessment and nematode counts. For each genotype the reproduction ratio is calculated. The advantages of this method are that a lower nematode inoculum and fewer plants per genotype are needed (since more roots per plant can be inoculated) and more hybrids can be screened at the same time. In each experiment, R. similis susceptible (Valery) and resistant (Yangambi-km 5) cultivars are included. Valery shows a high reproduction ratio with this protocol, while very low or no reproduction ratio is found in Yangambi-km 5.

Using this method, a number of promising hybrids have been identified in Uganda. These have low reproduction ratios, that are not significant different from Yangambi-km 5, indicating that they support low nematode densities, and these will be screened further for possible inclusion in the breeding programme and eventual release.

Currently, representatives of each clone set from the East African highland bananas are being screened. Hybrids derived from the East African highland bananas are also being evaluated. Progeny of the cross TMB2x 6142 × TMB2x 8075 look particularly promising. Both parents are also being screened against banana weevil, and first results indicate that they are resistant to the pest. They are also resistant to black Sigatoka and fusarium wilt. TMB2x 8075 has Pisang Jari Buaya (PJB) in its pedigree, which is highly resistant to R. similis.

Fungal Endophytes

Investigation of the potential of endophytic isolates of Fusarium oxysporum from banana for the control of plant parasitic nematodes began in 1997. The fungal cultures used in screenings for nematode control were isolated from Ugandan banana roots, cv. Mbwazirume, at the University of Bonn. Selection of candidate isolates was based on the nematode inactivating effects of their culture filtrates. None of the isolates screened to date were vegetatively compatible with the tester strain of the wilt fungus, Fusarium oxysporum f. sp. cubense, indicating that they are non-pathogenic. The fungal cultures were re-imported into Uganda through quarantine.

Standard protocols for screening fungal endophytes against the most damaging nematode, R. similis, have been established. Preliminary results of pot trials show that certain endophytic isolates reduced both nematode damage and reproduction in some banana clones. Plant growth-promoting effects have also been observed. Pot trials are on-going to determine the optimal combination of banana cultivar and fungal endophyte. The most promising endophyte, V5w2, is currently being tested in the field.

Clean Planting Material

Using pest- and disease-free planting material can reduce the spread of both plant parasitic nematodes and banana weevils. The objective of this technology-transfer project is the delivery of healthy banana planting material to farmer communities. In May 1999, a workshop was organized in Namulonge, Uganda to give training in hot water technology and to provide basic understanding of banana and yam pests and diseases.

Before planting, suckers are pared (to remove roots and infested rhizome tissue), after which they are hot-water treated (for 20 minutes at 53-55ºC). The use of clean planting material can increase production by 30-50% per cycle for at least four cycles compared with standard farmers' material in Uganda.

In Uganda, the project is carried out in four districts. The NGO Environmental Alert was involved in the mobilization and supervision of the farmers' activities and extension contact in Mpigi, while district agricultural officers performed these functions in Luwero, Masaka and Ntungamo. Baseline data were collected to establish initial damage and production levels. In total 1713 farmers were trained and 4487 suckers were treated and planted. Monitoring and data collection will be carried out in the areas where the planting has been done. A hot water tank has been constructed and the Minister of Agriculture for Buganda handed it over to farmers in that region. Similar projects are in operation in Rwanda and in Zanzibar.

Nematode-Root System Interactions

The spatial distribution of nematode population densities and damage in roots of the varieties Pisang Awak, Sukali Ndizi and Nabusa were investigated at three localities in Uganda, each with a distinctive nematode population composition. At Namulonge R. similis dominates, at Ntungamo the dominant species is P. goodeyi, while at Mbarara the two species coexist. At all three sites, suckers were removed from mats and assessed for nematode reproduction and damage. Nematode population densities were randomly distributed along the primary roots while nematode damage was significantly higher close to the corm than further along the primary roots, independent of cultivar and location. It was also observed that R. similis-infected banana mats had weaker plants and a reduced root system compared with those infected by P. goodeyi. The total length of dead roots was significantly higher when R. similis was present.

A study was also conducted on the relationship between banana nutritional status and nematode infection and damage. Nematode infection impairs nutrient absorption and distribution in the banana tissues. Potassium was shown to be the most impeded. When `complete' nutrients were supplied to nematode inoculated plants, it was observed that nematode populations and damage in roots were reduced. Excess or deficient nutrient supply resulted in increased nematode population and damage.

By: Carine Dochez,
IITA-ESARC,
P.O. Box 7878, Kampala, Uganda
Email:
Fax: +256 41 223459

Soil Suppression and Nematode Biocontrol in Australia

The Australian banana industry produces approximately 250,000 tonnes of bananas for Australian domestic consumption. The most damaging nematode to banana production in Australia and worldwide is the burrowing nematode (Radopholus similis). The nematode causes toppling of bunched pseudostems, reduced bunch weight and increased cycling time. Nematode damage on bananas has been managed by the routine use of chemical nematicides which are expensive and hazardous. Some nematicides are becoming less effective due to the development of enhanced biodegradation, reducing the chemical's efficacy.

However, the incidence of burrowing nematode attack in Queensland is patchy. Some farms have been producing bananas for up to 30 years with little loss to nematode damage, and they have no need for nematicides. We investigated some of these farms to see if their soil naturally suppressed burrowing nematodes and, if so, whether this could provide a basis for a biological control method. Burrowing nematodes feed and reproduce within the root cortex of banana plants, and this does not make them particularly amenable to biological or other any other form of control.

Glasshouse trials were conducted by Julie Stanton and Jenny Cobon (Queensland Department of Primary Industries (QDPI), Indooroopilly) and Tony Pattison and Caroline Versteeg (Centre for Wet Tropics Agriculture) to assess soil from ten banana crops in southeast Queensland and six in north Queensland, respectively, for natural suppression of burrowing nematodes. Banana cv. Williams were planted in either sterilized or unsterilized soil from each site, and three weeks later were inoculated with nematodes. Any biological effects in the soil were assessed by counting nematodes extracted from the banana root system in a misting chamber ten weeks after inoculation.

More nematodes were recovered from sterilized than unsterilized soil in eight of the ten soils from southeast Queensland, which indicated some kind of natural suppression due to a biological factor. However, only one soil from north Queensland showed similar suppression. Work is continuing to determine whether soil can be induced to become suppressive by the addition of organic amendments such as chitin and ash, and whether indigenous antagonistic organisms present in these soils may be encouraged to reduce nematode damage in bananas.

Seeking the Root Cause

The next stage in developing a successful biological control is to identify candidate organisms that are antagonistic to burrowing nematodes. Various microorganisms may be responsible, and the Queensland team have been looking at the range that may be involved.

Julie Stanton and Jenny Cobon conducted glasshouse trials with a number of fungi (all Fusarium oxysporum) they isolated from the cortex and the stele of surface-sterilized banana roots. The isolates included three from naturally suppressive sites and four from non-suppressive sites in southeast Queensland. Three more isolates were supplied by Kendle Gerlach of the University of Queensland. Banana cv. Williams plants were inoculated with fungi at repotting by placing inoculated grain sorghum under the roots. The plants were inoculated with nematodes 3, 6 and 9 weeks later. Ten weeks after inoculation, nematodes were extracted from the roots for 7 days in a misting chamber, counted and the fungi reisolated from the banana roots.

There was a reduced recovery of nematodes from a fungal isolate taken from a suppressive soil site in plants inoculated with nematodes 3 weeks after inoculation with fungus, but there was no reduction with the later inoculation dates.

Caroline Versteeg investigated endophytic actinomycetes she isolated from surface-sterilized banana roots taken from a site which showed suppression of burrowing nematode in north Queensland. The actinomycetes were obtained as pure cultures and initially tested for their ability to reduce nematode motility. Three isolates were found to significantly reduce nematode motility. These isolates were grown on sterilized wheat bran, which was then mixed with sterilized soil, in which a tissue-cultured banana plant, cv. Williams was grown. The banana plants were inoculated with motile burrowing nematode 7 days after inoculation with the actinomycetes.

In the pot trials no actinomycete isolates significantly reduced the number of burrowing nematode in the roots of the banana plants relative to an untreated control. However, more endophytic bacteria isolated from suppressive soils are currently being evaluated in pot trials. Modification of the screening procedures may be needed to determine the biocontrol potential of the candidate organisms.

Rhizobacterial isolates collected from banana fields in north and southeast Queensland are being investigated by Linda Smith (QDPI, Indooroopilly) for their role in suppression of Fusarium oxysporum f. sp. cubense. The isolates are also being assessed for their antagonistic potential toward burrowing nematode in bananas. Screening of isolates by Tony Pattison and Caroline Versteeg has commenced in north Queensland. Initial results suggest some suppression of R. similis by ten rhizobacterial isolates, and further screening is continuing.

Finally, Tony Pattison and Caroline Versteeg have been conducting pot trials to look at the antagonistic potential of mycorrhiza isolated from banana-growing soil in north Queensland. They found that the presence of R. similis appeared to reduce mycorrhizal colonization within banana roots. Conversely, there were significantly fewer nematodes in plants colonized by mycorrhiza, but there was no growth improvement in plants colonized by mycorrhiza. The use of mycorrhiza as a treatment to reduce R. similis infection for tissue-cultured banana plants is a good avenue worthy of further investigation.

Aiming for Suite Success

By conducting this research we hope to identify organisms that are antagonistic to burrowing nematodes and can be developed for biological control. Such organisms could be introduced to banana plants either at the repotting stage in a tissue culture nursery, as a dip for vegetative planting material prior to planting in the ground, or by injection into an established plant. The endoparasitic behaviour of burrowing nematode means that no single inundative organism is likely to provide the solution to nematode problems in bananas. By developing a suite of organisms with different niches on the banana root system, the chances of successful biocontrol of burrowing nematodes on bananas should be increased.

By: Tony Pattison, Queensland
Department of Primary Industries,
Centre for Wet Tropic Agriculture,
PO Box 20, South Johnstone,
Qld 4859, Australia
Email:
Fax: +61 7 4064 2249

IPM Leads to Increased and Sustained Yields in Ghana

Plantain is a primary staple food in Ghana, providing a source of food and cash income for resource-poor farmers throughout the year, thus contributing to national food security. In recent years, there has been a substantial decline and reduction in plantation life (<3 years) within the country. Severe plant toppling in the first season and rapidly diminishing yields in ratoon cycles mean that plantain is commonly treated as an annual rather than a perennial crop. This situation necessitates frequent land clearing which is costly to the farmer and the environment. The decline in plantain production is attributed to high levels of nematodes, banana weevils and foliar disease (black Sigatoka) together with poor soil fertility (resulting from shortened fallows) and the high cost of crop management. Plantain damage due to nematodes and weevils is frequently compounded by the use of infested planting material since farmers are usually unaware that suckers are the main source of inoculum for these pests. Lack of healthy planting material represents a major constraint to plantain farmers in Ghana. Suckers are costly, they are often infested with nematodes and weevils, and may not be available when the farmer is ready to plant.

Raising Clean Planting Material

These pest and disease problems have been addressed through farmer-participatory research. The development of a simple scheme for the production and rapid multiplication of healthy planting material in community nurseries has provided the foundation for improved plantain production. Plantain suckers collected from a healthy plantation are pared (roots and corm outer layers removed) with a cutlass to remove nematodes, and weevil eggs and larvae. The pared suckers are then multiplied using a `split-corm' technique, whereby they are split into several setts (4-10 depending on sucker size) which are germinated for 4-6 weeks in seed beds or boxes containing sawdust or other locally available sprouting media. The sprouted suckers are transferred to a field nursery at a spacing of 60 x 60 cm where they are maintained for 4-6 months before transfer to the farmers' fields. The advantage of the field nursery is that plants can be carefully maintained in a relatively small area at a stage when they are particularly susceptible to environmental stress. Prior to transplanting to the farmers' fields, suckers are pruned (pseudostem cut back to a height of 10 cm) and pared. If equipment is available, pared suckers are hot water-treated (53-55ºC for 20 minutes) before planting, as this is a highly efficient method for eliminating nematode inoculum in the suckers.

Farmer-Participatory Trials

On-farm trials were established in 1995 and 1997 to study the influence of planting material treatment (hot water treatment) and improved crop management (regular weeding) on pest and disease dynamics, growth and yield of plantain. At flowering, the density of Pratylenchus coffeae (the most damaging nematode on plantain in Ghana) in roots, was significantly lower in plants grown from hot-water treated suckers. The percentage of bunches lost prior to maturity (due to toppling, stem breakage, failure to flower or premature death) exceeded 75% for untreated materials under traditional management, compared with 43% for hot water-treated materials under improved management. The yield from hot water-treated materials that had regular weeding was more than trebled compared with untreated materials. Moreover, it was observed from the 1995 trial that planting material treatment, in combination with regular weeding, substantially lengthened plantation longevity. Farmer-managed plots with untreated materials were largely abandoned after the first season due to the high incidence of plant toppling. In contrast, hot water-treated materials that were well-managed continued to yield for 4 years.

Farmers' Perceptions

In 1999, Farmers' Fora held at three pilot sites gave farmers an opportunity to discuss and evaluate the technologies tested. Farmers stated that nursery production provided a means to generate and multiply clean suckers as required. Moreover, nursery-derived materials performed better than untreated suckers and could also be sold at a higher price than untreated suckers when required. The major constraints in nursery production were the need for watering during the dry season and the cost of labour for regular weeding. In addition, it was realized that use of the hot water tank for sucker treatment would be more feasible for farmers' groups than for individuals.

When asked to discus the performance of treated suckers, farmers' perceptions closely reflected the results of both agronomic and economic analyses. They observed that plots with hot water-treated materials under improved management (optimum plant spacing and regular weeding), required more labour but that this disadvantage was outweighed by better plant establishment, higher numbers of suckers, a shorter time to maturity, a higher number of bunches, longer plantation life and higher cash returns. For example, one farmer stated that "Plantain can survive on the farm for several years before dying-off, unlike previous years where the plant dies off after one harvest". Paring had been adopted by at least 40% of the plantain farmers at three pilot sites, representing good progress since a participatory rural appraisal in 1993 when farmers were unaware of the need for planting material treatment.

Economic Feasibility

Financial appraisal of plantain nursery production at three villages in Ghana showed clearly that the production and rapid multiplication of clean planting materials can be profitable even when there are adverse price fluctuations. Farmers producing clean suckers can benefit both through the use of their own clean planting material to produce improved yields and also by diversifying their farm income through the sale of clean suckers to other farmers.

Cost/benefit analyses showed that the use of clean planting material and improved management practices for plantain production was profitable compared with traditional practices. For example, hot water-treated suckers under improved crop management (optimum plant spacing and regular weeding) gave an economic return of US$800/ha over a 2-year period, representing net additional returns of $300/ha compared with traditional practices.

Dissemination

Technologies developed for the production and rapid multiplication of clean plantain planting material formed the basis of the curriculum for a plantain IPM Farmer Field School in Ghana (1997-1999). Twelve extension agents (from six major plantain-growing regions) were trained in methods for improved plantain production. Techniques learned were transferred to plantain farmers' groups in each of the officers' districts.

By: Kim R. Green, Project Coordinator

Research was funded by BMZ, Germany and executed by the International Institute of Tropical Agriculture (IITA) in collaboration with the Ministry of Food and Agriculture, Ghana (MoFA) and the Agricultural Research Station of the University of Ghana at Kade. The Plantain Farmer Field School was funded by the UN Development Programme (UNDP) (through FAO) with assistance from MoFA and GTZ (Gesellschaft für Technische Zusammenarbeit).

Contact: Kwame Afreh-Nuamah, National IPM Coordinator, Facility of Agriculture, University of Ghana, Legon-Accra, Ghana
Email:
Fax: +233 21 500184

Peter Neuenschwander,
Plant Health Management Division, IITA,
BP08-0932, Cotonou, Benin
Email:
Fax: +229 350 556

Post-Harvest Diseases of Banana

Diseases of Economic Importance

The most important post-harvest disease of banana is crown rot, which is caused by a complex of fungi with Colletotrichum musae being the main pathogen. Additionally, Fusarium spp. are ubiquitous secondary invaders and Acremonium sp., Botryodiplodia theobromae, Fusarium moniliforme, Fusarium pallidoroseum (formerly F. semitectum), Nigrospora sphaerica, Penicillium spp. and Verticillium spp. are of regional importance. These secondary invaders are often isolated from severely diseased fruit in higher frequencies than C. musae and therefore blamed as the causal agent. However, whenever quantitative inoculation studies were carried out in order to fulfil Koch's postulates, C. musae required at least one log unit less inoculum than the other fungi to evoke symptoms.

Colletotrichum musae establishes a latent infection in the field at early stages of fruit development. When the banana hands are severed from the rachis during harvest, spores of fungal inoculum enter the wound and initiate disease development at this window of opportunity. Crown rot symptoms usually only become visible during fruit ripening in the countries of destination. Then, however, the disease can progress rapidly and, in severe cases, the rot penetrates the pulp which renders the fruit unmarketable. Crown rot causes losses of 2-10% in all banana-exporting countries. Its incidence rises periodically in the rainy season.

Other post-harvest diseases include anthracnose, also caused by C. musae, cigar-end rot caused by Trachysphaera and Verticillium spp., finger rots caused mainly by B. theobromae, Ceratocystis paradoxa, Pestalozzia leprogena, Phomopsis sp. and Sclerotina sclerotiorum, and squirter disease associated with N. sphaerica in combination with physiological stress factors. Other fruit spots caused by Cercospora hayi and Deightoniella torulosa, and pitting disease caused by Pyricularia grisae are more accurately regarded as field diseases but symptoms exacerbate after harvest. Also, the boundary between field and post-harvest diseases is gradual because many post-harvest pathogens can establish asymptomatic infections in the field.

Crown Rot Control

Crown rot is commercially controlled by the fungicides thiabendazole (TBZ) and imazalil, alone or in combination, applied as a post-harvest dip or spray. Cross-resistance to benomyl, formerly used in Sigatoka control, renders TBZ relatively ineffective in many traditional banana areas. Other fungicides have been tested on an experimental scale but are not registered for use on export bananas. Several projects have addressed non-chemical control options for environmental concerns as well as the health of banana workers and consumers.

Breeding programmes, most notably at FHIA (Fundación Hondureña de Investigación Agrícola) in Honduras, have produced crown rot-resistant hybrids but these have different organoleptic characteristics from Cavendish clones and are not well accepted by the mainstream customer.

Cultural management options include reduction of inoculum and earlier harvesting which is sometimes combined with techniques that accelerate fruit development in order to minimize yield loss. Physical approaches require first and foremost a rapid cooling of the fruit after harvest and a continuous cooling chain throughout transit which should be of minimum duration. Due to the highly perishable nature of banana fruit, the chain of operations has to be well organized at all levels. Smooth crown cuts and immediate transfer of fruit into an alum solution for delatexing also reduces crown rot. During shipment, modified and controlled atmospheres have shown promise but many of them either proved too expensive for routine use or had negative side effects on other fruit quality characteristics. Natural and non-synthetic chemicals have been investigated, among them calcium preparations, plant extracts, organic acids and waxes. One of the most promising compounds, a citrus seed extract, is now rarely used because of its inconsistent effect, possibly due to the low shelf-life of the product. Alginate-calcium gels reduced crown rot under experimental conditions. They are likely to form part of an integrated approach with biological control rather than on their own.

Among the biological options, induced systemic resistance and antagonists have been tried. Induced systemic resistance is operational in a wide range of crops but has not yet been exploited in any post-harvest situation. However, culture filtrates or cell wall fragments of C. musae induce the production of antifungal components in the peel of green banana fruit. As a result, conidial germination of C. musae was inhibited on treated skins. Subsequent attempts to employ the more easily obtainable dead conidia of C. musae as a resistance-inducing post-harvest treatment were unsuccessful. However, as a pre-harvest treatment (injection of the rachis 1-2 weeks before harvest) high crown rot levels could be reduced to ca 70% of control. More data, however, are required to substantiate and quantify these tendencies.

A programme funded by ODA (the UK Overseas Development Administration, now DFID, the Department for International Development) and managed by NRI (Natural Resources Institute, UK) in the Windward Islands identified several indigenous organisms with potential for biological or integrated disease control. Whereas only few bacteria appeared effective, mycoparasites (fungi parasitizing other fungi) showed great promise. Some of them attacked the whole range of fungi involved in the disease complex, including structures which are very resistant to fungicidal attack such as conidia and haustoria. Others showed great tolerance to fungicides themselves and could thus be combined with reduced concentrations of fungicide in an integrated disease management system. The highly diverse population structure of C. musae renders a single-strain biocontrol agent unlikely to provide consistent crown rot control. However, mixtures of strains could overcome this problem. Each mycoparasite was found to act via a different main mechanism, i.e. parasitism, antibiosis, competition. Combinations of up to four strains of mycoparasites belonging to different species (mostly Gliocladium spp. and Paecilomyces spp.) complemented each other and progressively increased the biocontrol efficacy against mixed infection. Preliminary studies suggest that incompatibility is not a problem. Future research should thus focus on mycoparasite mixtures in compatible formulations such as alginate-calcium gels.

Further information: Krauss, U., Bidwell, R.; Ince, J. (1998) Isolation and preliminary evaluation of mycoparasites as biocontrol agents against crown rot of banana. Biological Control - Theory and Application 13, 111-119.
Krauss, U.; Matthews, P. (1997) Biocontrol crown rot of banana by strain mixtures of mycoparasites. In: Iglesias L., J. (ed) Primer Taller Internacional sobre Control Biológico y Producción Integrada en al Cultivo de Banano Fundación Ambio, EARTH, Guápiles, Costa Rica, November 1997, pp. 51-63.

By: Ulrike Krauss, CABI-CATIE Project, CATIE, 7170 Turrialba, Costa Rica
Email:
Fax: +506 556 0606

Alternatives to Chemical Control for Anthracnose in Guadeloupe

Anthracnose of bananas, caused by Colletotrichum musae, is the most important postharvest disease affecting the quality of exported fruits from the French West Indies. This disease develops during fruit conservation and ripening and it depreciates the fruit marketability. Anthracnose is in practice controlled by a postharvest fungicide. However, under Guadeloupe conditions, this chemical control has now reached deadlock for three main reasons: (a) aerial fungicidal sprays to control Sigatoka disease have resulted in the appearance of strains resistant to the active ingredients used for anthracnose postharvest control; (b) fungicidal treatments are not effective in all production zones, quite apart from the appearance of resistance; (c) consumer demand is for a reduction in pesticide use, especially those applied postharvest. So, there is a need for new control strategies, and these could be developed from a better knowledge of the bioecology of the pathogens.

Disease expression is dependent on a number of factors present at different steps in banana production, from fruit production in the field, to fruit packing, transport and conservation, maturation in ripening rooms, and marketing. Of these, the variation in potential fruit quality at the field level is particularly important, since it is responsible for seasonal (disease is more severe from September to January) and spatial (disease is more severe in low altitude lands of Guadeloupe) variations. Potential fruit quality is governed by a physiological (fruit susceptibility) and a phytopathological (level of fruit contamination) component. Recent work has been carried out in Guadeloupe on these two components in order to propose alternative strategies to chemical postharvest control.

The Physiology of Fruit Susceptibility

A diagnostic survey was conducted on 106 banana plots in order to identify the factors which might explain variation in fruit susceptibility to wound anthracnose as measured through artificial inoculation at flowering and wounding the fruit at harvest. This study showed that fruit susceptibility varies widely with pedoclimatic conditions and farming practices.

In the pedoclimatic area of halloysitic and ferralitic (low altitude) soils, where fruit anthracnose lesions developed most (54 plots), a relationship was found between the manganese (Mn) content of fruit and susceptibility to anthracnose: the plants producing the most susceptible fruit had higher foliar Mn concentrations and lower calcium (Ca) concentrations, and had grown on rather acid soils.

It is possible that the high Mn content of the fruit could have arisen from stress situations which could reinforce the ability of the fruit to synthesize ethylene, a hormone that can play a very important role at different levels in the host-pathogen interaction. It has been shown that ethylene activates germination, formation of appressoria and lesion extension. Anoxic conditions resulting from soil compaction or bad drainage can lead to a reduction of different forms of manganese into Mn2+ in the soil, and to a massive absorption of Mn2+ by the plant, accompanied by a lowering of the leaf Ca2+ content. For many plants, it has been shown that root anoxia causes ethylene synthesis by the plant shoots. It is then possible that the fruit from banana plants subjected to anoxic conditions may have a greater ability to synthesize ethylene. Work is still in progress to test this hypothesis experimentally in order to manage fruit susceptibility to anthracnose through optimized farming practices.

The Phytopathology of Fruit Contamination

Fruit pollution occurs in the field. Conidia germinate rapidly and form a melanized appressorium which remains inactive until the fruit ripens. A penetration hypha then develops and the mycelium enters the skin and later the fruit pulp, forming brown lesions. Once quiescent infections are formed, the pathogen is permanently installed on the host because dark appressoria can survive very adverse environmental conditions. So, potential fruit quality depends on the quantity of conidia that reach the fruit surface (fruit pollution) and the quantity of conidia that form a dark appressorium (fruit contamination).

Fruit Pollution

Colletotrichum musae does not sporulate on the green parts of the banana plant but only on senescent organs. We have conducted studies in order to (a) identify the inoculum sources contributing mostly to fruit pollution; (b) determine the dynamics of fruit pollution from flowering to harvest; (c) establish the mode of transport of this inoculum to the surface of the fruit; and (d) evaluate the effect of covering the bunches with a plastic sleeve on fruit pollution.

Fruit pollution occurs mainly from inoculum produced on the floral parts and the last bunch bract. Because they are closest to the fruit, the floral parts are the most effective inoculum source for fruit pollution. The elimination of the floral parts and of the last bunch bract at the flowering stage reduces considerably the quantity of conidia trapped, from flowering to harvest, in rainwater run-off under the bunches, and the level of fruit contamination measured at harvest. Moreover, C. musae is readily isolated from floral parts.

Fruit pollution occurs mostly during the first month after bunch emergence (the critical period) and strongly decreases thereafter. Most conidia were trapped in rainwater run-off under the bunches during this critical period. Most inoculum was isolated from the floral parts during this period. Lastly, the climatic conditions prevailing during this critical period were related to the levels of fruit contamination observed at harvest and to the cumulative number of conidia trapped in rainwater run-off from flowering to harvest.

In the absence of rainwater, inoculum is not dispersed and does not reach the fruit surface. All the floral parts of bunches from plants grown under rain-out shelters were inoculated with C. musae conidia and a large amount of inoculum was isolated from them. No anthracnose lesions were observed on the fruit at harvest.

Sleeving of bunches limits rainwater runoff and inoculum dispersal to the fruit surface. A reduction of more than 80% in the level of fruit contamination is observed on sleeved bunches compared with unsleeved bunches, even though there is no effect on inoculum production by the floral parts.

Fruit Contamination

We developed a methodology to assess the level of fruit contamination through the number of anthracnose lesions that develop on fruit (the technique is applicable to immature fruits aged 4 weeks). A good correlation between the number of anthracnose lesions and the quantity of appressoria is achieved when the fruit are conserved at high temperatures with elevated levels of ethylene.

The formation of melanized appressoria was evaluated, with a constant inoculum concentration, in controlled temperature and humidity conditions close to the natural contamination. The presence of free water is essential and appressoria formation does not occur within six hours. This indicates again the importance of rain or long dew periods for contamination.

Alternatives to Chemical Control

The above results suggest there are a number of strategies for managing anthracnose at various stages in the production system that could be investigated as alternatives to chemical postharvest control.

1. At the Field Level. Fruit pollution can be reduced by removing the floral parts and the last bunch bract. The operation must be carried out as soon as possible to minimize inoculum development. This practice, combined with sleeving bunches, gives a very significant reduction in the level of contamination of the fruit. Particular attention must be paid to early sleeving, which should be done before the fingers reach a horizontal position (or before all bracts have fallen). A better knowledge of the regulation of ethylene biosynthesis would allow farming practices to be developed which could improve fruit resistance to anthracnose.

2. In the Packing Station. Fruit must be handled carefully during transport to the packing station and packaging in order to avoid bruising, as this enhances disease expression. Maintaining good quality for the water used in de-handing and in the washing tanks is also important to avoid contamination of crowns. Lastly, packing the fruit in polybags allows the formation of a modified atmosphere (higher CO2 and lower O2 content), which is important to improve fruit conservation and slow down fungal development.

3. In Ripening Rooms. The quantity of ethylene used in the ripening rooms, as well as the time of ethylene contact and the temperature of fruit conservation can increase disease development during fruit maturation. Present practices should be reconsidered: high ethylene rates (>1000 p.p.m.) are used even though small rates (1 p.p.m.) can induce the climacteric rise.

Testing the Alternatives

The possibility of eliminating postharvest treatment through these different non-chemical measures will be evaluated in a trial on pilot farms. The level of fruit contamination at harvest will be forecast using the assessment test on 6-week-old fruits, and actual contamination will then be assessed. These results will provide useful feedback on the performance of the new measures.

By: L. de Lapeyre de Bellaire and
M. Chillet,
CIRAD-FLHOR,
Station de Neufchâteau,
Sainte Marie, 97130 Capesterre belle eau, Guadeloupe, French West Indies
and X. Mourichon, CIRAD-AMIS,
BP 5035, 34032 Montpellier, France

Contact: Luc de Lapeyre de Bellaire
Email: /
Fax: +590 86 80 77