Sorting Salvinia

An aquatic fern native to southern Brazil, giant salvinia (Salvinia molesta) is considered one of the most invasive plants in the world. Its rapid rate of spread is facilitated by environmental adaptability coupled with an ability to propagate vegetatively from plant fragments.

It has been the target of classical biological control programmes since the 1960s in Africa, Asia, and Australia. However, the first attempts were unsuccessful in Africa, India, Fiji, and Sri Lanka because of the misidentification of the plant as Salvinia auriculata. Researchers surveyed S. auriculata in Guyana and Trinidad and found a small weevil identified as Cyrtobagous singularis. Although C. singularis did establish in several areas, it had no effect on the infestations. Salvinia molesta was separated from S. auriculata in 1972 and its native range in Brazil was discovered in 1978. In 1980, what was thought to be a biotype of C. singularis from S. molesta was introduced at Lake Moondarra in Australia and proceeded to destroy more than 30,000 t of S. molesta in less than one year. Closer examination of the 'biotype' resulted in its elevation to species status, namely C. salviniae. This new species reversed earlier failures and successful programmes were conducted in Australia, Fiji, Ghana, Kenya, Malaysia, Papua New Guinea, South Africa, Zambia, Zimbabwe, India, Botswana, Namibia, and Sri Lanka, where control has been dramatic and rapid: in many cases S. molesta was reduced by more than 90% in less than a year following release of C. salviniae*.

Here we look at progress in two programmes to combat recent invasions. The first, in the USA, underlines how taxonomic uncertainties and confusion can still confound biocontrol. The second, in the Senegal River in Senegal and Mauritania, deals with management of invasive species to mitigate threats to both the environment and economic development.

*Tipping, P. (2000) Biological control programmes for giant salvinia: history and update on US efforts. Water Hyacinth News No. 2, p. 6.

❑

Salvinia: USA Begins Round Two

Cyrtobagous salviniae was released for the second time against giant salvinia (Salvinia molesta) in Texas and Louisiana in October 2001. Weevils for this release were obtained from cooperators in Australia where they have been used successfully to control Salvinia molesta. Following the second release in the USA, numerous adults and significant damage were found at most of the sites 2 months later in December. The winter season is expected to inhibit any further activity by the weevils but we hope to learn if they can overwinter at locations in east Texas and western Louisiana. Additional releases of C. salviniae will be conducted if necessary during the spring and summer of 2002.

Salvinia molesta has been established in the wild in the USA since at least 1998, but it is possible that it has been living free in the USA for rather longer, as it has been widely distributed as an ornamental plant and is easily obtained via the Internet. First discovered in eastern Texas, it now extends into western Louisiana.

The first release of C. salviniae was conducted in June 1999 using weevils collected from common salvinia, S. minima, in Florida rather than the tried-and-tested stock from Australia, as this obviated the risk of introducing new pathogens or parasites into the USA. Cyrtobagous salviniae had been introduced accidentally into Florida prior to 1960 and is now found throughout the state feeding on S. minima, and has also been found attacking S. molesta at one site in southwestern Florida. Unfortunately, the results of the first biocontrol attempt were unclear as many of the original Texas and Louisiana release sites were corrupted or destroyed by floods, droughts, saltwater intrusion, or landowner actions. Significant damage of the salvinia was noted at one release site, though, before it was destroyed by the landowner (despite a previous agreement not to do so).

Gene sequence studies in early 2000 found minor differences between 'Australia' and 'Florida' weevils in the number of base pairs in the D2 gene. Further releases of the 'Florida' weevils were suspended because of the taxonomic uncertainty this created, and instead efforts were redirected to populations of 'Australian' C. salviniae, collected originally from Brazil and used in successful biological control programmes in Australia, Papua New Guinea, South Africa, and other countries.

However, the release permit for the 'Florida' C. salviniae was not extended to cover the 'Australia' C. salviniae. Regulatory officials in the USDA-APHIS (Animal Plant Health Inspection Service) required that a new permit be issued, which involved a lengthy process made more lengthy when a petition written by another lab was rejected because of substandard research and reporting, thereby causing further delays in obtaining a general release permit. In cooperation with Wendy Forno of Australia (formerly of the Commonwealth Scientific Industrial Research Organisation, CSIRO) and Sharon Docherty and Martin Hill from South Africa (formerly of the Plant Protection Research Institute, PPRI), we wrote a new petition, which was approved and the permit was finally issued, thus facilitating the second releases.

Studies designed to sort out the differences between the populations of C. salviniae are continuing. Preliminary data shows that 'Florida' weevils are equally attracted to both S. molesta and S. minima, can repro-duce on S. molesta, and live longer and lay more eggs on S. molesta than on S. minima. In addition, the 'Australia' weevil has demonstrated the ability to suppress S. minima in tank studies.

By: P. W. Tipping and T. D. Center
USDA-ARS Invasive Plant Laboratory, 3205 College Ave., Fort Lauderdale, FL 33314, USA
Email: /

Fax: +1 954 476 9169

❑

Salvinia: West Africa Battles Invasives

The freshwater wetland systems of Mauritania and Senegal are crucial habitats for Palaearctic migrant birds crossing 200 km of the arid western Sahara Desert, and as such they are the focus of national and international conservation efforts. A series of national parks has been set up by the governments of Senegal and Mauritania in recognition of the importance of the ecosystems, many of which are designated World Heritage Sites and/or recognized as Wetlands of International Importance by Ramsar. The wetlands and the rivers that feed them, though, are also fundamental to maintaining local livelihoods and regional economies, providing fishing, irrigation for agriculture and potable water supplies for both rural and urban areas. Traditionally maintained by varying seasonal rainfall, advances in hydrology now create opportunities for water flow to be regulated to meet needs throughout the year. Such changes inevitably lead to disruption of the ecosystem, and restoring a balance that is sustainable in the long term is proving a challenge, and is having to overcome some unexpected obstacles including invasion by alien water weeds, with Salvinia molesta (giant salvinia) causing alarm most recently. Concerted efforts involving local, national, regional and international co-operation have been made to mitigate the Salvinia threat. Mechanical clearance provided limited and costly short-term relief, but improved water management and biological control through introduction of the weevil Cyrtobagous salviniae is providing a permanent solution. In a wider context, however, this programme provides a blueprint for both preventing the introduction of other invasive species, and implementing good management of those already present.

Damming a Problem

Historically, the wetlands were sustained by natural flooding, but this was variable and in some years of diminished rainfall (particularly since the 1960s) it failed to restore water levels. The erection of dykes, sluices, temporary dams and, more recently, permanent dams on the Senegal River by OMVS (Organisation pour la Mise en Valeur du fleuve Sénégal, a trilateral organisation grouping Mali, Senegal and Mauritania) formed the basis of a management plan for the Senegal River basin, intended to allow river navigation, and provide reliable irrigation for agriculture on hundreds of thousands of hectares as well as water and electricity supply for rural and urban areas in Senegal and Mauritania. The Diama Dam, built near the mouth of the Senegal River and 25 km upstream from the city of St Louis, prevents seawater incursion into agricultural land in the delta during the dry season. Since it became operational in 1988 it has wrought significant ecological changes in the lower Senegal River basin. Upstream, the formerly estuarine, seasonally dry conditions of the lower river have given way to permanent freshwater, and this has led to luxurious development of aquatic vegetation, and notably dense stands of Typha australis (reed mace) in shallow water, and mats of floating weeds, initially Pistia stratoites (water lettuce) and now Salvinia. Besides visible impacts on the riparian human and wildlife populations there have been less-evident effects, including a substantial increase in malaria, for example, owing to an increase in year-round standing water for breeding. Down-stream, in contrast, conditions became more saline and water supply virtually ceased during the dry season. These changes were enhanced by embankments built to separate the river from the floodplain, which led to large parts of the floodplain and estuary becoming drier.

The emergence of an invasive weed threat is thus part of a wider issue: the impact of changing land-use patterns and hydro-logical management on the natural ecosystems along the lower Senegal river, where salinity traditionally varied from nearly fresh (during inundation) to brackish as water levels fell through the dry season. The ecological disturbances above and below the Diama Dam are reflected in the national parks that bound the river to the north and south.

Djoudj National Park (Parc National des Oiseaux du Djoudj) in Senegal was created in 1971. It was designated a Wetland of International Importance under the Ramsar Convention in 1980 and inscribed on the UNESCO World Heritage List in 1981. This seasonally inundated wetland system covers some of 16,000 ha of brackish lakes and pools, linked by a network of channels stemming from the Senegal River. Djoudj forms a permanent sanctuary for some 1.5 million birds, and even more migrants: an estimated three million pass through between September and April, and more than 70 species of migrant birds were identified during an expedition to catch and ring Palaearctic birds in 1990. The sanctuary now lies upstream of the dam. Changes in vegetation since it was completed included an initial explosive growth of Pistia during the first half of each dry season, accompanied by a more insidious but no less damaging spread of the emergent weed Typha.

The Diawling National Park (Parc National du Diawling) lies on the Mauritanian side of the Senegal River delta, downstream of the Diama Dam in the former estuary. The area became completely cut off from fresh water after the Dam's construction, and for several years it was flooded only with seawater. Owing to the high evaporation rate, the area quickly became a salt desert, with high pyrite concentrations in the soil, which also made it unsuitable for cultivation. OMVS constructed four sluices to re-flood the area and restore the pre-dam hydrological scheme and to preserve the rich biodiversity. Since then IUCN (World Conservation Union) has built two more dykes and sluices to optimize water management, and sponsored artificial flooding by the park management. In consequence, 16,000 ha of salt desert are restored and the ecosystem is reviving, a unique example of the reconstruction of a natural environment. Diawling was made a National Park in 1991 with a mandate to integrate conservation and development (including fishing and pastoralism within its boundaries), and was designated a Ramsar site in 1994. It, too, has been invaded by Typha, and more recently by Salvinia.

Waves of Invasives

The invasive weed problem surfaced in the Djoudj Park soon after the Diama Dam became operational in 1988. Reports of Pistia were first received in 1989. The weed's encroachment over the next 5 years was relentless, covering part of the Djoudj River water surfaces within the Park in the early dry season, making navigation difficult and threatening the habitat of permanent and visiting birds. The fall in salinity following the completion of the dam and inadequate control of flooding (leading to an insufficient drying-out period) were cited as being responsible for this and other changes in vegetation in the Park, and similar changes in the nearby Senegal River and other water bodies including Lac de Guiers. This lake, some 50 km upstream of the Diama Dam, is the city of Dakar's major drinking water supply and also provides irrigation water for vast agricultural areas.

Fortunately, there was a good history of successful biocontrol of Pistia with the Neotropical weevil Neohydronomus affinis, beginning in Australia and subsequently elsewhere including Africa. Introductions in Senegal began in Lac de Guiers in 1994, with the support of IITA (International Institute of Tropical Agriculture) and GTZ (Deutsche Gesell-schaft für Technische Zusammenarbeit). Introduction of weevils from Lac de Guiers into the Djoudj Park took place in 1998 and subsequent years within the framework of an EU (European Union) funded research project (coordinated by (Koninklijk Instituut voor de Tropen/Royal Tropical Institute, Netherlands (KIT/RTI) in collaboration with the Finnish Environmental Institute, the University of Vienna in Austria and the Senegalese Ministry of the Environment). The Pistia vegetation became markedly stressed, but as the plants die off in the course of the dry season owing to the increased salt content of the water, the natural enemy populations collapse. However, Pistia proved to be a prolific seed producer (because of - some or all of - increased salinity, crowding and nutrient-rich water owing to bird colonies) and strong seedling regrowth each season meant the biocontrol agents needed to be reintroduced annually. Releases made immediately after inundation at the beginning of the dry season in September proved most effective. The introduction of the weevils was accompanied by improvements in water management, and together these measures have reduced Pistia populations to a level that does not impact adversely on the ecological function of the Park.

While the Pistia problem has been solved, though, Typha is still thriving within the national parks and outside. In less than 10 years, it has built an almost impassable wall between the river dykes and the open water on both sides of the river, clogged shallow waterways within the parks, invaded the economically important Lac de Guiers, and has caused particular problems by its spectacular take-over of a shallow freshwater reservoir just above the Diama Dam. Typha continues to have diverse negative impacts on drinking water, fishing, water-borne diseases and pests, which far outweigh any potential of the plant to combat erosion or be used in manufacturing.

A new threat, from Salvinia, was first observed in the Senegal River near the Djoudj Park in September 1999, the result of an accidental introduction from a nearby flooded field where it was under cultivation. By then, though, Salvinia had already invaded a stretch of more than 70 km of the river between the village of Rosso and the Diama Dam, approximately 20 km upstream from the city of St Louis*. It had formed thick mats along the Typha fields, which occur in shallow water along the southern shore in Senegal and the northern shore in Mauritania. Tributaries of the Senegal River as well as narrow channels through the Typha, which had been kept clear by the local population using laborious manual methods, were now entirely overgrown by Salvinia. Consequently, it had become very difficult for the people in the villages along the river to reach open water and even when they did fishing was impossible because their nets became clogged with Salvinia plants. Only the main flow of the river was still open. Dense plant masses also piled up against the Diama Dam, and these led to it being opened three times a week to flush down the Salvinia plants. Ironically, this unplanned release of water turned out to be very beneficial downstream of the dam. The remaining mangrove trees, craving freshwater during the dry season, staged a recovery. Other beneficiaries were the young marine fish that still try to use this former estuary as a nursery, and of course the local people.

As Salvinia spread through adjacent water bodies and basins, including the Lac de Guiers, it became increasingly apparent that this was a new threat, not just to the ecological equilibrium, but also to the economic stability and human health of the region. It provided a substrate for other weeds to encroach on water bodies, and there was a threat that it could spread into rice fields, as it has elsewhere in the world. The weed also impedes gas exchange, and as the plants decay they consume oxygen, which further disturbs the ecological balance and impacts negatively on aquatic fauna and potable water quality. Conversely, though, weed cover increases the available habitat for disease vectors such as snails and mosquitoes.

Weevils Win Again

The recognition of the threat from a new invasive floating waterweed precipitated prompt action in both Mauritania and Senegal, although fear of nontarget effects, even from the tried-and-tested Salvinia biocontrol agent, led to delays in funding and of biocontrol implementation in Senegal.

In Mauritania, Salvinia blocked water inlets to the Diawling Park completely, hampering the annual artificial flooding. Barriers built in the river in front of the sluices, intended to prevent Salvinia entering the Park, collapsed under the weight of plants as soon as the sluices were opened during the rainy season inundation in July-August 2000. Worse, the Park authorities could not now close the sluices because of the remaining plants. Thus water continued to leak into the former floodplain for 5 months before the sluices were repaired and closed. This extended period of freshwater in the Park also allowed Typha to establish itself in the areas around the sluices. Salvinia had serious impact on local people, as fishing had become impossible and some 40% of the fishermen left the area to work in the city or in the rice fields, thus jeopardizing the 7 years of ecosystem restoration efforts aimed at promoting remigration back to the delta by reestablishing the natural resources to sustain riparine livelihoods.

The Minister of Rural Development and the Environment in Mauritania gave a green light to biological control in February 2000. Donors were slower to react, however. Instead, contributions from the Minister himself, the Senator of Rosso, national environmental NGOs, the fishermen of the Diawling Park, and some DGIS (Dutch Development Aid Agency) workers paid for a shipment of 300 Cyrtobagous salviniae weevils from PPRI (Plant Protection Research Institute), South Africa in April 2000. Charmed by this private initiative, PPRI charged only transportation costs: a mere US$300. The weevils were used to establish a starter colony in plastic containers (purchased by an Austrian and Dutch development workers) on the shores of the Senegal River. Continuing the spirit of `do-it-yourself' biocontrol, local fishermen from the village of Zirét Takhredient became involved in rearing and maintaining the insect cultures, regularly refreshing the breeding culture with fresh Salvinia plants and removing dying ones. These people normally fish in the Park, migrating to the river as the Park dries up. Continuing support for Diawling Park staff and the fishermen was provided by a GTZ technical advisor working at the Ministry of Rural Development. First releases were made in June 2000 at various sites from the Diama Dam to Rosso, and 9 months later the impact was visible within the Typha stands: Salvinia had completely disappeared. More fishermen are now beginning to return.

The only international funding for this striking success was from China, which contributed to the state budget for mechanical removal of Salvinia in front of the main water inlets of the Park, just before the artificial flooding period in 2000. The speed with which control has been achieved is testament to the dedication of those involved, decisive action at every level, and the cooperation and generosity of all involved. Such qualities may not be enough to meet future invasive threats: Salvinia is acknowledged to be an `easy' target for biocontrol, and the international donor community needs to be ready to help Mauritania in the future.

Across the river in Senegal, the Djoudj Park had remained free of the weed. Sluices were kept closed to prevent Salvinia incursion, but clearly this bought only limited time. The weed was, in any case, impacting seriously on the livelihoods and health of the population outside the Park. The Ministry of the Environment quickly mobilized initiatives. The scientific committee (Groupe de Réflexion et d'Appui Scientifique et Technique; GRAST) of the National Parks Directorate (Direction des Parcs Nationaux; DPN), comprising scientists, decision makers and local people was tasked with developing, directing and evaluating a Salvinia management strategy. This met first in February 2000 with Dr Arnold Pieterse (KIT/RTI). Recognizing the need for regional coordination, the management strategies of Senegal and Mauritania were developed during and following further inter-country meetings held in April 2000 and June 2001.

Participants at the February 2000 meeting agreed that biological control would be the lasting lynchpin of Salvinia management in Senegal too, with mechanical control being used to provide rapid and short-term alleviation. While the biological control effort was being set in motion, mechanical removal began to clear important waterways and in particular keep backwaters and tributaries near the sluices for the Djoudj Park clean of Salvinia.

A joint civilian and military committee for mechanical eradication (Comité Civilo-Militaire d'Appui au Développement; CCMAD) was coordinated by the Société d'Aménagement et d'Exploitation du Delta de la Fleuve du Sénégal (SAED). CCMAD comprises representatives of local people, military personnel, Djoudj Park staff, regional water and forests inspection services, the tourist office and DPN. Funding for mechanical control came from both state and private sources and from the UN Development Programme (UNDP) Global Environment Facility (GEF). During 6 weeks in May-June 2000, 200 civilians and military personnel put in 6000 h to remove as much weed as possible (and particularly to clear infestations upstream of sluices), and to erect barriers to prevent further spread into backwaters feeding the Park. More than 20,000 m3 of weed was removed, which provided short-term alleviation of the weed problem and contributed to containment of its spread. Cost, however, was prohibitively high in terms of manpower and fuel costs (5000 litres of diesel fuel, for example, was used), and mechanical measures could delay but not indefinitely prevent Salvinia's invasion of the Park and were not practical on the larger scale needed for widespread control.

The biological control component of the Salvinia management strategy was placed under the management of DPV (Direction de la Protection des Végétaux), which was tasked with coordinating efforts by national and international bodies. The first batch of 300 weevils was imported from PPRI (shortly after weevils were introduced into Mauritania), financed by the Dutch government via IUCN. The release of these first weevils in Senegal ended in failure, as a so-called 'starter' population was not cultivated at a protected site. Instead, all the insects were released at unmarked sites in the river where the infested plants could not be kept together. Subsequent surveys failed to recover any weevils and it may be assumed that the small number of insects became too dispersed to build up a population. With the support of the Dutch government, IUCN imported 1200 more weevils from South Africa in March 2001, a year later. A rearing operation was started at the Biological Station (Station Biologique) in Djoudj Park in the framework of a new policy project of the EU, which is coordinated by RTI. In this context, two Senegalese students at the University of St Louis were trained in rearing techniques by PPRI, and subsequently insects from their rearing programme were released in the Senegal River in the vicinity of the Park and in the Lac de Guiers. A UNESCO (UN Educational, Scientific and Cultural Organization) funded mission to Senegal in April 2001 found that the rearing operations were being hampered by a severe lack of resources at the Biological Station of DPN, however. The mission concluded that the operational capacity in terms of logistics, materials and staff, could be improved.

Despite these misgivings, it soon became apparent that the weevils were doing a magnificent job. Within 6 months of this visit, Salvinia plants were dying throughout the Senegal River, and their colour was turning from green to dark-brown or black. Evidence for the weevils' wide dispersal was seen both in Senegal and in Mauritania. In Senegal, Salvinia plants were infested at distances far from the release sites near the Djoudj Park, and it may be assumed that the weevils had flown across from Mauritania, where they had been successfully released almost a year earlier. A workshop held in St Louis in October 2001 on the problem of invasive plant species made it possible for the participants, who came from all over Africa, to see the Salvinia die off in the Senegal River with their own eyes.

Battle Won, War Continues

The rapid and spectacular invasion of the Senegal River delta by Salvinia captured the attention and stimulated action by local, regional, national and international stake-holders. However, the Salvinia explosion in the River happened in the wake of the largely unchallenged invasion by the equally damaging but less conspicuous species, Typha, along the river, in reservoirs and inside the parks.

Even worse could follow: water hyacinth (Eichhornia crassipes) is not far away and threatens the ecosystem. During the St Louis workshop, participants learned that water hyacinth plants are being sold as ornamentals at a plant nursery in the city. This does not pose an immediate threat to St Louis, as the water in the Senegal River downstream of the Diama Dam is too brackish for water hyacinth to survive. However, if plants were somehow to be transferred to the river upstream of the Diama Dam, which is only 20 km away, a new ecological disaster would emerge. Water hyacinth is even more aggressive than Salvinia and the available biological control agents are less efficient than Cyrtobagous. The authorities announced that the plants would be destroyed. However, as large numbers of water hyacinth plants are present in the markets of Dakar, extreme watchfulness remains a priority.

Thus, although first Pistia and now Salvinia have been brought under control, invasive weeds continue to threaten the region. In this sense, what is happening in the Senegal River basin is an example of a problem facing the whole continent: land-use change owing to population pressure and agricultural development is allowing invasive species to threaten Africa's biodiversity, economy and health. The agriculture, river traffic, hydroelectric and other developments that the dams on the Senegal River were built to help are under threat from invasive weeds. The actions of the parties involved in the threat to the Senegal River basin, and in particular the concerted and cooperative actions of the countries in recognizing the need to include invasive species issues in management and development plans for the delta, provide a good model for other parts of the continent.

*Pieterse, A. (2000) Aquatic weed management. In: Rijn, P.J. van (ed) Weed management in the humid and semi-humid tropics. Amsterdam, The Netherlands; Royal Tropical Institute, pp. 169-176.

Websites:
Djoudj National Bird Park:
www.senegal-online.com/senega292.htm
UNESCO World Heritage List:
www.unesco.org/whc/sites/25.htm
Ramsar: www.ramsar.org/

Contact:
Dr Arnold Pieterse, Royal Tropical Institute, Mauritskade 63, 1092 AD Amsterdam, The Netherlands
Email: l
Fax: + 31 20 6684579

M Sara Diouf, Directeur Adjoint, Direction des Parcs Nationaux, PO Box 5135, Dakar-Fann, Senegal
Email:
Fax: +221 252399

M El Waled ould Mome, Director,
Diawling National Park, PO Box 3935, Nouakchott, Mauritania
Email

❑

Black Wattle Problem Emerges in Indian Forests

Black wattle (Acacia mearnsii) is a fast growing leguminous tree native to Australia. It is widely used as a source of tannin, fuel wood, charcoal, poles, green manure and windbreak. Suited to cooler tropics, this tree grows well in tropical areas where the annual rainfall is more than 1000 mm.

Extensive areas of black wattle plantations have been established in South Africa, South America, southern Europe and Southeast Asia. The main purpose of introduction was for the commercial production of tannins, which is used for leather tanning and in products such as wood adhesive.

In Kerala State in southern India, A. mearnsii was introduced in the 1980s and mainly grown in the high altitude areas (over 1000 m above sea level; masl). It was preferred over other candidate forestry species because of its fast growth rate and the minimum post-planting care required. However, attempts to grow A. mearnsii on a plantation scale were not successful in most places in Kerala owing to high seedling mortality, eco-climatic stress and other factors. Hence, it now occupies only a very small area in the State, and fresh planting is not undertaken because of recurrent failure in establishment.

The experiment with black wattle plantations has left an ominous legacy, however, for A. mearnsii has not simply gone away. Recent surveys conducted by the Kerala Forest Research Institute (KFRI) indicate that in certain isolated pockets in the high altitude areas, some trees of A. mearnsii survived against the odds and are now growing luxuriantly, forming small scrub jungles. At Vattavada (1800 masl) in Idukki District, it was noticed that, within a period of 3 years, A. mearnsii has penetrated and spread over a 1 km2 area in the dense subtropical montane (shola) forests, suppressing the native vegetation. Spread of the trees into the core areas of the highly diverse shola forests at Kolukkumalai (2480 masl) in the same district was also observed. The high competitive ability and seed production, prolonged seed dormancy and high rate of seed viability of the species probably helped the tree to spread like a wildfire into these forests. Collection of branches and logs of A. mearnsii by the local people for firewood purposes will also have helped spread of the tree species. Wild animals such as bison and deer also aid in seed dispersal. The allelopathic properties of leaves and branches are other possible factors favouring the gregarious growth of the species.

Needless to say, the biodiversity of the subtropical montane forests in Kerala is now under great threat owing to this unchallenged invasion by A. mearnsii. Control methods need to be considered urgently, and in this context it should be noted that this species is a serious weed in South Africa, where it was introduced much earlier than in Kerala.

By: Dr K. V. Sankaran, Kerala Forest Research Institute, Peechi - 680 653, Kerala, India
Email:

❑

Black Wattle: South Africa Manages Conflict of Interest

Australian Acacia (wattle trees) have been utilized in South Africa since the 1820s for sand stabilization, garden ornamentals, timber and pulp production and tannin extraction. Their use has been widespread and most of the 13 naturalized species form an integral, although not always welcome, part of South African socio-economic culture. Black wattle (Acacia mearnsii) is the most economically important Australian Acacia present in South Africa, both as a silvicultural crop plant and as an invasive weed.

Black wattle is indigenous to southeast Australia where it forms a common component of Eucalyptus forests. The species was utilized initially in South Africa as an ornamental as it appeared in the 1858 catalogue of the Cape Town Botanical Gardens. However, the commercial potential of black wattle was soon realized by John van der Plank who commenced the development of plantations in KwaZulu-Natal in 1864, primarily for the production of tannins, which are extracted from the bark. Black wattle bark is a rich source (31-51% dry weight of bark) of water-soluble tannins, primarily 3,7,3',4,'5'-pentahydroxy-2-phenyl chrom, which is used for tanning leather and the manufacture of water resistant resins or adhesives for reconstituted wood products. Black wattle wood is also an important export product of South Africa with the majority being processed as pulpwood for the production of paper and paperboard products. In 2001, South Africa exported 1.2 million tonnes of black wattle wood product worth around R360 million (US$31.5 million) from 130,000 ha of managed plantations centred in the provinces of Mpumalanga and KwaZulu-Natal in northeast South Africa, and from black wattle control programmes. In 1998, the industry directly employed between 10,800 and 13,000 people, mostly unskilled labourers. In addition to this, black wattle is a source of firewood utilized for cooking and heating in lower income rural communities where it is also used for informal housing and building construction.

Environmental Impact

Despite the economic virtues of black wattle in South Africa, the tree has serious environmental impacts, which are reflected in its status as a weed of national importance. Black wattle occurs on 2.5 million hectares in South Africa mostly in the southern and eastern sectors of the country where mean annual rainfall exceeds 500 mm. The invaded area is the equivalent of 131,000 ha of condensed infestation and this is expected to increase at 5-10% per annum without proactive intervention. The main negative impacts associated with black wattle are the reduction in surface stream flow with a net present value cost of R16,285 million (US$1425 million) (based on an annual water consumption of 577 million m3), loss of biodiversity, and increased soil erosion and destabilization of river banks. Black wattle has invaded grassland, fynbos, savanna and forest biomes in South Africa and is considered a threat to the species-rich Cape Floral Kingdom and many of the biodiversity `hot spots' of southern Africa. The potential for species' reduction and loss is therefore substantial. The key ecological traits that contribute to the success of black wattle are its ability to develop large soil-stored banks of long-lived seeds that are triggered to germinate en masse by fire (a characteristic that is shared by many other Australian acacias) and the development of a large, structurally-dominating crown.

Biological Control

The vast scale of black wattle invasion in South Africa, coupled with the species negative impacts, led in 1973 to the initiation of a biological control programme that targeted the seeds of black wattle. Organisms that could develop or feed on vegetative parts of the plant were not considered because of the direct negative impact these may have on the black wattle industry. However, resistance to research on seed-reducing agents for black wattle mounted over the subsequent years. The industry challenged the validity of the research programme by questioning the status of black wattle as a serious invader, which had not been adequately documented at the time. In addition, growers were concerned about the possible depletion of seeds within plantations that are required for natural regeneration of crops, and the protection of seed orchards used for mass production of seed from selected tree stock. The research effort shifted as a result of this conflict and focussed on the biological control of other invasive, mostly non-commercial Australian acacias, particularly A. longifolia, A. melanoxylon, A. cyclops, A. saligna and the related Paraserianthes lophantha (Mimo-saceae). As a result of this research, all these species are now considered to be under satisfactory, or partial, biological control in South Africa using seed-reducing insects and a host-specific gall-forming rust fungus. However, in 1987 the research programme found itself again in conflict with the black wattle industry when a seed-feeding curculionid, Melanterius servulus, proposed for the biological control of P. lophantha, was found to feed on seeds of black wattle in laboratory tests. The black wattle industry opposed the release of this insect from quarantine due to concerns of potential damage the insect may cause to black wattle seed supplies. The research programme was suspended, but then recommenced following pressure by environmentalist and farmers affected by P. lophantha. A negotiated agreement was achieved stipulating that releases of M. servulus could be made on P. lophantha providing it could be demonstrated that black wattle seed orchards could be protected with insecticides. Synthetic pyrethroids registered for use in wattle plantations were found to cause high mortality in the field on the analogous M. ventralis, which develops in seeds of A. longifolia. As in vitro tests showed M. ventralis and M. servulus had similar mortality responses to these insecticides, the wattle industry accepted that seed orchards could be protected, should the need arise. The beetle was formally approved for release in 1989 on the condition that releases were confined to the Cape Peninsula in the extreme southwest of the country, until it could be confirmed that M. servulus would not attack black wattle under field conditions. Subsequent field surveys showed that M. servulus does not use black wattle as a host and that feeding on this plant in laboratory tests was an artefact of confined, non-choice test conditions.

Conflict Resolution

Resolution of the potential conflict by a negotiated agreement, in retrospect, was the most pragmatic approach that could have been taken on this issue. Negotiations were founded on basic trust and the ability of both sides to acknowledge and understand each other's concerns. The eventual outcome was satisfactory to both sides of the conflict, but the fact that it took about 25 years to resolve the issue using an adequately represented industry and biological control research has been rightly criticized. In South Africa today, biological control research is governed by a process of public consultation and liaison with affected parties. In the case of black wattle biological control, a steering committee with industry representatives, researchers and other relevant organizations or individuals, share in information transfer. Potential conflicts of interest are identified in the early stages. This process has now been formalized as a mandatory procedure under the National Environmental Management Act 1998 (No. 107). Although at times the legislative process is administratively clumsy and inefficient, it enacts the principle of freedom of information and equitable consultation and is therefore a progressive move.

Melanterius Seed Weevil

The resolution of the conflict of interest associated with the release of M. servulus in South Africa opened the opportunity to re-commence the search for seed-reducing agents for black wattle. A small, univoltine beetle, M. maculatus, that feeds on developing seeds can be locally common on black wattle and closely related Acacia spp. (section Botrycephalae) in Australia. The beetle was approved for release in 1993 after it was proven that non-target species, particularly indigenous African acacias, were not accepted as hosts. Mounting public concerns over the environmental impact and continued spread of black wattle in South Africa, coupled with the knowledge that Melanterius could be controlled in seed orchards facilitated the process of approval for release. Once again, release approval was conditional, in that beetles would only be distributed in the Western Cape, to allow the wattle industry time to prepare for future incursions of the beetle into plantation areas. The insects have become abundant in the vicinity of their original release sites, but their natural dispersal has been slow. The beetle has not been detected anywhere near the commercial wattle growing regions of South Africa.

Gall-Forming Fly

Black wattle is a difficult target for biological control. The species inhabits a broad eco-climatic range in South Africa, has a high fecundity, sometimes aseasonal flowering and fruiting, and an early sexual maturation period. These attributes make it unlikely that satisfactory control will be achieved by any single seed-reducing species. Although M. maculatus is causing increasing levels of seed damage, sufficient seed reduction across the full range of black wattle in South Africa will probably only be achieved by the additive effect of compatible seed-reducing insects. We are examining the potential role of a gall-forming cecidomyiid midge as an additional seed-reducing agent for black wattle. Gall-forming insects have been spectacularly successful in reducing seed loads on A. longifolia and A. pycnantha in South Africa by committing the plants to allocate host resources to gall formation at the expense of fruits. This `forced commitment' acts as a resource sink and vegetative performance of infected trees is reduced in the process. Clearly, this indirect impact would not be accepted on black wattle in South Africa, and indeed other countries where the species is economically important. However, an undescribed cecidomyiid, Dasineura sp. has been discovered that forms flower galls (2-3 mm) by causing the ovary to swell soon after oviposition and form small, multi-chambered galls. Fruit production is then prevented. A low biomass of galls, despite high infection levels, together with a shorter activity period compared to fruit development, releases the host tree from resource commitment and could have beneficial consequences on vegetative growth. The cecidomyiid appears to have sensitive host-finding abilities and as adults are readily dispersed by wind, offers the potential for effective seed-reduction of black wattle in South Africa. As with most gall-forming cecidomyiids, the insect has a narrow host range and presents no threat to African Acacia, which belong to different subgenera from Australian Acacia.

Protecting Seed Orchards

The issue of protecting black wattle seed orchards from attack by the cecidomyiid is paramount to the debate on whether this insect could be approved for release in South Africa. Trials utilizing a range of insecticide formulations on trees in Australia will be undertaken to resolve this problem. Insecticide spraying for the suppression of M. maculatus or Dasineura sp. in black wattle seed orchards will involve an additional cost to the industry. Indigenous pests such as foliage-feeding Lepidoptera can warrant pesticide application in plantations, but outbreaks are sporadic and annual applications are mostly not necessary. Most wattle growers in South Africa (75%) regenerate trees following harvesting using line-seeding, where treated seeds are drilled into the soil, especially in colder areas, or use nursery-raised seedlings from improved tree lines. About 1000 ha of seed orchard services this market in South Africa and would require annual protection from seed-reducing biological control agents should these establish within seed production areas. All but 18 ha of this area is dedicated to the collection of seed for line-seeding that yields around 3 t of seed per annum. Melanterius maculatus and Dasineura sp. have a period of overlapping adult activity between September and October and insecticides applied to the tree canopy during this time should limit the impact of both insects. A single application of insecticide is likely to be sufficient for control of Melanterius, but several may be necessary for the Dasineura, if the total potential seed crop is to be protected, as adults emerge from the soil during the entire flowering period of black wattle. Systemic organophosphate insecticides may have a greater role in the protection of trees from Dasineura than foliar-applied synthetic pyrethroids. The projected cost to the wattle industry for the protection of 1000 ha of seed producing plantations from both Melanterius and Dasineura sp. is between R115,000-R300,000 (US$10,000-26,000) per annum, based on the aerial application of cypermethrin at R110/ha. This cost could be reduced with improved line-seeding techniques, which are considered wasteful of seed, or the phasing out of line-seeding in favour of replanting with nursery-raised seedlings. This has been the trend in many areas of South Africa, and the implementation of biological control in wattle growing areas could accelerate this rate of change.

Natural Regeneration of Wattle Crops

Wattle growers that rely on natural seedling regeneration as a method of crop re-establishment (25%) are unlikely to require protective insecticide sprays. Even in the presence of biological control agents, low numbers of seed will be produced that will accumulate in the soil over the 10-year crop rotation period. A single application of synthetic pyrethroid during a prolific flowering season would certainly guarantee seed supply for the next crop. Also, several silvicultural practices will reduce the impact of biological control agents. After clear-felling, there is a 2- to 3-year sexual maturation period for flower production and a 3- to 4-year period for fruit production. During this time, populations of Melanterius and Dasineura will become locally extinct, and recolonization must occur from neighbouring sources at the onset of flowering or fruiting. The rate of re-establishment of seed-reducing insects will depend on the size of founder populations and their intrinsic rate of increase. Very little is known about the latter for both Melanterius and Dasineura. Geographic and weather variables will influence the size of founder populations with distance from the neighbouring source, the nature of barriers between sites, and the direction of the prevailing wind being most important. Plantations that are isolated are likely to experience slow colonization by Melanterius and Dasineura and seed production in the first few years of the establishing crop should be close to normal. Alternatively, grouping two or three consecutive seasons' plantings together within a mosaic of non-Acacia plantation species could reduce migration rates and lower the impact of the insects on subsequent seed crops. As black wattle is often grown amongst blocks of Eucalyptus and Pinus, only minor changes to farm planning would be required to implement this model.

The `Sterile Tree'

Emerging technology is very likely to have an important impact on the way biological control of black wattle seed is managed in South Africa. In an innovative initiative sponsored by the wattle industry, a breeding programme has commenced to produce a `sterile tree' through manipulation of the tree's genetic composition using gamma irradiation, and the formation of triploid trees. If successful, the sterile tree will produce no flowers or seeds, and if adopted by the wattle industry, biological control could be practised in South Africa with minimal conflict of interest. In this situation, the benefits of the industry could be preserved while allowing wild black wattle trees to be freely suppressed using biological control. It is this situation that is expected to positively change the cost:benefit ratio of black wattle in South Africa from a current undesirable 0.4 to 4.3. If this were achieved, the divided perceptions of black wattle in South Africa would change to that of a generally welcomed `guest' with great utilitarian value. Also, effective implementation of biological control of black wattle seed in South Africa may greatly enhance the industry's success in seeking approval to expand plantation areas to meet market demand for its products.

Black wattle is grown commercially in many other countries and is reported to have escaped cultivation in Tanzania, Zimbabwe, Swaziland, India, Madagascar, Hawaii, La Réunion, Brazil and New Zealand. The impact of naturalized black wattle in these areas is largely unknown, but given the experience in South Africa, it is likely to be considerable or has the potential to become so. Introduction of seed-reducing biological control agents of black wattle to these countries, particularly in the early stages of invasion, is likely to result in massive long-term cost savings.

By: Robin Adair, Agricultural Research Council, Plant Protection Research Institute, Private Bag X5017, Stellenbosch 7599, South Africa
Email:
Fax: +27 21 883 3285

The author thanks S. Neser, B van Wilgen and J. Hoffmann for comments on an earlier version of this article.

❑

Harmonizing Weed Biocontrol in Australia

Over the coming issues of BNI you will be reading about a range of biological control of weeds projects that are underway in Australia with CSIRO (the Commonwealth Scientific and Industrial Research Organisation). [See the next two articles, and also Training News, this issue.] Here we give some background to these initiatives, and outline the development of the strategy of cooperative action that lies behind them.

Weed control is a problem that is faced by most farmers no matter what agricultural production regime they belong to. There would be few farmers in Australia, or indeed around the world, who could claim to be running weed free properties, whether they are raising ostriches or growing trees.

In Australia weeds are the most widespread problem faced by growers, much more so than salinity, for example. Weeds do not discriminate between natural environments or manipulated production systems, nor do they stop at the farm fence or any border. Realizing this is an important step in weed control, as is the realization that weed issues cannot be addressed one species at a time using only one technique. A holistic approach to weed management is required.

One of the initiatives that is helping to bring together a range of weed control options in Australia is the National Weed Strategy and the relevant strategic plans associated with the 20 Weeds of National Significance (WONS).

As well as thinking about weeds in a holistic manner, organizations involved in weed management need to work together. A major step towards greater collaboration was taken in 1995 when the Cooperative Research Centre (CRC) for Weed Management Systems was established and this has continued with the commencement of the new CRC for Australian Weed Management in July 2001. The original Weeds CRC formally brought together leading organizations involved in weed control and was a catalyst in the integration of various weed management practices. The new Weeds CRC takes this collaboration one step further and incorporates northern Australia, allowing a truly national approach to weed management.

One of the options in weed management is biological control, the use of naturally occurring invertebrates (such as insects) or pathogens (fungi) to control plants that have become weeds. Biological control does not eradicate a weed, but if successful it can restore Nature's balance to a point where the weed is no longer of economic importance. Currently it is estimated that agricultural weeds in Australia cost more than Au$3.3 billion per annum, so reducing that economic impact by even a small fraction would be a major achievement. For many years, CSIRO Entomology has been involved in the biological control of a range of agricultural and environmental weeds. Some of these weeds are on the WONS list as part of the National Weed Strategy.

There are currently a number of agricultural weeds that are targets for biological control by CSIRO. For temperate Australia, projects exist for Paterson's curse (Echium plantagineum), Onopordum thistles, nodding thistle (Carduus nutans), Scotch broom (Cytisus scoparius), Emex and blue heliotrope (Heliotropium amplexicaule). New projects have recently commenced on wild radish (Raphanus raphanistrum) and serrated tussock (Nasella trichotoma), although it will be some years before agents are released and established on these latter weeds. For environmental weeds, agents have been released on bridal creeper (Asparagus asparagoides), bitou bush and boneseed (Chrysanthemoides monilifera spp.), horehound (Marrubium vulgare), blackberry (Rubus fruticosus), St John's wort (Hypericum perforatum) and Scotch broom with work commencing on Montpellier broom (Genista monspessulana).

In tropical Australia agents have been released on mimosa (Mimosa pigra), parkinsonia (Parkinsonia aculeata), mesquite (Prosopis spp.) and sida (Sida acuta) and research is being conducted into control for Mexican poppy (Argemone mexicana), Hyptis and bellyache bush (Jatropha gossypiifolia).

The impact from some of these agents is only just starting to be felt because biological control is a long-term option. Some of the target weeds have been in Australia for over 100 years, and it would be blithely optimistic to expect the agents to reverse a century of damage in the space of a year or two. For instance, for most weeds there is a huge build up of long-lived seed in the soil and this stock of seeds must be depleted before any kind of control can be affected. In some cases this amounts to hundreds of thousands of seeds per square metre of soil, all waiting to germinate and all capable of surviving for up to 10 years or more.

As organizations work more closely together, biological control is increasingly being looked at as an important component in overall weed management, rather than as a last resort when all else has failed. Research into the integration of biocontrol agents with other management practices is allowing Weeds CRC researchers to develop best practice management guides that will ultimately provide landowners with a package of information that they can implement on their properties.

Contact: Kate Smith, CSIRO Entomology, GPO Box 1700, Canberra, ACT 2601, Australia
Email:
Fax +61 2 6246 4177

❑

Rollers Released on Australian Coastal Weed

Bitou bush (Chrysanthemoides monilifera rotundata) is a major conservation weed of coastal southeastern Australia and through the National Weeds Strategy is listed as one of Australia's 20 Weeds of National Significance. [See also BNI 20(4), 108N (December 1999) `Speedy seed fly'.]

One more biological weapon has now been added to the assault on this South African conservation menace and it comes in the form of a moth. The leaf-rolling moth (Tortrix sp.) is the most damaging insect feeding on bitou bush in its homeland, South Africa, and is the sixth biological control agent to be released on this weed along the New South Wales (NSW) coast.

Bitou bush was first recorded in NSW in the early 1900s and from 1946 through until 1968 it was deliberately planted along the NSW coast to aid control of erosion and post-mining rehabilitation. It was so successful in this role that it continued to invade coastal habitats in southeastern Queensland, NSW and Lord Howe Island. Bitou bush is particularly prevalent on the central and north coasts of NSW and the total area infested in Australia is now estimated to be over 70,000 ha.

CSIRO Entomology, in collaboration with the Cooperative Research Centre for Australian Weed Management and NSW Agriculture, has attracted funding through the National Heritage Trust to allow the rearing, release and evaluation of the leaf-rolling moth. The moth was first released near Grafton NSW in 2001 and subsequent releases have been made along the NSW coast from Moruya in the south to the Queensland border.

The life cycle of the leaf-rolling moth takes about 8 weeks from egg to adult, depending on the season. Eggs hatch after about 8 days and the larvae move to the shoot tips where they begin to feed. The larval stage lasts about 30 days and during this time the larvae feed on leaves, stems and surfaces of young shoots resulting in death of shoot tips. High larval populations in summer, when the insect is most active, may severely defoliate, weaken or kill plants. The larvae then pupate for around 10 days and then live as adults for about 14 days during which time they mate, lay eggs and the cycle begins again.

Biological control agents, like the previously released tip moth and seed fly, as well as the leaf-rolling moth, complement each other and increase pressure on bitou bush, making it less competitive. Biological control is a long-term strategy for control and is most often best used with a combination of other control methods. The best combination to achieve control is often site specific, but may include herbicide, manual removal or fire.

Once the weed has been removed from an area it is important to ensure that another weed does not take its place. Revegetation of heavily infested areas has a major role to play in the prevention of further problems.

Contact: Kate Smith, CSIRO Entomology, GPO Box 1700, Canberra, ACT 2601, Australia
Email:
Fax +61 2 6246 4177

❑

Australia's Two-Step Strategy for Blue Heliotrope

The mauve coloured flowers of the deadly South American blue heliotrope (Heliotropium amplexicaule) infest thousands of hectares in eastern Australia. Originally introduced in the 19th century as an ornamental garden flower, blue heliotrope has now spread from southern Queensland as far south as the Victorian border and into South Australia. It is poisonous to stock, causing liver damage that can result in loss of condition and often death.

The first biological control agent, the leaf-feeding beetle Deuterocampta quadrijuga, was released on 21 November 2001. The release of this beetle is due to the efforts of the Blue Heliotrope Action Committee of northern NSW, who continued to seek support for biological control and were successful in gaining funding through the Rural Industry Research and Development Corporation (RIRDC).

In the early 1990s CSIRO conducted surveys in South America that identified some insect species as potential biological control agents. However, at the time blue heliotrope did not attract sufficient industry funding, so no further work was done and the plant continued to spread. After a long delay CSIRO was able to begin in earnest in 1998. Professor Miguel Zapater of the University of Buenos Aires, Argentina came on board to study the biology of a number of the potential agents that were identified in the earlier study. As a result, D. quadrijuga and Longitarsus sp., a flea-beetle whose larvae feed on the roots, were prioritized for the biological control of blue heliotrope.

In March 2000, CSIRO received the first batch of D. quadrijuga eggs. The larvae and adults of the beetle feed on leaf tissue and can cause complete defoliation of the weed. They were reared in quarantine and spent a year and a half being tested to ensure that they would be safe to release into the Australian environment. Tests showed that the beetles only attacked the South American blue heliotrope and did not pose a risk to non-target plant species, including native Australian Heliotropium species.

Based on these results, in July 2001 Australian Plant Biosecurity authorities approved the release of the beetle for the biological control of blue heliotrope. Efforts are currently underway to obtain funding for host-specificity testing of the flea-beetle and eventual redistribution of these agents throughout the infested areas of southeastern Australia.

It is hoped that the planning and evaluation carried out thus far will bring the desired results of curtailing the unchecked spread of blue heliotrope and limiting its impact in the areas in which it has gained a presence.

Contact: Kate Smith, CSIRO Entomology, GPO Box 1700, Canberra, ACT 2601, Australia
Email:
Fax +61 2 6246 4177

❑

Cocktail for a Sharpshooter

The vineyards of California are under threat from an old enemy given a new lease of life by a new ally, the glassy-winged sharp-shooter (GWSS), Homalodisca coagulata. Pierce's disease, a serious malady of grapes (caused by the xylem inhabiting bacterium Xylella fastidiosa) was first identified in California over a century ago. The bacterium produces xanthan gum, which blocks the xylem vessels. The leaves of diseased plants typically develop drying or scorching symptoms, and the vines become unproductive and usually die within 1-2 years of infection. GWSS is a xylem-feeding insect that readily acquires and transmits X. fastidiosa. Currently, there is no known cure for eliminating the disease from infected vines.

In rapid response to this deadly alliance, an inter-disciplinary collaborative effort in-volving the US Department of Agriculture - Agricultural Research Service (USDA-ARS) and Animal and Plant Health Inspection Service (USDA-APHIS), the California Department of Food and Agriculture (CDFA), County-based Co-operative Extension Personnel, the University of California (UC) and industry and private organizations planned and launched a multi-pronged attack to simultaneously manage the threat posed by the GWSS-Xylella combination. Its goals are to contain the sharpshooter's spread, and at the same time develop a cocktail of control and curative measures to protect the wine industry in the south of the state from further devastation. A meeting held in San Diego, California in December 2001 provided an opportunity to review progress.

Until recently, Pierce's disease was spread by native sharpshooters, principally the blue-green sharpshooter (Graphocephala atropunctata). These are poor fliers and prefer other plants in preference to grapes for feeding. However, even G. atropunctata was able to bring about the destruction of the Orange County wine industry in the late 19th century. During the 1880s Pierce's disease decimated more than 16,000 ha of grapes in the Anaheim area. The incurable disease has appeared on and off ever since, but its spread was limited. Farmers in most parts of the state were able to control it by pruning infested branches, grubbing out infected vines, and replanting.

The status quo was shaken by the arrival in California of H. coagulata, a cicadellid native to the southeastern USA. First identified in the state in Ventura County in 1990, it began wreaking havoc in 1999 when the first disease outbreak occurred in the vineyards of Riverside County's Temecula region. Much larger, vagile, and more robust than native sharpshooters, it spreads the disease much more efficiently. Moreover, it has a recorded host range of over 70 species, and thrives in urban and rural environments. As well as Pierce's disease, it can transmit diseases such as oleander leaf scorch (in California), phoney peach disease (in the southern USA), almond leaf scorch and alfalfa dwarf. In California, GWSS has two generations per year and overwinters as adults. It can fly distances over 400 m and up to heights of 8 m, frequently appears in high numbers, and survives winter temperatures dipping to −6.5°C in citrus orchards. It spread throughout the wine producing areas of southern California, and into the south of the San Joaquin Valley threatening the table grape industry there. Current losses have been estimated at US$14 million dollars-worth of damage to the wine industry in just a few years. GWSS has now established in San José and threatens California's premier wine producing areas of Napa, Sonoma, and Mendocino counties.

A state-wide management programme includes survey activities to determine the distribution of GWSS in California and detect new infestations, and regulatory activities to prevent artificial spread to uninfested at-risk areas. In this context, UC Cooperative Extension (UCCE) has launched an information campaign to alert growers and enlist their assistance in monitoring for the disease.

Biological control is seen as the key ingredient in an IPM solution. One native egg parasitoid, Gonatocerus ashmeadi, is already abundant in California, but is primarily effective during the summer. Researchers at UC Riverside are evaluating another species from the same genus, G. triguttatus, which attacks earlier in the season, and could potentially depress numbers of first generation sharpshooters. In cooperation with CDFA, they have released G. triguttatus reared from stock imported from Texas and Mexico in vineyards and citrus groves in Riverside, Ventura, and Tulare counties, and are now waiting to see whether this species can survive the Californian winter. Additionally, UC Riverside researchers are looking at the preferences of parasitoids for GWSS of different ages to assist with mass rearing efforts of these natural enemies. The outcomes of competition between different species of egg parasitoids for GWSS egg masses of different ages are being studied to determine if new natural enemy additions to California will be complimentary or antagonistic to GWSS control. One other area that is being investigated is the importance of flowers and other sugar sources for helping increase the longevity and fecundity of GWSS parasitoids in citrus orchards and vineyards. This strategy may be particularly important for helping parasitoids survive through the winter to attack the spring generation of GWSS eggs. The spring GWSS egg population currently suffers from low levels of parasitism (around 30-60%) by G. ashmeadi. Summer levels of GWSS egg parasitism by G. ashmeadi are much higher, often exceeding 95%.

Scientists at the USDA-ARS Beneficial Insects Research Unit in Weslaco, Texas are looking at egg parasitoids found in south Texas, Louisiana and northeastern Mexico. Surveys in south Texas during 2001 showed that 86% of GWSS egg masses were attacked by the parasitic wasp, G. triguttatus. This species will continue to be released in California during 2002. Gonatocerus fasciatus is one species being targeted for importation into California from Louisiana this year. Natural enemies in the home range of GWSS are thought to be at least partly responsible for the low densities of sharpshooters in these regions. At the same time USDA is conducting DNA analysis on the Californian sharp-shooters, and comparing results with populations from elsewhere to try and determine the source of origin of the invasion. The results may help delineate where the most effective biocontrol agents might be found.

Exploration for sharpshooters and more natural enemies, though, is already being conducted in South and Central America. At the USDA-ARS South American Biological Control Laboratory in Argen-tina, scientists are surveying and collecting parasitoids from the eggs of the South American sharpshooter, Tapajosa rubrimarginata, collected from areas of Argentina with subclimates similar to those found in the grape-growing regions of California. Over a dozen species of egg parasitoids have been collected. Shipments to US quarantine for evaluation against GWSS are expected to begin in 2002. Parasitoids or other sharpshooter natural enemies from similar subclimates might outperform natural enemies imported from different climates. Chile, also, has some good climate matches and is being included in new explorations for new biological control agents.

Even good biological control will not provide a complete answer to the GWSS problem, as small numbers of insects can still transmit disease and wreak havoc. (Consequently, there is not a viable grape industry in areas of the USA where GWSS is native.) Biocontrol will therefore be just one part of the GWSS management strategy, and other possibilities including chemicals (insecticides and bactericides), cultural control (barriers), breeding programmes (traditional and transgenic), and monitoring strategies (for GWSS and Xylella) are being investigated.

Plant resistance is a good complementary tool for biocontrol, but currently there are no commercial vines resistant to Pierce's disease. Research is focused on investigating the molecular mechanisms of susceptibility to X. fastidiosa, and also on investigating mechanisms of resistance in wild grapevine relatives that do not develop the disease. With this information, classical breeding or biotechnology could be employed to create a vine more resistant to Xylella.

Making vines physically unattractive to the sharpshooter is another approach. USDA-ARS scientists in West Virginia have found that a coating of white kaolin particles makes them inhospitable. Field trials coordinated by USDA-APHIS in vineyards, some bordering citrus orchards, in Kern County, California gave encouraging results. Three treatments applied from mid-March to mid-April resulted in sharp-shooter numbers lower than those found on insecticide-treated vines. In addition, the kaolin treatments were cheaper than six insecticide treatments. It could provide a promising non-toxic early-season alternative to insecticides. It may not be suitable for use once the vines have bloomed, as visible white kaolin residues on grapes, although harmless, do affect wine quality and would probably be unacceptable to consumers.

Innovative biotechnology may allow the causal agent of Pierce's disease to be targeted. Armed with the full genome sequence of X. fastidiosa, scientists at UC Riverside are studying some 100 genes that, if removed, could render the bacterium harmless by preventing transmission or infection. To be successful, the modified bacterium would need to be able to outcompete the wild-type disease-causing form. An alternative strategy is to use another a bacterium or virus to kill Xylella. This would entail using bioengineering techniques to modify an antagonist from the gut of the sharpshooter to produce enzymes lethal to X. fastidiosa. Infected sharpshooter populations could potentially pass this on from generation to generation. A candidate bacterium has been shown to reproduce in the sharpshooter gut. Now a suitable lethal gene is being sought, which would need to be inserted into the bacterium, and then application technology would have to be developed, and of course rigorous safety standards would have to be met before any genetically modified organism could be released. Another potential method for disarming X. fastidiosa is to prevent it producing the xanthan gum, the substance that damages and eventually kills the plant. Some bacteria are known to break down xanthan, and the search is on for one that does this inside diseased grapevines. Although these tools are at an early stage of development, they could be a powerful boost for control options.

The wide host range of GWSS causes particular problems. A variety of common weeds can carry X. fastidiosa, and this reinforces the importance of good weed control to prevent disease spread. In addition, fava bean, a common cover crop in vineyards in northern California, can also host the bacterium, and legume crops are now contra-indicated as crops in vineyards.

The sharpshooter is especially problematic where citrus and vines are cultivated in close proximity. Although GWSS can transmit disease in citrus, these pathogens are not found in California, and the sharpshooters feed and reproduce in citrus without causing significant damage. They tend to overwinter in the citrus trees' protective foliage and move out to colonize vineyards in spring. Trials coordinated by UC Davis Extension Service staff in Kern County demonstrated that contact insecticide applied to citrus in winter reduces adult numbers, while a follow-up systemic treatment in spring targets the wingless nymphs, which are confined to the plant they hatch on. Thus fewer survive to move onto the vines when they mature. In separate trials conducted by the USDA-ARS Western Cotton Research Laboratory in Arizona and UC Riverside researchers, pyrethroids and neonicotinoids gave fastest knock-down of the pest, with one pyrethroid compound giving 100% kill in 6 hours. Compound residues from both classes also continued to give good knockdown beyond 28 days. However, resistance development is a major concern with sole reliance on insecticides for GWSS management.

An additional problem with this approach, however, is that significant areas of citrus are under IPM for primary pests or under organic production. Therefore, considerable grower resistance to paying for and using insecticides for a pest that is not a problem in citrus is being experienced and disruption of stable IPM programmes caused by broad spectrum pesticide use is resulting. GWSS control where organic or IPM citrus borders grapevines is thus particularly difficult and controversial. Although biocontrol by egg parasitoids and other compatible measures are expected to alleviate the problem to some extent, the introduced parasitoids will take time to have an impact in citrus, wilderness and residential areas. In the meantime, cage trials to test the efficacy of augmentative releases of commercially produced green lacewing larvae (Crysoperla rufilabris) conducted by CDFA scientists in the southern San Joaquin Valley gave promising results. Further trials are planned with other commercially available predators. Results of trials in organic lemons have reduced sharpshooter numbers, although not as much as in conventionally treated citrus.

USDA is also researching alternatives to chemical pesticides for killing the eggs without affecting natural enemies.

Contact: Mark S. Hoddle, Department of Entomology, University of California, Riverside, CA 92521, USA
Email:
Fax: +1 909 787 3086

Walker Jones, USDA-ARS, Kika de la Garza Subtropical Agricultural Research Center, Beneficial Insects Research Unit, 2413 East Highway 83, Weslaco, TX 78596, USA
Email:
Fax: +1 956 969 4888

❑

New Legislation Benefits Weed Biocontrol in South Africa

Biological control, by definition, does not result in the complete eradication of the target weed. The continuous presence of remnant populations of the target weed is a prerequisite of sustainable biocontrol; yet it is this aspect that often brings biocontrol into conflict with weeds legislation, because legislation usually requires the total and immediate eradication of declared or noxious weeds.

Until recently, this has also been the situation in South Africa, where biological control has not always been recognized by law as a viable, long-term control measure for alien invasive plants. Several potentially successful biocontrol projects have been impeded by the insistence of the relevant authorities that herbicides or mechanical control, often in an attempted 'eradication', be applied as soon as the populations of biocontrol agents go through a temporary depression that is part of their natural population cycle. Other biocontrol projects never got off the ground because the authorities were not prepared to risk the phasing out of the herbicidal regime in favour of biological control. Stands of declared weeds under complete biological control have frequently been cleared chemically or mechanically by order of the authorities or out of ignorance, resulting in the loss of valuable biocontrol material.

The recent amendment of the relevant legislation in South Africa has gone a long way to rectify this situation. During March 2001, regulations 15 and 16 of the Conservation of Agricultural Resources Act (CARA) were drastically revised. Despite a much tougher stand on weeds, as reflected by the substantial increase in the number of plant species that are now either prohibited or regulated, some imaginative changes in the regulations have significantly improved the prospects for biological control.

In the original 1983 version of CARA, harmful plants were divided into 46 species of declared weeds (which had to be eradicated) and 35 species of invader plants (which needed to be controlled in rural areas only, if they were threatening any agricultural resource). Of these invader plants, ten species were alien plants, while the other 25 were indigenous species. By making no distinction between alien invaders and indigenous species that became unnaturally abundant only as a result of inappropriate management practices, several valuable indigenous species were placed under unnecessary 'suspicion' in well-managed areas in which they were not causing problems. No provision was made for dealing with plant species that were clearly harmful in certain situations, yet were valuable to part of the population.

The amended CARA regulations separate alien problem plants (discussed under regulation 15) from indigenous ones (dealt with under regulation 16), differentiate between three categories of alien problem plants, and are unique in making special legal provision for biological control.

Classification of Problem Plants by the CARA Amendments

Bush encroachment indicators. The indigenous problem plants (44 species) are now called 'indicators of bush encroachment'. The amended regulations display an understanding of the ecology of bush encroachment by advising land users to take extra care in areas characterized by the listed species. In addition, they prescribe ecologically sound management practices aimed at preventing bush encroachment and at combating it where it already occurs.

Declared weeds and invader plants. Alien problem species are subdivided into three categories, based on their current utilization (or lack thereof) in South Africa.

• The greatest proportion of alien problem plants (124 species) are those that have no function to fulfil in South Africa or whose harmfulness outweighs any useful properties they might have. These were placed into Category 1, and are called 'declared weeds'. This denotes that they are prohibited and must be controlled wherever they occur. The only exception is that they may grow in biological control reserves (discussed later).
• A much smaller group (35 species) consists of invasive plants that nevertheless have certain beneficial properties that warrant their continued presence in certain circumstances. These were placed into Category 2, and (together with Category 3 plants) are called 'invader plants'. They have commercial, structural or utility value as timber, fruit or medicinal plants, soil stabilizers, wind breaks or sources of fodder, fuel or building material. Land users will be allowed to cultivate or retain these species on condition that they take the necessary responsibility for the water use and seed pollution by these trees. Permission has to be obtained from the Executive Officer, who will demarcate a specific area for the growing of a particular Category 2 plant species. Category 2 plants growing anywhere except in such demarcated areas or in biological control reserves will be regarded as declared weeds, and will have to be controlled. Category 2 plants may be sold only to, and purchased only by the users of areas demarcated for the particular species.
• Another group of 'invader plants', placed into Category 3 (39 species), consists mainly of popular ornamental plants. It was considered unnecessarily harsh to prohibit them outright, and instead a ban was placed on all new plantings. The existing specimens of Category 3 plants, except those growing in or near watercourses or wetlands, may be retained until they die naturally. In cases where these plants threaten any agricultural resource, the Executive Officer has the authority to enforce additional measures. No further trade in Category 3 plants will be allowed, plants may no longer be propagated and all seedlings must be controlled.

Another feature of the recent amendments to the CARA legislation is that certain sterile or less invasive forms of plants from Categories 1 to 3 may be grown and sold legally. Examples include the sterile double-flowered cultivars of Nerium oleander, the sterile form of Lantana montevidensis, all spineless cactus pear cultivars and selection of Opuntia ficusindica, the sterile cultivar `Rubrum' of Pennisetum setaceum, and all cultivars of Pyracantha angustifolia. Similarly, several plant species were placed into different categories in different provinces, based on the climate in which they are likely to become invasive. Examples are Acacia dealbata, which is a declared weed in the Western Cape Province and a Category 2 invader plant in the rest of the country; Ardisia crenata, which is a declared weed only in the Northern Province, KwaZulu-Natal and Mpumalanga, and Schinus terebinthifolius, which is a declared weed in KwaZulu-Natal and a Category 3 invader plant in the rest of the country. Most of these exemptions were negotiated in consultation with the nursery industry.

Implications for Biological Control

Control of weeds and invaders instead of eradication. The amended CARA regulations do not insist on the eradication of all declared weeds or invader plants that grow in areas where they are forbidden by law, but only on their control. CARA defines 'control' as: "the combating of plants by means of the prescribed methods" (including biological control), "to the extent necessary to prevent or to contain the occurrence, establishment, growth, multiplication, propagation, regeneration and spreading of such plants".

This formulation allows scope for biological control. Seed-destroying insects have the potential to prevent or contain the establishment and spreading of the target weeds, e.g. the snoutbeetles Melanterius spp. that destroy up to 100% of seeds produced by several Australian Acacia spp., aided in certain cases by inflorescence-galling wasps (Trichilogaster spp.); the flowerbud-feeder Trichapion lativentre together with the seed-feeder Rhyssomatus marginatus that almost obliterate seed production in Sesbania punicea and the combination of a fruit weevil, Erytenna consputa and a seed moth, Carposina autologa, which together destroy the vast majority of newly produced seeds of Hakea sericea, as well as seeds accumulated on the plant. Natural enemies that weaken their target plant by attacking its vegetative growth can be said to contain the plant's growth, propagation and regeneration. Examples are the 'nutrient sink' effect of the gall wasps (Trichilogaster spp.) mentioned above, and the leaf, bark and flower feeding beetles Leptinotarsa texana and L. defecta that reduce the photosynthetic surface and deplete the nutrient reserves in the rhizomes of Solanum elaeagnifolium, thus containing its growth and vegetative reproduction, and preventing fruit set by feeding on the flowers. In all the cases mentioned above, effective biological control complies with CARA's definition of control.

Recognition and safeguarding of biological control. The CARA regulations prescribe the following control methods, as long as they are appropriate for the particular species and ecosystem:

• uprooting, felling, cutting or burning
• treatment with registered herbicides
• biological control carried out in accordance with the stipulations of all the relevant Acts
• any other method recognized by the Executive Officer
• a combination of the above, except that biocontrol reserves and areas where biocontrol agents are effective, shall not be disturbed by other control methods to the extent that the agents are destroyed or become ineffective

Today this is probably the only weeds legislation that specifically mentions biological control as an acceptable control method.

CARA goes even a step further by protecting biological control from interference by other control measures. An example of such interference, which would now be regarded as illegal, was the enforced herbicidal treatment of Opuntia aurantiaca in areas where the cochineal Dactylopius australis was effectively controlling the cactus. This used to happen frequently in the Eastern Cape Province, especially while the cochineal populations were at the bottom of a natural curve, but still managed to keep the weed population below the economic damage threshold. A state subsidy on the registered herbicide, MSMA, encouraged land users even further to use herbicides instead of biological control. Nevertheless, the herbicide campaign against this weed was a failure, despite being the country's most expensive weed control project. Where biological control was given an opportunity to reach its full potential, O. aurantiaca is no longer a problem. The same applies to Hypericum perforatum in the Western Cape Province.

Biological control reserves. Biocontrol researchers need protected areas in which to monitor the establishment and performance of newly-released biocontrol agents, even though the agents cannot yet be expected to exercise effective control of their target weeds. Protected areas for biocontrol agents also serve the purpose of letting the agents build up their numbers to allow their collection and mass-release into new areas, or to enable the agents to colonize recently cleared surrounding areas as their target weed regenerates. Mindful of these requirements, CARA provides for the designation of biological control reserves.

Biocontrol reserves will be designated by the Executive Officer, on application in writing by a biocontrol researcher or practitioner ("biological control expert"). It will be illegal to destroy the biocontrol agents in such a reserve, or to do anything that will reduce their efficacy. This stipulation implies that a mechanism has to be created by which all persons involved in the control of alien invasive plants are informed of the whereabouts of biological control reserves.

Biocontrol as an Instrument for Resolving Conflicts of Interest

Biological control in the form of seed-destroying agents has the potential to restrict the invasive potential of a formerly invasive plant species without interfering with its utilization. This is illustrated by the use of the curculionid, Melanterius acaciae to destroy the seeds of the valuable Australian timber species, Acacia melanoxylon, without affecting the quality of the wood. Different Melanterius spp. have been released to reduce the invasive potential of several Australian Acacia spp. in South Africa [see also 'Black wattle: South Africa manages conflict of interest, this issue]. Similarly, the bruchids, Algarobius prosopis and Neltumius arizonensis, which destroy the embryos in the seeds of Prosopis spp. without significantly reducing the nutritional value of the pods, allow the continued utilization of this valuable shade tree and source of stock feed. In this case, the efficacy of seed destruction depends on the measures that are taken to keep the pods away from the livestock until the beetles have had the opportunity to complete their life cycle in the seeds.

The availability of biological control as a conflict resolving tool was one of the factors that has made it feasible for CARA to allow persons to continue farming with invasive plant species, such as Acacia mearnsii, Acacia melanoxylon, Prosopis spp. and several Pinus species. Category 2 plants may be grown only in legally demarcated areas and under controlled conditions. One of the few practicable ways for the land user to comply with the requirement of preventing or restricting the spreading of the invader plant from the demarcated area is by releasing host-specific seed-destroying natural enemies. Without this option, the present Category 2 plants would probably have had to be included under Category 1 (declared weeds). This step would have caused such an outcry by the present users of these plant species that Government would most likely have withdrawn all the species concerned from legislation altogether, leaving them to invade the country unchecked.

Seed-destroying biocontrol agents also have a function to fulfil with regard to Category 3 invader plants (mainly ornamental trees). Despite the general ban on further planting of Category 3 plants, a clause was included whereby the Executive Officer may grant exemption from this stipulation, amongst others. Such an exemption will only be granted if the Executive Officer is satisfied that the risk of invasion was minimal, e.g. if the climate is unsuitable for seedlings to survive, or if seed production is negligible. The city planners of Pretoria and Johannesburg might invoke this exemption clause to obtain permission to continue planting jacaranda trees along their roads on condition that a host-specific seed-feeding insect be introduced and released against this popular street tree.

Conclusion

The amended CARA regulations might be unpopular because of the inclusion of such a large number of popular plant species: every farmer, forester, landscaper and nursery owner - in fact almost every landowner - is now at risk of transgressing the law, albeit inadvertently. With regard to biological control of alien invasive plants, however, the amendments create an awareness of this control method amongst the authorities, promote its use and its integration into management strategies, and safeguard the valuable agents that have been released into the field. The country is sure to reap the benefits of this forethought in the form of improved, cost-effective and sustainable control of invading alien plants.

Further Reading

These three key publications give background information on the weed programmes referred to in this article.

Hoffmann, J.R. (ed) (1991) Biological control of weeds in South Africa. Agriculture, Ecosystems and Environment, Special issue, 37(1-3), 255 pp.

Olckers, T. & Hill, M. P. (eds) (1999) African Entomology Memoir No. 1. Biological Control of Weeds in South Africa (1990-1999), 182 pp.

Olckers, T.; Zimmermann, H.G.; Hoffmann, J.H. (1998) Integrating biological control into the management of alien invasive weeds in South Africa. Pesticide Outlook 9(6), 9-16.

See also the two PPRI handbooks described in `New Books', this issue.

By: Hildegard Klein, ARC-Plant Pro-tection Research Institute, Private Bag X134, Pretoria 0001, South Africa
Email:
Fax: +27 12 3293278

❑

Portuguese Leave British Standing!

Which would you least like to have to work with: an invasive impenetrable thicket-forming, intensely prickly shrub, or a minute, almost immobile insect control agent? Spare a thought for scientists involved in gorse (Ulex europaeus) biocontrol in New Zealand, then, for they have both to contend with. The gorse thrips (Sericothrips staphylinus) imported from the UK and released in New Zealand some 10 years ago has the dubious honour of being named "slowest moving biocontrol agent of all time" by Landcare Research.

As this 1- to 2-mm-long, usually wingless insect had not dispersed far beyond its original release sites, it was clear to the scientists involved that it would need a fairly massive redistribution exercise if the control programme were to succeed. Not entirely surprisingly, they found the public reluctant to help - not just because of the gorse prickles, but also because the size and cryptic nature of the insect made people less than confident in dealing with it.

However, dialogue with colleagues working on gorse biocontrol in Hawaii suggested that the British thrips was a bit of a non-starter. Hawaii had imported both British and Portuguese strains of S. staphylinus, and while they had also found the British strain to be slow-moving or worse, the Portuguese insects dispersed with encouraging rapidity. For example, one 6000-ha area of gorse was completely infested within 6 years.

Portuguese thrips from Hawaii were imported into quarantine in New Zealand, and were being reared for releases planned for early 2002. No one is quite clear why the Portuguese insects move faster. It may be that more develop wings... or could it be that the Portuguese have more of an explorer's sense of adventure than their stay-at-home British relatives!

Contact: Hugh Gourlay, Landcare Research, Lincoln, New Zealand
Email:

❑