December 1997, Volume 18 No. 4

General News

Weevils are a Thorny Issue

A stormy debate has broken out following the publication of a paper in Science by Savada Louda and colleagues at the University of Nebraska at Lincoln*. They argue that the flowerhead weevil Rhinocyllus conicus, introduced widely in North America since 1968 for control of thistles of Eurasian origin, may pose a serious threat to indigenous thistle species.

They documented an expansion of the host range of the weevil on native Cirsium species in national parks in Colorado and South Dakota and have published evidence that Cirsium species are being increasingly affected by the weevil: in 1992 the weevils infested fewer than 10% of flower heads but by last year up to 70% were infested. They found that the direct effect of the weevil on seed production was severe - with reductions of seed in Cirsium canescens of up to 86% and in Cirsium undulatum averaging 72%. They also found that native tephritids associated with Cirsium were ad-versely affected. The researchers argue thistles are particularly vulnerable because they are fugitive species with large seeds and generally depend on current seed production for establishment and persistence.

Louda et al. say that the original releases were made despite the results of initial feeding trials which indicated that the weevil's host range included the native North American genera Cirsium, Silybum and Onopordum; they say it was anticipated that a stronger oviposition preference for Carduus together with more successful larval development on this genus were expected to limit its spread to native plant species. What they say were unexpected were the frequency and magnitude of nontarget host dest-ruction, the time delay from insect introduction to host range expansion (where documented), the geographic extent of spread to native species, and the continuing increase in weevil feeding on native species.

Writing in the same issue of Science, Don Strong of the University of California argues** that too few resources have been directed towards studying the ecological basis of biological control, and in particular food web reticulations - the complex natural food webs. Louda et al. point out that successful and some unsuc-cessful programmes all leave nonin-digenous species in the environment, and their ecological costs have been insufficiently studied. They say that the complexity of the issue and the lack of data on post-release impact on nontarget species have left the issue unresolved - that there is a paucity of quantitative studies after deliberate introductions.

The two threads that run through these articles are an alleged failure to take note of the findings of pre-release studies in this specific case (and more generally whether the basic science is sufficiently well researched) and the general failure to follow up successful and unsuccessful biological control programmes to determine any long-term effects of introduced species on the ecosystem. The common practice of cutting costs by terminating programmes before the post-release phase is a decision not usually in the hands of the scientists conducting the programme. Funding agencies need to be made to understand the signifi-cance of including post-release studies in biological control programmes for the sake of both safety and good science.

To expand on the issues raised in these Science articles, we have asked for comments from Paul Boldt who was involved in the original importation of R. conicus into the USA.

*Louda, S. M; Kendall, D.; Connor, J.; Simberloff, D. (1997) Ecological effects of an insect introduced for the biological control of weeds. Science 277, 1088-1090.

**Strong, D. R. (1997) Fear no weevil. Science 277, 1058-1059.

Response of a Rhinocyllus Researcher

Biological control of weeds is the use of natural enemies to reduce weed populations to levels below their economic threshold. It is a non-chemical application of ecological principles to the economics of pest control. This method is not risk free but 762 biological control of weed projects, in 55 countries, during the past 160 years have never eliminated a plant species. In 1968, the European weevil, Rhinocyllus conicus was first released in North America as a biological control agent against musk thistle, Carduus nutans.

Musk thistle was accidentally intro-duced into the United States more than 100 years ago. During the 1950s and 1960s, this weed was a serious problem on pastures, fallow land and roadsides in at least 20 states. By 1973, economic infestations threatened one in ten counties in the United States. Musk thistle had become a major problem in both agricultural and native plant systems. Several states enacted noxious weed and seed laws requiring farmers to control their thistle populations. However herbicide control was expensive and contrib-uted to environmental contamination.

Rhinocyllus conicus promised beneficial control of musk thistle, but posed known risks then to native thistles. Host-screening tests showed that the weevil preferred musk thistle, but could feed on thistles in the native genus Cirsium. At that time native thistles were abundant and their habitat was not seriously threatened. Many of these thistles, such as Louda et al.'s study plant, Platte thistle, Cirsium canescens, were considered weedy and of little value.

The release of R. conicus was not made until the research data and potential consequences of the release were evaluated by a committee of agricul-tural and environmental professionals, federal agricultural regulatory personnel, and individual state administrators. Neither the National Environmental Policy Act (NEPA) nor the Endangered Species Act (ESA) had yet been enacted.

Louda et al. and Strong concluded that the presence of R. conicus on several species of Cirsium is unacceptable. Our position is that an expanding population of musk thistle presents a more unacceptable ecological and economic risk.

Two examples, of many, demonstrate the economic benefits from the decision to utilize R. conicus as a biological control agent on musk thistle. In the early 1970s, R. conicus was released in Virginia. By 1978, the weevil had cleared 100,000 acres [more than 40,000 ha] of pastures for a savings in herbicide and labor costs of US$ 1,500,000 per year or US$ 30,000,000 in the past twenty years, in 1977 US dollars.

By the early 1990s, 85 of 106 counties in Tennessee were infested with musk thistle. Establishment of R. conicus saved US$ 500,000 in chemical and labor costs in 1995 and US$ 1,000,000 in 1996. As in Virginia, these savings will continue to accumulate.

Seed damage on native thistles that was attributed to R. conicus is greatly inflated in the article of Louda et al.: (1) Plants were not selected randomly. The large plants evaluated in this study bloomed early and produced large heads. Both characteristics are attractive to R. conicus. The ovipositional period of R. conicus is shorter than the flowering period of Platte thistle. Randomizing would include plants with delayed flowering period and reduce damage estimates. (2) Seed mortality attributed to R. conicus should be compared to seed mortality by native insects, not to seeds in noninfested heads. (3) Plant sample size and the duration of Louda et al.'s studies were insufficient for a fugitive species. Platte thistle is subject to population swings and site relocations. Populations of juveniles often grow in areas away from the parents and therefore may elude damage.

Louda et al. and Strong suggest that the federally listed Pitcher's thistle, Cirsium pitcheri, may be attacked by R. conicus. Their concern is based on the connection that both thistles grow in a sandy habitats and are taxon-omically related. They did not consider the biology of R. conicus. This weevil avoids plants in open-wind habitats, as is Pitcher's thistle along the shores of the Great Lakes. Also the ovipositional period of the weevil is about half the length of the flowering period of the thistle. Were R. conicus to infest Pitcher's thistle, damage would likely be far from devastating. The Pitcher's thistle is more threatened by off-road vehicles and clearing of shoreline property than by R. conicus.

By: Paul E. Boldt, Research Entomologist, Agricultural Research Service USDA, 808 Blackland Rd., Temple TX 76502, USA

E-mail: boldt@brcsun0.tamu.edu

Bunny Business in New Zealand

When New Zealand's Deputy Director-General of the Ministry ofAgriculture, Dr Peter O'Hara, decided in July 1997 not to permit rabbit calicivirus disease (RCD) to be imported as a biocontrol agent for rabbits in New Zealand, he allowed that his decision would be seen as controversial. One wonders whether he could have imagined just how much more controversial the issue of rabbit control in New Zealand was about to become. Dr O'Hara gave three principal reasons for his decision: poor understanding of the epidemiology of both the virus and the disease it produces; uncertainty as to the effectiveness, and cost-effectiveness, of the virus as a control agent in New Zealand; and the inadequate legal basis of the proposed biocontrol management control programme. He said, "Our current lack of understanding of RCD as a tool for biological control is a serious impediment to rational decision making".

But he was not unsympathetic to the farmers' plight. One significant factor he highlighted was the unsustainability of the new `user pays' rabbit control policy which had replaced the government-subsidized system that had been in operation for more than one hundred years - a system which he suggested had made a major contribution to confining the rabbit problem to a relatively small area of New Zealand's land mass. In 1996 the cost of rabbit control had been shifted onto the landowners, and on rabbit-prone land these costs represented a major, and for many farmers unsus-tainable, financial burden. He said the fundamental problem was the high cost of currently available rabbit control technology, and that there was a pressing need for a new approach - he allowed that RCD might still be a candidate if its management could be assured and epidemiological questions answered.

Despite an incredulous and outraged reaction from farmers' groups, a revision of the decision was refused by the Director-General of Agriculture, Professor Bruce Ross, on 19 August. Fears were expressed that the virus would be illegally imported, and these were almost immediately proved justified: on 26 August, just one week after Professor Ross's decision, the first incidence of RCD was confirmed on South Island in the Cromwell region. Immediate and strenuous efforts were made to contain the disease, here and at a handful of other places where outbreaks were confirmed, by setting up strict quarantine measures. But as more and more cases were reported from an ever increasing area of the island over the next few days it became apparent that attempts to prevent the virus spreading were futile. The MAF Chief Veterinary Officer, Dr Barry O'Neil, said that it appeared that the disease had been deliberately and illegally introduced, an action which he described as "incredibly irresponsible", especially in view of the questionable provenance of the strain introduced, while other experts pointed out that late winter was not the ideal time to release the virus. Some public bodies also expressed concern: the Forest and Bird Society accused whoever had brought it in of committing ecological anarchy. However, some farmers said that the disease was present over most of South Island - that it had been present for two months, and spreading, before it was finally detected in Cromwell. But although many were proud to admit they had helped the virus to spread, none would admit to importing it.

Bizarre tales circulated of home recipes for brewing disease cocktails, and infected carrots and carcasses being distributed. Many farmers celebrated what they saw as a victory over bureaucratic procrastination, and they called for the virus to be given the official stamp of approval. At the same time, Government officials were raising doubts about the efficacy of the "Kitchen Whiz" viral strain and disputed that it was moving naturally between rabbits. However, MAF later issued a list of 12 technical reasons against the importation of a `pure' Australian strain and for continuing with the management of the illegally imported strain, arguing that it had no evidence suggesting that a pure strain would be any more effective and the time-lag in obtaining Australian seed-stock would have adverse effects.

Politicians were divided in their reactions - some sided with the Government in condemning the farmers' actions, but others gave the farmers their wholehearted support; the Parliamentary Commissioner for the Environment, Morgan Williams (who had been employed by the Ministry of Agriculture (MAF) and managed the NZ$ 25 million rabbit and land management programme between 1989 and 1995), was reported as saying that the illegal introduction had been an inevitable reaction by farmers who felt they had been let down by the system.

By the second week in September the Government were ready to throw in the towel and admit defeat. On 9 September they announced that they were faced with the reality that the illegally introduced RCD virus was clearly established in the South Island and that they would therefore move to legalize its on-going spread. They argued that this was essential to provide clear advice on how to handle the virus safely and effectively, and also to assist the Department of Conservation in its efforts to protect native predators deprived of rabbit prey, and to monitor risks to human and other nontarget species. But they made it quite clear that they were not condoning the farmers' do-it-yourself approach to biocontrol: they declared themselves still committed to ident-ifying where and how the disease had originated, and prosecuting those responsible for its introduction and initial spread.

Although spreading the virus became legal on 1 October, Dr O'Hara made it clear that there were still legal obligations: a free-for-all had not been declared. Under the Animal Protection Act there was a code of ethical conduct to be followed by people inoculating captured wild rabbits to produce viral material, and there were possible implications for the collection, preparation and spread of RCD under the Pesticides Act and the Resources Management Act.

Not made magnanimous in apparent victory, farmers' groups have refused to work directly with Government, national or local, to monitor RCD spread, but instead a farmer-led rural trust has been given a grant by MAF to monitor and give advice on the spread of the disease.

...And Mixed Messages from Australia

In Australia, the Government were able to announce some encouraging results. There, rabbit calicivirus (RCV) escaped from quarantine in 1995, [see BNI 17(1)] and was officially released in all mainland states in October/November 1996 and in Tasmania in April 1997. Although it is now found in all states, a report in February this year indicated that the virus was killing rabbits at only 90 of 350 sites where it had been released, and there was no evidence of viral activity at 184 sites. In April it was acknowledged that obvious declines in populations were evident at only a quarter of release sites. It was argued that high temperatures had slowed the spread of the virus and it was expected to pick up again in the autumn, when more releases were also planned. Initial impact in terms of rabbit kill was greater in the more arid regions, with more than 65% decreases in populations recorded at some sites. But in other areas, such as the Central Tablelands of New South Wales, little or no decline was recorded.

However, the preliminary findings of the National Monitoring and Surveil-lance Program released in August indicated that in the arid Flinders Range (where RCV first appeared in November 1995), the disease has played a major role in enhancing recruitment of perennial plants, even in areas where sheep grazing has increased as rabbit populations have declined. These perennial plants are important components of semi-arid ecosystems, providing stability especially during drought, support for other native vegetation and habitat for native fauna.

Dr Brian Cooke of the CSIRO (Commonwealth Scientific and Industrial Research Organisation) Division of Wildlife and Ecology has found evidence for a good suite of insects acting as vectors of RCV in the field, including seven fly species, two mosquitoes and European rabbit fleas. Circumstantial evidence points to insects being the main vectors: Dr Cooke notes that the optimum conditions for disease spread are around 24ºC; this is consistent with insect-mediated spread - if spread was mainly rabbit-to-rabbit, then it would be expected to be fastest in winter when the virus survives best. As it is, spread of the virus is expected to be poor at temperatures below about 15deg.C.

For further information on RCD/RCV see: Rabbit Calicivirus News on the Internet: http://www.csiro.au/communication/rabbits/rabbits.htm

Citrus Leafminer

Citrus leafminer, Phyllocnistis citrella, (CLM) is believed to have originated in Asia but has undergone a recent expansion of its range and is now found in most of the citrus-growing areas of the Mediterranean, the Caribbean, Central and South America, and in North America where it has disrupted IPM management prog-rammes. It is unclear quite why it did not colonize these regions earlier, what precipitated the recent spread and why it spread so fast, but few other pests have spread so fast over such a short time, and quarantine measures have had little impact. Programmes based on classical biological control of CLM have been initiated in many parts of the World, focused on the introduction of parasitoids from its probable area of origin in Asia. In addition, over 40 species of indigenous parasitoids have now been recorded as being recruited onto this pest as it has moved into the Mediterranean Basin, Africa and the New World. Marjorie Hoy and Ru Nguyen* have recently summarized efforts to deal with CLM by classical biological control intro-ductions, and have put forward arguments for how they consider the programme in Florida should proceed. In particular, they argue that great caution should be exercised before introducing ectoparasitic eulophids from Asia as biological control agents to Florida, and that the release of those known to be facultative hyperparasit-oids is undesirable. The following is taken from their analysis of events and the decisions behind them.

CLM reached Florida in 1993 and spread over 850,000 acres [approach-ing 350,000 ha] in a few months, despite all efforts to prevent this. Here, the need to develop an IPM approach was critical as the use of pesticides against it would have disrupted IPM programmes against other citrus pests. The encyrtid Ageniaspis citricola, which had already been reared, evaluated and released in Australia, was imported and released in Florida in 1994. Monitoring over the next two years indicated that the parasite established, multiplied, dispersed and overwintered in most of the Florida release sites; and up to 99% parasitism was recorded 15 months after release and rates of 60-80% were common. The parasite is now well established, widely distri-buted and abundant, although its ability to suppress leafminer popula-tions to below economic threshold has yet to be assessed. The story in Louisiana, the Bahamas and Honduras is similar, and in Honduras, A. citricola has replaced indigenous parasitoids to become the dominant species.

So are further parasitoid introductions needed? The authors argue not at present - not this early in the programme. They say that A. citricola appears to have all the attributes of a highly effective parasitoid: narrow host range, high reproductive rate, female-biased sex ratio, high dispersal rate and high searching rate. It is climatically adapted to the humid tropical and subtropical climates, but is also reported to have established in some Mediterranean countries. It may yet prove to be that rarity in classical biological control, the `silver bullet' - a single species capable of providing substantial population suppression, and if so, additional introductions in Florida could be redundant or, worse, disruptive owing to competition or hyperparasitism.

Fear of failure with A. citricola (or in some cases its failure to establish following initial releases) has led to further species introductions in Australia, Israel and other Mediter-ranean countries. In Spain, A. citricola looked initially promising as it multiplied and spread, but it apparently failed to survive the 1996-97 winter as it was not recovered the following spring. In Morocco, the parasitoid is thriving in coastal areas, but not in other hotter and drier areas, and they are introducing other parasitoids because of this. Most parasites of CLM are eulophids - from its area of origin in Asia and `recruited' as it spread. The authors suggest that pressure to release additional species - a cocktail of eulophids - has occurred because of the fear of the possibility that A. citricola may not establish or provide adequate control in mediterranean climates. Certainly it is known to need high humidity: adults die if held in relative humidities of below 95% for more than about an hour. Thus their behaviour would have to compensate if they are to survive mediterranean dry seasons, and whether they are able to this is not known.

The difficulties of locating and identifying what is likely to be the `best' natural enemy, and of predicting the effect of introductions are acknowledged. The authors looked at other classical biological control programmes to see if a pattern could be discerned, and evaluated parasitoid families on the basis of historical records. They concluded that although endoparasitic eulophids seem to be effective, ectoparasitic eulophids do not have a good record and are not a good prospect: many of them are facultative hyperparasites. In contrast, braconids, encyrtids and pteromalids have a higher rate of success in the historical record of classical biological control. They also note that the most effective natural enemy may vary over the pest range. They point to the paucity of information on biotypes of A. citricola in relation to the acknow-ledged importance of climate matching in classical biological control.

Extreme caution is countenanced in making releases of eulophid species which may prove to be either facultative or indirect hyperparasites of A. citricola. Detecting this requires detailed laboratory and field studies and access to historical records of the host ranges of each eulophid. The authors do not claim to predict the future, but they do point out that not all parasitoids collected from the CLM are necessarily desirable for classical biological control in regions where A. citricola is likely to be effective, and that, in Thailand for example, the parasitoid has been shown to be hyperparasitized by a large number of species.

*Hoy, M. A.; Nguyen, R. (1997) Classical biological control of the citrus leafminer Phyllocnistis citrella Stainton (Lepidoptera: Gracillariidae): theory, practice, art and science. Gainesville, Florida; Association for Tropical Lepidoptera. Tropical Lepidoptera 8, supplement 1, 19 pp.

It's a Peach for Yemen!

Swift and spectacular results recorded following the introduction into Yemen earlier this year of an exotic parasitoid, Pauesia antennata, to control the brown peach aphid, Pterochloroides persicae, (BPA) suggest that classical biological control may be scoring another success against aphid pests.

The BPA has reached pest status and caused widespread damage in a number of countries in southern Europe, the Mediterranean, North Africa and, more recently, the Arabian peninsula, but its area of origin is thought to be much further East. Distribution records suggest that it has gradually extended its range westwards and southwards, and more recently spread out from a focus in the eastern Mediterranean and Central Asia. However, although its main host, peach, originates from China where it has been grown for 4000 years, BPA has never been recorded from China. Peach and related fruit trees are assumed to have spread from China along trade routes - the ancient Silk Routes - over the centuries. The pattern of diversity of its known natural enemies supports this hypothesis: predators are the most commonly occurring natural enemies, and most are indigenous to a specific locality and have a wide host range - they have been `picked up along the way'; parasitoids are rarer, are more host specific and have been reported from Central Asia and Pakistan - as if they have slowly followed the aphid's westerly expan-sion.

The BPA is in the aphid subfamily Lachninae which contains a number of economically important genera. One of these, the cypress aphid, a Cinara species, is a devastating forestry pest in Africa for which IIBC initiated a biological control program-me in 1991 - successful establishment of another Pauesia species was reported in Malawi earlier this year [see BNI 18(2)]. The BPA is an aggregative aphid and forms large colonies; stress comes both from removal of sap and from the formation of sooty mould on the copious honeydew which reduces photo-synthetic capacity. There can be 12-27 generations per year. Fruit from infested trees is less marketable because of contamination from aphid honeydew and sooty mould, and because they are often abnormal in shape, size and colour. Heavy aphid infestations cause a decline in tree vigour and, over a number of years, even the death of the tree.

First reports of the pest in Yemen in 1993 described it as causing significant reduction in yield and quality of fruit, and heavy infestations were also causing severe decline of trees in the genus Prunus, but particularly to peach and almond which are grown by more than 70,000 farmers to whom they are both an important source of income and an important part of the diet. The Food and Agriculture Organization of the United Nations (FAO) funded an emergency Technical Cooperation Programme to find integrated pest management (IPM) solutions to the problem. IIBC were given the task of surveying for and selecting candidate biological control agents, quarantining the potential agents for introduction and conducting preliminary importation and release of the selected agents in the Yemen.

Drawing from the hypotheses about the area of origin and spread of the BPA, surveys were conducted in Pakistan on Prunus (peach, plum, apricot and almond) and apple trees, a complex of predators and parasitoids was identified and further studies conducted. Pauesia antennata, the only parasitoid recorded from late nymphal instar and adult aphids, was concluded to be the most promising for intro-duction to Yemen. Meanwhile in Yemen, monitoring studies gathered data on aphids at four sites for the whole of 1996 to give a baseline for comparison after parasite release. Rearing facilities were set up and staff given technical training in Pakistan.

Following quarantine in the UK, the introduction of P. antennata to Yemen was made in January 1997. In February release began, concentrated at three sites around Sana'a, since when more than 50,000 parasitoids have been released in total, with up to 1000 per day being released at the peak of production. The results have been spectacular: within two months the aphid populations in these areas and beyond have collapsed as a result of parasitoid attack and the parasitoid has dispersed up to 120 km away from this area.

The collection of baseline data in the year before parasitoid releases and the continued monitoring after will allow a very good appraisal of the situation before and after introduction of the agent.

For a review of BPA see: Kairo, M. T. K.; Poswal, M. A. (1995) The brown peach aphid, Pterochloroides persicae (Lachninae: Aphididae): prospects for IPM with particular emphasis on classical biological control. Biocontrol News and Information 16, 41N-47N.

For further information contact: Tony Cross, IIBC, Silwood Park, Buckhurst Road, Ascot, SL5 7TA, UK [E-mail: t.cross@cabi.org] or Moses Kairo, IIBC, Gordon Street, Curepe, Trinidad & Tobago

[E-mail: CABI-IIBC-CLAS@cabi.org] or Ashraf Poswal, PARC-IIBC Station, PO Box 8, Rawalpindi, Pakistan

[E-mail: CABI-IIBC-pakistan@cabi.org].

Biological Control of Bacterial Wilt Disease of Potato in Kenya

The International Mycological Institute (IMI) has been involved in developing a biocontrol agent against bacterial wilt of potato over the past five years. Funding has been through the UK Department for International Dev-elopment (DfID, formerly ODA) under two projects of three years' duration, that focus on Kenya and potato cultivation by small scale farmers. The research in Kenya has been coordinated through the Kenya Agricultural Research Institute (KARI).

Bacterial wilt is caused by Ralstonia (syn. Pseudomonas) solanacearum and is widely distributed in tropical, subtropical and warm temperate regions. The disease is a major constraint on the production of several important crops besides potato, particularly other solanaceous crops (tomato, chilli), bananas (Moko disease), groundnuts and ginger. However, within what is a very diverse species a particular sub-set of the pathogen is recognized as causing bacterial wilt disease almost exclusively on potato. This is commonly referred to as the potato race, and is synonymous to race 3 or biovar 2 under the current typing schemes. This race predominates in cooler climates typical of potato cultivation and forms the altitudinal and latitudinal limits of R. solanacearum distribution.

In the Kenyan context, potato cultivation occurs between 1200 and 2800 m above sea level as small scale practices where socio-economic pressures result in near continuous potato cultivation. When coupled with the poor availability of certified seed, the persistence and spread of diseases like bacterial wilt is cause for concern. Losses in Kenya owing to bacterial wilt have been serious in recent years, particularly at mid-altitude growing areas. These environs are infected predominantly by race 3 isolates. The spread of bacterial wilt in potato is not entirely understood, but in the majority of cases latently infected tubers are responsible. To date, no effective control methods exist for bacterial wilt disease. Plant breeding, field sanitation, crop rotation and use of bactericides have met with only limited success. Alternative methods for control, such as biological control, involving appropriate technology and low cost to the farmer are urgently needed.

Various recent studies have indicated that biological control of bacterial wilt diseases could be achieved using antagonistic bacteria such as non-pathogenic mutants of the wild type pathogen. However, as a consequence of the heterogeneity within the species R. solanacearum, no one biological control agent or methodology is likely to be universally effective. For this reason, IMI has focused the research on race 3, as the relative homogeneity of this race linked to its localized geographic distribution and limited host range affords a simple model in which the efficacy of a biocontrol agent can be assessed.

To date, a highly sensitive genomic fingerprinting procedure (rare-cutting restriction analysis by pulsed-field gel electrophoresis) has enabled the rational selection of R. solanacearum isolates for development as biocontrol agents through mutagenesis to a non-pathogenic form. Two mutagenesis protocols have been adopted for the induction of non-pathogenicity: method 1 (developed by the French organization Institut National de la Recherche Agronomique: INRA) involved transformation by an exogenous DNA element that shared homology to the pathogenicity genes; method 2 involved a deletion event induced by the insertion and deletion of the sacB gene of Bacillus subtilus. Using these methods non-pathogenic mutants have been produced for the selected race 3 isolates. The initial biocontrol assessments with these mutants under growthroom conditions at IMI have consistently demonstrated significant levels of protection against the onset of the disease. Current research continues to establish the agricultural value of the biocontrol agents developed and their environ-mental impact on the indigenous micro-flora of soils. This research is becoming increasingly based in Kenya, in line with the requirements of the research and the attachment of Mr Kinyua Mirimi of KARI to the project.

The need for the research continues to gain precedence in Kenya with new and more serious outbreaks of the disease centred on new regions being reported. In the national newspapers KARI researchers have likened the disease to the AIDS of potato!

For more information contact: Julian Smith, International Mycological Institute, Bakeham Lane, Egham, Surrey TW20 9TY, UK

E-mail j.smith@cabi.org

Immunocontraception in African Elephants

A humane and practical method of controlling elephant populations is being pioneered by a project involving the University of Pretoria (South Africa), Zoomontana (Montana, USA) and the University of Georgia (USA) in collaboration with the National Parks Board of South Africa.

They are developing a contraceptive method based on the immunization of female animals against pregnancy. Females are given a vaccine derived from the proteins forming the zona pellucida, the layer which surrounds the egg; the proteins in this case are derived from pigs. These proteins are crucial to fertilization, as the male sperm must bind with them before penetrating and fertilizing the egg. However, in the immunized female, antibodies to the pig zona pellucida proteins are formed, and when the animal comes on heat and ovulates the antibodies bind to the sperm receptor sites and block them. The female does not become pregnant after mating but will otherwise follow a normal sexual cycle. Once immun-ization is discontinued, the blood antibody levels fall, and in all but two animal species tested the animal gradually recovers normal fertility. Scientists also note that the system uses no hormones, and presents no danger to a foetus (in the event of a pregnant cow being treated inadvert-ently) and no potential for contam-inating the food chain.

Initial studies indicated that the antibodies raised against the pig zona pellucida proteins recognized the elephant proteins, and would thus work as a contraceptive. Further experiments were initiated to assess the dose-response characteristics, and so far results are similar to those in horses where the efficacy of the method has been extensively demon-strated.

Field trials were begun in the Kruger National Park in South Africa in October 1996, when female elephants with young calves (chosen because they were therefore unlikely to be already pregnant) were immobilized and assessed for absence of pregnancy by ultrasound, and then vaccinated and fitted with a radio collar for monitoring. Booster vaccinations were administered in November and again in June and October 1997 by `darting' the elephants from a helicopter. Vaccine administration by this method was shown to be practical and effective, an important factor for treating free ranging elephants. In addition, the height-at-shoulder of the calf was found to be a good indicator of calf age and thus the likelihood of the mother being pregnant; this would allow accurate selection of non-pregnant females in the field for remote vaccine delivery.

The laudable search for an alternative to culling has not proved to be easy. Earlier this year, a separate project assessing the prospects for using slow-release oestrogen implants, analogous to the human contraceptive pill, found that these induced a state of permanent false oestrus, and the behavioural complications this caused made the approach untenable. However, the scientists involved in the immunocontraception project stress that no behavioural or other side-effects have been recorded with this treatment despite extensive research in some 35 other animal species as well as elephants, and they are optimistic of success.

For further information contact: Douw Grobler, Ian Whyte and J. J. van `Altena, Shukuza, Kruger National Park, South Africa

E-mail: mailto:dougw@parks-sa.co.za

Also: Professor Henk Bertchinger
e-mail: henkbert@op1.com.za

Farmyard Tales

Two separate teams of researchers in the USA have shown some unexpected results for natural enemies of synan-thropic Diptera.

Black Dump Fly in Dairies

Biological control studies for manage-ment of flies on dairies have been limited primarily to the use of house fly parasitoids, and pathogens. In a first study of its kind, the potential of using the black dump fly, Hydrotaea aenescens (Muscidae), a predator of house fly larvae was investigated. This fly has been evaluated and in many cases used successfully to control house flies in poultry and swine operations. The ability H. aenescens to control house flies on dairies has not been considered because the manure of dairy cows, which is the habitat of many house fly larvae, is not conducive to the development of black dump fly larvae.

Previous long term sampling indicated approximately 30% percent of the flies on the test farm could be eliminated through manure management. How-ever, the majority of flies on the dairy were being generated in the calf barns where manure management was not possible. Parasitoids were abundant, and more than four species were commonly found. However, the parasitoids were relatively ineffective because the majority of house fly puparia were found buried under the fine silty sand within the barns. Consequently, other biological control alternatives were considered. Since unweaned calves are physiologically different from the adult cows (initially non-ruminating), it was thought their manure might support the develop-ment of H. aenescens. Results from preliminary laboratory tests using pure manure and manure mixed with bedding were encouraging. In all the treatments, the larvae developed to adults.

The first field release of H. aenescens larvae was made in mid January 1996. Adults were observed several weeks later at the release site, thus indicating the ability of H. aenescens larvae to complete their cycle in the calf barns. Eighteen additional releases were made from 22 April 1996 through 22 July 1996. Sticky card data were collected to show the ratios of house flies and black dump flies during the period when laboratory-reared dump fly larvae were released into the calf bedding.

The black dump fly did become established in the calf barns at the test dairy, but its hold in this new environment was somewhat tenuous. The 1996 sticky card data revealed that the percentage of H. aenescens from April through June decreased from 80% to less than 5%, with the onset of warm weather and the associated rapid development of house fly populations.

Surprisingly, the black dump fly not only survived the winter, but its population rebounded and maintained a strong presence in the calf barns. Whereas the 1996 sticky card data indicated a steady decline in the population densities of the black dump fly, the 1997 data showed that populations actually increased from ca. 50% in April to ca. 80% in May and June. These data indicate that the use of H. aenescens to control house flies on dairies does in fact show some promise, but more testing is needed to successfully integrate this fly with other environmentally friendly house fly control techniques.

By: Reginald R. Coler, Coler Entomol-ogical Services, Inc., 2405 NW 66th CT, Gainesville, FL 32653-1633, USA, Jerome A. Hogsette, Center for Medical, Agricultural and Veterinary Entomology, USDA-ARS, PO Box 14565, Gainesville, FL 32604, USA and Robert Farkas, University of Veterinary Sciences, Budapest, Hungary.

For further information contact: Reginald Coler,

e-mail: afn04330@afn.org

Pteromalids in Illinois

The results of a three-year study conducted by researchers at the University of Illinois at Urbana-Champaign to provide baseline data on the parasitoids of stable flies and house flies in Illinois have identified parasitoids with promise for biological control of these flies in the US midwest.

Spalangia endius, a well-known parasi-toid, reared, sold and used widely for biological control of nuisance Diptera in dairy systems, especially in California and in the northeastern USA, was common throughout the south of the State, despite the fact that it had been written off as climatically unsuited to this region because of the seasonal extremes of cold winters and harsh summers. A rethink on its potential is strongly indicated linked with consideration of the local climatic variations.

Spalangia nigoraenea, which dominates in other parts of the USA, was the most common species across the State as a whole and was also suggested to have potential as a biocontrol agent, while Spalangia nigra was common on stable flies in the north.

For further information see: Jones, C. J.; Weinzierl, R. A. (1997) Geographi-cal and temporal variation in ptero-malid (Hymenoptera: Pteromalidae) parasitism of stable fly and house fly (Diptera: Muscidae) pupae collected from Illinois cattle feedlots. Environmental Entomology 26, 421-432. Also see the Internet: http://www.uiuc.edu.