Rubber Vine in Terminal Decline

Successful control of rubber vine, Cryptostegia grandiflora, in Australia is being trumpeted as another triumph for classical biological control. The speed of the success probably owes as much to the development of mass production and field release techniques as it does to the efficacy of the rust pathogen. Following mass releases of the pathogen in 1995 the omens are now good for the long-term control of C. grandiflora, which has been described as the single biggest threat to natural ecosystems in tropical Australia and has cost Australian farmers millions of dollars.

Rubber vine is a Madagascan endemic genus comprising two species. Both are naturalized in northern Australia. They were introduced in the 19th century, particularly by mining communities in central Queensland, as ornamentals and also for the commercial potential of the latex they produce. Only C. grandiflora has become weedy, but it now has an estimated range of 40,000 km2 in Queensland.

Although a number of herbicides are effective, infestations are so vast and remote that wide-scale spraying is neither economically nor environmentally acceptable. A biological control programme funded by the Queensland Meat Growers and Packers and implemented by Queensland Department of Natural Resources (DNR) and CABI Bioscience began in 1985. Since then, two agents have been introduced: the pyralid caterpillar Euclasta whalleyi and the rust fungus, Maravalia cryptostegiae. It is the rust fungus that is contributing most spectacularly to the weed's control.

This success of this project has been hard-won, for the optimum conditions for rust infection require wet season conditions, when releases are most difficult on the practical level. Classical weed biocontrol generally uses an inoculative approach, but the huge area infested together with the weed's rapid advance towards prestigious national parks in the Northern Territory meant an inundative approach was adopted. For this to be successful, three practical hurdles had to be cleared:

• Mass production of an adequate supply of viable inoculum
• Access to remote sites when roads were impassable
• Effective and portable means of application

Bulking up of quality-controlled spores for aerial releases was conducted in pot-grown plants kept in simulated-dew conditions in a growth chamber to promote inoculation, and subsequently transferred to a green-house. Spores were harvested using a cyclone spore collector [see BNI 20(2), 55N (June 1999) 'Cleaning up'], and air-dried for temporary storage. Releases were made during the wet season within 3-4 days of spore harvesting, during or just before rain to maximize the chances of infection.

Releases at 29 sites in January-March 1995 were made by light aircraft, as roads were generally flooded and/or impassable. This still presented hazards and access to some sites required exceptional flying skills. At some sites it was not possible to land at all and applications were made from the air.

Three application techniques were developed:

• A motorized knapsack mistblower, modified so it could be dismantled for carriage in a light aircraft. If applied before rain, spores were sprayed in aqueous suspension; if during or after rain, dry spores were introduced into the airstream at the end of the nozzle. This method was fast and effective and reached foliage up to 20 m above the ground (and similar distance horizontally). Areas of ca. 0.25 ha could be sprayed in 10-15 minutes.
• Suspensions were also applied to the underside of foliage using two types of hand-held hydraulic sprayers adjusted to deliver a fine spray pattern. Infection rates were as good but application rates were slower than with the mistblower.
• Inaccessible sites were sprayed from a light aircraft flying at ca. 100 m at an air speed of ca. 130 km/h by adapting equipment designed for dingo baiting. Spore solution released from a 100-mm diameter tube passing through the rear port-side of the aircraft became atomized owing to air shear, leading to precipitation on the foliage.

More ad hoc methods were used to give sufficient inoculum for further inundative releases. Pot-grown plants, infected by placing them around an infected plant in the open air, were deployed in rubber vine infestations. This method was ideally suited to a farmer participatory release programme initiated by DNR, as timing was not critical and plants could be watered as necessary and continued to produce spores for 3-4 weeks.

Field collection and re-distribution of fresh rust-infected foliage (placed onto or suspended from the canopy) when dew formation or rainfall was forecast, was also adopted widely by farmers. Although best infection rates were achieved when material was redeployed within 24 h, cool storage extended its 'shelf-life'.

All inoculation methods proved reliable in establishing the rust in the field, and dispersal was rapid. Dispersal increased exponentially over time, reaching a plateau 10-15 months after the releases when individual areas of infection merged. Twenty months after the releases, most of the rubber vine in the Gulf and Peninsula areas were infected by the rust. The rate of spread over subsequent years became difficult to estimate, but just over 3 years after release it had dispersed at least 550 km.

Within a few months of the rust's introduction, heavy leaf infection was evident at most release sites (most leaves were completely covered with pustules) and there was severe premature leaf fall. The significant levels of defoliation at most release sites at the end of the first wet season were accompanied by growth of native grasses amongst the rubber vine plants (and this has subsequently allowed the integration of fire as a management tool for the weed). Within 12 months, secondary effects became apparent. There were fewer flowers, and dieback of shoots and seedlings was observed. Four years after rust introduction, only very few seedlings were found amongst heavily infested plants and the rubber vine population had decreased 25-65%.

The effects of the rust do vary with groundwater availability. On seasonally dry soils, rubber vine has a semi-deciduous to deciduous habit. In a riparian situation and where the water table is high, the weed continues to grow during the warm dry winter season. The rust is virtually dormant during the dry season, so plants growing year-round can compensate during the dry season for rust-induced premature defoliation and have more flowers and fruit and less dieback than those growing in seasonally dry soils.

With substantial to complete control over its entire range in Australia predicted over the next 4-6 years, what has made this programme such a resounding success? The strategy of inundative release adopted by the programme and the successful negotiation of implementation hurdles outlined above has been crucial, but so has the impact of the pathogen on its host. Experimental results suggested that the rust could reduce leaf cover by 73% over a 12-month period, and flowering by 48% and pod production by 85%, and these predictions have been borne out by field observations. The invasiveness of rubber vine in Australia is attributed in part to prolific seed production (a single large vine may produce 8000 seeds from a single flowering) but the seeds rarely survive in the soil for more than a year, so reduced seed production had an immediate impact. In addition, heavy infection in the field led to widespread seedling mortality and death of older plants, especially in seasonally dry sites. Thus the rust's multiple impacts have been able to bring about a rapid reduction in the invasive ability of rubber vine.

Contact: Allan Tomley, Alan Fletcher Research Station, Queensland Department of Natural Resources, PO Box 36, Queensland 4075, Australia
Email:
Fax: +61 7 3379 6815

Harry Evans, CABI Bioscience UK Centre (Ascot), Silwood Park, Buckhurst Road, Ascot, Berks. SL5 7TA, UK
Email:
Fax: +44 1491 829123

❑

Paterson's Curse, a A$100M-a-Year Weed, Faces Biocontrol in Australia

Biological control is frequently claimed to be both environmentally and economically more sustainable than alternative methods, and chemical control in particular. However, rigorous assessments of biological control programmes are rare. Here we report on a benefit/cost (B/C) analysis of the programme for biological control of Echium spp. in Australia, which was begun by CSIRO in 1972. The new analysis projects economic gains from biological control based on observed values for the rate of spread of an insect agent and its impact on the target weed in Australia.

Echium plantagineum (Paterson's curse) is an introduced winter annual pasture weed of Mediterranean origin. Free of native Mediterranean plant and insect communities, it has become one of the dominant pasture weeds of temperate Australia. Although three other introduced Echium species (E. vulgare, E. italicum and E. simplex) occur as weeds in Australia, E. plantagineum is the most important as a pasture weed there, but in this study the four species are referred to collectively as 'Echium'. Although relatively nutritious in terms of digestible nutrients, and valued as a pasture plant in some places, Echium contains pyrrolizidine alkaloids that are poisonous to livestock, reducing weight gain and wool clip and in severe cases leading to death. From its start as a garden flower in the 1840s, Echium is now estimated to occur as a weed on over 30 million hectares in Australia.

Echium was first suggested as a candidate for biological control at the Australian Weeds Council in 1971. From its base in Montpellier, France, CSIRO Entomology soon started surveys in the weed's native range. Of the hundred or more insect species recorded on Echium, eight were selected as possible biological control agents, with the first imported into quarantine in Canberra in 1979. In 1980, a small group of graziers and apiarists lodged an injunction in the Supreme Court of South Australia to stop the biological control programme as they considered the loss of Echium a threat to their livelihoods. The Biological Control Act 1984 established procedures for assessing and authorizing biological control programmes in Australia; a subsequent inquiry and B/C analysis was conducted by the Industries Assistance Commission (IAC), which concluded with the judgement that a biological control programme on Echium should go ahead¹.

The Supreme Court injunction was eventually lifted and the importation of insects into Australia resumed and rigorous specificity testing undertaken. Six insect species have been successfully released: a leaf mining moth, Dialectica scalariella, crown and root weevils, Mogulones larvatus and M. geographicus, a root beetle, Longitarsus echii, a stem boring beetle, Phytoecia coerulescens and a pollen beetle Meligethes planiusculus. Of these insects, D. scalariella and Mogulones larvatus were introduced first and have been released across the geographic range of the weed. Mogulones larvatus is known to be limiting the Echium population at two of the earliest release sites and approaching control at many of the younger release sites.

Based on the positive population trend of M. larvatus and its ability to limit the weed at an increasing number of sites, the economic analysis of the IAC report was revisited so projected economic gains from biological control could be quantified. Unlike previous B/C analysis of biological control, where an insect is given an arbitrary impact and rate of spread, the current analysis incorporates observed values based on the biology and ecology of M. larvatus and its weedy host, Echium, over the past 9 years.

Building the Model

Of some 1000 releases of M. larvatus, 400 have been confirmed successful in terms of insect survival to subsequent seasons. Of these successful releases, 189 were in the state of New South Wales (NSW), 143 were in Victoria, while South Australia (SA) and Western Australia (WA) had only 34 each. For this analytical model, the rates of spread of insects and development of attack rates on Echium, and the rates of expected progress of geographic coverage of maximum attack, based on field data and observations by scientists on the project, are described as functions of time. Function parameters differ according to climate zone in terms of the date of the autumn season break; both attack and spread rates are highest with an early autumn break (March) and lowest with a late break (May). This variation occurs because late breaks tend to decouple the occurrence synchrony of Echium and M. larvatus.

The study uses the district location, grazing area and stocking rate information supplied by the IAC, updating and correcting an earlier analysis², and overlaying the new insect release location and date data. Autumn break date classifications were assigned to districts according to the month in which greater than 25 mm median rainfall is received, based on long-term monthly median rainfall maps from the Bureau of Meteorology. This allowed projected extents of insect spread to be mapped.

We assumed for districts in which there had been more than one release, the maximum spread of insects from each release was to the area defined as the district total divided by the number of releases in the district. This is a conservative assumption given the fact that the earliest insect releases will have spread over greater surface areas and reached greater densities than later releases, and the fact that insects are not limited by administrative boundaries. These conservative assumptions were made to limit the computational burden posed by 400 insect releases distributed over an 8-year period across 44 districts of varying size and divided into 130 sub-districts, depending on year of release. For each sub-district, year-by-year sequences of areas with partial relief were simulated, then aggregated back to the 44 districts as area equivalents with full economic loss relief.

There are several other conservative assumptions in our analysis. One is that all long-term biological control of Echium results only from the activity of M. larvatus even though there is good reason to anticipate complementary successes of the other agents released against the weed. The model conservatively assumes no further releases beyond the 400; in reality, state departments of agriculture continue to respond to farmers' requests, and the Australian Wool Innovation and Meat and Livestock Australia continue important support for releases of biocontrol agents against Echium. The model focuses on the valuation of increased pasture productivity and ignores reductions in conventional spraying costs. While reductions in pasture spray costs may be anticipated, these are likely to be replaced with the costs of measures taken by farmers to facilitate the success of the biological control agents and to limit reinvasion by other pasture weeds. The model also ignores control costs and losses attributable to Echium as a weed in crops, which amount only to some A$1.2 million annually and may be assumed to continue indefinitely.

The economic damage caused by Echium in pastures is assumed to remain unaffected by M. larvatus at attack levels below 50%. Attack levels above this are assumed to result in increasing reductions in economic loss.

The attack and spread simulation model, set for the particular size, release dates and autumn break parameters of each sub-district, was used to generate a time series of areas with varying degrees of partial economic relief from Echium. The years required to reach these limits differed according to district size and number of releases. For each year in each district, a ratio was calculated of the (weighted) relieved area to the total area. These ratios were multiplied times the maximum proportions by which total stocking rates were assumed to be increased in the absence of Echium in the IAC report, district by district (these ranged from a maximum of 0.2 to a minimum of -0.1). Total stocking rates for each district were expressed as dry sheep equivalents (DSE) where 1 DSE relates to 1 wool sheep, 1.5 DSE for each meat sheep, 10 DSE for each beef animal and 15 DSE for each dairy cow.

In order to express the aggregate economic relief in dollar terms a conservative value per DSE recovered was wanted. The lowest gross margin per DSE in NSW is A$8.80 for wool-producing wethers. A value of A$8 per DSE was chosen as a conservative base for modelling, though values double this are recorded for sheep and cattle enterprises in NSW where the greatest infestations of Echium occur. The year-by-year estimates of dollar value loss relief were aggregated across districts by state.

Adding Up Results

The simulated time paths of the benefits for each state are projected in Figure 1. Illustrated are the projected four-state aggregate benefit streams for the case of A$8 per DSE (undiscounted in panel 'a' and discounted at 10% in panel 'b'). The greatest benefits from biocontrol of Echium are anticipated in NSW, followed by Victoria and Western Australia. Comparatively little benefit is expected for South Australia, where the late autumn breaks put M. larvatus at a disadvantage.

The biological control research and development programme was begun by CSIRO in 1972. Total R&D expenditures on Echium biocontrol by CSIRO and its partners from 1972 to 2001 have reached A$14 million. The sum of the undiscounted benefits (Figure 1) minus the cost stream, results in a time series of undiscounted net annual benefits. Several such series were created using a range of DSE values (A$4 to A$16 per DSE) and discounted at a range of rates (5% to 20%) to produce the results in Table 1.

Table 1 Net present (2002) value of the Echium biocontrol programme, with 400 successful insect releases in 1993-2000, viewed over the 1972-2050 period (A$ millions).

Discount rate	Value of recovered pasture productivity (A$ per DSE)
	A$4	A$8	A$12	A$16
5%	526	1074	1623	2171
6%	423	871	1319	1767
7%	343	714	1084	1454
8%	280	590	900	1210
9%	229	491	753	1015
10%	186	410	634	858
11%	151	343	536	729
12%	120	287	455	622
13%	92	239	385	532
14%	67	196	325	454
15%	43	158	272	386
16%	20	122	223	325
17%	-3	88	179	270
18%	-27	54	136	218
19%	-53	20	94	168
20%	-81	-14	52	118

With A$8 DSE's, annual benefits in terms of increased productivity of grazing lands are projected to increase from near-zero in 2000 to some A$75 million by 2015, and A$90 million by 2025. The discounted (5%) net present value (NPV) of the benefit-cost stream from 1972 to 2015 is projected at A$287 million, for a B/C ratio over 14:1. For the 1972-2050 period the NPV is A$1074 million for a B/C ratio of over 50:1. The internal rate of return (discount rate that drives the B/C ratio to zero) exceeds 19%.

Summing Up

The success story projected for biological control of Echium in Australia is likely to be at a slower pace than envisaged by the IAC (1985). Nevertheless, the return on investments is expected to be very respectable indeed. Keeping in mind that just over A$14 million has been spent on the biocontrol programme for Echium, the high net present values anticipated with all but the most extreme combinations of low DSE values and high discount rates (lower left corner of Table 1) give strong assurance of success.

Further analysis is needed to determine the value of targeting additional insect releases beyond the 400 of the 1993-2000 period where there may be gaps in populations that would otherwise take many years to fill.

Acknowledgements

This report represents a collaborative effort under the Cooperative Research Centre for Australian Weed Management (Weeds CRC). It corrects errors in an earlier version, which understated the contributions of Echium biocontrol in Western Australia². It also updates all NPV estimates to a 2002 basis. The authors wish to thank David Vincent and David Pearce, of the Centre for International Economics (CIE), for raising a number of important questions over the course of the analysis. We thank those collaborating at the state level with CSIRO Entomology in releasing and monitoring biological control agents on Echium, in particular, those who provided the geo-referenced release location and date information required for this analysis: Kerry Roberts, Agriculture Victoria, KTRI, Frankston; Ross Stanger, SARDI, Entomology Unit, Adelaide; Paul Sullivan, NSW Agriculture, Tamworth; and Paul Wilson, Agriculture WA, South Perth. Funding support for such work from Australian Wool Innovation and Meat and Livestock Australia is also gratefully acknowledged.

References

¹IAC (1985) Biological control of Echium species (including Paterson's curse / Salvation Jane). Canberra, ACT; Australian Government Publishing Service, IAC Report No. 371.

²Nordblom, T.; Smyth, M.; Swirepik, A.; Sheppard, A.; Briese, D. (2001) Benefit-cost analysis for biological control of Echium weed species. In: Centre for International Economics (eds) The CRC for Weed Management Systems: an impact assessment. University of Adelaide, South Australia; Weeds CRC, pp. 36-43. Available as Technical Series No. 6, under 'Resources' at:
www.waite.adelaide.edu.au/CRCWMS/

By: Thomas L. Nordblom, Farrer Centre, School of Agriculture, Charles Sturt University, Wagga Wagga, NSW 2678, Australia [ ]
and Matthew J. Smyth, Anthony Swirepik, Andrew W. Sheppard and David T. Briese, CSIRO-Entomology, GPO Box 1700, Canberra, ACT 2601, Australia
[]

❑

Update on the Status of Water Hyacinth on Lake Victoria

The water hyacinth (Eichhornia crassipes) invasion of Lake Victoria brought both invasive waterweeds and biological control to the world's attention in the late 1990s [see BNI 21(1) (March 2000) 1N-8N 'Harvesters Get That Sinking Feeling']. Since then, media interest has been intermittent with occasional stories concerning apparent resurgence of the weed. However, the work of the scientific community has been relentless, with efforts focused on measures to bring about sustainable long-term control of water hyacinth in the Lake Victoria basin. No one involved imagines that the end is in sight, but significant advances have been made in building regional cooperation. Here, we look first at reasons for the water hyacinth resurgence in the lake, and what can and is being done to mitigate its impact. Second, we describe the beginning of a new collaborative biological control initiative in Rwanda, which is releasing Neochetina weevils into the main source of Lake Victoria's water hyacinth and, through public awareness exercises, publicizing both the dangers water hyacinth poses and what biological control is about.

❑

Resurgence in Lake Victoria: a Case for Optimism

Ugandan scientists of NARO's Biological Control Unit based at NAARI (National Agriculture Research Organization - Namulonge Agriculture and Animal Production Research Institute) under the leadership of James Ogwang were instrumental in the initiation and implementation of the successful water hyacinth (Eichhornia crassipes) biological control in Uganda's water bodies including Lake Victoria. Since the introduction of Neochetina weevils in 1995, an 80% reduction in water hyacinth populations has been recorded in the Ugandan waters of Lake Victoria. During their studies in 1997 and 1998, Ogwang and his team found, on average, some 25 adult weevils per plant at various sampling sites around the lake, and these huge weevil populations brought about one of the most spectacular crashes in a weed biomass in the history of biological control. However, since the second half of 2000 there have been reports of a resurgence of water hyacinth. This was an obviously cause for concern both for the people living around the lake and politicians, who feared that the control initially effected by the weevils might be short-lived and that the water hyacinth might be returning to plague them again.

Further studies on the recently appeared water hyacinth fringe showed that the young, healthy and rapidly growing plants were the result of the germination of seeds which had been deposited in the sediment before the previous mat crashed. Their germination was stimulated by the mat's collapse, which allowed easier light penetration of the water, while seedling growth was enhanced by high levels of nitrates and phosphates in the water due to runoff from agriculture and to the release of nutrients from the decaying water hyacinth mat. In contrast, the weevil populations in the area were very low, presumably because eggs, larvae and pupae sank with the previous mat and drowned, while adults would have dispersed as the plant quality of the old mat declined. Therefore the new growth was able to proliferate in the absence of weevils.

The weevils will, in due course, disperse naturally back onto these fringes of plants in the western arm of the Lake. Indeed, a survey conducted in October 2000, which recorded an average of three adult weevils per plant at one site, indicated that the weevils were already moving back. However, there are still pockets of water hyacinth that are devoid of adult weevil feeding scars, and it may be necessary to conduct some augmentative releases of weevils in these areas to speed up the biological control process.

In the long-term it is hoped that the situation on Lake Victoria will become more stable, that the resurgence of the weed will be attenuated and that continuing augmentative releases of the weevils will not be needed. Introduction of some of the other biological control agent species being used elsewhere in the world, especially those that have short generation times, are capable of rapid population increases, and are good dispersers might further reduce the extent of water hyacinth resurgence.

However, the long-term biological control of water hyacinth on Lake Victoria will rely heavily on a reduction in the nutrients entering the lake

Source: Ogwang, J. (2001) Is there resurgence on Lake Victoria? Water Hyacinth News No. 4, pp. 1-2.

Contact: James Ogwang, Biological Control Unit, Namulonge Agricultural and Animal Production Research Institute, P.O. Box 7084, Kampala, Uganda
Email:

❑

Rwanda: Regional Cooperation in Action

Although water hyacinth (Eichhornia crassipes) in Lake Victoria is heavily attacked by Neochetina weevils, the Kagera River feeds large quantities of uninfested weed into the lake in the form of mats torn away from riverbanks or as individual plants. The implementation of a biological control programme within the headwaters of this river system in Rwanda to reduce the water hyacinth biomass entering the lake is therefore a key part of long-term control. The urgency of the situation is underlined by James Ogwang, Ugandan National Agriculture Research Organization - Namulonge Agriculture and Animal Production Research Institute (NARO-NAARI), who found on a visit in February 2002 that Rwandan highland water hyacinth colonization is increasing. New infestations are being reported from previously clear water bodies with alarming frequency. Regional collaboration is currently supporting both biological control and public education initiatives in Rwanda.

The Kagera River system has as one of its primary sources the Mukungwa River, high in the Virunga mountain range near Ruhengeri, Rwanda. This river flows south to join the Nyabarongo River, which then joins a smaller tributary near the Burundi border to form the Kagera River. The Kagera turns north, flowing between first Rwanda and Tanzania and then, turning east, through Uganda and Tanzania before finally reaching Lake Victoria. In total, the river system is some 500 km long and includes waterfalls, lakes and swamps. It is infested along its entire length by water hyacinth, but it is the last 160 km, below the final waterfalls at Kikagati in Uganda, where the weed grows unaffected by turbulent waters associated with elevational drops. Downstream of this point the river flattens and, as it flows across the Tanzanian floodplain, water hyacinth flourishes along riverbanks out to about 2 m from the shoreline. Water currents and velocity prevent water hyacinth from growing much further out, except in some bends, inlets and sloughs, or during periods of drought or flood. However, the two banks of the river produce 320 km of linear shoreline growth potential for the weed to a width of approximately 2 m, which is equivalent to some 64 ha.

The extent of the problem this creates is clear from estimates of the volume of weed pouring into the lake by Clean Lakes, Inc. (CLI). Within 1 km of Lake Victoria, the daily rate of weed flowing down the Kagera River ranges from 0.2 ha/day to more than 1.5 ha/day (an average 0.75 ha/day or 300 ha/year), depending on seasonal river volume conditions. If a growth rate model of 1% per day were assumed, then these 64 ha growing along the shoreline would generate about 0.64 ha of new weed growth/day - and this is, on average, equivalent to rates documented elsewhere.

The governments of Kenya, Tanzania, Uganda, Rwanda and Burundi have begun to coordinate water hyacinth control efforts through regional organizations such as the East African Community, the Lake Victoria Fisheries Organization, the Lake Victoria Environment Management Programme (LVEMP) and the Nile Basin Initiative, and through bilateral agreements. Recent activities in Rwanda provide an example of this in action. A Memorandum of Understanding on Common Agricultural Issues signed in 1997 by the governments of Uganda and Rwanda paved the way for the countries to commit to full collaboration in the management of water hyacinth, beginning high in the Kagera River basin watershed. Funding and technical support for the implementation of the biological control programme for water hyacinth in Rwanda is being provided by CLI through a 2-year Cooperative Agreement with the United States Agency for International Development, Greater Horn of Africa Initiative (USAID-GHAI) through the Regional Lake Victoria Water Hyacinth Management Program, and by the Institut des Sciences Agronomiques du Rwanda (ISAR).

In the first part of the programme, training activities and visits were carried out in Uganda and Tanzania. In November 1999, CLI facilitated the training of Rwandan and Burundian government officials in Uganda, led by James Ogwang together with staff of the Uganda Ministry of Agriculture, Animal Industries and Fisheries/Water Hyacinth Unit (MAAIF/WHU). Subsequently, ISAR staff visited the LVEMP-operated weevil rearing facilities at Bukoba and Kyaka in Tanzania in mid 2000. Ogwang followed this training up during a visit to Rwanda in February 2002 and provided fine-tuning advice for optimizing rearing.

In the meantime, a site was identified and the first weevil rearing station established in 2000 at the Karama Animal Husbandry and Fisheries Unit, an ISAR branch located some 70 km southeast of Kigali on the shores of a small water body, Lake Kilimbi, near to the Nyabarongo river. Also during this period, authorizations to export and import weevils, respectively, were obtained from the relevant Ugandan and Rwandan authorities.

In September 2000, some 850 weevils (both N. bruchi and N. eichhorniae) were collected from the NAARI weevil rearing tanks by NAARI, ISAR, and CLI staff. They were transported to Rwanda by air, and then by road to the ISAR/Karama rearing facility where they were used to inoculate water hyacinth in previously established tanks. First releases into the Kagera River system were made later that month in the seasonal Lake Kiruhura, a small depression near the Nyabarongo River 20 km south of Kigali. This 1+ ha site was 60% infested with water hyacinth. Weevil impact monitoring using remote satellite imagery was also initiated, with the assistance of the US Geological Survey, to support documentation of weevil release efficacy. This is using IKONOS 1-m PAN and 4-m multispectral band data in coordination with the US Geological Survey-EROS Data Center (USGS-EDC).

Exchange country visits provided opportunities to identify further needs and make plans. In December 2000, CLI supported a visit by MAAIF/WHU to Rwanda to review biological control efforts and to interact with counterparts in ISAR. Amongst suggestions made following this visit were needs for (a) two more rearing stations for weevil mass rearing and distribution, and (b) sensitizing the public to the dangers from water hyacinth. There were also calls for capital equipment, for coordination between institutions, and for Rwanda, especially the Kagera River system, to be included in the planned regional surveillance system. A report made following this visit by MAAIF/WHU to the LVEMP regional water hyacinth group meeting also laid the groundwork for a larger regional study visit which took place in July/August 2001.

In June 2001, a baseline survey of various water hyacinth parameters was carried out at four locations in the Kagera River system: Ruhengeri on the Mukungwa River, near Kigali on the Nyabarongo River, Kibungo on the Kagera River, and Lake Mihindi on the Kagera River within the Akagera National Park. Two additional sites for water hyacinth weevil rearing station establishment were selected at ISAR/Ruhengeri and Lake Ihema within the Akagera National Park. A follow-up visit in July 2001 was made to establish these stations and to improve facilities at the ISAR/Karama rearing station. The new stations were stocked with weevils imported from Uganda in July-August 2001.

During his visit in February 2002, Ogwang together with ISAR staff began monitoring the establishment and impact of the weevils at four points along the river system: at Ruhengeri, Karama, Lake Mihindi and upstream of the Rusomo Falls. They recorded plant growth parameters and weevil numbers and feeding scars per plant for each location. Accumulated information will allow the project to assess whether any changes are needed in the strategy of weevil use for water hyacinth control.

In late October 2001 CLI and ISAR staff looked for the uppermost infestation of water hyacinth within the Kagera River system, evaluated weevil status in the rearing stations, made more weevil releases, and finalized plans for a public awareness campaign. The highest point of infestation was found at an elevation of 1649 m on the Mukungwa River in Ruhengeri Prefecture. CLI and ISAR staff made an historic, but symbolic, release of two weevils from the Ruhengeri rearing station here. Weevil life cycles were found to be longer at the high altitude of the Ruhengeri rearing station where dry season temperatures rarely exceed 25ºC, and damage to water hyacinth plants was correspondingly less than at lower altitude sites. This is consistent with findings elsewhere that the weevils are less effective control agents in cool/high altitude conditions.

Temperature is probably not the only factor to affect weevil performance. Throughout the Kagera River system, and especially in the middle to lower system, it is suspected that high year-round sediment loads in the main river will be limiting to weevil populations due to sediment adhering to cocoons and suffocating pupating larvae or adhering to roots and reducing available pupation sites. In February 2002 Ogwang found that although weevils were established close to the highest point of water hyacinth infestation and there was slightly more feeding damage than in December 2001, their populations remained low. He agreed that this was probably partly temperature-related but, noting the turbidity of the river water (a result of massive soil erosion), suggested that the weevils were unlikely to perform well here. Instead, he suggests that the local population should be encouraged through local political organizations to participate in manual removal of the weed, an activity which could be promoted in conjunction with the purchase and distribution of appropriate tools.

Project participants currently consider that other areas in the middle and lower Kagera basin with clearer waters (i.e. dry season floodplain ponds, swamps and lakes) are likely to provide better habitat for weevils. Ogwang noted that the waters lower down the river system were fresher and cleaner in February 2002 than the previous October. This coincides with the end of the short rainy season that typically runs from October through December.

The ISAR/Karama site was not visited during the October 2001 visit, but a tank of weevils from the Ihema rearing facility was released into Lake Ihema so that all life cycles of the weevil would be present to begin dispersing. Two tanks were also transported to Lake Mihindi, where the infested plants were placed among shoreline infestations and in open water adjacent to a large mobile infestation covering several hundred hectares. In February 2002, Ogwang found weevil activity at this site significantly better than higher in the watershed. He found that cultures contained predominantly N. bruchi; this reflects the situation around Lake Victoria, where N. bruchi predominates over N. eichhorniae. Ogwang also found that local villagers were keen to become involved in biological control activities, wanting to know what the weevils looked like and participating in another redistribution exercise to Lake Mihindi. He suggests that where weevil rearing facilities exist, the local population should be integrated in the weevil release process. He notes that this will be possible only after establishment of organized village committees headed by keen leadership. Experience in Uganda, he says, shows that local villagers become active if such a programme gives them a sense of responsibility and ownership.

The importance of the project and involving the public is underlined by the recent confirmation of water hyacinth infestations in Lake Mpanga, where dense mats occur along the shore closest to the Kagera River. Ogwang suggests that this would be a good site strategically for a fourth weevil rearing facility: with clear water and a well-organized fishing community keen to be involved, it is an ideal location for distribution of weevils to this and nearby lakes.

Spreading the Message

A public awareness campaign was begun in November 2001 at six locations in Rwanda: a fishing community outside the Akagera National Park boundary, three villages on the Kagera River system in different parts of the country, the ISAR/Ruhengeri weevil rearing station, and a trading centre near Kigali. Each meeting followed the same format, which focused on demonstrating the impact of water hyacinth and the weevils through visual aids (water hyacinth posters, weevil-free and weevil-infested plants, and blown-up aerial and land-based photographs of affected areas and the various control options). At each meeting, a local leader introduced the water hyacinth topic. Next, CLI and ISAR team members were introduced, and they talked about the origin, distribution and impacts of water hyacinth, both in the Kagera River and in Lake Victoria. They went on to explain control options, and in particular biological control using the weevils. The implementation of control in Rwanda was discussed and the weevils' life cycle explained. The meeting then looked at water hyacinth posters, after which there was a question and answer session. Following a closing session, large and colourful posters (in Kinyarwanda, French and English) were distributed, and finally people were invited to observe the release of weevil-infested plants into water hyacinth-affected water bodies.

Additional weevil releases and monitoring activities are set to continue. More public awareness exercises are also being conducted. Encouragingly, efforts so far appear to be bearing fruit. For example, Ogwang found that villagers including children at the village he visited near ISAR/ Karama in February 2002 were aware of the biological control project's activities. At Lake Mpanga, Ogwang found local people keen to talk about the weed and the posters were popular.

For the future, though, there is more to be done. First, the public awareness campaign should be intensified to educate people throughout the country about the weed - this could take place at universities, local government offices, key ministries, and in major towns or trading centres.

Second, collaboration between ministries and institutions responsible for management and control of water hyacinth in Rwanda and in its East African country counterparts, which has underpinned the progress made so far, should be further strengthened. In addition, the Government of Rwanda should be encouraged to put in place a mechanism to stop or limit the spread of water hyacinth, especially by ornamental plant sellers.

To accelerate the release programme, additional rearing centres or increased capacity of existing stations are needed. A complete survey of the major rivers and lakes of Rwanda would allow the extent of water hyacinth distribution in Rwanda to be determined accurately. Alternative biological control agents, such as the mite Eccritotarsus catarinensis, should be considered, but this would need consensus from countries of the Lake Victoria basin.

Sources: Moorhouse, T.; Agaba, P.; McNabb, T. (2000) Recent efforts in biological control of water hyacinth in the Kagera River headwaters of Rwanda. In: Julien, M.H.; Hill, M.P.; Center, T.D.; Ding Jianqing (eds) Proceedings, 2nd Meeting of the Global Working Group for the Biological and Integrated Control of Water Hyacinth, Beijing, China, 9-12 October 2000.
www.cleanlake.com/rwanda_biological_control_program.htm

Moorhouse, T.G.; Asiimwe, T.; Agaba, P.; McNabb, T.J. (2001) Management heads upstream of Lake Victoria. Water Hyacinth News No. 5, pp. 1-3.

Ogwang, J.A. (2002) Regional Lake Victoria Water Hyacinth Management Program. Biological control monitoring and training. Report of a visit, 3-9 February 2002.
www.cleanlake.com/Rwanda3rd-9th.htm

Contact: Tom Moorhouse, Clean Lakes, Inc. - Uganda, Nile International Conference Center, Room 235, P.O. Box 7057, Kampala, Uganda
Email:
or Tom McNabb,

❑

Buckthorn Biocontrol for North America

Exploratory entomology carried out in the 1960s is underpinning a new North American-funded CABI Bioscience project based at the Switzerland Centre in Delémont. CABI Bioscience, in partnership with the State of Minnesota Department of Natural Resources, aims to find biocontrol agents for the buckthorns Rhamnus catharticus and Frangula alnus, which are invasive in North America and cause significant economic and environmental damage.

Buckthorns (Rhamnaceae) are small trees or shrubs and occur worldwide. Eurasian species were introduced to North America in the mid 19th century and both R. catharticus and F. alnus have spread widely in Canada and the USA. Rhamnus catharticus has a vast range bounded by Nova Scotia in the northeast, Saskatchewan in the northwest, Kansas in the southwest and North Carolina in the southeast. Frangula alnus is found in northeastern USA and southeastern Canada. Probably because of its history as an ornamental, F alnus' distribution pattern tends to be associated with urban centres and nearby wetlands, although it is spreading into natural and agricultural areas. Rhamnus catharticus, in contrast, was widely planted in hedges and farm shelterbelts and less as an ornamental, and it is more widely distributed and more common in rural areas than F. alnus.

Both species are spread primarily by seed in the wild, with many animals (particularly birds and small mammals) who eat the ripe fruit acting as vectors, although in wetland areas seeds are also dispersed by water. The buckthorns have long growing seasons, rapid growth rates and resprout vigorously following top removal. In North America, their leaf break occurs before most other woody deciduous plants, and senescence occurs later. They form dense, even-aged thickets, and the large leaves and continuous canopy formed by lateral crown spread create dense shade. These characteristics often obliterate native species and alter the composition of the forest communities these species have invaded. Allelopathy may also contribute to invasiveness.

In North America, R. catharticus invades many types of habitat including moist and dry upland sites, but grows best in well-drained soils and is less vigorous in dense shade. It has spread into floodplain and riparian forests, oak forests, woodland edges, ravines, fence rows, old fields and prairies. By contrast, although F. alnus invades some of these habitats, it is above all an aggressive invader of wetland habitats, including wet prairies, marshes, calcareous fens, sedge meadows, sphagnum bogs and tamarack swamps. Frangula alnus colonization is enhanced by drainage intervention that favours woody growth.

These invasive buckthorns, though, also pose an agricultural threat. They are alternate hosts of crown rust fungus, Puccinia coronata, which can severely reduce oat yields, and also of the soyabean aphid, Aphis glycines, a major pest in eastern Asia which has recently become established in North America.

From previous studies, we know that more than 200 species of arthropods and 40 species of fungi are associated with buckthorns in Eurasia, of which about 18 and 14, respectively, are likely to be restricted to the genera Rhamnus and Frangula. Some dozen of these have been prioritized by the project for further study, which will initially focus on host plant testing in Switzerland. Host range testing of potential arthropod biological control agents carried out in the 1960s involved very little testing of nontarget buckthorns, however, and a priority of the project is to remedy this. Already, a proposed test list has been drawn up which includes over 40 related species, including native American species, closely related species and sympatric species as well as economically and culturally important species.

A number of fungi have also been identified as potential biological control agents. Field surveys in Europe will assess their impact in the field, so promising candidates can be prioritized for further study.

Contact: André Gassmann, CABI Bioscience Switzerland Centre, 1 Rue des Grillons, CH-2800 Delémont, Switzerland
Email:
Fax: +41 32 421 4871

❑

Dr S.P. Singh Retires after a Distinguished Career

After nearly four decades in the service of biological control and entomology, Dr. Surinder Pal Singh retired in February 2002. His name has been closely linked with biological control in India and he has been instrumental in building a strong institution for biological control in the country.

Born on 11 August 1941 in Faridkot in the land of five rivers, the Punjab - a state with well-developed agriculture, and also noted for its men of valour and industriousness and its fun-loving people - Dr S. P. Singh was the eldest in a family of three brothers and one sister. His early education was in Faridkot itself but he went on to complete his Bachelor's degree in Agriculture from College of Agriculture, Ludhiana during 1961. His early passion for insects saw him complete his Master's degree in Entomology from the same college in 1963. After a brief stint in the Punjab Agriculture Department, he started his research career in entomology as a Research Assistant in the Ludhiana College in 1963. In 1967 he commenced his service in the Indian Council of Agricultural Research (ICAR) by joining the Central Potato Research Station, Patna and remained in the service of ICAR until his retirement.

His strong interest in entomological research in general, and biological control in particular helped him gain a Ministry of Education, Government of India scholarship from 1968 to 1973 for advanced postgraduate research in bioecology and biological control of cotton bollworm. This earned him his PhD from the Kuban Agricultural Institute, Krasnodar in Russia. It also helped him pick up the Russian language, which he remembers to this day.

Back in India he was appointed in 1974 as Scientist in Agricultural Entomology at the Central Horticultural Research Station (Indian Institute of Horticultural Research (IIHR) - ICAR), Chethalli, Karnataka State and is continuing to live in this state even after his retirement. He went on to become the Head of the Station from 1980 for more than 3 years. He was appointed as Project Co-ordinator, All India Co-ordinated Research Project on Biological Control of Crop Pests and Weeds at IIHR, Bangalore in 1984. In 1988, he was transferred as Project Co-ordinator (Biological Control) and Head, Biological Control Centre, which was located at the erstwhile Commonwealth Institute of Biological Control (CIBC) Indian station, Bangalore. The Centre was upgraded to the Project Directorate of Biological Control (PDBC) and Dr Singh was appointed Project Director in November 1993, in which capacity he worked until his retirement.

Biological control has been his forte from the time he started his work on biological control of cotton bollworm for his doctoral thesis in Russia. His early years of research on citrus pests were spent in quantifying the role of natural enemies in suppression of several citrus pests. The significant role played by him in evolving production and release technology for Cryptolaemus montrouzieri and its effective transfer to the farmers of Kodagu district under the Lab to Land programme from 1978-83 is still remembered by farmers in the district. This also paved the way for commercialization of technology for suppression of mealybugs on crops such as citrus, coffee and grapes. He has been a champion of basic research on biological control and has led this in many areas, including biology and ecology of natural enemies, rearing techniques for natural enemies, development of temperature- and insecticide-tolerant strains, enhancing the potential of natural enemies, tritrophic interactions between natural enemies, host insects and host plants, and introduction of exotic natural enemies. Dr Singh has also been instrumental in developing biocontrol-based technologies for the management of mealybugs, scales, psyllids and lepidopteran pests of maize, sugarcane, tomato, cabbage, cotton and other crops.

He was quick to react to changing pest scenarios especially in relation to the problems due to introduced pests. One such case was when the Leucaena psyllid was accidentally introduced in the late 1980s. His vision and pragmatic approach, which led to classical biological control through the introduction of Curinus coeruleus from Mexico via Thailand, saved the subabul plantations from the ravages of this pest and today the coccinellid is well established in India. To Dr Singh also goes the credit for playing a significant role in developing the first endosulfan-tolerant strain of Trichogramma chilonis in the world. The technology was taken up immediately by private industry for large-scale production and field use in crops. This product is now in use on more than 24,000 hectares in India.

The credit for developing and sustaining the phenomenal growth of PDBC should go to Dr Singh's enthusiasm, zeal and vision. The institute started with just six scientists and today has a core of 25. The coordinated project also grew from ten centres initially to 16 centres now. To achieve and sustain this he encouraged his team of young scientists to explore new vistas in biological control. He also took pride in the excellent capabilities of his team of scientists and undertook strong confidence building measures in young workers. Dr Singh was instrumental in opening new laboratories to initiate work in areas such as entomopathogenic nematodes, biological control of nematodes, plant disease antagonists and weed pathogens. His dynamic leadership and perseverance enabled PDBC to secure funds from a wide variety of national and international funding agencies, which helped to equip laboratories and also enabled scientists to visit leading institutions in other parts of the world and bring themselves up to date on the latest advances in the field. The ultimate recognition of his institution-building capability came when PDBC was adjudged as the Best Institution in ICAR in 1998, just 5 years after its establishment.

Dr Singh's capabilities in research, leadership and institution building helped gain biological control recognition as a primary tool in pest control in India, and resulted in the World Bank-NATP recognizing PDBC as a Team of Excellence in biological control in India for developing human resources in biological control.

He has guided the research work of many postgraduate students and has contributed significantly to the literature on biological control through published papers and books. His books include titles on production technology for natural enemies and 15 years of research on biological control in India. Dr. Singh has travelled widely across the globe and has visited Russia, the USA, Tunisia, Taiwan and Malaysia.

His outstanding contribution to the field of entomology has been widely recognized in the many awards and fellowships conferred on him, and has led to his inclusion as a member of national review teams and expert panels. He has been a leading figure in prestigious scientific societies, and is the President of the Society for Biocontrol Advancement and Editor of the Biocontrol Newsletter. He has also been a member of the editorial boards of a number of national and international scientific journals, including, of course, Biocontrol News and Information.

His retirement was marked by entomologists in Bangalore joining the staff of PDBC in February to wish him a long and fulfilling retirement to be spent with his wife and son. Those who worked with Dr Singh will remember him for his infallible memory, independent and strong opinions about people, and the unflagging zeal, enthusiasm and energy with which he carried out his work.

By: S. Ramani and S. K. Jalali (PDBC, Bangalore)

❑

Australian Moves in Europe

John Scott, the current Director of the CSIRO's European Laboratory in Montpellier, France, will be finishing his term in July 2002. CSIRO's European laboratory has a traditional role in the exploration and testing of biological control agents for Australia, having been the source lab for very successful control projects against skeleton weed, ragwort, nodding thistle and Paterson's curse. The role of the laboratory is expanding to act as a European portal for a range of CSIRO's other activities (see website ) . John will be replaced by Andy Sheppard for a 3-year term in July. Like John, Andy is an ecologist working on invasive plants and their management strategies and has an active international collaborative project on brooms and gorse.

Websites:
CSIRO: www.csiro.au
CSIRO European Lab:
www.csiro-europe.org/lab.html

❑