Session 3: Non-Traditional Biological Control Agents 102 Session 3 Non-Traditional Biological Control Agents

XIII International Symposium on Biological Control of Weeds - 2011 Session 3 Non-Traditional Biological Control Agents 103

The Case for Biological Control of Exotic African Grasses in Australia and USA Using Introduced Detritivores

D. Sands1 and J. A. Goolsby2

1 CSIRO Division of Ecosystem Sciences, PO box 2583, Brisbane, Queensland 4001 Australia Email: [email protected] 2 U.S. Department of Agriculture-Agricultural Research Service (USDA-ARS), Beneficial Insects Research Unit, 2413 E Hwy 83, Weslaco, Texas 78596 USA Email: [email protected]

Abstract

Many species of African grasses were introduced into Australia and the USA to improve the quality and biomass of green pastures for domestic livestock. However, a range of non-target environmental impacts eventuated from these introductions, including spatial displacement of indigenous ecosystems, and induced changes to soils and nutrient re-cycling. The accumulation in biomass of flammable and senescing grasses predisposes grassland and woodland ecosystems to increasing impacts from wildfires, threatening indigenous plants and animals, and the recovery after fires of natural ecosystems. Very few detritivores have adapted to feed on and decompose detritus from African grasses in Australia and the USA, resulting in accumulation of dead leaves and a build-up of fuel increasing the risks from wildfires. In grasslands and woodlands of the Northern Hemisphere, epigeic detritivores and leaf-shredders include groups of invertebrates such as earthworms and isopods. In Australia, larvae of leaf-shredding moths, beetles and several other insects are important detritivores including oecophorid moths, cryptocephaline beetles, termites and cockroaches; some specifically adapted to breakdown of sub-surface plant materials in dry and moist ecosystems. In grasslands and woodlands the range of epigeic insects are likely to reduce accumulating dead biomass and fuel loads that contribute to flame height and intensity of fires. We propose that detritivores of African grasses may be potential biological control agents for senescing and dead biomass and meet the specificity requirements as agents, to control invasive African grasses in the USA and Australia.

Introduction indigenous grasses have co-existed with grazing by abundant large herbivores and respond positively to regular schedules of burning. After fires and rain African grasses were introduced into Australia copious re-growth referred to as “green pick” and (Tothill and Hacker, 1983), the USA and other favoured by livestock, is more palatable and contains countries in the mid to late 1900s, to improve the higher concentrations of nutrients when compared quality of pastures and forage for grazing by livestock. with unburned, unpalatable and senescing grass or Introductions aimed to increase green biomasses its detritus. to enable high stocking rates, or maintain seasonal African grasses spread into natural ecosystems growth of pastures with drought tolerant species, in Australia and USA where layers of senescing to suppress weeds and produce protective covers to leaves and detritus accumulate, accompanied by reduce soil erosion and nutrient losses. In Africa, little or no decomposition attributable to indigenous

XIII International Symposium on Biological Control of Weeds - 2011 104 Session 3 Non-Traditional Biological Control Agents

herbivores. Grass seeds or vegetative fragments whole natural ecosystems (Everitt et al., 2011). In the are often carried into natural areas by livestock, USA, east African guineagrass, Megathyrsus infestus on clothing or on vehicle tires. They spread from (Andersson) B. K. Simon & S. W. L. Jacobs, and P. rangelands into native grasslands, woodlands, and ciliare are similarly invasive, as are other invasive rainforests (Table 1), where they displace non- species in many parts of Australia (Table 1). flammable or weakly flammable plant communities with understories of highly flammable grass mono- Impacts of African grasses on invertebrate stands. Invasive grasses (especially signal grass, biodiversity Brachiaria decumbens Stapf.) advance growth from the margins of paths and roads, into undisturbed natural vegetation. The high flammability of some African grasses Introductions of African species have been (e.g., molassesgrass, M. minutiflora) (Tothill and mostly beneficial as fodder for livestock but several Hacker, 1983) adds to detrimental impacts on have detrimental impacts or have become serious ecosystems, but the combination of competition weeds, e.g., molassesgrass (Melinis minutiflora and flammability (e.g., of guineagrass, M. maximus; Beauv.) is unattractive to stock (Elliott, 2008) buffelgrass P. ciliare) is not well documented. Root and guineagrass, Megathyrsus maximus (Jacq.) R. balls and rhizomes of African grasses modify the Webster (= Panicum maximum, Urochloa maxima) soil texture, chemistry and inhibit activity by sub- (Henty, 1969) can be poisonous (cyanogenic). surface invertebrates. buffelgrass (Pennisetum ciliare (L.) Link) is a host Some indigenous animals use African grasses, and potential transmitter of sugarcane whitefly particularly in the absence of indigenous species, (Neomaskella bergii Signoret) in Australia (Palmer, for example in Australia, wallabies and kangaroos 2009) while gamba grass (Andropogon gayanus will feed on and use dense stands for alluding Kunth) produces dense, tall and flammable stands predators. Small vertebrate animals and birds will (Table 1) that limits mustering or location of stock use the grasses for food or shelter, for example, some in northern Australia. Several grasses (e.g M. finches will feed on seeds of M. maximus and wrens maximus, B. decumbens) invade forestry plantations will construct nests in the dense thickets. A few in Australia and New Guinea (Henty, 1969) and are polyphagous invertebrates feed on African grasses difficult to manage. In eastern Australia, African and maintain their abundance in the absence of grasses increase the access to hosts by the paralysis indigenous plant hosts, for example, grass-feeders tick (Ixodes holocyclus Neumann), facilitating the including the larvae of moths and butterflies (Braby, higher transfer of nymphs and adults from tall grass 2004; Zborowski and Edwards, 2007) will feed on species to humans and livestock. African grasses; especially when the densities of indigenous grass food plants cannot alone sustain African grasses displace plant communities breeding. While having serious impacts on many plants, In Australia and USA in the absence of natural invasive grasses reduce or prevent invertebrates enemies, African grasses spread and will out- occupying the shrub and ground cover plants, compete most indigenous grasses, shrubs and low using their fallen limbs and rock shelters, or the plant communities. African grasses inhibit vegetative decomposing organic materials. Most insects, for growth and reduce or prevent seedling recruitment, example, occupy and breed in the understory on by shielding them against entry of light. Those shrubs, “sub-surface” plants, under bark or logs grasses with rhizomes compete for root space and and in fallen leaves. Representing the majority moisture, replacing indigenous grasses and sedges of invertebrate species in temperate and tropical with surface roots. For example, African lovegrass ecosystems, they regulate the architecture of plants, (Eragrostis curvula (Schrad.) Nees) in southern break down decomposing vegetation and re-cycle Australia, and Buffelgrass (P. ciliare) in Australia nutrients, and form part of the food webs and food and southern Texas, USA, displace beneficial and chains for small vertebrates. Some grasses (e.g., B. indigenous plants and often destroy the integrity of decumbens) appear to be repellent to indigenous

XIII International Symposium on Biological Control of Weeds - 2011 Session 3 Non-Traditional Biological Control Agents 105 invertebrates that normally breed on, visit or shelter and Hacker, 1983) adds to detrimental impacts on in grasslands. ecosystems, but the combination of competition Most indigenous insects require spatial access and flammability (e.g., of guineagrass, M. maximus; to their understory food plants for oviposition and buffelgrass P. ciliare) is not well documented. Root particular phenotypic expressions of their food balls and rhizomes of African grasses modify the plants. In Australia, phytophagous Lepidoptera of soil texture, chemistry and inhibit activity by sub- grasslands are mostly monophagous or oligophagous surface invertebrates. and require particular densities of their food plants. After each fire event, re-growth of exotic grasses They are often very ‘local’ and restricted in their is invariably more aggressive than prior to being breeding sites and patrolling sites, such as hilltops, burned; African grasses out-compete the slower re- to undisturbed small areas containing their food growth and recruitment by the indigenous plants. plants or supporting specific vegetation types. For After being burned and re-sprouting from root example, some Hesperiidae always congregate on stocks, African grasses grow more rapidly than isolated hilltops with particular rock formations or indigenous grasses and can prevent re-sprouting of plant communities. Hilltops and slope ecosystems indigenous plants or germinating seedlings. The most are particularly vulnerable to displacement from serious impacts on ecosystems occur when African signalgrass (B. decumbens), buffelgrass (P. ciliare) grasses are burned frequently and the indigenous and guineagrass (M. maximus) and there is strong plants have no time to recover. After each fire, evidence of losses of rare species from protected invasive grasses advance progressively into intact areas in south-eastern Queensland (unpublished). areas taking advantage of entry of light to stimulate Buffelgrass is a particularly serious threat to rapid seed germination, rhizome advancement and many insect species in the inland areas of Central to shade out of indigenous plants. Queensland, and across Central Australia to Plant species and plant communities vary in their northern Western Australia. responses to fire ranging from highly flammable to weakly flammable, and fire adapted to fire sensitive. African grass and fire interactions Several grasses (e.g buffelgrass) will invade the understory of dry rainforests and other weakly and Frequent-lit fires in indigenous ecosystems are non-flammable plant communities (e.g., brigalow, known to have significant impacts on some species Acacia harpophylla F. Muell.), replacing indigenous of invertebrates in Australia (Greenslade and Smith, flora with mono-stands of highly flammable grass 2010), but the exacerbating effects from invasive and its senescing products. The biomass becomes flammable grasses have until recently (Sands more uniform in flammability in any given area and Hosking, 2005) not been well documented. without the variability in height and architecture African grasses are either much more flammable of most indigenous ecosystems. Most Australian than indigenous species (Table 1) or the densities grassland species are adapted to being burned where and uniformity of their communities enhances the recovery of species and rate of recovery varies fuel loads. The heat generated and flame height with the frequency, season and the scale of each mostly exceeds those from flammable indigenous fire event. The intensity of each fire also determines species, for example, the flames from gamba grass biodiversity recovery and may limit the densities of often exceed 100 m, sufficient to kill trees, prevent plants and animals able to re-colonize. their re-sprouting and change the structures When burned, the vegetative re-growth of the of all woodland into grassland species. When exotic grasses replaces indigenous re-growth, or is accompanied by sufficient soil moisture, re-growth followed by stagnation, senescence and damping off. of African grasses after fires is more rapid and By displacing whole plant communities, the exotic vigorous than indigenous species, with mono-stands grasses can extirpate or reduce the distribution out-competing indigenous species for light, root and densities of indigenous arthropod food plants, space and by changing soil composition and texture. and eliminate the hosts, prey, and shelters used as The high flammability of some African grasses habitats. For example, repeated burning of gamba (e.g., molasses grass, Melinis minutiflora) (Tothill grass has had widespread impacts on biodiversity

XIII International Symposium on Biological Control of Weeds - 2011 106 Session 3 Non-Traditional Biological Control Agents

including losses of threatened species including the control of dicotyledonous plants. Several insects iconic Leichardt’s Grasshopper, now nearly extinct in are known to be grass specialists, for example the Northern Territory (D. Rentz, pers. comm.). Absence Elachistidae (Lepidoptera) with larvae that mine of unburned refuges enabling escape by mobile leaves and stems of grasses, may contain candidate stages makes most surface-dwelling invertebrates agents for the vegetative parts of grasses (Mercadier even more vulnerable (Lindenmayer et al., 2011), et al., 2010). Documented biological control projects especially if they are not sufficiently mobile to escape targeting exotic grasses are few (Julien and Griffiths, from the radiant heat by sheltering underground. 1998) but there are notable recent targets, including The intensity, scale, frequency and seasons of fires Mediterranean giant reed, Arundo donax L., currently directly affects invertebrate survival if they have invading riparian parts of the Rio Grande Basin in no fire-adapted strategy, unburnt refuges, or are the southwestern USA (Moran and Goolsby, 2009) not sufficiently mobile to escape from being burnt. and Chilean needlegrass, Nassella neesiana (Trin, & African grasses when frequently burned can lead to Rupr.) Barkworth, in southern Australia (Anderson extirpations and have possibly caused extinctions of et al., 2010; Faithful et al., 2010). A conventional terrestrial invertebrates including the invertebrates approach to biological control has been used against that depend on shrubs, sedges and grasses for food the vegetative parts of these grasses using insects or and shelter. fungi as control agents. Frequent burning may promote seed germination Very few detritivores have adapted to feed on but it rapidly exhausts the seed bank when recruiting and decompose detritus from African grasses in plants have no time to mature between fire events. Australia and the USA, resulting in accumulation Frequent fires in plant communities occupied by of dead leaves and a build-up of fuel that increases African grasses may rapidly deplete seed banks the risks of wildfires. In grasslands and woodlands of and lead to the development of “landscape traps” the Northern Hemisphere, epigeic detritivores and (as defined by Lindenmayer et al., 2011) and are leaf-shredders include groups of invertebrates, such likely to cause in-breeding depression in some as earthworms. In Australia, predominant groups fauna and flora. The flammability of some African include oecophorid, tortricid and hepialid moths grasses has been recognized (Tothill and Hacker, and cryptocephaline beetles (Clytrinae), termites 1983) but the effects on ecosystems, especially small and cockroaches, springtails (Collembola), Protura, animals (e.g., by molasses grass, Melinis minutiflora, and Diplura. Epigeic earthworms are not significant guineagrass, M. maximus; buffel grass P. ciliare) have detritivores in Australian ecosystems (Baker, 1996) not been well documented. Within 12 months of but many insect groups adapted to breakdown of being burned, regrowth of several grasses (e.g., B. sub-surface plant materials are well adapted in decumbens) will exceed the pre-burning biomass. indigenous Australian ecosystems (Table 2). Other Burning for “fuel reduction” or “hazard reduction” potential detritivore agents would appear to include becomes impractical when burning schedules favour isopods, earthworms and fungi. While the roles of re-growth of the African grasses and thus, frequent Australian detritivores are not well understood, burning schedules can increase the fuel loads in these observations are useful for revealing the hosts short time periods following burns. and habitats of indigenous species, and might aid the prospects for locating suitable grass-adapted agents The case for biological control of African in Africa (Table 2). grasses using detritivores There are more than 5,000 species of oecophorid moths, comprising about ¼ of the known Targeting African grasses by biological control moth fauna, in Australia. The larvae break down is not without difficulties when considering fallen leaves in all Australian eucalypt ecosystems the importance of green forage to farmers and and in many types of grasslands (Common, 1996; equestrian groups. Classical biological control Zborowski and Edwards, 2007). The larvae of many for some grass targets using insects (Moran and species decompose leaves and reduce accumulation Goolsby, 2009) or fungi (Anderson et al., 2010) and thus the flammability of dead leaves. Their is showing promise comparable with biological activity binds sub-surface organics with soils and

XIII International Symposium on Biological Control of Weeds - 2011 Session 3 Non-Traditional Biological Control Agents 107 prevents erosion; they re-cycle nutrients (first stage prolific growth and seasonal senescence of exotic breakdown), some may reduce the surface tension grasses, followed by the accumulation of copious of soil surface layers of decomposing leaves and help dry leaves and detritus, is providing a major threat to retain moisture (unpublished). Leaf litter oecophorid biodiversity, particularly of invertebrates, in Australia moths provide an essential source of ground-level and the USA. For more than 10 years extreme weather prey for insectivorous animals and birds throughout events in eastern Australia have influenced the Australia. They are winter breeders, the adult moths growth of African grasses and build up in fuel loads are poorly mobile and susceptible to extirpations in dry periods, exacerbated by increases in periods of from too-frequent burning of their habitats. The prolonged drought and heavy rainfall events. Some larvae of leaf eating beetles, Cryptocephalini (> grass species have become more invasive in natural 500 species) (Coleoptera: Chrysomelidae) are also ecosystems, for example Megathyrsus maximus is an important group of detritivores in Australian becoming more competitive and Melinis minutiflora ecosystems. appears to be increasing in distribution in northern Epigeic agents would appear to be desirable New South Wales (unpubl.). As a result of climate candidates for biological control of African grasses changes, these two species as well as B. decumbens, but assessment for non-target impacts such as are becoming the most serious invasive threats to competition with native species is as important as understory-breeding invertebrates in the coastal, for any conventional biological control agents. For subtropical parts of Australia. example, in Tennessee an epigeic Asian earthworm, Targeting for biological control the decomposing Amynthas agrestis Goto and Hatai, competes with leaves (sometimes reaching 2 m in depth in M. indigenous detritivores for forest floor organic litter minutiflora) and detritus from African grasses, and reduces millipede biodiversity, both species and is least likely to be of concern when the livestock their abundance (Snyder et al., 2011). Some isopods industries promote grazing of green pastures. are known to reduce the leaf litter of indigenous Prospects for finding leaf shredders or detritivores grasses and show degrees of host preferences, for to control dead and flammable African grasses is example in California, the isopod, Porcellio scaber comparable to the challenge of finding African Latreille, had a substantial effect on litter loss of dung beetles to control dung breeding flies using annual grasses but it had no effect on an exotic dung from introduced large animals in Australia species, common velvetgrass, Holcus lanatus L. (Waterhouse, 1974; Waterhouse and Sands, 2001). (Bastow et al., 2008). Control methods for African grasses are often difficult, costly and polluting. Moreover, fuel reduction practices, including micro-mosaic Discussion methods recommended for short-term fire management to protect invertebrate biodiversity Increased frequency of fire is believed to have (New et al., 2010; Greenslade and Smith, 2010), are had a major impact on small animals in northern becoming more difficult in the presence of rapidly Australia (Woinarski et al., 2010), a trend probably accumulating detritus from African grasses. In the Australia-wide that has led to extirpations and longer term, biological control of the grasses is the sometimes extinctions of fire sensitive species desired and only option for controlling invasive (unpublished). Increasing impacts by fire can be grasses. In the case of detritivores to be tested as attributed to African grasses wherever the scale is agents, before introductions they must first be shown sufficient to displace and increase the flammability to have no detrimental impacts on indigenous of natural ecosystems. In all natural areas throughout detritivores, or interfere with other ecological Australia, threats to small animal biodiversity are processes exacerbated by increases in the scale, frequency and inappropriate seasons (e.g., cool seasons when insect References stages are immobile) of fires, wherever African grasses are invasive. Anderson, F.A., Gallego, L., Roth, G., Botto, E., Without natural re-growth following each burn, McLaren, D. & Barton, J. (2010) Investigations

XIII International Symposium on Biological Control of Weeds - 2011 108 Session 3 Non-Traditional Biological Control Agents

into biological control of Chilean needle grass Julien, M.J. & Griffiths, M.W. (1998) Biological (Nassella neesiana) in Australia and New Zealand. Control of Weeds: A World Catalogue of Agents In Proceedings Seventeenth Australasian Weeds and their Target Weeds, 4th Ed. CABI Publishing, Conference (ed. Zydenboss, S.M.), pp. 215-218. Wallingford, UK. 223p. New Zealand Plant Protection Society, Auckland Lindenmayer, D.B., Hobbs, R.J., Likens, G.E., Krebbs, New Zealand. C.J. & Banks, S.C. (2011) Newly discovered Baker, G.H. (1996) Seasonal activity of the earthworm, landscape traps produce regime shifts in wet Gemascolex lateralis (Megascolecidae), in forests. Proceedings of the National Academy of a Eucalyptus woodland in South Australia. Science of the USA 108, 15887-15891. Transactions of the Royal Society of South New, T.R., Yen, A.L., Sands, D.P.A., Greenslade, P., Australia 120, 173-174. Neville, P.J., York and A. & Collett, N.G. (2010) Bastow, J.L., Preisser, E.L. & Strong, D.R. (2008) Planned fires and invertebrate conservation Holocus lanatus invasion slows decomposition in south east Australia. Journal of Insect through its interaction with a macroinvertebrate Conservation 14, 567-574. detritivore, Porcellio scaber. Biological Invasions Mercadier, G., Goolsby, J. A., Jones, W. A. & Tamesse, 10, 191-199. J. L. (2010) Results of preliminary survey in Braby, M.F. (2004) The Complete Field Guide Cameroon, Central Africa, for potential natural to Butterflies of Australia. CSIRO, Canberra, enemies of Panicum maximum Jacq. (Poales: Australia. 339 p. Poaceae), guineagrass. Subtropical Plant Science Common, I.F.B. (1996) Oecophorinae. In Checklist 61, 31-36. of the Lepidoptera of Australia, Monograph of Moran, P.J. & Goolsby J.A. (2009) Biology of the Australian Lepidoptera Vol. 4 (eds. Nielsen, E.S., galling wasp Tetramesa romana, a biological Edwards, E.D. & Rangsti, V.), pp. 1-529. CSIRO, control agent of giant reed. Biological Control Canberra, Australia. 49, 169-179. Elliott, M. (2008) Grasses of Subtropical Eastern Palmer, C.M. (2009) Buffel grass (Cenchrus ciliaris L.) Australia. An Introductory Field Guide to is a host for the sugarcane whitefly Neomaskellia Common Grasses – Native and Introduced. bergii (Signoret) (Hemiptera: Aleyrodidae) in Nullum Publications, Murwillumbah, NSW, Central Australia. Australian Entomologist 36, Australia. 105p. 89-95. Everitt, J.H., Drawe, D.L., Little, C.R. & Lonard, Sands D.P.A. & Hosking, C.M. (2005) Ecologically R.I. (2011) Grasses of South Texas. A Guide to Sustainable Fire Management: an Advisory Code Identification and Value. Texas Tech University for Brisbane’s Western Suburbs. Moggill Creek Press, Lubbock, Texas, USA. 321p. Catchment Group, Pullen Catchments Group Faithful, I.G., Hocking, C. & McLaren, D.A. (2010). Inc. and The Hut Environmental Community Chilean needle grass (Nassella neesiana) in the Association Inc., Brookfield, Queensland native grasslands of south-eastern Australia: Australia. 38p. biodiversity effects, invasion drivers and impact Snyder, B.A., Callam, M.A. Jr. & Hendrix, P.F. (2011) mechanisms. In Proceedings Seventeenth Spatial variability of an invasive earthworm Australasian Weeds Conference (ed. Zydenboss, (Amynthas agrestis) population and potential S.M.), pp. 411-434. New Zealand Plant Protection impacts on soil characteristics and millipede in Society, Auckland New Zealand. the Great Smoky Mountains National Park, USA. Greenslade, P. & Smith, D. (2010) Short term effects Biological Invasions 13, 349-358. of wild fire on invertebrates in coastal heathland Tothill, J.C. and Hacker, J.B. (1983) The Grasses in southeastern Australia. Pacific Conservation of Southern Queensland. Tropical Grasslands Biology 16, 123-132. Society of Australia, University of Queensland Henty, E.E. (1969) A Manual of the Grasses of New Press, St Lucia, Australia, 475p. Guinea, Botany Bulletin No 1. Department of Waterhouse, D.F. (1974) Biological control of dung. Forests, Division of Botany, Lae, New Guinea, Scientific American 230, 101-108. 215p. Waterhouse, D.F. & Sands, D.P.A. (2001) Classical

XIII International Symposium on Biological Control of Weeds - 2011 Session 3 Non-Traditional Biological Control Agents 109

Biological Control of Arthropods in Australia. the Great Smoky Mountains National Park, USA. ACIAR Monograph No 77, Australian Centre for Biological Invasions 13, 349-358. International Agricultural Research, Canberra, Tothill, J.C. and Hacker, J.B. (1983) The Grasses Australia. of Southern Queensland. Tropical Grasslands Woinarski, J.C.Z., Armstrong, M., Brennan, K., Society of Australia, University of Queensland Fisher, A., Groffiths, A.D, Hill, B., Milne, D.J., Press, St Lucia, Australia, 475p. Palmer, C.M. (2009) Buffel grass (Cenchrus Waterhouse, D.F. (1974) Biological control of dung. ciliaris L.) is a host for the sugarcane whitefly Scientific American 230, 101-108. Neomaskellia bergii (Signoret) (Hemiptera: Waterhouse, D.F. & Sands, D.P.A. (2001) Classical Aleyrodidae) in Central Australia. Australian Biological Control of Arthropods in Australia. Entomologist 36, 89-95. ACIAR Monograph No 77, Australian Centre for Sands D.P.A. & Hosking, C.M. (2005) Ecologically International Agricultural Research, Canberra, Sustainable Fire Management: an Advisory Code Australia. for Brisbane’s Western Suburbs. Moggill Creek Woinarski, J.C.Z., Armstrong, M., Brennan, K., Catchment Group, Pullen Catchments Group Fisher, A., Groffiths, A.D, Hill, B., Milne, D.J., Inc. and The Hut Environmental Community Palmer, C., Ward, S., Watson, M., Winderlich, S. Association Inc., Brookfield, Queensland & Young, S. (2010) Monitoring indicates rapid Australia. 38p. and severe decline of native small mammals Snyder, B.A., Callam, M.A. Jr. & Hendrix, P.F. (2011) in Kakadu National Park, northern Australia. Spatial variability of an invasive earthworm Wildlife Research 37, 116-126. (Amynthas agrestis) population and potential Zborowski, P. & Edwards T. (2007) A guide to the impacts on soil characteristics and millipede in Australian Moths. CSIRO Australia. 214p.

XIII International Symposium on Biological Control of Weeds - 2011 110 Session 3 Non-Traditional Biological Control Agents

Table 1. Highly flammable African grasses affecting biodiversity in the tropics and subtropics of Australia and USA ESTIMATED DISPLACE- GRASS REGIONS AFFECTED MENT + FUEL VALUES * Tropical northern & inland; gamba grass moist & dry tropical woodlands +++ (Andropogon gayanus ) & grasslands

buffelgrass Tropical & subtropical northern +++ ( Pennisetum ciliare) & inland: central & western; woodlands & grasslands

lovegrass Subtropical & temperate sub- ++ (Eragrostis curvula) coastal & inland; woodlands & grasslands

guineagrass (Megathyrsus maximus) and east Tropical & subtropical coastal: ++ African guineagrass (Megathry- eastern, wet woodlands & forests sus infestus) ++ signalgrass Subtropical coastal: moist & wet (Brachiaria decumbens) woodlands

molassesgrass Tropical & subtropical; eastern & +++ (Melinis minutiflora) coastal wet woodlands & forests

South African pigeongrass Tropical & subtropical eastern, +++ (Setaria sphacelata) (Schumach.) dry & coastal woodlands & Stapf & C.E. Hubb. grasslands

giant Parramatta grass Tropical & subtropiical inland; +++ (Sporobolus fertilis) (Steudel) woodlands & grasslands Clayton

* estimated values contrasted with the widespread Australian kangaroo grass, Themeda triandra Forssk. (rated +)

XIII International Symposium on Biological Control of Weeds - 2011 Session 3 Non-Traditional Biological Control Agents 111

Table 2. Insect detritivores and leaf shredders commonly associated with Australian ecosystems Ecosystem Associated Detritivores/Leaf Shredders leaf litter moths (Oecophoridae, Tortricidae: Epitymbiini) Dry eucalypts & woodlands leaf beetles (Cryptocephalini), Isoptera (Microcerotermes, Ephelotermes, Hesperotermes, Nasutitermes)

cockroaches (Blattoidea; Geoscapheus, Cryptocercus), Rainforests & moist forests moths (Oecophoridae: Barea)

leaf litter moths & “mallee” moths (Oecophoridae), termites (Isoptera: Drepanotermes, Lophotermes, Nasutitermes, Tumuli- Grasslands termes)

leaf litter moths (Oecophoridae & Tortricidae); Isoptera & others but not

well documented Heathlands

XIII International Symposium on Biological Control of Weeds - 2011 112 Session 3 Non-Traditional Biological Control Agents

Rhizaspidiotus donacis (Hemiptera: Diaspididae), an Armored Scale Released for Biological Control of Giant Reed, Arundo donax

P. J. Moran1, J. A. Goolsby2, A. E. Racelis2, E. Cortés3, M. A. Marcos-García3, A. A. Kirk4 and J. J. Adamczyk5

1U. S. Department of Agriculture-Agricultural Research Service (USDA-ARS), Exotic and Inva- sive Weeds Research Unit, 800 Buchanan St., Albany, CA 94710 USA Email: [email protected] 2USDA-ARS, Cattle Fever Tick Research Laboratory, Moore Air Base, 22675 N. Moorefield Rd., Edinburg, TX 78541 USA Email: [email protected] 3Instituto de Biodiversidad CIBIO. Universidad de Alicante, Campus Universitario San Vicente del Raspeig, 03080, Alicante, Spain Email: [email protected] 4USDA-ARS, European Biological Control Laboratory, CS 90023 Montferrier-sur-Lez, 34988 Sant Gely du Fesc (Montpelier), France Email: [email protected] 5USDA-ARS, Thad Cochran Southern Horticultural Laboratory, 810 Hwy 26 West, PO Box 287, Poplarville, MS 39470 USA Email: [email protected]

Abstract

Non-native, invasive perennial grasses have not been widely targeted for classical biological control with insects, despite their global prevalence and damaging effects, in part because of a perceived paucity of host-specific insect herbivores. Armored scales (Hemiptera: Diaspididae) have not been used for biological weed control. However, over 250 armored scale species occur on grasses, of which 87% feed only on this family and 58% feed on only one grass genus, suggesting that armored scales may have unrealized biological control potential. We selected the armored scale Rhizaspidiotus donacis Leonardi as a candidate agent against the exotic, invasive, water-consuming grass known as giant reed (Arundo donax L.), based on literature records and our collections indicating host-specificity to the genus Arundo and its broad geographic range in the Mediterranean basin. Observations of reduced giant reed vigor at scale-infested sites in Spain and France were confirmed in native range field studies showing a 50% reduction in lateral shoot growth rate and rhizome weight on scale- infested versus non-infested A. donax, as well as significant reductions in photosynthesis rates in quarantine laboratory studies. Specialized procedures were developed using gelatin capsules to isolate females from host tissues and neonate crawlers from females. Based on laboratory and native range field studies, the host range of R. donacis is limited to Arundo spp., and the life cycle requires 5–6 months. A scale accession from eastern coastal Spain established larger populations on rhizomes of two invasive A. donax accessions from this same region than on rhizomes representing a separate, geographically isolated introduction. In 2011, R. donacis became the first armored scale released for biological weed control, establishing robust rearing colonies and reproductive field populations with evidence of lateral shoot deformities at the first release site along the Rio Grande in Texas, USA.

XIII International Symposium on Biological Control of Weeds - 2011 Session 3 Non-Traditional Biological Control Agents 113

Introduction known, most with broad host ranges (Miller and Davidson, 1990). No armored scales have been released for weed control, although two adventive Arundo donax L, known as arundo, giant reed, or carrizo, is native from the Mediterranean Basin Chionaspis species are widespread and damaging to India, and invasive in North and South America, on saltcedars (Tamarix spp.) in North America South Africa, and Australia. Arundo has colonized at (Wiesenborn, 2005). Armored scales are commonly least 30,000 ha along rivers, reservoirs and irrigation found on grasses (Evans and Hodges, 2007), with canals in the Lower Rio Grande Basin of south Texas over 250 species known (http:// www.sel.barc.usda. and northern Mexico (Yang et al., 2009, 2011), gov/scalenet.html), and 222 of these (87.5%) occur removing economically-significant amounts of only on hosts in the Poaceae; 58% (128) are found water, displacing native plants, altering flood and fire on only one grass genus. Armored scales may thus regimes, harboring cattle fever ticks, and hindering have unrealized potential for biological control of border security activities (Moran and Goolsby perennial grasses. 2009; Racelis et al., 2011), with similar ecological A biological control program targeting A. donax, impacts in California (Coffman et al., 2010; Lambert led by the USDA-Agricultural Research Service, is et al., 2010) and in other arid riparian ecosystems. the first to release multiple non-native insect agents Non-native, invasive perennial grasses (Poaceae) to control a grass weed. We selected Rhizaspidiotus and sedges (Cyperaceae) are among the world’s donacis Leonardi as a candidate agent on the basis of most widespread weeds (Witt and McConnachie, literature records (reviewed in Goolsby et al., 2009a) 2003), attaining densities in riparian and rangeland and our collections indicating specificity to the grass habitats that exceed the capacity of chemical and genus Arundo, which has no native members in mechanical control. However, few biological North or South America. The geographically-broad control projects have targeted perennial grasses, due native range of this scale, as well as the thin, brittle to a perceived paucity of genus- or species-specific arundo shoots observed at sites with dense scale insect herbivores (Evans, 1991). Pathogenic fungi populations, also favored its selection for further have been considered for inoculative (Yobo et al., testing. 2009; Anderson et al., 2011) or inundative releases (Yandoc et al., 2004). Surveys of herbivorous insect Methods and Materials fauna on common reed (Phragmites australis (Cav.) Trin. ex. Steud.) (Tewksbury et al., 2002), giant reed Determination of native distribution (Arundo donax) (Tracy and DeLoach, 1998; Kirk et al., 2003), Sporobolus spp. (Witt and McConnachie, Collections to survey insects on A. donax were 2003), Calamagrostis spp (Dubbert et al., 1998) and made at over 330 sites in 19 countries between multiple grass genera (Tscharntke and Greiler, 1995) 2000 and 2007 during the spring, summer and fall. have shown that oligo- and monophagous herbivores exist on grasses, but exhibit feeding patterns and Rhizome samples were taken to the USDA-ARS taxonomic affiliations that are non-traditional in European Biological Control Laboratory to detect weed biological control. Examples include shoot- the presence of R. donacis and natural enemies (Kirk galling wasps (Hymenoptera: Eurytomidae), stem- et al., 2003). boring flies (Diptera: Chloropidae), galling and non- galling leafminers (Diptera: Cecidomyiidae) and, in Evaluations of impact the case of giant reed, an armored scale (Hemiptera: Diaspididae). In the Province of Alicante, in southeastern Armored scales are immobile besides the short- coastal Spain, 15 shoots at each of five sites were lived neonate crawler (first instar) and adult male treated with foliar and root drench insecticide stages. These small, sometimes cryptic insects monthly for 12 months and 15 shoots were left generate cumulative damage to perennial plants untreated. Growth rates were compared across the over several generations of an often prolonged life two treatments and between monthly measurements cycle (McClure, 1990). Over 200 pest species are (Cortés et al., 2011a). To examine arundo scale

XIII International Symposium on Biological Control of Weeds - 2011 114 Session 3 Non-Traditional Biological Control Agents

effects on rhizome weight, nine sites with and nine Pepper, J. Goolsby, P. Moran, A. Contreras Arquieta, without R. donacis were selected in the Languedoc A. Kirk, and J. Manhart, unpublished), and from region of southwestern France and the Province of one site (Balmorhea) with a distant point of origin. Alicante in Spain (Cortés et al., 2011b). In a three- month quarantine laboratory study, potted, fertilized Mass-rearing and field release arundo stems were infested with R. donacis at a rate of 158 crawlers per week, and also received arundo A permit from the USDA-APHIS-PPQ to wasps (Tetramesa romana Wallker) at a rate of 18 per release the arundo armored scale into the field was week. Shoot and selected leaf lengths were compared received on 16 December 2010. Mass-rearing was to wasp-only and control treatments (Goolsby et conducted in 700-L plastic inner tubs filled to one- al., 2009b). In a six-month lab study, 500 to 700 half depth with pea gravel and elevated 10 cm above R. donacis crawlers were released per stem and an outer tub modified with drain holes to maintain leaf gas exchange measured 24 weeks later to infer the water level below the rooted arundo rhizomes, effects on photosynthetic processes (Moore et al., which were kept dry and partially exposed on the 2010). In a six-month greenhouse study, rhizomes gravel surface. A similar subsurface watering and crowns of young arundo shoots were infested system was used for rhizomes planted in outdoor with 1,000–5,000 crawlers and shoot elongation and greenhouse trenches. Releases at a field site in and biomass examined (Racelis et al., this volume). Del Rio, Texas on the floodplain of the Rio Grande In greenhouse,studies, crawlers were released onto began in January 2011 (Goolsby et al., 2011). rhizomes either fertilized with urea plus slow-release nitrogen-phosphorous-potassium pellets, or given Results only pellets (Moran and Goolsby, this volume).

Determination of biology and host range Native distribution

Adult females were shipped on arundo Our collections and literature records indicate rhizomes from 15 sites in southwestern France and that the geographic range of R. donacis includes Mediterranean Spain. Females were removed from eastern and southern Spain, extreme southern rhizomes and stored in 1.5-cm gelatin capsules at 27 France, Italy, Crete, southern and western coastal °C, 60% RH, 14:10 L:D. Crawlers were collected in Turkey, and coastal Algeria, with no collections gelcaps, which were screened microscopically for in the Balearic Islands, Corsica, Sicily, Croatia, Aphytis acrenulatus Rosen and DeBach ectoparasites Bulgaria, Israel, Egypt, Morocco, the Canary and other contaminants. Capsules with crawlers Islands, China, Nepal, or India. Some of the most were pinned to shoots of 2-month old arundo shoots robust R. donacis populations, in eastern and or non-target plants. Destructive dissections at southern Spain, were found on arundo accessions varying time points and isolation of crawlers, adult matched using microsatellites to the invasive males and females were used to determine duration arundo genotypes in the Lower Rio Grande Basin. and survival of life stages (Moran and Goolsby 2009). For host range studies, shoots of A. donax, 40 Impact on target weed other grass species, and 5 non-grasses received 200 crawlers each and were dissected three months after Infestation by R. donacis reduced both lateral infestation (Goolsby et al., 2009a). Field studies shoot growth and rhizome weight of A. donax by were conducted in Spain as a follow up to examine 50% in the native range (Table 1). In quarantine, the laboratory non-target development on Leptochloa arundo armored scale slightly enhanced the negative spp. and Spartina alternifolia Loisel grasses. Sub- effect of the arundo wasp on main shoot growth, and specific host range was examined by infesting independently reduced the ability of the plant to absorb arundo rhizomes from two Texas sites (Austin and light energy to convert nitrogen into protein by over Laredo) that genetically match several populations 60% (Table 1). In a 6-month greenhouse study, high in eastern and southern coastal Spain (D. Tarin, A. levels of arundo scale infestation (5,000 crawlers)

XIII International Symposium on Biological Control of Weeds - 2011 Session 3 Non-Traditional Biological Control Agents 115 reduced main shoot growth by 50% or more (Table by the 50th day post-emergence, with a brief period 1). Urea fertilization increased female reproduction beforehand during which adult females had eclosed and settling of second-generation crawlers, but but were still inside the second instar nymphal results were inconsistent across two rearing cuticle. Adult females fertilized by males required a environments (Moran and Goolsby, this volume). total of 170 total days from emergence as crawlers to reproductive maturity. Survival to adulthood (both Biology sexes combined) was 20-25%. Each reproductive female produced an average of 85 crawlers. More At 26 °C, arundo scale crawlers lived less than two details may be found in Moran and Goolsby (2010). days without food, or, on arundo shoots, settled to an immobile ‘whitecap’ phase to complete the first instar Host range within 10–13 days of emergence. Males completed second-instar development within an additional The arundo scale readily completed development 20 days and emerged as adults by the 40th day after on A. donax, with significantly less development emergence. Non-feeding winged males lived less on A. formosana, and none on closely related than two days. Females completed the second instar common reed, Phragmites australis or 37 other

Table 1. Evaluations of impact of the arundo armored scale Rhizaspidiotus donacis on giant reed (Arundo donax), based on pre-release observational and manipulative field studies in Mediterra- nean Europe, quarantine laboratory studies, and a post-release greenhouse study in Texas, USA. Length of Time Reduction Asso- Study Environ- of Scale Infesta- Variable ciated with Scale Reference ment tion Infestation

Daily rate of shoot Field-Spain 12 months 61% Cortés et al., 2011a growth (cm day-1)

Field-France and Unknown Rhizome weight (g) 46% Cortés et al., 2011b Spain

Stem and leaf Laboratory 3 months 5–10% Goolsby et al., 2009b length1

Maximum rate of Laboratory 6 months 61% Moore et al., 2010 electron transport2

Non-quarantine 6 months Shoot elongation 50%–60% Racelis et al., this volume greenhouse rate and aboveg- round biomass

1Effect of scale over and beyond that of the arundo wasp Tetramesa romana. Differences between arundo wasp infesta- tion alone and wasp+scale were not significant. 2Measure of efficiency of conversion of light energy in photosynthesis.

XIII International Symposium on Biological Control of Weeds - 2011 116 Session 3 Non-Traditional Biological Control Agents

native grasses. Very limited development to adult stands along the Rio Grande and decreases in water occurred on three species in the genus Leptochloa consumption by giant reed are key benchmarks in this and on Spartina alterniflora in quarantine, but no biological control program, and are being examined settling by R. donacis was observed in dissections of as arundo armored scale populations develop. these grasses when found in populations sympatric with A. donax in Spain and France, even when Acknowledgments potted plants were maintained in stands of heavily scale-infested arundo for 6 months. More details We thank Crystal Salinas, Ann Vacek, and may be found in Goolsby et al. (2009a). Sub- Connie Graham for technical assistance. This specific host specificity tests in quarantine indicated research was supported in part by the U.S. development of significantly larger second- Department of Homeland Security and the Lower generation populations on invasive Texas rhizomes Rio Grande Valley Development Council. matched to Spanish locations from which crawlers were obtained than on Texas rhizomes representing a geographically distinct source of introduction References (J. Goolsby, E. Cortés, P. Moran, J. Adamczyk, M. Marcos García and A. Kirk unpublished). Anderson, F.E., Díaz, M.L., Barton, J., Flemmer, A.C., Hansen, P.V., & McLaren, D.A. (2011). Mass-rearing and field release Exploring the life cycles of three South American rusts that have potential as biocontrol agents of Mass-rearing of R. donacis began in December the stipod grass Nassella neesiana in Australia. 2010 and open field releases in January 2011. On Fungal Biology 115, 370-380. 20 July 2011, whitecaps indicative of reproduction Coffman, G.C., Ambrose, R.F., & Rundel, P.W. (2010) in the field were observed at the Del Rio, Texas Wildfire promotes dominance of invasive giant site. Reproductive females and a new generation reed (Arundo donax) in riparian ecosystems. of immatures were found on rhizomes at this site in Biological Invasions 12, 2723-2734. August 2011 (Goolsby et al., 2011). Colony-reared Cortés, E., J., Goolsby, J.A, Moran, P.J., & Marcos- females produced similar numbers of crawlers as did García, M.A. (2011a) The effect of the armored females collected in Europe (P. Moran, unpublished). scale, Rhizaspidiotus donacis (Leonardi) (Hemiptera: Diaspididae), on shoot growth of the invasive plant Arundo donax (Poaceae: Discussion Arundinoideae. Biocontrol Science and Technology 21, 535-545. The arundo armored scale R. donacis exhibits Cortés, E., Kirk, A.A., Goolsby, J.A., Moran, P.J., a broad native geographic range with substantial, Racelis, A.E., & Marcos-Garcia, M.A. (2011b) measurable impacts on the growth of giant reed, but Impact of the arundo scale Rhizaspidiotus narrow fidelity to the target genus Arundo, with full donacis (Leonardi) (Hemiptera: Diaspididae) population growth potential only on A. donax. The on the weight of Arundo donax L. (Poaceae: biological attributes of this scale are suitable and Arundinoideae) rhizomes in Languedoc southern adaptable for mass rearing, given sufficient attention France and Mediterranean Spain. Biocontrol to geographic matching of scale accessions from the Science and Technology 21, 1369-1373 . native range to the known geographic sources of Dubbert, M., Tscharntke, T., Vidal, S. (1998) Stem- the giant reed populations that are invasive. After boring insects of fragmented Calamagrostis only one generation in greenhouse studies, the habitats: Herbivore-parasitoid community arundo scale reduced lateral and main shoot growth structure and the unpredictability of grass shoot and alters photosynthetic processes. Feeding by abundance. Ecological Entomology 23, 271-280. multiple generations is likely necessary to reduce Evans, G.A. & Hodges, G.S. (2007) Duplachionaspis rhizome weight and shoot recruitment in the field. divergens (Hemiptera: Diaspididae), a new exotic Improvements in visibility through dense giant reed pest of sugarcane and other grasses in Florida.

XIII International Symposium on Biological Control of Weeds - 2011 Session 3 Non-Traditional Biological Control Agents 117

Florida Entomologist 90, 392-393. Elsevier, Amsterdam. Evans, H.C. (1991) Biological control of tropical Moran. P.J. & Goolsby, J.A. (2009) Biology of the weedy grasses, In Tropical Weedy Grasses (eds. galling wasp Tetramesa romana, a biological Baker, F.W.G. & Terry, P.J.), pp. 52-71. CAB control agent of giant reed. Biological Control 49, International, UK. 169-179. Goolsby, J.A., Moran, P,J., Adamczyk, J.J., Kirk, A.A., Moran, P. J. & Goolsby, J.A. (2010) Biology of the Jones, W.A., Marcos, M.A., & Cortes, E. (2009a) armored scale Rhizaspidiotus donacis (Hemiptera: Host range of the European, rhizome-stem Diaspididae), a candidate agent for biological feeding scale Rhizaspidiotus donacis (Hemiptera: control of giant reed. Environmental Entomology Diaspididae), a candidate biological control agent 103, 252-263. for giant reed, Arundo donax (Poales: Poaceae) Moran, P.J. & Goolsby, J.A. (In this volume) Effect of in North America. Biocontrol Science and nitrogen addition on population establishment of Technology 19, 899-918. the arundo armored scale (Rhizaspidiotus donacis). Goolsby, J.A., Spencer, D., & Whitehand, L. (2009b) Moore, G.W., Watts, D.A., & Goolsby, J.A. (2010) Pre-release assessment of impact on Arundo Ecophysiological responses of giant reed (Arundo donax by the candidate agents Tetramesa donax) to herbivory. Invasive Plant Science and romana (Hymenoptera: Eurytomidae) and Management 3, 521-529. Rhizaspidiotus donacis (Hemiptera: Diaspididae) Racelis, A.E., Davey, R.B., Goolsby, J.A., Pérez de under quarantine conditions. Southwestern León. A.A., Varner. K.& Duhaime. R. (2011) Entomologist 34, 359-376. Facilitative ecological interactions between Goolsby, J.A., Kirk, A.A., Moran, P.J., Racelis, A.E., invasive species: Arundo donax (Poaceae) Adamczyk, J.J., Cortés, E, Marcos-García, M.A., stands as favorable habitat for cattle ticks (Acari: Martinez Jimenez, M., Summy, K.R., Ciomperlik, Ixodidae) along the US-Mexico border. Journal of M.A., & Sands, D.P.A (2011) Establishment of the Medical Entomology, in press. armored scale Rhizaspidiotus donacis, a biological Racelis, A.E., Moran, P.J, Goolsby, J.A., & Yang, C. control agent of Arundo donax. Southwestern (In this volume) Estimating density dependent Entomologist 36, 373-374. impacts of the arundo scale, biological control Kirk, A.A., Widmer, T., Campobasso, G., agent for the invasive giant reed. Carruthers, R.A., & Dudley, T.L. (2003) The Tewksbury, L., Casagrande, R., Blosssey, B., Häfliger, potential contribution of natural enemies from P. & Scwarzländer, M. (2002) Potential for Mediterranean Europe to the management of biological control of Phragmites australis in North the invasive weed Arundo donax (Graminae: America. Biological Control 23, 191-212. Arundinae) in the USA, In Proceedings of the Tracy, J.L. & DeLoach, C.J. (1998) Suitability for California Invasive Plant Council Symposium (ed classical biological control for giant reed (Arundo Porosko, C), pp. 62-68. California Invasive Plant donax) in the United States, In Arundo and Council, Berkeley, California. Saltcedar Management Workshop Proceedings Lambert, A.M., Dudley, T.L., & Saltonstall, K. (2010) (ed Bell, C.E.), pp. 73-109. University of California Ecology and impacts of the large-statured invasive Cooperative Extension, Ontario, California. grasses Arundo donax and Phragmites australis Tscharntke, T. & Greiler, H.J. (1995) Insect in North America. Invasive Plant Science and communities, grasses, and grasslands. Annual Management 3, 489-494 Review of Entomology 40, 535-558. McClure M.S. (1990) Influence of environmental Wiesenborn, W.D. (2005) Biomass of arthropod factors, In Armored Scale Insects: Their Biology, trophic levels on Tamarix ramosissima Natural Enemies, and Control, Vol. A. (ed Rosen, (Tamaricaceae) branches. Environmental D.), pp. 319-330. Elsevier, Amsterdam. Entomology 34, 656-663. Miller, D.R. & Davidson, J.A. (1990) A list of the Witt, A.B.R. & McConnachie, A.J. (2003) The armored scale insect pests, In Armored Scale potential for classical biological control of Insects: Their Biology, Natural Enemies, and invasive grass species with special reference to Control, Vol. B. (ed Rosen, D.), pp. 299-306. invasive Sporobolus spp. (Poaceae) in Australia, In

XIII International Symposium on Biological Control of Weeds - 2011 118 Session 3 Non-Traditional Biological Control Agents

Proceedings of the XI International Symposium Journal of Applied Remote Sensing 3, 033530. on Biological Control of Weeds (eds Cullen, Available at http://spiedigitallibrary.aip.org). J.M., Briese, D.T., Kriticos, D.J., Lonsdale, W.M., Yang, C., Everitt, J. H. & Goolsby, J. A. (2011) Using Morin, L., & Scott, J.K.), pp. 198-202. CSIRO aerial photography for mapping giant reed Entomology, Canberra, Australia. infestations along the Texas-Mexico portion Yandoc, C.B., Charudattan, R., & Shilling, D.G. of the Rio Grande. Invasive Plant Science and (2004) Suppression of cogongrass (Imperata Management 4, 402-410. cylindrica) by a bioherbicidal fungus and plant Yobo, K.S., Laing, M.D., Palmer, W.A., & Shivas, competition. Weed Science 52, 649-653. R.G. (2009) Evaluation of Ustilago sporoboli-indici Yang, C., Goolsby, J.A., & Everitt, J.H. (2009) Using as a classical biological control agent for invasive Quickbird satellite imagery to estimate giant reed Sporobolus grasses in Australia. Biological Control infestations in the Rio Grande Basin of Mexico. 50, 7-12.

XIII International Symposium on Biological Control of Weeds - 2011 Session 3 Non-Traditional Biological Control Agents 119

Fergusonina turneri/Fergusobia quinquenerviae (Diptera: Fergusoninidae/Nematoda: Tylenchida: Sphaerulariidae), a Bud-Gall Fly and its Obligate Nematode Released for the Australian Paperbark Tree, Melaleuca quinquenervia

T. Center1, K. Davies2, R. Giblin-Davis3, P. Pratt1, M. Purcell4, S. Scheffer5, G. Taylor2 and S. Wright1

1Invasive Plant Research Lab., USDA-ARS, Fort Lauderdale, FL USA [email protected] 2University of Adelaide, Glen Osmond, South Australia 3University of Florida, Fort Lauderdale, FL USA 4Australian Biological Control Lab., USDA-ARS & CSIRO, Brisbane, Australia 5Systematic Entomology Lab., USDA-ARS, Beltsville, MD USA

Abstract

The gall fly Fergusonina turneri Taylor and the nematode Fergusobia quinquenerviae Davies & Giblin-Davis form a mutualist association on Melaleuca quinquenervia (Cav.) S.T. Blake, an Australian tree that has invaded south Florida. Together they form multi-locular galls that compromise meristems thereby curtailing growth and reproduction of the targeted plant. Flies oviposit and nemaposit into vegetative and reproductive M. quinquenervia buds. Nematodes initiate cedidogenesis producing hypertrophied tissue prior to fly egg hatch. The maggots then feed on the primed nutrient-rich tissue while presumably inducing further enlargement of the galls. Meanwhile, the parthenogenetic nematodes produce a second generation of amphimictic individuals. The mated female nematodes invade the hemocoel of fully grown female (3rd instar) fly larvae. The fly pupates and the female nematodes produce juveniles that invade the rudimentary ovaries of the developing female flies. The adult fly then emerges from the gall carrying juvenile nematodes in their ovaries. All female flies contain nematodes, which are deposited in buds during oviposition allowing the cycle to begin anew. Molecular analyses of related Melaleuca species and host range studies proved this fly- nematode combination to be specific to M. quinquenervia. A permit for their release as biological control agents was subsequently granted. They were released in south Florida beginning in 2005 and temporarily established, but disappeared completely after about three generations. To date, these two agents have not established self-sustaining field populations in Florida. The more recent release and establishment of Lophydiplosis trifida Gagné, a gall-forming midge, produces similar effect thus precluding the need for the fly/ nematode combination. Nonetheless, this is the first time that a mutualistic combination of two agents has been approved and attempted for use in a biological control program.

XIII International Symposium on Biological Control of Weeds - 2011 120 Session 3 Non-Traditional Biological Control Agents

Tetramesa romana (Hymenoptera: Eurytomidae), a Parthenogenic Stem-Galling Wasp Released for Giant Reed, Arundo donax

A. E. Racelis1, P. J. Moran1, J. A. Goolsby1, A. A. Kirk2 and J. J. Adamczyk1

1USDA-Agricultural Research Service, Kika de la Garza Subtropical Agriculture Research Center, Weslaco, TX USA [email protected] 2USDA-Agricultural Research Service, European Biological Control Laboratory, Montpelier France

Abstract

Plant development and growth are controlled by a series of independent processes that are determined by physiological and genetic mechanisms. Gall-inducing organisms can interrupt these processes and modify normal plant growth, often via plant species- specific interactions, and thus should be considered seriously for biological control. In the case of the arundo wasp, Tetramesa romana Walker (Hymenoptera: Eurytomidae), larval development on giant reed (Arundo donax L.) induces gall formation, interrupting meristematic activity and stimulating lateral budding. Since egg laying females prefer to oviposit on phenotypically labile tissues, gall formation has a negative influence on stem elongation, which in turn can limit the competitive ability of aggressive giant reed. Apparent host range of the arundo wasp was tested in quarantine by recording the occurrence of ovipositor probing events of the female, which occurred on 15 of 35 different test plants. However, its actual host range, indicated by successful larval development, was restricted to two species in the Arundo genus. With its host specificity, relatively the short generation time (an average of 33 days) and prolific parthenogenic reproductive output (an average 21 offspring per female), T. romana has promise as a control agent for giant reed and was released in the spring of 2009 in the Lower Rio Grande Basin of Texas and Mexico. The USDA-ARS has developed an adaptive, mass-rearing protocol for multiple arundo wasp genotypes based on responses to ambient light, temperature, and host-plant conditions. A series of sequential greenhouse studies reveal superior performance by Spanish genotypes of the wasp over French genotypes, in terms of greater reproductive output and compatibility with abiotic conditions. A mass release program for T. romana has been initiated using fixed-wing aircraft, and post-release field evaluations, including field studies of population establishment based on genotype and climate matching, are ongoing.

XIII International Symposium on Biological Control of Weeds - 2011