Just a couple of years ago, sceptics suggested that natural enemies were not up to coping on the grand scale of the water hyacinth (Eichhornia crassipes) disaster in Lake Victoria. They may have to eat their words, as that is just what South American Neochetina weevils have done to the water hyacinth. Furthermore, impatient stakeholders who favoured a supposedly speedier mechanical solution may have landed themselves with a large embarrassment. Massive harvesting machines ordered expressly (and at vast expense) to deal with the weed arrived in late 1999 to an uncertain future. There is now little weed to be seen on the surface of the lake, and there may be little left for the harvesters to do.
Although it reached Lake Victoria only 10 years ago, water hyacinth, a native of the Neotropics, was first recorded in Africa from the River Nile in Egypt in the 1890s. During the next century it spread from the Nile and later points of introduction in a series of expansions through the continent, until it colonized most of the major freshwater lakes and rivers of Africa. Lake Victoria was one of the last major waterways to be invaded by water hyacinth. It was reached only after the weed invaded in the 1980s the headwaters of the Kagera River in Rwanda, which empties into Lake Victoria at the Tanzania/Uganda border. However, infestations in Lake Victoria have another possible and more direct source: water hyacinth was apparently in cultivation in urban areas around the lake during the 1980s, and so could also have been introduced directly (which is how it is presumed to have reached Lake Kyoga in Uganda).
Water hyacinth has long been perceived as a problem for shipping and fishing industries, and only more recently as a threat to biodiversity. It can undergo rapid proliferation, forming dense mats of plants that cover water bodies, so reducing light and oxygen and changing the water's chemistry, fauna and flora. It poses a real risk to human activities or environmental conservation in the tropical regions of Africa, and in particular in Lake Victoria where the weed has thrived in the eutrophic conditions, notably in the shallow Ugandan coastal waters. Satellite remote sensing technology has recently allowed ICRAF (the International Centre for Research in Agroforestry, Kenya) and Future Harvest (Washington DC, USA) to confirm what many have long been suggesting: nutrients pouring into the lake fed the carpet of water hyacinth that was choking the rest of the life in the lake.
By the mid 1990s the effects of water hyacinth in Lake Victoria had become catastrophic. An unconfirmed report suggested that 12,000 ha of the lake were covered. The weed showed signs of significantly altering the ecological balance of the lake and posed an enormous threat to the region's socio-economy, health and industrial output. It choked the shorelines and created, in the stagnant water, an ideal breeding ground for malarial mosquitoes and the snail hosts of bilharzia. According to James Ogwang, NARO (National Agricultural Research Organization, Uganda), it became a common sight to see large mats of water hyacinth blocking Port Bell and, more seriously, the Owen Falls Dam. Hydroelectric production is particularly vulnerable to water hyacinth as the flow of water into the turbines tends to enhance the aggregation of the weed. Despite the best efforts of operators in four boats fitted with rakes and conveyor belts, the weed periodically clogged the dam, and the Ugandan capital, Kampala, was plunged into darkness. Ogwang points out that in addition to causing domestic inconvenience frequent power cuts had an effect ultimately on industrial output. By blocking ports and landing areas, the weed also threatened fishing industries and disrupted trade as vessels had to wait for wind to clear channels. Fishing communities all around the lake faced a complete loss of livelihood. Passenger services were cancelled and Kenya Railways, who grounded a five-vessel fleet operating between Kisumu, Port Bell and Mwanza in Tanzania, estimated that the weed had caused a 70% reduction in lake transport. At Mwanza, large tankers remained in port, blockaded by metre-high water hyacinth. Water hyacinth even became a threat to security as, for example, Kenyan police and fisheries patrol boats became trapped in weed-locked ports.
The pan-continental extent of the problem was highlighted, says Jeff Waage (CABI Bioscience), at the CAB International Triennial Review Conference in September 1999. CABI's African member countries actively raised their concerns regarding water hyacinth in the discussion of CABI's growing role in invasive species management, and formulated a recommendation encouraging CABI to expand its activities.
Biological control of water hyacinth is nothing new. Indeed, it has been one of biological control's resounding success stories. Initial exploration by staff from CIBC (the Commonwealth Institute of Biological Control - now part of CABI Bioscience) and the US Department of Agriculture (USDA) at Fort Lauderdale in Florida led to the discovery of natural enemies, including two Neochetina weevil species and the moth Niphograpta albiguttalis (= Sameodes albiguttalis) from Argentina. These species have since been distributed worldwide. Water hyacinth was first controlled in Florida using these species. Interestingly, biocontrol there is currently adversely affected by other control measures such that biocontrol cannot reach equilibrium. In Louisiana, by contrast, water hyacinth coverage has been reduced dramatically, although it took many years to reach current low population levels.
Another key player has been CSIRO (the Commonwealth Scientific and Industrial Research Organisation, Australia) who implemented successful control programmes in Australia (in the 1980s) and subsequently in Papua New Guinea (PNG) in 1993-98. The PNG project, funded by ACIAR (the Australian Centre for International Agricultural Research) involved the release of both species of Neochetina weevils. On the lagoon system of the Sepik River, water hyacinth covered some 27 km2 in 1995, but this has been reduced to a fairly stable 5-10 km2 by the action of the weevils. Weed cover in the worst-affected lagoons was reduced from 80% to less than 10% fringing cover. According to Mic Julien (CSIRO Entomology) temperatures in PNG were not limiting and best control was achieved in eutrophic situations where it took 2.5 years; in less eutrophic situations it took longer, up to 5 years.
In collaboration with USDA and PPRI (the Plant Protection Research Institute, South Africa) and recognizing that the current control agents were not suited to all habitats and may be affected by seasonal events such as monsoonal floods, CSIRO studied and released the moth Xubida infusella in Australia and PNG. This insect is established but has not yet had time to demonstrate if it will complement other agents and have a significant impact. In addition, field and laboratory studies conducted on the two Neochetina weevils demonstrated their different nutrient requirements; N. eichhorniae does better on the lower nutrient plants while N. bruchi performs best on higher nutrient water hyacinth. Such information has important management implications.
Biocontrol in Africa began in the 1970s with release of Neochetina weevils and Niphograpta albiguttalis on the White Nile in the Sudan. Water hyacinth first appeared there in 1955 and rapidly became a major weed throughout the Judd region, extending over 1700 km from Juba to the Jebel Aulia Dam south of Khartoum. Cooperation between the government of the Sudan and USDA led to the release of Neochetina eichhorniae in 1976. Under a joint project involving the British and Sudan governments and CIBC, N. bruchi and Niphograpta albiguttalis were introduced and large-scale releases were made in 1979-81. All three agents were established in 1980-81 and this led to dramatic reduction in the growth rate of the weed. Since the 1960s, the dam had suffered from annual accumulations of over 11,000 ha of water hyacinth, but since 1982 there has been no build up of floating mats. The formerly intensive chemical control programme was terminated in 1983, although some manual clearing of canals continued. Follow-up studies to evaluate the relative contributions of the three natural enemies were curtailed by civil unrest, but anecdotal evidence suggests that control has been maintained.
PPRI is currently conducting enterprising research into water hyacinth biocontrol. Water hyacinth has been South Africa's most damaging aquatic weed since it was first recorded there around 1900. A biocontrol programme was first launched in 1973 with the release of the weevil Neochetina eichhorniae, but terminated in 1977 when the authorities opted for quick solutions and resorted to incompatible chemical measures. The biocontrol project was restarted in 1985, since when four more insect natural enemies (another weevil, N. bruchi, the moth Niphograpta albiguttalis, the mite Orthogalumna terebrantis and the mirid Eccritotarsus catarinenesis ) have been released and established.
PPRI have concentrated recently on investigating reasons for variability in water hyacinth biocontrol. Martin Hill (PPRI) says that they are asking why biological control of water hyacinth has worked in some areas of the world, and some areas of South Africa, but not others. He says that their results suggest that the following may be important in South Africa:
Water hyacinth has been
present in the Shire River in Malawi since the early 1960s. It came from
the Zambezi River and spread slowly north along the lower Shire River,
colonized the upper Shire and the southern part of Lake Malawi in the
mid 1990s, and is still spreading in Lake Malawi. A project funded by
DFID (the Department for International Development, UK) to control the
weed using biological control was implemented from 1996 to 1999 by the
Fisheries Department, Malawi, with assistance from CABI Bioscience and
PPRI, South Africa. It had four parts:
Novel aspects of this project included the active engagement of local communities in biological control implementation and the development of extension messages as radio jingles in three local languages. This is the first project in Africa to contain both socio-economic and environmental monitoring components in an attempt to quantify the effects of the weed and measure the success of control procedures.
Both the Neochetina spp. have been established, and they and the mite O. terebrantis have been widely distributed along the river, starting with the lower Shire. Large-scale impact of the biological control agents upon the abundance of the weed is not expected for another 1-2 years. However, once water hyacinth reached the upper Shire River, biological control agents were released, and since then there have been consistent reports of low incidence of water hyacinth. One feature of this programme is that it covers control of the weed from relatively oligotrophic waters, in a cooler climate at the start of the Shire River as it leaves Lake Malawi, to eutrophic waters in the hot Shire basin. Hence there is an opportunity to monitor the impact of biological control in different environments in different parts of the river which may give clear indications of how host plant nutrient status and temperature affects the impact of biological control agents.
In the early stages of the project, glyphosate was used strategically to knock down a new and heavy infestation of water hyacinth in the Lilongwe River, which flows into Lake Malawi. After spraying, biocontrol agents were released onto the weed so that the surviving weed would become infested with the biocontrol agents prior to being washed down the river during flooding. Since then, heavy infestations have not been noted on the river.
Work continues now through the water hyacinth component of the Environmental Management Project within the Environmental Support Programme with funding from the World Bank. The Fisheries Department will introduce further biological control agents, including the mirid E. catarinensis and the moth Niphograpta albiguttalis, with continuing assistance from CABI Bioscience and PPRI.
Water hyacinth appeared in the coastal lagoons of West Africa in 1984-85 and quickly spread, causing serious disruption to local communities who lived on or beside the water and relied on it for their livelihood. A programme led by IITA (the International Institute of Tropical Agriculture) at Cotonou, Benin concentrated on the release of two species of Neochetina weevils and the moth Niphograpta albiguttalis, obtained from CSIRO, Australia. The first releases were made in Benin in 1989, and subsequent releases were made in other countries including Nigeria, Ghana, Côte d'Ivoire, Niger and Burkina Faso. It is still too early to assess results in many of these countries, but Peter Neuenschwander (IITA) says that in Benin the success of the programme is beginning to be recognized as some waterways are opening up to fishing again. In 1999, IITA staff were invited to a fishing festival by fisherman of one formerly covered lake that can now be fished after a gap of 7 years. This festival was shown and explained on national television. However, some of the infestations are still resisting control, and a fourth agent, the mirid E. catarinensis, was imported from South Africa in late 1999. In January 2000, another fishing community invited IITA to a similar festival. Economic studies show a benefit/cost ratio for this programme well in excess of 100, which is, Neuenschwander points out, in the same range as other famous biological control programmes against agricultural arthropod pests. Thus, he says, after more than 10 years of painstaking monitoring and advocacy, the successful reduction of water hyacinth is beginning to be recognized by fishing communities and politicians alike.
Yet while biocontrol programmes in other African countries were progressing, it was not until 1997 that the problem on Lake Victoria was seriously addressed. At that point it was suggested that the severity of the infestations that had developed meant that biocontrol could take 5, or even up to 10, years to bring the weed under control. Too long, said many. Others argued that natural enemies would never succeed in controlling the weed and that mechanical and chemical methods stood a better chance of providing a much-needed solution. The World Bank made available funding to manage the weed through the Lake Victoria Environmental Management Project (LVEMP). In the first phase of a 5-year sub-regional programme involving Kenya, Uganda and Tanzania, US$77 million were allocated for research on fisheries and water quality in the lake, and management of the wetlands and the environment, and of this $8.3 million were allocated to controlling water hyacinth.
Why, with such a notorious invasive weed as water hyacinth invading the lake, did it take so long for the problem to be recognized and solutions sought? According to Hans Herren of ICIPE (the International Centre for Insect Physiology and Ecology, Kenya), the machinery of international cooperation simply ground too slowly. He argues that specialists in Africa (South Africa, Benin, Zimbabwe and Kenya) with experience of water hyacinth programmes were warning of impending disaster as early as 1991, but were ignored until the problem had reached catastrophic proportions. As a result, the cost and time needed to solve it escalated.
In June 1991, a workshop in Harare, Zimbabwe organized by the Commonwealth Science Council, UK (CSC) in collaboration with IIBC (the International Institute of Biological Control, now part of CABI Bioscience) and CSIRO discussed the control of Africa's floating water weeds. Delegates from eight African countries were joined by water hyacinth control experts and representatives of other concerned bodies. The workshop identified water hyacinth as one of Africa's three worst water weeds. It was described as increasing at an alarming rate in rivers, lakes, swamps and lagoons. The severe socio-economic and environmental impact of the weed was also recognized.
Timothy Twongo, UFRO (Uganda Fisheries Research Organization), speaking at the workshop, reported that aerial surveys of Lake Victoria the year before, in June 1990, had revealed extensive occurrence of water hyacinth in bays and inlets on the Ugandan shoreline and among offshore islands. He warned of casual reports since then suggesting that the weed was spreading rapidly, and he argued that another aerial survey would confirm an escalating picture of colonization of the lake. At the same meeting, Doug Taylor of IUCN (the International Union for Conservation of Nature and Natural Resources) described how the pattern of water hyacinth movement and colony location led IUCN to the mouth of the Kagera River. They found substantial quantities of water hyacinth in the form of large clumps and individual plants continually entering the lake and spreading, mostly northwards, from the mouth of the river. Taylor accurately identified future trouble spots on Lake Victoria: the Owen Falls Hydroelectric Dam at Jinja in Uganda (because of calm water on the lake side), the innumerable fish-landing sites of Uganda's shore and islands, and Kenya's Winam Gulf (because of wind blown weed) and thus Kenya's main lake port at Kisumu. He also noted that water hyacinth had been reported to be covering Speke Gulf east of Mwanza in Tanzania. Taylor said that control of water hyacinth in Lake Victoria would depend upon the introduction of a biological control agent, and that "No other control measure will work because the spread of the weed [in Lake Victoria] is now too wide for application of chemical or mechanical means."
The workshop also recommended that control should be based on integrated methods, but noted that solutions were likely to be site-specific. They recommended that biological control be implemented immediately an infestation was confirmed. They noted that biological control was the only cost-effective, sustainable and environmentally friendly method of combating water weeds; although physical and/or chemical measures should be implemented if these were judged necessary for short term relief, this should not jeopardize the long-term goal of biological control.
With hindsight, the disastrous consequences of the subsequent delay in implementing control measures in Lake Victoria are all too apparent, but at the time the issues were not as clear cut. Advice and information coming in from other parts of the world with experience of water hyacinth invasions gave conflicting messages. Although to some it was a serious threat to utilization of water resources, to others it was a valuable resource for fertilizer, biogas production, fibre and other local industries. Mic Julien comments that unfortunately the decision makers had not conducted the search for information that would have shown that, of the many utilization schemes, large and small, attempted in many countries, only cottage industries have ever continued. The main problems, he points out, are that water hyacinth is about 95% water, and that it grows in inaccessible places; energy and costs for gathering and transporting it for processing are prohibitive. Localized cottage industries are unable to use sufficient weed to offset the problems it causes.
Cooperation between countries was a problem: the weed spread more easily and quickly than information across borders, and inter-country consensus was not always easy to get, despite the facilitation role of FAO (the Food and Agriculture Organization of the UN). The three countries bordering Lake Victoria's shores held divergent views on the nature of and solution to the spread of the weed. In addition, the problem facing Kenya, Tanzania and Uganda on Lake Victoria was traced back to the Kagera River in Rwanda. How could a problem for these countries be addressed by action taken in a fourth? Even within one country, water hyacinth management fell uneasily between camps: weed control is normally the remit of agriculture ministries, and impedance of shipping and ports falls outside their area of influence. New concerns about its effects on biodiversity brought in environment ministries. This meant that no single ministry was completely informed about the problem, or had responsibility for dealing with it. The result was a delay in recognizing the problem, and then in dealing with it while inter-ministry committees were set up to organize a national programme. Jeff Waage points out that some of the greatest tragedies associated in recent years with alien species have occurred when the problem fell snugly between the remits of different agencies. He says that with water hyacinth in Lake Victoria, only when this problem was resolved could real progress be made with limited national resources. Even then, the need for better inter-ministerial and inter-governmental coordination still constrained implementation.
Once the need for some form of control on Lake Victoria was agreed, there was still conflict over the best approach. Some recognized biocontrol as desirable, but others were dubious about importing an alien species to control an alien species - especially to Lake Victoria, where the legacy of Nile perch introduction looms large. And this, says Julien, was despite evidence that biological control had controlled significant water hyacinth problems in other countries without unwanted effects. Mechanical control meant the use of large pieces of equipment of questionable appropriateness and efficacy - the first delivered to Uganda sank on the day it was launched. However, harvesters were popular with aid programmes, especially if they could be sourced from the donor country. Agrochemical companies that had effective marketing and distribution networks in the countries concerned advocated chemical control.
Biological control of water hyacinth in Uganda is spearheaded by the Biological Control Unit of NARO - a semi-autonomous research organization under the Ministry of Agriculture, Animal Industry and Fisheries. James Ogwang says that it is not an overstatement to say that one of the greatest scientific achievements in Uganda in recent times, which seems to have gone largely unnoticed, is the seemingly `sudden disappearance' of water hyacinth from Ugandan waters, especially Lakes Victoria and Kyoga.
Ogwang explains that this started in 1992, when the Ugandan government, through NARO, formulated the National Task Force on the Control of Water Hyacinth. This body of Ugandan experts in various scientific fields was charged with advising the government on the best way to combat the weed. The task force recommended an integrated approach using mechanical measures, natural enemies (biological control) and, pending conclusive tests, possibly herbicides. However, says Ogwang, following a public outcry and debate about biosafety, the National Environment Management Authority of Uganda (NEMA ) decided that herbicide use should be deferred. To date, Ogwang emphasizes (and a point we return to later), no herbicides have been used in Ugandan water bodies against water hyacinth other than in small area controlled trials. The control options that have been used are mechanical harvesters (mainly at Port Bell and the Owen Falls Dam, where they kept the turbines free of weed and reduced blackout time and maintenance) and biological control agents. In 1992, the National Task Force recommended the importation of the weevils N. bruchi and N. eichhorniae from IITA in Benin.
However, it took some while longer before the weevils could be released in Lake Victoria. There was initial opposition from Kenyan quarantine authorities to releasing biocontrol agents into Lake Victoria until further safety tests had been conducted. Ogwang says that despite their safety record elsewhere, NARO scientists were required to carry out safety tests on bananas (a staple food crop in Uganda), rice, onions and egg plants/aubergines to reassure doubters in both Uganda and neighbouring countries before releases were allowed in Lake Victoria.
In the meantime, a successful `pilot' programme on Lake Kyoga was funded by the CSC and implemented by Ogwang and Ken Harley (CSIRO). Ogwang received training from IITA in Benin and brought back Neochetina weevils, which were reared at Namulonge Research Institute, where they subsequently adopted rearing techniques developed by CSIRO. Weevils were harvested monthly and released onto dense weed mats in Lake Kyoga, beginning in 1992. Additional rearing stations were constructed at three other sites around the lake, and fisheries staff and local fisherman were trained to manage these stations. Together, these increased the rearing capacity and numbers of weevils released, and the distribution and spread of the weevils.
Over the first 5 years of the project, the weevil populations built up progressively, reaching up to 16 weevils per plant by the end of 1997. Plant weight and other parameters of plant growth were also found to have fallen dramatically, and although no formal measurements of infestation surface area was made, it was evident that a significant reduction occurred over the 5 years since the weevils had been introduced. Regrowth and seedling growth was heavily attacked, as the weevils prefer young plants. However, the water gradient of the White Nile was also crucial. Reduced growth and weevil-damage meant that the weed was unable to form dense mats, and those that formed broke up more easily and were swept downstream. Weevil-damaged plants also appeared unable to attach to the bank and were swept away. Problems experienced by fishermen at landing sites were considerably reduced, and severe weed blockages were rare.
Following Kenya's agreement, weevils were released into Lake Victoria in December 1995. Ogwang says that since 1995, weevil rearing has continued at NARO's Biological Control Unit based at Namulonge Research Institute and at on-shore rearing facilities constructed at seven sites along the shores of Lakes Victoria and the mouth of the Kagera River. He describes how the biocontrol programme again recruited and trained fisheries extension staff and local fishermen in lakeside communities for the mass rearing programme. "They were", he says, "instrumental in the fast distribution of the initial weevil population in our lakes."
Ogwang also pays tribute to the weevils: "These silent warriors have been slowly but surely causing irreversible damage by reducing leaf area, root length, petiole length and reducing plant vigour and weight". He says that the weed has been dying back at a stupendous rate in remote parts of both lakes, since 1992 in Lake Kyoga and 1995 in Lake Victoria. But, he intimates, most of the world remained in ignorance: "The only witnesses to this noble process have been fishermen themselves," and he observes that the fishermen were convinced and appreciative long before alleged `experts'.
Water hyacinth has been present on the Sigi and Pangani Rivers in northeastern Tanzania since the 1950s. Environmentalists did not recognize it as a threat although it had been interfering with hydroelectric power generation at Hale on the Pangani River since 1964. The Tanzanians employed control measures of manual removal and prevention of weed movement that complemented each other; 200 litre drums were used as booms to protect the water intakes from the floating weed, and these were found to be more effective when the accumulated material was manually removed.
The two Neochetina weevil species were imported into Tanzania in April 1995 from IITA. The national programme team, led by Gaspar Mallya, was given training in rearing and monitoring techniques by CSIRO. Rearing units were established at the National Biological Control Centre at Kibaha, near Dar es Salaam. Harvesting of the biological control agents started in August 1995 and 2000 weevils were released on the Sigi and Pangani Rivers. By June 1996, 9460 weevils had been released and their establishment and impact on the weed has since been confirmed. Manual removal of the weed at Pangani Falls water intake has not been required since late 1997. (However, as a consequence, the local community has been deprived of their previous employment of manually removing the weed.)
Water hyacinth was first reported from Tanzanian Lake Victoria in 1991. Mobile weed mats were less of a problem here than in the northern part of the lake. The main infestation was a ribbon of weed along the shore. By 1998 an estimated 2000 ha of lakeshore was covered with the weed, which impeded landing sites and access to the lake. Manual removal was extensively employed. In the Mara and Mwanza regions, regional and district authorities organized village communities into self-help groups, and over 200 tonnes of the weed were removed from the landing beaches in one year in the Mara region. Where manual removal was not organized, fishermen occasionally employed casual labour to clear the landing sites and fish was often used as payment. Equipment used in manual removal included rakes, hoes, wheelbarrows, forks, etc. These tools, and protective gear, were made available by LVEMP to the communities involved. Prior to this, manual removal was carried out barehanded.
The most likely source has been confirmed as the Kagera River. Large clumps of the weed are today seen drifting downstream some 50 km above the mouth of the Kagera, but these are largely dismembered as the river goes over two waterfalls en route to the lake. Surviving stem fragments anchor to the banks and grow into plants, and then mats, which break away to drift into Lake Victoria.
Since August 1997 Neochetina weevils have been released in Tanzania at 20 points around Lake Victoria and along the Kagera River. More than half of the weevils released was harvested from the Sigi and Pangani Rivers where many weevils were present on the severely damaged water hyacinth. The number of weevils so far released in Lake Victoria has been small in comparison with the magnitude of the problem, so the Water Hyacinth Control Component of the LVEMP has developed simple techniques for increasing redistribution. The Mobile Rearing Unit uses plastic buckets and basins that are moved from one site to another along the shoreline. Mated females are allowed to lay eggs in water hyacinth plants in buckets, after which adult weevils are harvested from the plants. The plants with eggs and larvae are placed in the field amongst water hyacinth and adults are placed on new plants in buckets to lay more eggs. In addition, Adult Weevil Multiplication Centres have been constructed around the lake for rearing and harvesting of weevils. Using these methods many thousands of adult weevils and plants containing immature stages have been released in the lake.
The Kenyan biocontrol programme is being conducted by KARI (the Kenya Agricultural Research Institute) and training has been provided by CSIRO Entomology. Neochetina weevils imported from Benin (IITA), Uganda, South Africa (PPRI) and Australia (CSIRO) were released directly into the field or reared in quarantine facilities at the National Agricultural Research Centre at Muguga, where host specificity testing was completed. Subsequently, they were reared for mass release at a newly constructed facility at the National Fibre Research Centre at Kibos, close to Lake Victoria, by a team led by Gerald Ochiel (KARI). A combination of container and pool rearing methods allowed a monthly production of some 2000 weevils. Infested plants and a total of some 36,000 adult weevils were released at 27 sites around the Kenyan shoreline of the lake between January 1997 and August 1998. Releases were made more than 50 m from the shore on stationary and floating mats; the latter helped disperse the weevils to non-release sites. Releases by canoe reached some otherwise-inaccessible sites.
Preliminary monitoring surveys conducted in 1998 indicated that the weevils had established at release sites, and they were also recovered at non-release sites up to 50 km from the nearest release site. Sampling indicated that they were reducing leaf size and weight, and that larval mining and feeding and adult weevil densities were generally increasing.
To biological control experts such as Mic Julien, the initial results of all three country programmes were not surprising and were immensely encouraging, with the weevils established in all the countries. However, and in line with predictions, there was no obvious early evidence of the weed mats retreating on Lake Victoria. Julien says that although he had deliberately given a conservative time frame of 5+ years for achieving control, the expectations of people unfamiliar with biocontrol and those agitating for other methods to be used were still that instant changes should be seen. Biological control was on track, he says, but people ignorant of plant-insect interactions or with other agendas could not understand, or did not want to know. Other methods were argued by them to be the only option for at least short-term alleviation. Under the World Bank project, $1.3 million was awarded to an American firm to supply machines to Kenya that would chop the hyacinth into small pieces, a move opposed by environmentalists and lakeside residents concerned about leaving the chopped weed in the lake as another source of pollution.
The first `Swamp Devil' harvester, costing $300,000, arrived in September 1999. But as it was being unloaded on the shore at Kisumu in Kenya, something had happened out on the lake. While the machines were in cumbersome transit, the water hyacinth began to turn brown and shrivel - not in fear of the harvesters, but because the accumulated effects of the introduced weevils had taken their toll over the previous few years. The changes becoming visible in the water hyacinth populations did not come as a surprise to Julien, who had predicted such an outcome a year before.
During a mid term assessment of the biological control component of LVEMP in November 1998 in Uganda, Julien and James Ogwang had found very little water hyacinth remaining in bays that previously had been clogged. The weed, says Julien, had sunk during the preceeding few months. An assessment was made of data collected by Ogwang and Richard Molo (NARO) and chronologies of activities and changes provided by a wide range of people (fishermen to recreational boaters to scientists). The results indicated that the activities of the weevils had contributed to the demise of the weed, and this, Julien points out, within 2.5 years of the release of the first weevils on the lake by Ogwang. Julien had also visited Kenya as part of the same assessment. He and Gerald Ochiel had found that although the surface area covered in that part of the lake remained unchanged, the entire floating mats were severely damaged and were showing signs of water logging (floating lower in the water as a result of larval tunnelling in the crown and in the lower portions of the petioles which reduces flotation through water logging). Every plant they looked at had a number of adult weevils, larvae and eggs. Julien predicted then that the mats would sink within the following 12 months.
Julien says that the results in Uganda and the impending control in Kenya were relayed to the LVEMP secretariat in Kenya and the World Bank, but the decision to import mechanical harvesters was upheld. Yet while the harvesters were on the high seas to Mombasa, being transported to Kisumu and made ready for their task, the mats of water hyacinth were sinking.
A survey of Uganda by Timothy Twongo in November 1999 found very little water hyacinth along the Ugandan mainland coast and around the islands. Peter Neuenschwander, who visited the area with Ogwang in late 1999, said that the impact of the released weevils was much faster than they had experienced in West Africa. He says that the results from Uganda now are absolutely spectacular. Around Jinja, where until recently bays were clogged with the weed, the waterways are clear. He and Ogwang found only a few clumps of water hyacinth and these were heavily infested with the weevils. Even the Owen Falls dam was clear, and Neuenschwander says that this was an unexpected bonus, as remaining weed could have been expected to accumulate here. Ogwang says that at one time there were massive dry mats of dead weed with millions of stranded weevils at the Owen Falls Dam, and that it was common to see many birds, especially the cattle egret, on the dry dead mats, feeding on the stranded weevils. He likens the weevils to gallant soldiers who have successfully accomplished their assignment, and says that many have since perished because there is simply no water hyacinth left to feed on.
In Kenya, Julien and Ochiel inspected the water hyacinth in December 1999, and confirmed that at least 75% of the weed had sunk. Bays that had been locked in with water hyacinth for years (and in some cases abandoned by the fishing communities) were now essentially clear of the weed, and fishing activities had recommenced. The remaining fringing mats were heavily damaged and harboured many weevils. The remaining floating material consisted of small (up to 3 m " 3 m but mostly much smaller) mats of water hyacinth held together and prevented from sinking by hippo grass (Vossia species) that had broken away from fringing mixed species mats. Julien anticipates that as the water hyacinth dies, from insect damage and/or competition from the hippo grass riding on its back, the hippo grass will be deposited in the lake to sink to the bottom.
Neuenschwander reports a similar story from the southern side of the lake at Mwanza, which he visited in late 1999. Steamers are no longer port-bound by the weed in Mwanza Bay and the remaining fringe of water hyacinth around the shore is heavily infested by both species of Neochetina weevil.
It still remains to convince some, both that water hyacinth has been brought under biological control, and that the weevils are responsible. At Kisumu, for example, where the mats have floated out into the lake, the reduction in weed cover is widely attributed to the wind. At Mwanza, fishery authorities were reportedly mystified as to the cause of the sudden clear waterways. These views are perhaps understandable, given the seasonal changes and movements of water hyacinth mats on the lake in previous years.
However, James Ogwang is irritated by a sudden plethora of alternative explanations for the disappearance of the weed, which he said are being proffered by alleged `experts' on biocontrol. He dismisses these as ill-informed, describing their purveyors as "prophets of doom" and he is keen to see any apparent mystery cleared up. The most scurrilous explanation currently propounded is that herbicides were used to control water hyacinth in Uganda. Ogwang explains that this belief originated because of the practice of unscrupulous fishermen who used poison to increase the harvest of fish from the lake, for what he describes as their selfish economic gains. This led to the European Union (EU) slamming a ban on fish imports from the East African region, but he points out that herbicide residues were never detected in EU fish samples, and two different stories became confused. In truth, he says, no herbicides were used to reduce the water hyacinth biomass in Uganda. Ogwang points out that: (1) Herbicides such as 2,- 4-D and glyphosate do not kill fish instantly and were not used by fishermen. (At most, given the low dosage commonly used in aquatic systems, there might be some long-term cumulative effect.) (2) In Uganda there are many other aquatic plants such as papyrus, hippo grass, Nile cabbage, etc., in the water bodies. If herbicides had been used at all against water hyacinth, all these other aquatic plants would have perished as well - but they are still healthy and flourishing. (3) It would require a massive operation, possibly involving aerial sprays, to rapidly kill the weed and therefore it would virtually be impossible to spray the weeds secretly, even at night, without being detected. Mic Julien also points out that experience elsewhere has shown that for large infestations repeat spraying is required for the initial kill and rapid regrowth requires repeated applications to maintain control. Control on the lake is being maintained and there is no evidence that herbicides are being applied on any scale let alone on the scale that would be needed.
Ogwang is equally dismissive of those who have ascribed the disappearance of weed to El Nino. Perpetrators of this explanation argued that the abnormally high amounts of rain during the El Nino added so much water to the lake that it became diluted to the extent that the water hyacinth could not get enough nutrients, and hence it disappeared. But, as Ogwang puts it, this explanation holds little or no water. He points out that high rainfall would lead to more rather than less nutrient being washed into the lakes, owing to increased erosion, in which case the weed growth would have been, if anything, enhanced rather than decreased. Elsewhere in Africa and the world water hyacinth grows prolifically in waterways less polluted than Lake Victoria.
Harvesters have been in use in Uganda for a number of years, where they did sterling work in keeping the Owen Falls Dam turbines largely free of the water hyacinth. They were also used to remove weed from Port Bell, although with probably less impact. They have been idle since late 1999.
In Kenya, the `Swamp Devils' were launched in late 1999, and did some clearing of weed at Homa Bay. Then the weed moved, but it did not settle in Kisumu Bay, as expected, or indeed anywhere else. This, says Mic Julien, was because a lot of it sank due to biological control, leaving just small mats of water hyacinth supporting other species. Julien suggests that although these will sink eventually (when the water hyacinth becomes sufficiently damaged), the harvesters are meanwhile doing a useful job around Kisumu, chopping these mixed mats and also papyrus mats. However, the future employment prospects of the Swamp Devils are uncertain: they are not capable of chasing around Winam Gulf looking for remaining water hyacinth.
The harvester contract, to chop 1500 ha of water hyacinth and leave it in the lake, cannot be fulfilled, says Julien, and this is also good news for those concerned about the environmental impact of such an activity. He estimates that there is not one third of that amount of water hyacinth left and the amount is decreasing daily as the weed succumbs to the weevils. Julien says that on his recent flights to and from Kampala's Entebbe Airport and Kisumu he saw no mats in Ugandan waters and a few limited mats off the Kenya coast, and they are in the process of disintegration.
Julien is surprised only by the speed at which the control of water hyacinth has been achieved. Successful control using solely these weevils has been achieved in a number of other countries, and based on that knowledge and experience, he fully expected biocontrol to be useful in Lake Victoria. He indicated a time span of at least 5 years so as to be realistic and not oversell biocontrol. Julien says that such good results after 2.5-3 years suggest that temperature and nutrition have not been limiting, and the results are similar to those achieved in nutrient-rich waters in PNG.
Julien is in the process of determining the feasibility of using remote sensing to verify the change in surface area of the weed mats. However, ground-level assessments to check on the weevil presence and plant density have already been conducted by Gerald Ochiel and Julien in Kenya in November 1998 and December 1999 and by James Ogwang in Uganda in December 1999. Their results provide solid evidence for the key role of the weevils in the disappearance of the weed.
Mic Julien and Hans Herren are agreed that there is a good chance that the harvesting could be abandoned and the weevils left to reduce the weed to mere decorative matter. Herren adds that it is time urgently to review the project and decide on the best follow-up, both for what is now being confirmed as the successful biological control of water hyacinth on Lake Victoria and for the associated problem of eutrophication of the lake waters. However, Herren is anxious that people should not see biological control as a cheap last resort, to be turned to after large sums of money have been spent ineffectively on other methods. He points out that biological control, if done well, costs money, and not least for monitoring activities, for which funding is all-too-often cancelled once a problem seems to have gone away. On the other hand, because biological control, in the classical sense of the term, provides a permanent solution, it offers huge savings over recurrently needed methods, such as pesticides or mechanical removal.
Ogwang is quick to point out that although most of the mats of water hyacinth and their weevil populations have all but disintegrated, there is a real possibility that remnants of the weed and young plants that sprout will re-grow and pose some threat. The plants would be growing in an environment with few or no weevils. Additionally, fresh massive amounts of the weed are daily being fed into Lake Victoria from the Rwanda highlands via the River Kagera. "In effect," he says, "we have won the battle but the war is still much around." Future activities planned by the Uganda programme include: the construction of more weevil rearing facilities for maintenance of weevil releases; Rwandese and Burundian nationals trained by Uganda to initiate weevil release so that weeds that enter Lake Victoria would carry weevil infested water hyacinth; research on other potential natural enemies including pathogens; and monitoring any changes in weed biomass to determine appropriate action.
The water hyacinth biocontrol experts who argued back in 1997 that giant harvesters were not the answer were not Luddites lying down in the path of progress. They really did know better, and there was already plenty of evidence of their better judgement in the clear waterways around the world where natural enemies had been allowed to feast in peace. As the British Guardian newspaper put it, "Forces of Weevil Trump World Bank" [16 January 2000]. Will rusting harvesters be left to line the shores of Lake Victoria as manifest evidence of international development's continuing folly? Or will someone somewhere see sense, send the monsters home, and put the money to better use? Herren suggests that the enormous and probably unneeded funding for mechanical harvesting would be better directed towards more biological control training. Julien says it should be used to develop better integrated and sustainable strategies for other weeds and pests, including training and infrastructure based on catchment and regional approaches. This idea is supported by Herren who advocates the establishment of a permanent regional facility, and also suggests that publicizing the success of water hyacinth biocontrol can assist with prevention activities in the lake's watershed.
Organizations involved in international programmes often disagree over the finer points of who did what, and history tends to be treated as a creative subject when it comes to apportioning credit for success. However, it is clear in this instance that most credit for the success of biocontrol of water hyacinth on Lake Victoria is due to the national programme staff who persevered, on occasion in the face of indifference and even outright opposition, against what they were told by so many alleged experts were hopelessly overwhelming odds. With the vital support of experts in water hyacinth biocontrol from around the world (CABI, CSIRO, ICIPE, IITA and PPRI), yet with often little money and inadequate resources, they succeeded in rearing and releasing the weevils, albeit not always in enormous numbers, and in training and mobilizing local communities. The exciting results now being reported are their vindication.
Contact: Matthew Cock,
Roger Day, CABI Bioscience
Centre, CABI Africa Regional Centre,
Martin Hill, PPRI, Private
Bag X 134,
Mic Julien, CSIRO
Peter Neuenschwander, IITA,
James A. Ogwang,
Japanese knotweed (Fallopia japonica syn. Reynoutria japonica) is an unwelcome intruder in the UK, not least because it is capable of bringing down walls and gravestones, and pushing through concrete paving. It has even been found growing through the foundations of a house in Wales and sprouting up from behind a sofa. Its main threat, however, is to riparian habitats. Its formidable strength is attributable to a mass of rhizome that can extend to a depth of over 15 m from which the plant can grow at a peak rate of more than 40 mm per day to reach well over 3 m in height. This otherwise handsome perennial is a classic example of the escape of a `desirable' garden plant into a new environment in which it has spread exponentially. Remarkably, research has shown that the whole UK infestation originates from one `mother plant' and, since knotweed is currently restricted to asexual reproduction, the whole population could be considered one organism - a fact that has led to one national newspaper headline, "Largest female on earth could strangle Britain".
Knotweed is extremely difficult to control since herbicide use is restricted on its favoured riparian habitats and a fragment of rhizome weighing just 0.7 g left in cleared land can regenerate a new plant. Its ability to push through tarmac and concrete means that infested land earmarked for development must be cleared of all knotweed remnants and the soil disposed of at specialist landfill sites. The cost of this process is considerable and can add 10% to a developer's overall budget. Knotweed also disrupts river protection works, increases the risk of flooding, and restricts access to waterways whilst excluding native vegetation under its dense canopy. Although estimates are hard to come by, Japanese knotweed control costs are likely to measure in the millions of pounds per year in the UK alone.
In its native range of Japan, south China and Korea, Japanese knotweed exists as a component species in the succession process. For example, it acts as a colonizer on the inhospitable volcanic fumaroles of Japan where it is soon outcompeted by other vegetation. The scientific literature shows a considerable range of natural enemies inflicting damage, in some cases severe, to Asian populations including a promising chrysomelid beetle and two primary pathogens whose specificity is likely to be high.
Steps are underway to implement classical biological control programmes against Japanese knotweed in the UK and the United States. CABI Bioscience has been developing biological control programmes against this, and other introduced riparian weeds, for over a decade. This groundwork, along with a scoping study funded by the UK Environment Agency, has resulted in biological control being identified as the only sustainable solution to this `knotty problem'. A meeting was recently held at CABI Bioscience's Ascot site in order to present the proposal, and to clarify the principles and safety of weed biocontrol to potential funding organizations in the UK including the Welsh Development Agency, the South West Development Agency and the Environment Agency. The proposal is also under consideration by concerned parties in the eastern United States where biological control is increasingly the first line of defence. This situation contrasts sharply with the UK where classical biocontrol is considered `novel and controversial' since there has never been a complete programme against a weed target in the UK.
A collection of over 50 newspaper articles on the subject of Japanese knotweed in Britain stands testament to public opinion on this unwelcome invader, and only funding stands in the way of this plant being faced with the old enemies it has done so well without.
By: Dick Shaw, CABI
American scientists have been looking into the possibility of harnessing the fungi Beauveria bassiana and Paecilomyces fumosoroseus for control of two major crop pests: diamondback moth (DBM) (Plutella xylostella) and Russian wheat aphid (Diuraphis noxia). USDA Agricultural Research Service (ARS) scientist John Vandenberg says that fungi are a promising alternative for pests developing resistance to Bacillus thuringiensis (Bt). A significant feature of the research has been extending laboratory work into the field, and assessing the commercial and large-scale efficacy of fungal preparations.
This lepidopteran pest of brassica crops has become problematic in recent years because of the development of resistance to insecticides, including the microbial pesticide Bt. Because of this, a significant proportion of the control treatments applied worldwide, which in total cost some US$1 billion annually, is of limited efficacy.
In laboratory studies, Vandenberg's team working with Cornell University scientists found dramatic differences between strains of P. fumosoroseus in their ability to infect and kill DBM. There were some highly virulent strains, and these were characterized by having large spores that attached easily to the insect cuticle and germinated quickly. They were thus able to penetrate the cuticle and initiate infection more quickly. All larval stages were susceptible to infection by B. bassiana in laboratory tests, but those that were exposed shortly before moulting avoided infection because the contaminated cuticle was shed before the fungus could penetrate it.
Moving into a farm-based setting, Vandenberg and his collaborators looked at the efficacy of B. bassiana for DBM control on cabbage seedlings in screenhouses (the conventional way of raising plants for the field in the USA), using a commercial formulation of B. bassiana, Mycotrol (Mycotech; Butte, MT, USA). Weekly or twice-weekly applications significantly reduced insect populations and seedling damage compared with water-sprayed controls. The control achieved was as effective as conventional insecticide treatments, and persisted for more than two weeks. Mycotrol as a wettable powder and an emulsifiable concentrate also reduced DBM populations when applied to larger plants in the field, and multiple applications improved performance.
Further field trials investigated the potential for B. bassiana in the season-long control of DBM and other lepidopteran pests (Pieris rapae, the imported cabbage worm and Trichoplusia ni, cabbage looper) in fresh market cabbage. Treatments assessed included B. bassiana strain GHA (Mycotrol ES) at 1.25-5 × 1013 spores/ha, Bt var. kurstaki (Javelin WG) at 0.3 and 1.2 kg/ha, and simultaneous or sequential combinations of the B. bassiana and Bt treatments. Six applications were made between 4 August and 15 September 1998, insects were sampled five times during the season, and damage and marketability were assessed at harvest. The use of B. bassiana alone reduced DBM numbers at all rates used, and P. rapae and T. ni numbers were reduced at higher rates. Both rates of Bt reduced numbers of all three pests. Four applications of B. bassiana followed by two late-season Bt treatments reduced larval counts of all three pests also. However, the results of these sequential treatments, and simultaneous treatments in which B. bassiana and Bt were applied simultaneously as tank-mixes, indicated that combined treatments were no better at reducing larval counts than Bt alone at either rate. Laboratory tests of the efficacy of B. bassiana-treated leaved for DBM indicated about 80% efficacy for freshly treated leaves, but this dropped to about 40% for samples taken three days after treatment. There were no differences in cabbage head weights between treated and untreated plots, but cabbage head marketability was best among Bt-treated plots and reduced in plots treated with B. bassiana.
In conclusion, although B. bassiana worked well alone against DBM and provided some control of other pests, reliable control of this complex of cabbage pests may require the development of other B. bassiana strains with better activity against Lepidoptera. However, just one or a few large larvae can inflict late-season damage to cabbage heads, and therefore late-season application of a highly toxic faster-acting insecticide may remain necessary to preserve marketability. However, combining early season B. bassiana with late applications of Bt or chemical insecticides may help to slow or limit the development of insecticide resistance.
Russian wheat aphid invaded the USA in about 1986 and since then has cost some US$850 billion in lost yields and control and other costs. It spends the winter primarily on wheat and barley throughout its North America range, which now extends through 16 American states and two Canadian Provinces.
In field studies run over three seasons, Vandenberg together with University of Idaho scientists tested the ability of spores of B. bassiana and P. fumosoroseus to control the pest on artificially infested spring-planted wheat. They found significant reductions in aphid populations in all three years in plots sprayed with B. bassiana as Mycotrol. Sampling four days after spraying revealed that 52% of aphids on wheat tillers were infected. In 1997-98, Mycotrol was applied to larger (0.4 ha) plots using a moveable pipe irrigation system. Aphid populations dropped significantly within 2-3 weeks of spraying. It has thus been proved that the treatment works on a large scale and has commercial potential, and this is the first time that the efficacy of Mycotrol against Russian wheat aphid has been demonstrated.
Tests with P. fumosoroseus proved inconclusive, however. The fungus reduced aphid populations in 1995 but not 1996 after one or two applications. The team is now looking at ways of improving the field stability of this fungus in collaboration with an ARS microbiologist.
Source: Agricultural News,
Before P. fumosoroseus can be deployed for Russian wheat aphid control, its effects on nontarget organisms need to be evaluated. Preliminary studies have been carried out by Vandenberg in collaboration with Judith Pell from IACR-Rothamsted, UK. They have examined the impact of P. fumosoroseus on convergent ladybirds, Hippodamia convergens.
Laboratory studies have shown that under certain conditions ladybirds may become infected with the fungus, but how far this physiological susceptibility in the laboratory might translate into ecological susceptibility in the field remains to be determined. Laboratory bioassays found some individuals succumbed following P. fumosoroseus sprays, and this was thought to be related to the degree of pre-trial stress to which they had been exposed. In a relatively unstressed batch of ladybirds, no mortality was observed following exposure to P. fumosoroseus, apart from a low (6.7%) level where ladybirds were exposed to 10× the field concentration and incubated for 72 h at 100% RH.
In general, the ladybirds avoided feeding on aphids showing signs of fungal infection. Starved ladybirds ate recently killed cadavers, but only when these were still green and showed no visual sign of infection. However, ladybirds became contaminated with conidia while foraging in the proximity of sporulating aphid cadavers, and were found to be able to vector the conidia to healthy aphids. Infection was initiated in an average of 53.1% of aphids exposed to contaminated ladybirds, whilst four out of seven contaminated ladybirds also became infected. Ladybirds foraging in the environment of sporulating cadavers are therefore most at risk from infection.
Contact: John Vandenberg,
USDA-ARS Plant Protection Research Unit,
Judith Pell, Entomology and
Nematology Department, IACR-Rothamsted, Harpenden, Hertfordshire, AL5
Over the last nine years, the biological control of locusts in Africa has become a reality with the development of Green Muscle® by the LUBILOSA (LUtte BIologique contre les LOcustes et les SAuteriaux) programme. Green Muscle® consists of a virulent strain of Metarhizium anisopliae that specifically infects and kills locusts and grasshoppers, and it can do this within 6-12 days of application. Locust feeding is markedly reduced within 3 days of infection, and over a 3-week period, the biopesticide's killing power is greater than that of conventional knock-down insecticides. It can be applied using conventional ULV spray equipment and has exceptionally good storage characteristics for a biological agent. Infected locusts can also provide a source of inocula for re-infection. It has been extensively tested and shown to be effective against all the major locust and grasshopper pests of Africa, but has no adverse effects on non-target organisms. Green Muscle® is recommended by an FAO panel of experts for operational use to control locusts in conservation and environmentally sensitive areas, and the product was registered in South Africa in December 1998 [See BNI 20(1), 5N (March 1999)] where it is manufactured by Biological Control Products SA Pty in Durban. The product is also expected to be registered in West Africa in early 2000.
Phase 3 of LUBILOSA was concerned with completing the process of taking research findings and developing from them a viable product, and this was achieved by the end of 1998. Phase 4 of the programme, which began in January 1999, has a different brief. It is concerned with Product Stewardship - the process by which sufficient support is provided to complete the commercialization of Green Muscle® and ensure its sustained purchase, uptake and use for locust and grasshopper control in Africa. LUBILOSA is committed to ensuring that Green Muscle® eventually becomes the first choice for the strategic control of locusts and grasshoppers in Africa.
The outlook is promising, with NGOs in West Africa increasingly turning to the product. SECAMA (Société Catholique du Mali) in Mali have purchased Green Muscle® and sprayed it on 200 ha - the first time that Green Muscle® has been purchased for operational use without technical or financial support of LUBILOSA. In Maradi, Niger, CARE International has trained their extension personnel in the use of Green Muscle®, and PROSOPAS (Projet d'Aménagement de Zone Pastorale) have taken part in demonstration trials in Taouah. In Benin, an extension programme is being developed by CRDB (Centre de Recherche pour un Développement Intégré à la Base) in collaboration with IITA (the International Institute of Tropical Agriculture).
Contact: David Dent,