J. Appl. Entomol. ORIGINAL CONTRIBUTION Role of olfaction in host plant selection and local adaptation of a polyphagous herbivore, Eucolaspis Sharp P. R. C. Doddala1,*, M. A. Minor1,a, Q. Wang1, D. J. Rogers2, E. M. Koot1 & S. A. Trewick1,a 1 Institute of Agriculture and Environment, Massey University, Palmerston North, New Zealand 2 The New Zealand Institute for Plant and Food Research Limited, Havelock North, New Zealand Keywords Abstract attraction, bronze beetle, feeding preference, host plant volatiles Host plant cues are known to shape insect–host plant association in many insect groups. More pronounced associations are generally manifested in Correspondence specialist herbivores, but little is known in generalist herbivores. We used P. R. C. Doddala (corresponding author), a polyphagous native beetle from New Zealand, bronze beetle, Eucolaspis Institute of Agriculture and Environment, sp. ‘Hawkes Bay’ (Chrysomelidae: Eumolpinae) to explore the role of Massey University, 4442 Palmerston North, olfaction in locating host plants and local adaptation. We also tested the New Zealand. E-mail: [email protected] role of other cues in the degree of acceptance or rejection of hosts. Adult *Present address: Division of Chemical Eucolaspis beetles were attracted to fresh leaf volatiles from apple and Ecology, Department of Plant Protection blackberry (Rosaceae). Male and female beetles responded similarly to Biology, Swedish University of Agricultural olfactory cues of host plants. An indication of evolutionary affiliation was Sciences, 23053, Alnarp, Sweden observed in olfactory preferences of geographically isolated conspecific populations. We found that geographically isolated populations of the Received: June 7, 2015; accepted: September beetles differ in their olfactory responses and exhibit some degree of local 23, 2015. adaptation. However, irrespective of geographical and ecological associa- doi: 10.1111/jen.12279 tions, blackberry was preferred over apple as a feeding plant, and another novel plant, bush lawyer (Rubus australis), was readily accepted by aContributed equally. 53.25% of the tested beetles. We show that plant volatiles play an impor- tant role in host location by Eucolaspis, but the acceptance or rejection of a particular host could also involve visual and contact cues. (e.g. Cymbidium) to conifers (e.g. Pinus). It appears Introduction that the beetles have a particular preference for exotic Eucolaspis, a native Eumolpinae (Coleoptera: fruit plants of the family Rosaceae, and orchards often Chrysomelidae) genus of New Zealand, comprises support large populations resulting in significant eco- polyphagous species commonly called as ‘bronze bee- nomic damage. Fruit crops from families other than tles’. Recent findings on morphology, distribution and Rosaceae such as avocado (Lauraceae), black currant phylogeny of Eucolaspis suggest fewer endemic species and gooseberry (Grossulariaceae), blueberry (Eri- (~4) than formerly described (~15) (Doddala 2012). caceae) and feijoa (Myrtaceae) are also damaged. Eucolaspis beetles were first reported on flowers of a Early records suggest that host-range expansion in New Zealand native shrub, manuka (Leptospermum sco- Eucolaspis may have occurred soon after European col- parium, Myrtaceae) (Fabricius 1781 in White 1846). onization and introduction of fruit crops in New Zeal- The beetles of the genus have since been recorded and (Huntley 1867). feeding on at least 67 other plant species in New Zeal- Infestation of host plants by different species of and, including 27 natives and 40 exotics (Table S1). Eucolaspis depends on local and regional distribution, Among the 33 plant families represented, the most as revealed from host plant use by different genetic common hosts are Myrtaceae and Rosaceae (11 and 6 lineages in New Zealand (Doddala 2012). We also sus- species, respectively), but host plants range from small pect some level of intraspecific variation in host plant pasture herbs (e.g. Trifolium) and terrestrial orchids use by bronze beetles, as seen in other leaf beetles 444 J. Appl. Entomol. 140 (2016) 444–452 © 2015 Blackwell Verlag GmbH P. R. C. Doddala et al. Host plant associations of Eucolaspis such as Chrysomela lapponica (Coleoptera: Chrysomeli- Materials and Methods dae) (Zvereva et al. 2010) and Agelasa nigriceps (Coleoptera: Chrysomelidae) (Kohyama et al. 2012). Normal seemingly healthy (showed no impairment We are particularly interested in the Eucolaspis sp. in walking, flying and feeding) adult beetles were ‘Hawkes Bay’ lineage, which infests apple trees and used for all bioassays. The test beetles originated also occurs on exotic blackberry, linden, white clover, from two different populations – an ‘apple popula- broad-leaved dock, native manuka and totara trees tion’ from an organic apple orchard (‘Royal Gala’) and possibly other plant species (Doddala 2012). in Havelock North (Hawkes Bay, New Zealand) and Long-term association with a host plant might lead a ‘blackberry population’ from a blackberry scrub in to stronger preference, through tolerance, detoxifica- Waikanae (Kapiti Coast, New Zealand). The two tion and recognition, which can have important populations used will be referred as ‘Havelock North implications for infestation of crop species and other apple’ and ‘Waikanae blackberry’ populations from hosts (Jermy 1984; Mostafa et al. 2011). Some geo- this point forward. Conspecificity of these two pop- graphically isolated conspecific populations of Euco- ulations was confirmed through the analysis of laspis sp. ‘Hawkes Bay’ persist on different host plants mitochondrial DNA and male genitalia (Doddala (apple and blackberry), suggesting local adaptation. It 2012). Insect collection, rearing and maintenance is not known whether these allopatric conspecific have been followed as previously described (Dod- populations exhibit differences in sensory abilities in dala 2012). Beetles used in the study were starved locating and selecting respective host plants. Any such for 4 h prior to bioassays. Starving is necessary to differences could impact pest control methods such as induce host finding or feeding and is a widely used behavioural control exploiting host plant stimuli. practice in similar experiments with herbivorous Hence, from both economic and scientific perspec- insects (Garcıa-Robledo and Horvitz 2009). tives, it is important to assess intraspecific differences The role of olfaction in host detection was tested in host plant choice and location in Eucolaspis sp. initially with Havelock North apple population beetles ‘Hawkes Bay’ beetles leading to local adaptation. in Y-tube bioassays (no-choice: host plant vs. clean In Chrysomelidae, cues that mediate attraction to air) using four plant species: apple (Malus domesticus host plants include plant size (Colorado potato beetle CV ‘Royal Gala’, Rosaceae), blackberry (Rubus fruitico- Leptinotarsa decemlineata – (Hoy et al. 2000)), other sus, Rosaceae), white clover (Trifolium repens, Faba- visual cues (e.g. L. decemlineata – (Szentesi et al. 2002) ceae) and broad-leaved dock (Rumex obtusifolius, and Phyllotreta striolata – (Yang et al. 2003)) and plant Polygonaceae). White clover and broad-leaved dock volatiles (e.g. L. decemlineata – (Bolter et al. 1997; are predominant under-storey plants in organic apple Schutz€ et al. 1997) and the strawberry leaf beetle, orchards in New Zealand, and both these plant species Galerucella vittaticollis – (Hori et al. 2006)). Among dif- are infested by bronze beetles in the field (P. Doddala, ferent host-finding cues, olfactory cues are the most pers. obs.). Fresh undamaged leaves of test plants frequently used (Stenberg and Ericson 2007). At least were used as host plant stimulus, and clean air was twelve leaf beetle genera show attraction to host plant used as control. volatiles (Fernandez and Hilker 2007; Stenberg and Intraspecific differences in olfaction were tested, Ericson 2007; El-Sayed 2014). None of the previously using Havelock North apple and Waikanae black- studied genera belongs to the subfamily Eumolpinae, berry populations that have different geographical which remains poorly studied in this respect. and host plant association, by dual-choice bioassays We examined preference for host plants among dif- (apple vs. blackberry) in the Y-tube olfactometer. In ferent populations of bronze beetles and assessed the addition, two sets of feeding bioassays were con- extent of preference as determined by evolutionary ducted either as a dual-choice (apple and black- association. The specific objectives of this research berry) or a tri-choice (apple, blackberry and bush were to evaluate whether the bronze beetles show lawyer) assay with the two test beetle populations. attraction to host plant odours and to explore whether Bush lawyer (Rubus australis) is a New Zealand allopatric conspecific populations vary in sensitivity to native Rosaceae that has never been reported as a olfactory, gustatory and other contact cues in host host plant for the beetles; we included this plant to plant preference. We conducted olfactory and feeding test how a novel host would influence meal compo- bioassays involving various known host plants and a sition. All the experiments were conducted during novel plant, with two conspecific populations of Euco- the active seasons (New Zealand spring – summer) laspis sp. ‘Hawkes Bay’ adult beetles. 2009–2013. J. Appl. Entomol. 140 (2016) 444–452 © 2015 Blackwell Verlag GmbH 445 Host plant associations of Eucolaspis P. R. C. Doddala et al. Post-feeding,