A Dacoit Defoliator
The reasons behind the puzzling sudden yet devastating outbreaks of the teak defoliator (Hyblaea puera) in India's teak plantations are beginning to be unravelled - the result of 20 years' painstaking research at the Kerala Forest Research Institute (KFRI). These results, together with substantial progress in developing microbial control methods, are improving the prospects for the successful management of what is probably the best-known forestry pest in India.
Teak has been recognised for centuries as the finest hardwood in the world because of its strength, durability, pest and rot resistance, attractiveness and workability. These outstanding attributes led to its demand in a wide range of industries, from shipbuilding to fine carving. Teak's native range includes India, Myanmar, Thailand and Laos, but teak plantations are now found in many countries of tropical Asia, Africa, Latin America and the Caribbean. It is, for example, the most widely planted tropical hardwood in Africa. In India, natural teak forests are restricted to peninsular India south of latitude 24º North. Indian teak forests are remarkably diverse and the genetic diversity exhibited by Indian teak is unparalleled. Here, it grows in both moist and dry deciduous forests, in different tree associations and on a variety of soils.
In the early days of commercial exploitation, natural teak was harvested from forests by either selective or clear-cut felling. Elephants were used to bring the timber from remote forest to collecting points from where it could be transported on (by river, road or rail), and this practice continues even today in remote areas. In India, early reckless exploitation (largely to supply naval dockyards) of teak from the natural forests led to depletion of teak and eventually to worries about the sustainability of the supply. It was soon recognised that world demand for teak could not easily be met by harvesting timber from natural forests and therefore attempts were made to raise teak in plantations. After initial setbacks, success came in the 1840s when India's first teak plantation was established at Nilambur in Kerala. Further pioneering research conducted here led to the development of the stump-planting technique, which increased the ease with which plantations could be established. Now there are some 1.4 million ha of teak plantations in India, and nearly 9 million ha are established worldwide.
With the expansion of teak plantations, pest problems also arose. The most serious is the teak defoliator. It was first recognised in 1898 in Konni Forest Division in Kerala State, and since then outbreaks have been recorded every year across India in extensive areas of teak plantations. Major outbreaks occur suddenly and are often spectacular, with millions of caterpillars feeding in the teak canopy. Trees can be defoliated within days, and healthy new growth can be stripped from thousands of hectares of teak forest in as little as a fortnight. However, although the outbreaks are spectacular when they occur, they are sporadic, so whether or not the defoliator causes significant long-term damage was the subject of some argument, until the matter was settled unequivocally by KFRI scientists. They undertook a five-year study in four- to eight-year-old teak plantations at Nilambur, where there are now 10,000 ha of teak plantations. The results showed that exposure to defoliator infestations led to a 44% loss of potential volume increment of these young trees. Extrapolating from this, trees protected from infestation would be ready for harvest in 26 years, rather than the usual 60 years, provided that other necessary inputs are given. This gave a clear indication of the economic desirability of controlling the teak defoliator.
Early attempts to control the pest concentrated on the use of insect parasitoids. By the 1940s, a package of practices was developed for biological control that relied on augmentation of the natural enemies by manipulating the natural vegetation within the teak plantation and the surrounding areas to promote some species of plants useful for the survival of the parasitoids. These recommendations were pushed aggressively by the forest entomologists but success was hard to achieve. Recent work has shown that successive generations of the teak defoliator do not inhabit the same place, so parasitoids are unable to cope with outbreak populations of the pest. Dr K. S. S. Nair, Director of KFRI, says that in many ways the teak defoliator control problem resembles a dacoity (banditry) situation. Even if they have the guns ready, they cannot pull the trigger unless they know where the dacoit will strike. So at KFRI they have been developing and refining methods of control, while at the same time trying to develop a method to predict the onset of an outbreak, by elucidating the mechanisms of outbreak initiation.
Studies on population dynamics showed that, contrary to expectations, small populations of the teak defoliator were present in teak plantations and natural forest during the non-outbreak period when food (i.e. young foliage) was scarce. What allows these to develop into massive outbreaks? It was initially supposed that numbers build up over several generations, but then, in large plantation areas in Nilambur, it was found that small (0.5-1.5 ha) widely separated epicentres were the forerunners of a more widespread outbreak as populations built up and spread. The appearance of these isolated treetop infestations proved to be correlated with the arrival of pre-monsoon showers, and represented the transitional stage between very sparse endemic populations and high density outbreak populations.
However, the reasons for the appearance of the epicentres were still not clear: they were not constant over the years and they did not represent highly favourable local environments. Furthermore, the results indicated that moth populations originating from the epicentres alone could not account for all the subsequent outbreaks within Nilambur, and an influx of moths from elsewhere was indicated. To explain where these might originate, scientists considered also the observations of newly emerged moth aggregations, synchronized flight activity and orientated movements of moths. KFRI scientists compared their findings with published accounts of research into aerial displacement of the spruce budworm, and suggested one of two hypotheses (or a combination of them) to explain the outbreaks: monsoon-linked long-distance displacement of airborne moth populations, or wind-aided concentration of dispersed local populations of moths.
Given the increased understanding of the outbreak mechanism, it was possible to consider control options. KFRI began by re-assessing the long-advocated biological-cum-silvicultural control strategy described above. At Nilambur, researchers recorded 15 of 40 parasitoid species known to occur in India, together with 79 species of predator, ranging from insects and spiders to birds. However, studies on the spatial dynamics of outbreaks showed that successive generations of the defoliator do not inhabit the same place, but emerging moths disperse before laying eggs, thus evading the impact of local build-up of parasitoid populations. It was argued that a control strategy based on naturally occurring parasitoids was unlikely to be effective, because this spatial separation of host and natural enemies during outbreaks would prevent a numerical response by parasitoid populations to host density. Parasitoids are thus unlikely to have an impact on outbreak populations, although they may play a role in suppressing endemic populations.
There also appears to be little prospect at the moment for using tree resistance to combat the problem. Many isolated trees often remain unaffected in the midst of a defoliated area. However, marking studies showed that the apparent resistance was phenological rather than genetic, and a tree not attacked in one year may be attacked in another year. There is, however, one early-flushing variety from Karnataka State, `Teli', which may be a useful genetic resource if the hypothesis of outbreak origin being linked to the onset of monsoon rains proves correct.
More promise currently is being shown by microbial agents. Some commercial preparations of Bacillus thuringiensis (Bt) have been shown to be effective in laboratory and field trials at Nilambur. However, a naturally occurring baculovirus has generated most interest. Natural infections of this pathogen usually account for large-scale mortality of later generation larvae, which leads to the collapse of the outbreak phase. The disease was characterized in a project funded by the Department of Biotechnology, Government of India as a nucleopolyhedrovirus (NPV). These viruses have a long history of safe and effective use in pest management worldwide and are highly specific to their target hosts. As the name implies, virus particles (the infectious units) are held within crystals of protein known as Polyhedral Inclusion Bodies (PIBs). Research demonstrated the defoliator NPV to have poor field persistence, which precludes inducing disease epizootics, but disease can be induced by spraying PIBs onto foliage. In field trials at Nilambur, 70-76% of foliage loss was prevented by a timely one-off spray of viral preparation during each outbreak. A collaborative programme with the UK Forest Research Agency (FR), led by Dr H. F. Evans, Head of Entomology, (funded by the UK Department for International Development, DFID) allowed protocols to be developed for field application of the virus. Use of a `Control Window' concept developed by FR enabled rapid progress to be made on the essential parameters required to use the virus in the field. The `Control Window' included rates and locations of larval feeding (target area), dosage-mortality relationships for all larvae (the dose), loss of virus from field factors such as ultra-violet light (attrition) and rates of coverage of foliage using selected sprayers and formulations (droplet density). These data were combined to determine precise field application rates for both virus concentration and droplet densities; preliminary field trials indicated that the parameters gave good predictions of expected target mortality rates. Further field trials have been conducted using the refined methods, and current work is focusing on developing mass production protocols and formulations for storage as well as field application, and determining optimum dosage and method of application to achieve maximum control of the infestation.
Preliminary trials in 1993 used a high volume sprayer and a rocker sprayer to reach the canopy of eight-year-old trees, and about one litre of spray fluid was required to cover each tree adequately. The recent collaborative work between KFRI and FR concentrated on ultra-low volume (ULV) sprayers. These have been standardized for delivery so that excellent coverage of foliage at the precise feeding sites of the teak defoliator larvae can be achieved, both with battery operated spinning disk atomisers and with motorized spinning cage atomisers. Spray parameters have also been calibrated to determine the optimum droplet sizes to deliver the desired dosage of PIBs per unit leaf. An oil-based emulsion was developed to prevent rapid evaporation under the very high temperature conditions typical in teak plantations.
After two decades of steady progress, the prospects for teak defoliator control appear to be very promising, but Dr Nair points out that much research is still needed. Testing the hypotheses of how outbreaks originate requires interdisciplinary efforts from the biological and physical sciences and coordinated data collection from the Asia-Pacific region. A critical mass of funds, expertise and technological inputs are necessary to address this problem, and KFRI hope to be able to catalyse such an effort. If it were established that outbreaks are indeed associated with northward progression of monsoon rains, forecasting methods can be developed to allow timely intervention, and the early flushing teak variety described above can be investigated further. On the other hand, recent studies have indicated that many of the late-season large outbreaks may be caused by moth populations originating from early epicentres. In a large teak area, such as Nilambur, it would therefore be useful to investigate whether rapid detection and destruction of early epicentres would prevent the late season widespread outbreaks.
Contact: Dr K.S.S. Nair [1/9/99-31/9/99]
Visiting scientist, Centre for International Forestry Research (CIFOR),
Latin America and the Caribbean are producing a fascinating array of biocontrol solutions and practices to suit the diverse agro-ecosystems and the variety of problems and farmer needs in this region.
Classical and augmentative biocontrol strategies are well established in sugarcane production in Venezuela, one of the countries where beneficials production in the private sector is strongest. Augmentative releases of Cotesia flavipes against Diatraea spp. borers on 35,000 ha of sugarcane estates in western Venezuela started in 1988, to complement earlier introduction of tachinid flies in the 1950s. Infestation levels have dropped from around 14% to 2% with average cost:benefit ratios for the industry in the order of 1:28. The success of biocontrol in sugarcane is largely due to the good organization of the sector and a network of rearing laboratories across the country. In contrast, maize is predominantly a smallholder crop and implementation of biocontrol has proved far more difficult in this sector, for socioeconomic rather than technical factors. Fall armyworm, Spodoptera frugiperda, is the key pest in maize and current reliance on chemical control has led to massive production costs and increasing pest damage. Biocontrol agents available for S. frugiperda include Bacillus thuringiensis (Bt), the fungus Nomuraea rileyi and the exotic egg parasitoid Telenomus remus, introduced in 1979. In the last two decades private and public sector Venezuelan organizations have developed and refined rearing systems for mass production of T. remus, including artificial diets for the host. New stocks of T. remus were obtained from Trinidad in 1986, which enabled production efficiency to increase and field releases to become feasible. Semi-commercial releases were made in 1989 and commercial releases have expanded from 120 ha in 1990 to over 1000 ha in 1997, with 50% reduction in maize input costs due to savings on insecticides. The private sector is now collaborating with NGOs and farmers' cooperatives to promote the use of natural enemies by smallholder farmers and one project involves farm family rearing of T. remus for local use [see BNI 19(3), 76N].
In Cuba, biocontrol has been implemented on a national scale since the early 1990s, mainly in response to the withdrawal of aid, including agrochemicals, from the former Soviet Union. Cuba has established a unique system of state and public sector beneficials and microbials rearing units known as Reproduction Centres for Entomophages and Entomopathogens (CREEs). Trichogramma production is carried out on a manual, cottage-industry scale on state farms and cooperative CREEs, where a two- room Sitotroga cerealella production unit can produce 50-70 million wasps per month for local supply to around 1000 ha. There are currently over 250 CREEs throughout the island and these are supervised by national research institutes who supply cultures and implement the national quality control system. In 1995 Trichogramma production was 94 million, treating 1.6 million ha of crops including cassava, pasture, cabbage and watercress, maize, solanaceous crops, beans, sweetpotato and other vegetables. The main targets for Trichogramma are the cassava hornworm Erinnys ello (26% area treated), Mocis spp. borers in pasture and Diaphania spp. borers on squash and cucumber. The tachinid Lixophaga diatraeae is produced in 49 rearing centres for stemborer control over one million ha of the country's important sugarcane industry. Predator conservation and augmentation methods used traditionally by smallholders have been reclaimed and refined under the current biocontrol programme. Predatory ants Tetramonium guinense and Pheidole megacephala can successfully control banana weevil Cosmopolites sordidus and sweetpotato weevil Cylas formicarius through a combination of artificial refuges and nest relocation. In 1995 over 96, 000 ha were treated with P. megacephala. Lacewings are used to a limited extent in hydroponic cultivation and in cassava and roses, and predatory mite production is now underway for mite control in citrus and other crops. Cuban laboratory production of predators includes two mite species, six coccinellids, three lacewings and the two predatory ants.
These are just two examples from `Nuevos Aportes del Control Biologico en la Agricultura Sostenible'*, a collection of state-of-the-art syntheses of biocontrol potential and practice in Latin America and the Caribbean presented at the 2nd International Seminar on `Biological Control within Sustainable Agriculture', held in Lima, Peru in May 1998. The seminar was organized by the Peruvian NGO Action Network for Alternatives to Agrochemicals (RAAA) in collaboration with the Peruvian National Plant Protection Service (SENASA) and the Neotropical Regional Section of the International Organization for Biological Control (IOBC). Twenty-three papers on biocontrol practice and implementation in the continent are included in the proceedings, representing seven different countries (Colombia, Honduras, Venezuela, Peru, Cuba, Mexico and Argentina). The seminar aimed to focus on current field practice rather than research and to disseminate experiences and analyse impact where possible, and contributions range from overviews of predator, parasitoid and pathogen control agents to summaries of crop-specific methods, country reports, commercial production, training and extension and the new paradigm of biotechnology in relation to biocontrol.
*Lizárraga Travaglini, A.; Barreto Campodónico, U.; Hollands, J. (1998) Nuevos Aportes del Control Biologico en la Agricultura Sostenible. Lima, Peru; Red de Accíon en Alternativas al uso de Agroquímicos, 397 pp.
The proceedings can be obtained (in Spanish
only) from: Action Network for Alternatives to Agrochemicals (RAAA), Apartado Postal
11-0581, Lima, Peru
Serious damage to indigenous plant species of the Galápagos Islands by the cottony cushion scale, Icerya purchasi, has led the Galápagos National Park Service and the Charles Darwin Research Station to consider the use of biological control in the archipelago for the first time. The cottony cushion scale was first discovered in the Galápagos in 1982 and has since spread to ten islands. At least 48 plant species are damaged by this species. High infestations of this scale pest debilitate the plant sufficiently to cause mortality in some species. In response to this, an advisory committee was set up in 1996 to discuss the pros and cons of using Rodolia cardinalis to control this species. The committee was primarily concerned with the impact that R. cardinalis might have on the Galápagos fauna and that it might feed on endemic scale insects, most notably, an endemic margarodid. Although R. cardinalis has been used as a biological control agent of the cottony cushion scale for over 100 years and has been introduced into many countries, to our knowledge, the feeding range of R. cardinalis has not been studied and only a few anecdoctal observations exist regarding its feeding behaviour (please enlighten us, if information exists to the contrary!). After two years of deliberations it was decided to host test R. cardinalis in a newly constructed insect containment facility at the Charles Darwin Research Station, while concurrently conducting experimental trials to determine the impact of the scale on endemic plant species. CSIRO Entomology in Brisbane kindly donated beetles and provided training, and in April 1999, a colony was established and research initiated to develop a suitable host testing methodology. Forays into the field to find the endemic scale insects, some of which have not been collected since the Beebe expedition (1923), have been successful so far and feeding trials are now underway. If all the species can be found, we hope to test endemic and native representatives of Margarodidae, Ortheziidae, Pseudococcidae, Eriococcidae and Diaspididae, among others. Results of the trials are expected to be available by December 1999 and will be presented to the advisory committee and the Park Service for a decision to be made as to whether the liberation of R. cardinalis is feasible in the Galápagos Islands or not.
We are indebted to Veronica Brancatini and Richard Vickers at CSIRO Entomology, Brisbane, for their continual support and advice, and of course for the beetles. Many thanks to Dug Miller at SEL/USDA for the scale insect identifications. We would also like to thank the Embassy of the Netherlands and Fundación Galápagos Ecuador for making this research possible.
By: Charlotte Causton,
Congratulations to the Project Directorate of Biological Control (PDBC) at Bangalore in India and its Director, Dr S. P. Singh. The PDBC has been chosen for the ICAR Best Institution Award - 1998 by the Indian Council for Agricultural Research