TEAMwork against Leafy Spurge
Biological control programmes for leafy spurge (Euphorbia esula complex) in North America have made significant progress since BNI's last report on the problematic Eurasian perennial [BNI 18(3), 65N (September 1997)].
Although 15 biological control agents for leafy spurge have been released in the USA since 1963, two in particular are well established throughout the continental 'spurge belt' in the northern Great Plains (North and South Dakota, Montana, Wyoming and the Prairie Provinces of Canada). The two European flea beetles - Aphthona nigriscutis and Aphthona lacertosa, first released in the USA in 1988 and 1994, respectively - have established large populations at numerous locations across a wide geographic area, and are now providing significant levels of leafy spurge control. Perhaps more importantly, this build-up of numbers and sites is allowing for easy collection and redistribution of flea beetles, which further promotes the concept of biological control.
The past 2 years have been especially exciting for those involved with leafy spurge management. Although reductions in spurge densities due to flea beetle activity have been documented at many locations for several years, never before have there been so many dramatic reductions spread across such a wide area. At some sites, extremely heavy infestations were completely eliminated. Biocontrol worked so well in some places, in fact, that collection efforts were hampered: people returned to sites that had historically produced large numbers of flea beetles to discover that both the spurge and flea beetles were gone. This phenomenon can be expected to become more common as flea beetle populations increase and more sites are established, and it emphasizes the need to educate people about properly managing their biocontrol resource.
A major development in leafy spurge management occurred in 1997, when the US Department of Agriculture (USDA) Agricultural Research Service joined forces with a sister agency, the USDA Animal & Plant Health Inspection Service, to create 'TEAM Leafy Spurge' (TLS).
This biologically based Integrated Pest Management (IPM) programme focuses on researching and demonstrating effective, affordable and ecologically sustainable leafy spurge management techniques. The cooperative programme stresses teamwork, resulting in a vast network of partnerships between the two USDA agencies, land grant universities and numerous other local, state and federal entities. Fields of study include (but are not limited to) classical biological control, multi-species grazing, herbicide use, cultural controls (such as tillage, reseeding and burning), the integration of various control tools, socio-economic impacts, the application of GPS/GIS and hyperspectral imaging technologies to weed management, and foreign exploration to investigate new leafy spurge biocontrol agents.
Data collected by TLS programme participants in the past 3 years are especially promising, and suggest that widespread control of leafy spurge with biologically based IPM strategies is no longer a question of 'if' but 'when'. This is a brief summary of some TLS research efforts:
Flea beetle establishment has improved dramatically the past 3-5 years, primarily because people now know more about using biocontrol and have access to large numbers of insects. A few years ago, releases of 100-500 flea beetles were common, with roughly one-third of all releases successfully establishing a population; releases of 3000 flea beetles are now considered the minimum, and establishment success has more than doubled. In an effort to quantify flea beetle establishment, population expansion and the resultant impact on leafy spurge, TLS established 264 'inventory and assessment' sites in 1998. Each site was inventoried - i.e. extensive data regarding soil type, moisture, topography, species composition, etc. were collected - prior to being seeded with 6000 Aphthona flea beetles (3000 A. lacertosa and 3000 A. nigriscutis). Data collected from these sites in the summer of 2000 document an establishment success rate of 85-95%, a seven-fold increase in flea beetle numbers and spurge canopy cover reductions in the range 40-95%. In addition, grass production and species richness (diversity) increased by averages of 47% and 27%, respectively.
TLS is demonstrating that a combination of multi-species grazing and biocontrol works well as an effective spurge management tool. The concept is simple: use sheep to graze heavy, dense patches of spurge, thus giving flea beetles increased opportunities to establish populations in the resultant thinner stands of spurge. For this demonstration, TLS selected a site that originally featured an extremely dense, widespread spurge infestation (more than 50% of the total demonstration area was infested with spurge exceeding 200 stems/m2). In just 3 years, the combination of multi-species grazing and biocontrol has reduced spurge densities by 31-50%; native vegetation and desirable grasses are responding by reestablishing in areas formerly occupied by spurge. Based on previous research, significant reductions are expected to occur in the 4th and 5th years (i.e. 2001 and 2002) of the demonstration. The demonstration clearly shows the economic and environmental advantages offered by combining the two biologically based IPM strategies.
Herbicide use is declining in areas where biologically based IPM programmes have been implemented. Ward County, located in northwestern North Dakota, provides an excellent example. Ten years ago, weed managers there invested considerable resources (both fiscal and personnel) into spraying 8000 of the 12,000 acres [3250/4850 ha] infested with spurge within the county. Last year, after just 3 years of aggressive biocontrol efforts, the county had 9500 acres [3850 ha] of spurge (a 20% reduction in total acreage) and used herbicides on just 400 acres [160 ha] (a 95% reduction). Similar examples can be expected from across the region as weed managers learn how to use biological control as a stand-alone tool and in combination with other tools.
In addition to creating cooperative partnerships and funding various research and demonstration projects, TEAM Leafy Spurge also stresses information and education. Although a variety of tools (press releases, magazine articles, pamphlets, CDs, etc.) are used to increase awareness of the problem and potential solutions, TLS has discovered that low tech, grass-roots efforts such as field day events are extremely valuable in regard to teaching people about biocontrol and IPM.
The programme's information and education efforts will be highlighted at 'Spurgefest II', set for June 2001. The 3-day event will include a leafy spurge symposium, tours of TLS research and demonstration sites, and demonstrations of flea beetle collection and redistribution techniques. Its precursor, 'Spurgefest '99', was wildly successful, capturing a considerable amount of media interest across the northern Great Plains and drawing participants from across the USA and Canada.
By: Steve Merritt, TEAM Leafy Spurge technology transfer
specialist, USDA-ARS Northern Plains Agricultural Research Laboratory,
1500 N. Central Ave., Sidney, MT 59270, USA
Weevils' Success against Canadian Rangeland Weeds
Progress on projects involving the classical biocontrol of two rangeland weeds, Dalmatian toadflax (Linaria dalmatica) and houndstongue (Cynoglossum officinale), was significant in 2000. Both projects involve success with relatively recent introductions of European insects, which were screened by the CABI Bioscience Centre in Switzerland with funding from Canadian and US sponsors.
This perennial with the bright yellow, snapdragon-shaped flowers is an invasive weed of grasslands, open forests, roadsides and waste areas of western North America. Originally introduced as an ornamental from eastern Europe in the 1900s, Dalmatian toadflax has since become a serious problem in Canada, particularly in the southern interior of British Columbia (BC) and contiguous areas of southwest Alberta, where it currently infests thousands of hectares of range and forest lands. An extensive root system and strong, early spring growth allows Dalmatian toadflax to outcompete surrounding rangeland vegetation. The weed also is a prolific seed producer, thus contributing to its continued spread and invasion into new areas. Cattle and wildlife generally avoid grazing on Dalmatian toadflax. Cattlemen in BC have listed the weed as their third priority for control after the knapweeds (Centaurea spp.) and houndstongue, because of the loss in grazing potential of toadflax infested lands.
The options for control of Dalmatian toadflax are few. Chemical treatment is uneconomical, difficult because of the plant's deep roots and waxy leaves, and potentially environmentally damaging when applied to large weed stands on grasslands. Furthermore, habitats where toadflax often grows, such as coarse-textured soils or near water, restrict the use of effective herbicides (e.g. picloram). Mechanical control (i.e. hand-pulling or mowing) also has not proven feasible. Biological control is considered the best long-term control option.
A European stem-boring weevil, Mecinus janthinus, first released in 1991 in Canada against Dalmatian toadflax, has established well and has produced major attack on the weed in southern BC. Monitoring of several 5- to 6-year-old release sites in 2000 revealed 100% attack of toadflax shoots for at least 50 m around the original release points. In some instances, this level of attack was evident for several hundred metres. Most of the damage is attributed to spring feeding on shoot tips by mass-emerging adults, thus causing significant stunting and a complete loss of flowering on reproductive shoots. The insect's ability to cause such impressive damage to toadflax was not predicted from earlier European studies. Although it has been too soon to detect actual reductions in toadflax density at most sites, we continue to monitor permanent plots to document any changes in vegetation.
Houndstongue is a biennial or short-lived perennial weed of mountainous rangelands in northwestern North America. Originally from Eurasia, the weed is thought to have been accidentally introduced in the 1800s and has since spread considerably. The weed thrives particularly in forest openings created through logging or mining activities, sometimes forming dense monocultures in these habitats.
In British Columbia houndstongue is a major concern to cattlemen because it hinders establishment of forage on new pastures and its barbed seeds or 'burrs' attach to cattle, causing irritation, potential reductions in auction price of animals, and a negative impact on the rancher's reputation. Houndstongue also is highly toxic to livestock, possessing pyrrolizidine alkaloid levels which are much higher than those found in another toxic range weed, tansy ragwort (Senecio jacobaea). Normally, livestock avoid feeding on green houndstongue, but problems may arise once the plant dies back in late summer or fall, or if it happens to get into hay.
As with Dalmatian toadflax, options for houndstongue control are limited, hence, a biocontrol programme was initiated in 1988 with European exploration for potential agents. The root weevil, Mogulones cruciger, was the first agent to be released in Canada after 9 years of host specificity testing. Since the initial releases in 1997, the weevil has established at all sites so far (i.e. 67 by the end of 1999) and already is showing evidence of reducing houndstongue density at the oldest sites. Monitoring of target and any potential nontarget impact will be of upcoming emphasis in our programme.
By: Rose De Clerck-Floate,
Azolla Biocontrol in South Africa
Azolla filiculoides or red water fern is a free floating aquatic fern native to South America. It was first recorded in South Africa in 1948. For many years the fern was confined to small streams and farm dams in the Colesburg area in the centre of the country. However, the lack of natural enemies and the presence of enriched waters have contributed to its inevitable spread by man, waterfowl and floods to many sites around the country. By 1998 the weed had been recorded at some 176 sites.
The increasing abundance of A. filiculoides in conservation, agricultural, recreational and suburban areas over the last 10 years is cause for great concern. Among the major consequences of the dense mats (5-30 cm thick) of the weed on still and slow-moving water bodies in South Africa are: reduced quality of drinking water caused by bad odour, colour and turbidity; increase in waterborne, water-based and water-related diseases; increased siltation of rivers and dams; reduced water surface for recreation (fishing, swimming and water-skiing) and water transport; deterioration of aquatic biodiversity; clogging of irrigation pumps; drowning of livestock; and reduced water flow in irrigation canals.
In South Africa, no herbicides are registered for use on red water fern, and this, in conjunction with the fact that its rapid rate of increase renders manual and mechanical control ineffective, suggested that biological control was the most suitable option for this weed. The frond-feeding weevil Stenopelmus rufinasus was collected from Azolla caroliniana in Florida (USA) and imported into quarantine in South Africa in late 1995. Following host specificity testing, the weevil was released in December 1997. The first release, of 900 weevils, was made on a one-hectare dam in a bird sanctuary in Pretoria, which was 100% covered by the weed. By February 1998 (2 months later) the red water fern mat had collapsed, and some 30,000 weevils were reared from one 2 m2 sample of decaying material.
To date the weevils have been released (usually in batches of 100 adults) at some 110 sites throughout South Africa. The information that we have on these sites is that the weevil has been responsible for clearing 72 of them completely. For the remaining 38 sites, either the weed has been washed away during flooding, or we have not revisited them, or it is too early to tell. All of the sites that have cleared have done so in less than one year. In addition to this, the weevils have migrated to other sites, sometimes up to 80 km away from the release site. We are uncertain if the weevils have been transported on weed by waterfowl, or if there has been short distance hop dispersal onto other dams of the weed, or if it is as a result of long-range dispersal by the adults. At around 40% of the sites the weed has returned up to 2 years after the initial clearing. Interestingly the weevils have located 90% of these and the weed is under control.
A thorough cost benefit analysis of this project has been completed and is in the process of being published; the preliminary benefit to cost ratio is 1200:1.
There are several interesting aspects to this project. Firstly, this appears to be a new association: although A. filiculoides has been recorded as native to the southwestern USA, this weevil is associated with A. caroliniana in Florida. Secondly the speed with which weevils were able to control even the largest mats in the most eutrophied waters. Thirdly, the weevils have been able to locate red water fern mats up to 18 months after the original mats had collapsed. It remains to be seen if the current level of control will be sustained over the long-term.
Contact: Andrew McConnachie,
Search for Rhinocyllus conicus references in recent literature, and you are likely to come up with a host of publications discussing the nontarget effects of this introduced weevil on native thistles in North America. However, down in New Zealand, it is a different story, for there the weevil has itself become the subject of nontarget attack.
The braconid Microctonus aethiopoides was introduced to New Zealand in 1982 as a biocontrol agent for the weevil Sitona discoides on lucerne, and has been useful in reducing damage in this important crop. Now, however, it has been found attacking R. conicus in some agricultural environments. Rhinocyllus conicus was released in New Zealand as a biocontrol agent for nodding thistle (Carduus nutans), and in conjunction with other agents often provides excellent control. The results of preliminary studies suggest that M. aethiopoides is likely to have limited impact on R. conicus in the field, but this is not the first instance of this parasitoid exhibiting a catholic taste for hosts in New Zealand. It has previously been recorded from three other weevil species in the field here (Nonotus albicans and two Irenimus spp.) with parasitism rates of 40% or more.
The nontarget effects recorded with Microctonus aethiopoides contrast with another species of the genus, M. hyperodae, introduced more recently for control of another weevil pest of lucerne, Listronotus bonariensis. This second parasitoid has so far proved to be restricted to its target host. The difference reflects at least in part the increase in awareness of environmental risk that has occurred in recent years: greater rigour in testing procedures meant that the narrow host range of M. hyperodae was predicted with a good degree of reliability.
Sources: Watch out, Microctonus is about.
Florida Fast Food
Our 'Florida Stripper' [BNI 21(3), 57N (September 2000)], the melaleuca weevil Oxyops vitiosa, has developed a taste for an artificial diet. This does not mean it will be invading hamburger bars, but it does make it much easier to rear. Ultimately, the success of a biocontrol programme relies on both having appropriate agents, and being able to mass rear and release them. This rearing technique includes both a diet and a pupation medium. Developed by USDA scientists at the ARS Invasive Plant Research Laboratory in Fort Lauderdale, it gives a real boost to the programme for biocontrol of Melaleuca quinquenervia in Florida.
The rearing technique has two key features: first, it contains the right balance of nutrients (including sucrose, glucose, maize starch, vitamins and minerals) and feeding stimulants (M. quinquenervia leaf extracts) to support the complete development of the insect from egg to adult; second, a substrate is described that provides a relatively high rate of pupation. Although larval development requires nothing more than the right nutrients, the pupae are more choosy. The substrate that works best included a mixture of sand and water with an absorbent material such as crushed florist's foam or peat moss; such a mixture retains enough moisture and allows enough air exchange for successful pupal development. Artificially reared weevils have already been released at sites infested with melaleuca in south Florida and appear to be performing well.
Contact: Gregory S. Wheeler,
Mist Flower Flagging?
Mist flower (Ageratina riparia) spreads at a phenomenal rate, and in New Zealand the public has been engaged to help to monitor this. From the results of a questionnaire, researchers at Landcare Research have drawn up a map of its current distribution and, by identifying the environmental conditions in which it grows, have also drawn up a 'worst-case' map showing the areas it could potentially invade. Presently most common in the north of North Island, it could spread to other parts of North Island and the top of South Island. However, with the help of two exotic biocontrol agents with a track record of success in Hawaii, its spread may soon be slowed if not halted.
The white smut Entyloma ageratinae was the first agent to be released against mist flower in New Zealand [BNI 20(4), 107N (December 1999)], and now an insect agent has also been approved for release by the Environmental Risk Management Authority. A shipment of the gall fly Procecidochares alani has been imported from Hawaii, and releases are expected to begin this summer. It is hoped the insect will complement the striking success of the pathogen.
The smut, which was released in November 1998, has been spreading even faster than its mist flower host. It has been recorded up to 56 kilometres from release sites in the north of North Island. In the Waitakere Ranges, the spread has been so effective that by December 1999 no uninfected areas could be found for studies into long-term vegetation changes. Even more unexpectedly, an intrepid pathologist who ventured out to Great Barrier Island to release the smut found it had beaten her there. Although it may have arrived on the wind (the nearest release site is on Waiheke Island, 77 km away), Jane Fröhlich suggests it is equally likely to have been taken there by an unwitting human carrier.
There are also encouraging signs that the fungus is causing significant damage to the weed. At the nine sites where the fungus was released, 30-50% of the mature leaves last summer were destroyed in the first wave of attack, and 30-90% of the regrowth that season met a similar fate. Monitoring will continue this summer. However, there are signs that the defoliation at release sites is already allowing the recovery of some native plants, including orchids and ferns, that had been choked out by mist flower.
Sources: History sometimes repeats. Lincoln, New
Zealand; Manaaki Whenua - Landcare Research. Patua te otaota - Weed
Contact: Jane Fröhlich, Landcare Research, Private Bag
92170, Mt Albert, Auckland, New Zealand
Rust for Bridal Creeper
Bridal creeper (Asparagus asparagoides) is one of Australia's most damaging and persistent environmental weeds, and a search for effective control is well underway [BNI 20(4), 108N-109N (December 1999)]. In the last 2 years two defoliating biocontrol agents have been released against it by researchers from CSIRO (Commonwealth Scientific and Industrial Research Organisation) and the Cooperative Research Centre (CRC) for Weed Management Systems.
The leafhopper Zygina sp. was released in May 1999 across southern Australia. It successfully oversummered at the majority of sites. At two research sites the insect has spread 300 m from the release site after 18 months. Schools and community groups are involved in rearing and release of the insect. It has now been released at over 200 sites, and is beginning to impact on the weed at many sites, where the onset of senescence is advanced by several weeks. Studies are underway to monitor reserve accumulation in the tubers.
Leafhopper releases were followed in 2000 by releases of the rust Puccinia myrsiphylli. There are high hopes for this pathogen, as it causes considerable damage to the creeper in its native South Africa. It was released at 50 sites across all states of southern Australia from mid July until the plants began to senesce with the onset of the hot summer weather in November. The rust established well at most sites, and the epidemic developed steadily until the onset of senescence. Spread, though, was slow from the artificially inoculated release areas. At one site, the rust spread about 30 m within 4 months of release, while at another site along the side of the road it spread as far as 100 m. It caused premature defoliation of the cladodes at the release points at these two sites. The slowness of spread may be related to the microclimate, which is characterized by very little wind movement because the weed tends to grow below other shrubs. However, teliospores (sexual spores) have been produced in abundance. This autumn (February/March) researchers will be able to see if and how the rust reappears at the release sites, and field trials have been set up to monitor the spread and epidemic development of the rust over the next 2 years.
Contacts: Louise Morin, CRC for Weed Management Systems,
CSIRO Entomology, GPO Box 1700, Canberra, ACT 2601, Australia
Tim Woodburn, CRC for Weed Management Systems, CSIRO
Entomology, Western Australian Laboratory, Private Bag 5,
Mimosa invisa: a Growing Menace in South India
Mimosa invisa, a noxious fast growing invasive weed of South American origin has recently emerged as a new threat to natural forests, forest plantations and agricultural systems in Kerala State in southwest India. The weed is an annual but wherever water is available year-round it can also grow as a biennial. Mimosa invisa has a scrambling stem bearing four or five rows of sharp prickles. Leaves are small, bipinnate and fluorescent green in colour. The inflorescence is a condensed spike (capitate) which is pinkish in colour. The weed produces a large number of seeds which have a long period of viability. It is heliophytic in adaptation and cannot grow well under closed canopy. Mimosa invisa is moderately drought resistant. The fact that it can invade and cover the ground completely, competing with other plants, smothering herbaceous growth implies habitat degradation and loss of biodiversity. Mimosa invisa growing areas are impenetrable because of the characteristic thick growth and the stem being armed with sharp prickles. It is known to be toxic to cattle.
A preliminary survey conducted in Kerala indicated that M. invisa is widespread in the central and southern parts of the State. Of the 52 sites surveyed in seven districts in the State, over 50% were heavily infested. In central Kerala, natural moist deciduous forests are heavily infested. In evergreen and semi-evergreen forests infestation is seen only on the fringes where canopy is partially or fully open. Among forest plantations, teak is seriously affected. The weed grows luxuriantly on roadsides and fallow lands. Mimosa invisa is found to overgrow and smother Mikania micrantha in most parts of Kerala. It appears that at the present rate of growth and spread, M. invisa may even exceed Mikania micrantha in posing serious threat to natural forests, forest plantations and agricultural systems in Kerala. The potential combined impact of these two weeds cannot be over-estimated.
Mechanical control of M. invisa by manual weeding is difficult and labour intensive. Moreover, the weed can sprout vigorously from the cut base soon after the onset of monsoon. An attempt to control it using 2,4 D and dinitrobutyl phenol (denoseb) in Brazil was not very successful. According to Rachel McFadyen, biological control using insect enemies is showing success in Queensland (Australia), and attempts to control this weed through biological means are highly warranted in India.
By: K. V. Sankaran,
DNA Fingerprinting: Pointing the Way
Finding biocontrol agents, whether for old or new pests, may superficially seem a straightforward business, but in practice it is very complex. Factors that make it so include (a) incomplete taxonomy of both pests and natural enemies; (b) wide native distribution of pests; and (c) limited information on associations between natural enemies and the target and its related species. Unravelling which are the best natural enemies to introduce, and how many species are needed is a time-consuming and therefore costly process, yet projects are generally severely limited by both. Projects also have a habit of coming to an untimely end: once the agents are out, it can be hard to persuade funding agencies that their progress needs to be monitored.
Molecular biologists have developed a range of techniques, based on multiplying up small variations in DNA to detectable levels, to allow them to distinguish between species, and between populations of the same species. An account of some of these techniques is given in a previous news item1, and more are being developed. They form a powerful battery of investigative tools with a lot to contribute if applied in a targeted intervention way in combination with other biocontrol research methods. Here we look at what kind of information DNA analysis can provide (drawing particularly on work with whiteflies). We see how this can inform decision-making in biocontrol projects. At this time, we are on a learning curve, and the usefulness and limitations of DNA techniques are still becoming apparent. However as they begin to be more widely understood, they will be used more often and more effectively.
In the last ten years, the whitefly Bemisia tabaci has become a serious worldwide pest, and one of the foremost whitefly vectors in field and glasshouse crops. It is as a vector that it poses the most severe threat, as it takes very few individuals to infect and devastate a field with viral disease (many of which are little studied as yet). A particularly polyphagous, fecund and hardy population of B. tabaci was identified in the USA in the early 1990s, which proved also to be a vector for a new arsenal of viruses. It spread rapidly through the New World, and to Europe, Australia, Southeast Asia and the Pacific. It was designated B. tabaci biotype-B based on esterase profiles (while the B. tabaci population native to the southwestern USA and northwestern Mexico was designated biotype-A). In 1994, biotype-B was given species status as B. argentifolii on the basis of allozyme and RAPD-PCR (random amplified polymorphic DNA - polymerase chain reaction) analysis, and viral transmission, morphological and mating studies. However, the status of the biotype/species has remained contentious, as indeed has that of B. tabaci as a whole. Many studies of many kinds have indicated wide variation in numerous parameters between different B. tabaci populations, and possibly more than would be expected within many species, yet it has not been possible to establish a reliable phylogeny. Recent DNA analyses (PCR amplification and sequencing of two mitochondrial markers (portions of the 16S ribosomal subunit and cytochrome oxidase I genes, COI) and one of the ribosomal internal transcribed spacers (ITS1)) for B. tabaci from different locations around the world have provided the first molecular evidence of important genetic divergence between geographically isolated populations2,3. The authors argue that it is more realistic to consider B. tabaci as a whole as a species complex, for any alternative would require the description of a separate species for each unique population.
The taxonomy of whitefly parasitoids provides another challenge. Some of the problems were outlined by Andrew Polaszek in an earlier news article4. A number of teams are now using DNA techniques to clarify the taxonomy of some groups. RAPD-PCR markers were first developed to help sort out the large number of foreign collections made as part of an interagency whitefly programme (involving the US Department of Agriculture, Agricultural Research Service (USDA-ARS) and Animal and Plant Health Inspection Service (USDA-APHIS), state agencies and university researchers). The USDA-APHIS team based at Mission, Texas has conducted studies using classical taxonomy backed up by RAPD-PCR analysis of Encarsia and Eretmocerus parasitoids from around the world5,6. Work is also being conducted by USDA-ARS researchers at Fargo, North Dakota, who are using satellite DNA sequences as species-specific markers. Satellite DNA contains highly repetitive non-coding sequences that accumulate mutations quickly as compared with other parts of the genome. This has the potential to give better information about geneflow between populations and to distinguish strains in rapidly evolving systems.
Conveniently for this morphologically indistinct group, RAPD banding patterns of Eretmocerus and Encarsia species differ. These unique patterns were found, by comparing them with minute species differences in antenna morphology, to correspond directly to distinct species within the two genera. In this instance, RAPD is a useful quarantine tool for making preliminary separations of geographic populations or strains of Eretmocerus, and possibly new species. However, there is not always agreement between traditional taxonomic and DNA techniques. RAPD analysis at the Mission laboratory consistently distinguished between several geographically isolated collections of Encarsia sophia (= E. transvena) that could not be distinguished by morphological methods.
DNA sequencing (of the D2 expansion region of the 28s rDNA gene) of Encarsia species has been carried out by Chris Babcock & John Heraty (University of California at Riverside, USA). Sequencing (of the ITS 1 and 2, mitochondrial CO1 and D2 and D3 regions) of Encarsia and Eretmocerus parasitoids of B. tabaci and Trialeurodes vaporariorum in Australia has been carried out by Paul De Barro & Felice Driver (CSIRO Entomology, Australia). The aim of these studies is to successfully characterize several species and species-groups of these parasitoids, as well as contribute to phylogeny reconstruction. Their studies have shown a strong relationship between morphological characters and molecular phylogenetic structure. This work is continuing at Imperial College, UK (Shahab Manzari, Andrew Polaszek, Robert Belshaw & Donald Quicke), focusing especially on the Encarsia inaron species group. As with the Mission work, sequence data have already shown the presence of undescribed species in the inaron-group which are morphologically extremely difficult to distinguish.
Contacts: Paul De Barro, Project Leader,
John A. Goolsby,
DNA analysis has potential for helping in one of the great conundrums of biogeography and classical biocontrol - where species originated and how they have moved, naturally or otherwise, around the globe. By analysing the genetic variation in populations of a pest in its indigenous range and comparing them to samples from its introduced range, it is often possible to pinpoint where the introduced material originated. For an introduced pest with a very large home range, this is probably a good place to start the search for a biocontrol agent, as classical theory tells us that this is where effective co-evolved natural enemies are most likely to be found.
The area of origin of the target pest is not always easy to pinpoint, particularly for pandemic species. For example, cypress aphid (Cinara cupressivora) was introduced to Africa in the late 1980s, with immediate and devastating impact on plantation and smallholder trees throughout the continent. Classical biological control was deemed to have the best chance of providing an effective long-term solution, but there was a catch. Cypress trees are grown all over the world, in subtropical and temperate climates, and exactly where the pest in Africa had come from was a mystery. To add to the confusion, the identity of the species in Africa was unclear (indeed, originally thought to be Cinara cupressi, it was later described as a new species). Beginning in 1992, CABI and collaborating scientists from many countries searched exhaustively (and exhaustedly) on cypress trees throughout North and Central America, and from the Atlantic coast of western Europe, through Central Europe, the Mediterranean, North Africa and the Middle East to Pakistan. Eventually, the probable source was tracked down to Syria, to cypress trees nestling below the ruins of castles built during the Crusades many hundreds of years ago, which proved to be the centre of origin of C. cupressivora (although it is now also found in the southern Mediterranean region). Doubtless, the castle walls have sheltered many a weary traveller in their time, but probably never before a footsore exploratory entomologist. For the search had taken 5 years, and during this time, although the biological control efforts had continued in tandem with the searches, they were hampered by the uncertainties. DNA analysis would not have obviated the need for all the exploration, nor for climate matching that directed the searches, nor for the morphometric studies that finally clarified the taxonomy. But by using DNA techniques in combination with conventional ones, not only might taxonomic uncertainties be resolved more easily, but promising areas to focus on for natural enemy exploration may be pinpointed more quickly.
Taking up the Bemisia tabaci story again, we can see how. Although, based on species and natural enemy diversity, the centre of origin of B. tabaci was believed to be either the Indian subcontinent or the Middle East/Africa, there was no indication where biotype-B had sprung from. Host plant affiliations suggested perhaps an origin somewhere in the Eastern Hemisphere, and this was supported by preliminary DNA sequence analysis (of the mitochondrial 16S ribosomal gene). Recent studies using DNA techniques have made a significant contribution to unravelling the convoluted biogeographical relationships within the species and of its parasitoids. In two of these2,3, putting ITS1, 16S and COI profiles into analytical models gave consistent results that distinguished New and Old World populations, and separated Old World populations into a number of groups, including Africa; Sahel and Sahel-like regions; Australia; and several distinct clades in Asia. Biotype-B was found to fit with the 'Sahel' group, and this provided the first definitive molecular evidence to support the hypothesis that biotype-B is an introduction into the New World from the Old World. It further suggested an area of origin in the eastern Mediterranean/North Africa. This fitted with features of its ecology, and with observations that only populations from the `Sahel' group produced the phytotoxic symptoms associated with biotype-B in the New World. Further, the molecular data supports the hypothesis that the unique physiological changes induced through the feeding of biotype-B is a recently evolved trait.
In another study5, DNA analysis and classical taxonomy were demonstrated to be a powerful combination. They helped to establish the identities and predict the putative centres of diversity of both B. tabaci hosts and their parasitoids, using material from southern Spain, Thailand and Texas (a target site for the biocontrol programme). Morphological analysis indicated the Spanish whiteflies to be a non-biotype-B, and those from Texas to be biotype-B, while those from Thailand were identified as 'B. tabaci species complex'. Molecular classification (based on sequencing of the mitochondrial COI gene) also suggested that the Spanish material was a distinct local variant. However, comparison with similar analyses of other B. tabaci populations indicated that it fitted into a clade containing reference B-biotypes from several locations, including Texas. Molecular identification of the Thailand whiteflies placed them in a Far East-Southeast Asian clade. The Eretmocerus data from the Australian study3 suggests that the species of effective Eretmocerus species from Spain, Pakistan, the Middle East and Australia are all very closely related. These results can be used in combination with other criteria such as climate matching to make decisions on priorities for surveys and testing of potential agents.
The area of origin of natural enemies may be far from apparent. It is not uncommon for a natural enemy to appear serendipitously, having been introduced to a new area along with the pest. The diamondback moth (DBM), Plutella xylostella, is the most important pest of crucifers worldwide. It is notorious for developing pesticide resistance and, therefore, biological control has been widely and successfully deployed in highland growing areas in Asia. The most important parasitoid is an introduced ichneumonid wasp of European origin, Diadegma semiclausum. This parasitoid is now the target of a new biocontrol attempt in East Africa.
Earlier efforts by the GTZ/ICIPE (Gesellschaft für Technische Zusammenarbeit/International Centre for Insect Physiology and Ecology) IPM Horticulture Project in cooperation with a number of national research organizations in eastern and southern Africa - and collections by others - have consistently yielded an apparently indigenous parasitoid identified as D. semiclausum. However, parasitization rates were very much lower (<15%) than the ones reported from Asia. Owing to this, and as there is no record of an introduction to this region of Africa, doubt always persisted about the correctness of the identification. Now, molecular taxonomic methods are being employed to solve the problem. The African material has recently been declared identical with Diadegma mollipla, a potato tuber moth (Phthorimaea operculella) parasitoid7. Preliminary molecular work confirms this separation from D. semiclausum: material collected from two populations in Kenya and one each from Ethiopia and Tanzania can be distinguished from the D. semiclausum introduced to Taiwan by their mitochondrial DNA.
Molecular taxonomy will also be used to help keep track of the establishment and spread of D. semiclausum and its impact on the local species once an introduction has been made. For this purpose, a simple and cheap yet safe method will need to be developed as thousands of samples from all over the region will have to be screened. [Progress in this field with other pest species is described below.] Finally, as the taxonomy of the genus Diadegma is so notoriously difficult, the project will also attempt to make a contribution towards the molecular classification of species.
Bernhard Löhr and Barbara Wagener are requesting fresh material from anywhere in the world for inclusion in this study.
It is important not to allow the results of DNA analysis alone dictate the path of a biocontrol project. AFLP (amplified fragment length polymorphism) is proving a particularly useful technique to distinguish between populations of the same species, and give a measure of the relatedness of them. It was used to analyse the relatedness of weed populations in a CABI Bioscience/Kerala Forest Research Institute (KFRI) project (funded by the UK Department for International Development, DFID) to develop a biocontrol strategy for Mikania micrantha (mile-a-minute) in India. Mikania micrantha is a Neotropical weed with an indigenous range extending from Mexico to Paraguay. In India, it is an invasive weed of tropical moist forest regions. It is a serious pest of tree crop/agroforestry systems in the Western Ghats and is spreading through Kerala towards Karnataka. It is also having a severe impact on tea production in Assam in northeastern India. AFLP analysis indicated the Indian weed to be of Central American origin. Surveys for pathogens revealed only mildly pathogenic species to be present in India, while 29 species were collected in Brazil, Mexico, Trinidad and Costa Rica, and four were considered to have potential as biocontrol agents. Of these, the rust Puccinia spegazzinii was identified as most promising, yet laboratory host testing of 11 strains indicated one from Trinidad to be the most pathogenic to most populations of the weed from India. This illustrates the importance of using DNA analysis in combination with other techniques, in this case with host testing. The results indicated that the most pathogenic isolate is a new association, information that would not have been apparent if either technique had been used in isolation.
Contact: Sean Murphy,
Needless to say, the Bemisia story has something to contribute here. The molecular analysis of Bemisia hosts described above was supplemented with morphological and RAPD-PCR analyses of parasitoids from these host populations. The new knowledge gained about the biogeographical relationships of the host meant that resources could be focused on the Spanish material as likely to provide the most co-evolved parasitoid strains5. In particular, attention was focused on a strain of Eretmocerus mundus. However, although this strain of E. mundus established successfully and seems to be spreading, the most successful introduced parasitoid in the Lower Rio Grande Valley, Texas appears to be E. hayati from Pakistan. This parasitoid attacks B. tabaci from the Asia group of populations, which are quite distinct from those found in the Middle East. In this latter case, climate matching appears to have been more important than co-evolution.
Contact: John Goolsby (details above)
Too fine a focus for exploration for natural enemies may be undesirable for other reasons. The specificity requirements of the recipient country (which may be affected by the pest status of the target organism) may dictate otherwise. For example, if the target site is an island with no other indigenous members of the same genus, or the target pest is causing catastrophic damage, a less than host-specific natural enemy might be acceptable to its authorities, and a wider search area may be appropriate. The former scenario is illustrated by a current CABI Bioscience biocontrol programme against the invasive privet species Ligustrum robustum ssp. walkeri in La Réunion. Although molecular taxonomy confirmed the area of origin of the subspecies as Sri Lanka, surveys have also been carried out in India, Vietnam and China on other Ligustrum species, since La Réunion has no desirable congenerics and only two other native members of the same family. This approach should maximize the chances of finding suitable biocontrol agents in this instance.
Contact: Richard Milne,
The Mikania project illustrates a decision-making and potentially time-saving use of the AFLP results. Although the Trinidadian Puccinia spegazziniii isolate was highly pathogenic to most samples (indeed, more so than to the host from which it was isolated), some Assam populations proved resistant. These were also separated by AFLP analysis, and all proved susceptible to a Mexican isolate of another pathogen, Dietelia portoricensis. Traditionally, the only way of testing for promising host - natural enemy associations is to conduct independent matching of potential natural enemies for the entire invasive population. Given the practical and financial constraints of most biocontrol programmes, such comprehensive testing was and is seldom done, with the result was that aberrant populations were and are missed.
However, using AFLP analysis, it is now possible to identify 'representative' populations and 'hot spots' of genetic variation. The outcome of this is that natural enemy testing can be done on a more representative selection of the pest in its introduced range, and, depending on the genetic homogeneity of the introduced material, the extent of the testing necessary may be reduced. In the case of Mikania, material from Assam shows variation which testing protocols need to take into account, but that from the Western Ghats is more homogenous.
Rationalizing culturing and host testing was seen as a priority by the US whitefly programme. This programme was unique in the availability of so many species of natural enemies from intensive directed foreign exploration, combined with extensive rearing facilities. Exploration for natural enemies was carried out by staff of the USDA-ARS European Biological Control Laboratory in Montpellier (France) together with other collaborating scientists, at sites around the world selected on the basis of climate matching with target areas in the USA. USDA-ARS scientists at Mission, Texas developed a unique quarantine protocol, integrating DNA (RAPD-PCR) analysis and morphologically-based systematics, to ensure "the maximum amount of species diversity with a minimum amount of duplication in cultures"6. This allowed them to assess a vast array of candidate parasitoid populations/biotypes on a large scale simultaneously.
Foreign collections were categorized in quarantine by plant type, collection site and macrotaxonomic characters of both hosts and parasitoids. Only parasitoids of 'the Bemisia tabaci complex' were accepted. Eretmocerus and Encarsia were separated on the basis of pupal and adult female characters. Individuals from each accession were then characterized using RAPD-PCR (primers CO4 and A10). Typically, material was characterized by both methods within three days, and unique accessions were set up in pure cultures on local B. tabaci, while duplicates of existing cultures were combined with them to increase their genetic diversity. USDA-ARS scientists used this process to allow them to evaluate multiple species on a large-scale. First, fecundity of candidate agents was assessed on hosts on selected crop plants. Promising parasitoids were reared and cage released onto the same crops to measure parasitism under field conditions. All species approved for release were tested by field release in an establishment evaluation. Then the 'sentinel plant' technique was used to test whether populations became widely established. In total, 38 exotic and two native parasitoids were evaluated8.
The role of DNA techniques in biocontrol programmes is not limited to the pre-release phase. Probes are now being developed to make post-release monitoring of natural enemies more effective. However, morphological analyses continue to provide the framework within which the more expensive and time-consuming molecular methods can be targeted.
DNA analysis has been used in the Bemisia programme to help evaluate the success of the released agents and conclusively identify field recoveries of cryptic species/populations in Texas, desert valleys of Arizona and California, and the San Joaquin Valley, California. For example, species-specific probes developed from satellite DNA were used to identify Eretmocerus species from material collected post-release5. However, the cost of RAPD analysis is the limiting factor, and to overcome this and time limitations, research has been initiated to develop specific DNA probes from the satellite DNA for development into a 'squash blot' kit. Such refinements are not always necessary. In an Australian study, molecular methods were used to help provide a structure within which the morphological systematist can operate. On the basis of this, they developed a user-friendly morphological key to Encarsia species that removes the need for any molecular analysis.
DNA techniques can also be used to obtain reliable information on the dynamics of parasitoid populations in the field. The Mission laboratory conducted RAPD-PCR analysis on parasitoid material recovered from uninoculated plots after agent release in Imperial Valley, California. The most promising agents in this desert climate appeared to be Eretmocerus emiratus, E. hayati, E. mundus and an Eretmocerus species from Ethiopia, while Encarsia transvena from Pakistan also looked promising. Morphological characters were used to determine that exotic Eretmocerus were in the majority (80%) at a time of normally peak native population levels. Encarsia populations also peaked, and DNA analysis of recoveries was used to show that all E. transvena recovered in the plots were of the newly released Pakistani strain, replacing the Spanish strain of the same species, which was not capable of such population increase during the summer months in this region.
Biocontrol of some other key crop pests in the USA is receiving a helping hand from DNA fingerprinting. Cornell University scientists, in collaboration with ARS researchers at the Beneficial Insects Introduction Research Laboratory in Newark, Delaware have modified an AFLP method for identifying parasitoid larvae within their pest host9. This method can even detect and identify early stages of parasitoid larvae which are difficult to distinguish by conventional means. A similar procedure was developed and used for blackflies (Simulium spp.) infected with Onchocerca in the 1980s, and this was an important development for assessing the prevalence of river blindness.
At a stroke, this does away with some lengthy and painstaking monitoring procedures. Once adult parasitoids have been reared in the laboratory and identified to species (a process which can take months), DNA probes can be developed. Then there is no longer any need to go through the lengthy 'rearing out' of parasitoids before they can be identified. Estimating percentage parasitism no longer requires the skilled dissection of large numbers of field-collected hosts. However, it is important to note that the DNA evaluations also require special equipment, skill, and time; the resources of a given laboratory (DNA machinery vs. growth chamber availability) would help determine the most practical method for that situation.
So far, the new method identifies two introduced European parasitoids: Peristenus digoneutis, which attacks the tarnished plant bug, Lygus lineolaris, a pest of many crops, and P. conradi, which attacks the alfalfa plant bug, Adelphocoris lineolatus, and it can distinguish them from the native P. pallipes. The Newark laboratory introduced both European species, and they have shown that the wasps are permanently established and are spreading in the northeastern USA. The bugs are pests of crops grown for seed, vegetables, fruits, cotton and seedling trees throughout the USA, and annually cause tens of millions of dollars in losses and control costs. Research is continuing to determine whether other related parasite species can also be identified with the new technique. DNA probes do not invariably supply all the answers, but provide another tool in the taxonomic kit. In this project, determining the phylogenetic relationships of the various species of Peristenus is one of the continuing objectives. In cases where clear morphological differences between species exist, their relationships have been confirmed by DNA information. More work on the most closely related species is needed to advance further toward the ultimate goal of phylogenetic reconstruction, a process requiring appropriate quantities of fresh material of all species in the genus, and the time to perform the analyses.
Contact: William H. Day,
The technique is not limited to insects, for ARS scientists elsewhere have developed similar methods to monitor different strains of weed pathogens following their release into the environment as biocontrol agents. Methods are emerging for detecting and identifying several isolates of Myrothecium verucarria, a soil fungus that kills morning glories, which plague sugarcane and other crops. In field studies, spraying redroot- and smallflower-morning glories with an oil-based carrier containing Myrothecium spores proved as lethal to these weeds as the herbicide atrazine.
The DNA fingerprinting technique will help biocontrol scientists keep close tabs on the spore growth and spread, host range and effectiveness of different strains of biocontrol pathogens such as Myrothecium following release. In this way, it can give genetic evidence linking a specific microbial release to a specific disease seen in target weeds. It also reveals the spread of biocontrol microbes and demonstrates their effectiveness in reducing invasive weed populations.
Contact: Jan Suszkiw,
Douglas G. Luster,
Programmes in future may be able to use information from DNA analysis in formulating biocontrol and IPM management strategies. One project that intends to capitalize on this is aimed at developing an integrated strategy for management of the coconut mite, Aceria guerreronis. This pest, which is of undetermined origin, causes significant problems for coconut growers around the world; crop losses of more than 30% have been reported in the Caribbean. It has long been a problem in the Americas, from where it was first recorded, and more recently from Africa (while coconuts are probably indigenous to Melanesia). The mite is unlikely to have been transported on coconut as it is not found on the mature nut, which is both the natural mode of dispersal and the form transported by humans. It is therefore assumed that the mite's original host belongs to the indigenous flora of the Americas. However, outbreaks in Asia (Sri Lanka) have only been reported recently, and it is unclear whether these have followed a recent introduction, or indicate the breakdown of natural population regulation; outbreaks tend to occur in pockets separated from each other by long distances. This issue could be addressed by comparing DNA samples from populations in the Caribbean and Africa with multiple samples from Sri Lanka. If the origin of the mite in Sri Lanka can be ascertained, an IPM strategy can be designed specifically either to deal with a newly introduced pest, or to identify and correct whatever has caused the upsurge.
Contact: Dave Moore,
Although there are too many strategic applications of DNA analysis to cover here, we include one that has particular significance for disease quarantine and management. In Canada, probes are being developed (US patent #5792611) to identify tree root rots and scleroderris canker in the absence of symptoms. One problem with such diseases is that they may remain latent and asymptomatic for long periods, during which time they are hard to detect but can still spread. Often, by the time symptoms are visible, it is too late to manage the disease. The DNA probes will have a two-fold application: first, to ensure the health of nursery seedlings before planting, which will help to prevent the spread of established diseases; second, to monitor imported stock for diseases, which will be a useful quarantine tool, enabling forest pathogens to be detected and diagnosed at the point of entry and thus prevent their introduction.
Contact: Richard Hamelin,
Finally, in a world where funding is hard to come by, being able to convince a potential donor that you have a success story waiting to happen could be what clinches the deal. Molecular techniques can sometimes identify in advance certain situations where classical biocontrol is likely to be successful. Where there is little variation within a population, host-specific natural enemies are less likely to be challenged by resistant varieties and therefore stand more chance of succeeding. For example, plant cytologists at Leicester University proved that all the invasive Japanese knotweed (Fallopia japonica) in the UK (and probably USA and Europe) appears to be same male-sterile clone, which would make it extremely susceptible to biological control - as well as providing the press with opportunities for headlines such as 'Largest female on earth set to swamp Britain'!
Contact: Richard Shaw, CABI UK Centre, Silwood Park,
To return for a final time to the USDA Bemisia biocontrol programme, what makes it particularly striking is that DNA techniques were used at all stages of the programme, to inform decisions, to increase efficiency, and to contribute to the scientific basis. As a result of introductions of Eretmocerus mundus and E. hayati, sentinel plant sampling indicated a dramatic increase in the numbers of introduced Eretmocerus spp. Before releases of these, native E. tejanus formed more than 95% of recoveries, yet within three months, exotic populations had risen to 85% of the Eretmocerus spp. recovered. Evaluations of their impact on B. tabaci populations are now underway, and are expected to confirm that they are making a significant contribution to biocontrol6. The authors sum up "We hope these results will encourage other biological control programs to develop predictive methods and test their predictions in field settings. The information gathered... might further the theoretical aspects of our science and in turn increase the likelihood of success in future biological control programs".
1Reid, A.; Murphy, S. (1999) Biogeographic fine-tuning helps weed biocontrol. BNI 20(2), 51N-54N.
2 Frohlich, D.R.; Torres-Jerez, I.; Bedford, I.D.; Markham, P.G.; Brown, J.K. (1999) A phylogeographical analysis of the Bemisia tabaci species complex based on mitochondrial markers. Molecular Ecology 8, 1683-1691.
3De Barro, P.J.; Driver, F.; Trueman, J.W.H.; Curran, J. (2000) Phylogenetic relationship of world populations of Bemisia tabaci (Gennadius) using ribosomal ITS1. Molecular Phylogenetics and Evolution 16, 29-36.
4Polaszek, A. (1999) Identification of whitefly parasitoids: some advice. BNI 20(4), 117N-119N.
5Kirk, A.A.; Lacey, L.A.; Brown, J.K; Ciomperlik, M.A.; Goolsby, J.A.; Vacek, D.C.; Wendel, L.E.; Napompeth, B. (2000) Variation in the Bemisia tabaci s.l. species complex (Hemiptera: Aleyrodidae) and its natural enemies leading to successful biological control of Bemisia biotype B in the USA. Bulletin of Entomological Research 90, 317-327.
6Goolsby, J.A.; Ciomperlik, M.A.; Kirk, A.A.; Jones, W.A.; Legaspi, B.C., Jr.; Legaspi, J.C.; Ruiz, R.A.; Vacek, D.C.; Wendel, L.E. (2000) Predictive and empirical evaluation for parasitoids of Bemisia tabaci (biotype 'B') based on morphological and molecular systematics. In: Austin, A; Dowton, M. (eds) Hymenoptera: evolution, biodiversity and biological control. Canberra, Australia; CSIRO Publishing, pp. 347-358.
7Azidah, A.A.; Fitton, M.G.; Quicke, D.L.J. (2000) Identification of Diadegma species (Hymenoptera: Ichneumonidae, Campopleginae) attacking the diamondback moth, Plutella xylostella (Lepidoptera: Plutellidae). Bulletin of Entomological Research 90, 375-389.
8Goolsby, J.A.; Ciomperlik, M.A.; Legaspi, B.C., Jr.; Legaspi, J.C. ; Wendel, L.E. (1998) Laboratory and field evaluation of exotic parasitoids of Bemisia tabaci (biotype 'B') in the Lower Rio Grande Valley of Texas. Biological Control 12, 27-135.
9Tilmon, K.J; Danforth, B.N.; Day, W.H.; Hoffmann, M.P. (2000) Determining parasitoid species composition in a host population: a molecular approach. Annals of the Entomological Society of America 93, 640-647.