Fourteenth Australian Weeds Conference

The role of natural enemies in regulating populations of biocontrol agents on gorse ( L.)

Jamie T. Davies1, John E. Ireson1 and Geoff R. Allen2 1 Tasmanian Institute of Agricultural Research, 13 St. Johns Avenue, New Town, Tasmania 7008, Australia 2 School of Agricultural Science, University of Tasmania, GPO Box 252-54, Hobart, Tasmania 7001, Australia

Summary Natural enemies can regulate popula- a Tasmanian fi eld study (Davies unpublished data). tions of phytophagous . In weed biological Natural enemies that have the potential to reduce control programs, natural enemies may either prevent the efficacy of T. lintearius have been identified agents from establishing or reduce their impact on in Australian surveys (Ireson et al. 2003). One of the target weed. A study of the development time for these was a specialist Tetranychid predator, the Chil- Phytoseiulus persimilis at two temperatures on gorse ean predatory mite, Phytoseiulus persimilis Athias- ( lintearius) and the pest mite Henriot (: Phytoseiidae). This species is a vora- T. urticae indicates P. persimilis will develop as fast, cious predator capable of reducing pest Tetranychus or faster, on the gorse spider mite as it does on pest spp. populations below economically damaging levels. mites. This suggests that P. persimilis is likely to have As it is such an effective predator, P. persimilis is com- a negative impact on gorse spider mite populations. monly used as a biocontrol agent of the polyphagous In addition, a survey of the fauna on gorse pest Tetranychus urticae Koch (Acari: Tetranychidae) identifi ed a range of potential natural enemies, con- and other pest Tetranychus spp. in a variety of crops sisting mainly of generalist predators, which could (McMurtry and Croft 1997). reduce the impact of the gorse thrips and provide an Phytoseiulus persimilis has been shown to sig- explanation for its slow dispersal rate. nifi cantly reduce the size of T. lintearius colonies Keywords Biological control, gorse, natural enemies, in predator exclusion studies in the USA (Pratt et Sericothrips staphylinus, Tetranychus lintearius, Ulex al. 2003). Studies have previously been conducted europaeus, Phytoseiulus persimilis. on the development of P. persimilis on diets of pest Tetranychus spp. including T. urticae (e.g. Galazzi and INTRODUCTION Nicoli 1996), T. kanzawai (Hamamura et al. 1976) and Host specific phytophagous arthropods are often T. pacifi cus (Perring and Lackey 1989). However, data introduced as biological control agents for invasive linking the development of P. persimilis on a diet of weeds. Natural enemies often play a key role in the benefi cial T. lintearius is lacking. If P. persimilis regulating populations of phytophagous arthropods can feed and develop on a diet of T. lintearius at the and are extensively utilised for the biological control same or similar rate as it does on pest Tetranychus of agricultural and horticultural pests. However, in spp., then it would be expected that the impact of the weed biological control, natural enemies may reduce T. lintearius on gorse will be greatly reduced by the the population size and therefore the effectiveness of action of this predator. weed biocontrol agents (Goeden and Louda 1976, The gorse thrips, Sericothrips staphylinus Haliday McFayden and Spafford-Jacob 2004). (Thysanoptera: Thripidae) was introduced to Australia Gorse, Ulex europaeus L. (Fabaceae), is a legumi- from New Zealand and released in 2001 (Ireson et nous European woody shrub that has become invasive al. 2004). In a glasshouse environment this species in many temperate regions of the world. In Australia, reduced the growth of gorse seedlings and also reduced gorse is a Weed of National Signifi cance (Thorp 1999) seedling survival when combined with other manage- seriously affecting agricultural and environmentally ment practices (Davies unpublished data). Although signifi cant regions in south eastern Australia. As part populations of S. staphylinus developed rapidly in the of an integrated management strategy, a guild of host protected environment of a glasshouse, populations specifi c biological control agents is currently being have been very slow to increase to high densities in introduced to gorse in Australia (Ireson et al. 2004). the fi eld and disperse (Ireson et al. 2004). One possible The gorse spider mite, Tetranychus lintearius explanation is that an ecological factor, such as natural Dufour (Acari: Tetranychidae), was introduced to enemy attack, is responsible. Australia from New Zealand and released in 1998. This paper provides a summary of two studies that This species is now well established in Tasmania and investigated potential natural enemies of T. lintearius Victoria (Ireson et al. 2003). T lintearius has been and S. staphylinus. In the P. persimilis development shown to reduce the growth of mature gorse bushes in study we compared the development of two strains of

101 Fourteenth Australian Weeds Conference

P. persimilis on diets of a pest and a benefi cial Tetrany- 147°01´E), was destructively sampled at an interval chus spp. (T. urticae and T. lintearius respectively). of 2–4 weeks for 12 months between December 2002 The natural enemy survey identifi ed potential natural and December 2003. To ensure that a range of micro- enemies of S. staphylinus on the fauna inhabiting habitats were sampled (including fl owers/fruit, new gorse. foliage and older growth) lower, mid and upper plant parts were collected from each of six plants on each MATERIALS AND METHODS sampling date. Phytoseiulus persimilis development study Two Plant material was then subjected to Tullgren strains of P. persimilis were reared from egg to adult funnel extraction for three days to remove arthro- on two diets in a randomised factorial design. The pods, which were collected into 120 mL plastic experiment was conducted in two separate controlled tubes containing 30 mL of 70% alcohol and a drop environment cabinets, which maintained the chosen of glycerol. temperatures (14 ± 0.7°C and 24 ± 0.7°C) and a daily Samples of mites (Acari) and thrips (Thysanop- photoperiod of 16 hours. tera) were cleared in Konos solution and slide mounted Rearing was conducted in perspex arenas similar in Berlese medium. All other arthropods were identi- to Perring and Lackey (1989). Individual P. persimilis fi ed from alcohol preserved specimens. All arthropods, eggs, less than two hours of age, were placed into except for microhymenoptera (several families in the each arena. Arenas were housed within 35 × 27 × 19 hymenopteran suborder Apocrita), were identifi ed to cm translucent plastic boxes containing 2 litres of family. Insects were identifi ed using the relevant keys saturated NaCl solution, which maintained relative in Naumann (1991) and mites (Acari) using keys in humidity between 75 and 76% at both temperatures Krantz (1978). Those families (and microhymenop- (Winston and Bates 1960). tera) that included other arthropods in their diet (Nau- The two strains of P. persimilis differed in their mann 1991, Krantz 1978) were classifi ed according source location and original host Tetranychus species. to their feeding habits: generalist predators – feed The ‘Tas’ strain was sourced from a fi eld site at Stone- across a range of arthropod genera; specialist preda- henge, Tasmania (42°24´S, 147°37´E) from within tors – feed on arthropods within a genus; parasitoids T. lintearius colonies on gorse. This strain appears – feed and develop within an arthropod host resulting to have naturalised and had been observed feeding in its death; omnivores – feed on both arthropods and on T. lintearius for more than 18 months (Ireson et plant material. al. 2003). The ‘NSW’ strain was supplied by the ‘Benefi cial Bugs Co.’ (www.benefi cialbugs.com.au) RESULTS AND DISCUSSION and reared on T. urticae colonies on bean (Phaseolus Phytoseiulus persimilis development study There vulgaris) at Richmond, NSW. Eggs of P. persimilis was a signifi cant difference in the development rate were collected using the method detailed in Perring and between the two strains at 14°C (P <0.05). However, Lackey (1989). At each temperature 40 P. persimilis separation of means showed that a signifi cant dif- eggs of each strain were used. ference between the strains occurred only on a diet Diets consisted of eggs of either T. lintearius or of T. urticae (Table 1). On this diet the ‘Tas’ strain T. urticae, which were reared and extracted from either completed its development approximately half a day gorse or bean respectively using a method detailed in earlier than the ‘NSW’ strain. Diet had no effect on Ireson et al. (1999). Half of each P. persimilis strain the development rate at this temperature. In contrast, (20) were allocated to each diet. the development rate at 24°C was signifi cantly affected Development times from egg through to adult were by diet (P <0.05, Table 1) and not by strain. At this determined by counting the number of cast skins in temperature, both strains completed their development each arena at each observation every 12 hours. Mites half a day earlier on a diet of T. lintearius. that died during or just after a moult were considered One of the reasons that particular specialist natural to have achieved the more advanced life stage. enemies such as P. persimilis, other phytoseiid mites Statistical tests were performed using GENSTAT and certain parasitic Hymenoptera are successful 6th edition. Data on the development time of P. persi- biological control agents of agricultural pests is be- milis were independently subjected to ANOVA for each cause they develop at a rapid rate relative to their prey temperature and LSD’s were calculated to separate (Dixon et al. 1997). Tetranychus lintearius completes treatment means. its pre-adult development in 38.2 days at 15°C and 15.3 days at 25°C (Stone 1986). However, we have Natural enemy survey Gorse plant material, from shown that P. persimilis completes its development in a field site near Lymington, Tasmania (43°11´S, less than half this time at slightly lower temperatures.

102 Fourteenth Australian Weeds Conference

Table 1. Mean ± SE egg to adult development time (days) for ‘NSW’ and ‘Tas’ strains of P. persimilis reared on diets of T. urticae (two-spotted mite) and T. lintearius (gorse spider mite) at 14°C and 24°C. The numbers in parentheses represent the number of individuals comprising the mean. Diet Temperature Strain T. urticae T. lintearius 14°C NSW 18.0 ± 1.2 a1 (13) 18.0 ± 1.3 a (10) Tas 17.4 ± 1.0 b (16) 17.7 ± 1.1 ab (15) 24°C NSW 6.0 ± 0.6 a (18) 5.6 ± 0.6 b (15) Tas 6.1 ± 0.6 a (18) 5.6 ± 0.6 b (17) 1 Means with the same letters for each temperature are not signifi cantly different (P <0.05; LSD).

Table 2. Potential arthropod natural enemies collected from gorse at Lymington December 2002–December 2003. Natural enemy classifi cation1 Arthropods GP SP Om Par Insects: Phlaeothripidae X X Thripidae X Microhymenoptera X Mites: Tydeidae X Phytoseiidae X X Cheyletidae X Anystidae X Erythraeidae X Ascidae X X Bdellidae X Cunaxidae X 1 GP = generalist predator; SP = specialist predator; Om = omnivore; Par = parasitoid.

Pels and Sabelis (1999) found predation of T. urticae The interaction between plants, phytophagous ar- by P. persimilis resulted in localised extinction, but in thropods and generalist predators or omnivores is highly a natural environment prey dispersal and asynchrony complex and therefore diffi cult to predict. However, in predator and prey populations will result in a patchy these interactions could affect the population dynam- equilibrium, with alternating prey overpopulation and ics of phytophagous arthropods used as weed biological localised extinction occurring. As P. persimilis will de- control agents. Both individual species and groups of velop on a diet of T. lintearius at a similar or more rapid generalist predators have been shown to reduce both rate than it would on a diet of the pest T. urticae, we pre- arthropod numbers and plant damage (Symondson et dict similar predator-prey dynamics to occur. Therefore, al. 2002). Similarly, omnivores that include arthropods we expect that P. persimilis is having a negative impact as part of their diet can contribute to the stability of on T. lintearius populations, resulting in a reduction of phytophagous arthropod populations (Coll and Guer- the effi cacy of this weed biological control agent. shon 2002). In fact, there is evidence to suggest that local generalist predators can result in the extinction Natural enemy survey A number of potential natural of certain introduced weed biological control agents in enemies were collected from gorse during the course the country of release (Ireson et al. 2002). of this study (Table 2). Of these, members of the fol- Populations of S. staphylinus have established in lowing families are reported as predators of members Tasmania, however, they have been slow to spread and of the family Thripidae, of which the biocontrol agent have been found only within a few metres of the central S. staphylinus is a member: Phlaeothripidae, Thripidae, release point up to three years after release (Ireson Phytoseiidae, Cheyletidae, Anystidae and Erythraeidae et al. 2004). Perhaps the range of potential natural (Sabelis and Van Rijn 1997). In addition, hymenopteran enemies identifi ed during this study are having an egg and larval parasitoids also attack members of the impact on the effi cacy of this agent and could explain family Thripidae. Those reported are all members of the slow spread if they are attacking newly established the superfamily Chalcidoidea (Loomans et al. 1997), S. staphylinus colonies. A natural enemy exclusion many species of which are microhymenopterans. experiment is needed to test this hypothesis.

103 Fourteenth Australian Weeds Conference

ACKNOWLEDGMENTS Australia. Proceedings of the XI International We thank Richard Holloway (Tasmanian Institute of Symposium on Biological Control of Weeds, ed. Agricultural Research) for technical support, Andy J.M. Cullen (in press). Ryland (Benefi cial Bugs Co.) for providing the NSW Krantz, G.W. (1978). ‘A manual of Acarology’, 2nd strain of Phytoseiulus persimilis and Lynne and David edition. (Oregon State University Book Stores, Dowe for providing access to the study site for the Oregon, USA). natural enemy survey. Funding assistance was provided Loomans, A.J.M., Murai, T. and Greene, I.D. (1997). by the CRC for Australian Weed Management. Interactions with hymenopterous parasitoids and parasitic nematodes. In ‘Thrips as crop pests’, REFERENCES ed. T. Lewis, pp. 355-97 (CAB International, Coll, M. and Guershon, M. (2002). Omnivory in ter- Wallingford, UK). restrial arthropods: mixing plant and prey diets. McFayden, R.C. and Spafford-Jacob, H. (2004). Insects Annual Review of Entomology 47, 267-97. for the biological control of weeds: Predicting para- Dixon, A.F.G., Hemptinne, J.L. and Kindlmann, P. sitism levels in the new country. Proceedings of the (1997). Effectiveness of ladybirds as biological XI International Symposium on Biological Control control agents: Patterns and processes. Ento- of Weeds, ed. J.M. Cullen (in press). mophaga 42, 71-83. McMurtry, J.A. and Croft, B.A. (1997). Life-styles of Galazzi, D. and Nicoli, G. (1996). Comparative study Phytoseiid mites and their role in biological con- of strains of Phytoseiulus persimilis 1. Develop- trol. Annual Review of Entomology 42, 291-321. ment and adult life. Bollettino dell’Istituto di En- Naumann I.D. (Chief editor) (1991). ‘The insects of tomologia ‘Guido Grandi’ della Universita degli Australia – a textbook for students and research Studi di Bologna 50, 243-52. workers’, 2nd edition, Volumes 1 and 2. (Mel- Goeden, R.D. and Louda, S.M. (1976). Biotic inter- bourne University Press, Melbourne). ference with insects imported for weed control. Pels, B. and Sabelis, M.W. (1999). Local dynamics, Annual Review of Entomology 21, 325-42. overexploitation and predator dispersal in an Hamamura, T., Shinkaji, N. and Ashihara, A. (1976). acarine predator-prey system. Oikos 86, 573-83. The relationship between temperature and devel- Perring, T.M. and Lackey, L.J. (1989). Temperature opmental period, and oviposition of Phytoseiulus and humidity effects on mortality and pre-adult persimilis. Bulletin of the Fruit Tree Research development of two Phytoseiulus persimilis strains Station 1, 117-25. (Acari: Phytoseiidae). International Journal of Ac- Ireson, J.E., Davies, J.T., Chatterton, W.S. and Hollo- arology 15, 47-52. way, R.J. (2002). Attempts to establish biological Pratt, P.D., Coombs, E.M. and Croft, B.A. (2003). control agents for boneseed in Tasmania. Proceed- Predation by Phytoseiid mites on Tetranychus ings of the 13th Australian Weeds Conference, eds lintearius, an established weed biological control H. Spafford Jacob, J. Dodd and J.H. Moore, pp. agent of gorse (Ulex europaeus). Biological Con- 407-11. (Plant Protection Society of WA, Perth). trol 26, 40-7. Ireson, J.E., Gourlay, A.H., Kwong, R.M., Holloway, Sabelis, M.W. and Van Rijn, P.C.J. (1997). Predation by R.J. and Chatterton W.S. (1999). Progress on the insects and mites. In ‘Thrips as crop pests’, ed. T. rearing, release and establishment of the gorse Lewis, pp. 259-354. (CAB International, UK). spider mite, Tetranychus lintearius Dufour, for Stone, C. (1986). An investigation into the morphol- the biological control of gorse in Australia. Pro- ogy and biology of Tetranychus lintearius Dufour ceedings of the 12th Australian Weeds Conference, (Acari: Tetranychidae). Experimental and Applied eds A.C. Bishop, M. Boersma and C.D. Barnes, pp. Acarology 2, 173-86. 320-4. (Tasmanian Weed Society, Hobart). Symondson, W.O.C., Sunderland, K.D. and Green- Ireson, J.E., Gourlay, A.H., Kwong, R.M., Holloway, stone, M.H. (2002). Can generalist predators R.J. and Chatterton W.S. (2003). Host specifi city, be effective natural enemies? Annual Review of release and establishment of the gorse spider mite, Entomology 47, 561-94. Tetranychus lintearius Dufour (Acarina: Tetrany- Thorp, J.R. (1999). Weeds of National Signifi cance chidae), for the biological control of gorse, Ulex – guidelines for developing weed strategies. europaeus L. (Fabaceae) in Australia. Biological (National Weeds Strategy Executive Committee, Control 26, 117-27. Australia, available on www.weeds.org.au) Ireson, J.E., Kwong, R.M., Gourlay, A.H., Davies, Winston, P.W. and Bates, D.H. (1960). Saturated so- J.T., Chatterton, W.S. and Holloway, R.J. (2004). lutions for the control of humidity in biological Progress on the biological control of gorse in research. Ecology 41, 232-7.

104