The molecular basis of host plant selection in Melaleuca quinquenervia by a successful biological control agent

Melaleuca quinquenervia (Cav) S.T. Blake (broadleaf paperbark) is an Australian tree that has become a serious weed in many places around the world. Two insects Oxyops vitiosa (the melaleuca weevil), and Boreioglycaspis melaleucae (the melaleuca psyllid), which were introduced to Florida as part of a biological control programme, have been very effective in reducing survival and reproduction of this weed. There are two terpene chemotypes of M. quinquenervia; one rich in the sesquiterpene E-nerolidol whereas the other is rich in viridiflorol. Viridiflorol is a strong feeding deterrent for the melaleuca weevil and retards larval development. The larvae therefore avoid the viridiflorol-rich chemotype, in contrast, female melaleuca psyllids prefer to oviposit on these leaves. To identify the molecular basis of these preferences, we isolated and characterised two terpene synthases from the viridiflorol-rich chemotype, both of which utilise farnesyl pyrophosphate and have the same product profile. Chemotypic variation in terpenes in M. quinquenervia is under strong genetic control and the reproductive potential of each chemotype is limited by a different insect. These insects could, therefore, be selective agents for the maintenance of chemotypic variation in M. quinquenervia.     

Chemical variation in the leaf essential oil of Melaleuca quinquenervia (Cav.) S.T. Blake

An examination of the leaf oils of Melaleuca quinquenervia over its geographical range in Australia and Papua New Guinea has shown wide variation in chemical composition but only two major chemotypes. Chemotype 1 is comprised of E-nerolidol (74–95%) and linalool (14–30%) and is found from Sydney, north along the east coast of Australia to Selection Flat, New South Wales, with an isolated occurrence near Maryborough, Queensland. Two divisions occur in this chemotype which are based on the presence or absence of significant proportions of linalool (14–40%). Chemotype 2 contains 1,8-cineole (10–75%), viridiflorol (13–66%), α-terpineol (0.5–14%) and β-caryophyllene (0.5–28%) in varying proportions and order of dominance in the oils. It is found throughout the distribution of the species, from Sydney to Papua New Guinea and New Caledonia. Within chemotype 2 there appears to be a continuous spread of oil composition without formation of any further discrete divisions as in chemotype 1.Analyses have shown that M. quinquenervia trees that occur at latitudes south of 25°S have high oil yields (1–3% w/w%, fresh leaves) and comprise chemotypes 1 and 2. North of 25°S, however, chemotype 1 does not occur and oil yields amongst the Australian populations are uniformly low (0.1–0.2%).     

Maintenance of a narrow host range by Oxyops vitiosa; a biological control agent of Melaleuca quinquenervia

Host range expansion in insect herbivores is often thought to be mediated by several factors, principal among them are secondary plant metabolites. In weed biological control, the host range of a prospective agent is one of the most important considerations in its implementation. Extensive host testing tests seek to determine the behavioral acceptance and nutritional value of different test plant species to the potential agent. A list of test plants is compiled that comprises species that are close taxonomic relatives of the target weed plus other species of economic or ecologic importance. The host testing of the Melaleuca quinquenervia biological control agent Oxyops vitiosa indicated that larvae would accept and complete development on the Australian target weed M. quinquenervia, two Australian ornamental species, Callistemon citrina, Callistemon viminalis (all Myrtaceae). However, the larvae did not complete development when fed a North American species Myrica cerifera (Myricaceae). The study reported here confirms these results and examines the nutritional and performance differences in O. vitiosa larvae fed leaves of these species. The leaf quality factors, percent moisture, percent nitrogen, toughness, and terpenoid content were related to larval survival, performance and digestive indices. The results indicate that plant quality among the Myrtaceae species was generally similar and correspondingly larval survival, performance and digestive indices differed little when larvae were fed leaves of these species. However, significant differences occurred in the plant quality of the North American M. cerifera compared with the Australian species which had leaves with the lowest percent moisture, lowest leaf toughness, highest percent nitrogen. This species, however, is not a physiological host as none of the neonates survived to pupate. When third instars were switched to M. cerifera from their normal host M. quinquenervia reductions were found in survival, biomass gain, digestive efficiency, and conversion of digested food to insect biomass. The marginal acceptance of this North American native plant in laboratory bioassays appears related to the terpenoid chemistry that has similarities to the taxonomically unrelated host M. quinquenervia. However, the high larval mortality corresponds to several novel terpenoids that are not present in the host. For weed biological control host testing these results indicate that M. cerifera is a poor host for O. vitiosa. Additionally, future test plant lists should include plants with secondary metabolites similar to the target weed as these compounds may constitute behavioral cues that are relevant to these specialized herbivores.     

Host Plant Quality Factors That Influence the Growth and Development of Oxyops vitiosa, a Biological Control Agent of Melaleuca quinquenervia

The Australian weevil Oxyops vitiosa was released in 1997 in Florida as a biological control agent of Melaleuca quinquenervia. The larvae of this agent are flush-feeders, found only on the growing tips of their host. Knowledge of this restriction to feeding on the growing tips and other nutritional requirements may assist in the establishment and dispersal of this species. Therefore, O. vitiosa survival was assessed when neonates were fed M. quinquenervia leaves from branches that had dormant buds or emerging bud leaves. Additionally, the influence of leaf quality from different sites and within sites was determined by the feeding of neonates emerging bud leaves collected at three sites and from three leaf qualities (poor, intermediate, and high). Within-site leaf qualities were described in the field by leaf color and in the laboratory by percentage dry mass and nitrogen. Larval survival was lowest when fed leaves from branches that had dormant buds. Associated with this low survival were high leaf toughness and percentage dry mass. When larvae were fed emerging bud leaves, most of the variation in larval survival and performance was attributed to differences in within-site plant quality. Generally, the highest-quality leaves had relatively low percentage dry mass and high percentage nitrogen. Larval survival generally decreased when fed the poor-quality leaves, and in one site, the intermediate-quality leaves. Larvae required less time to develop to adults when fed the high-quality leaves. Development time increased in females but not in males when the larvae were fed the poor-quality leaves. Adult biomass of both females and males generally increased when the larvae were fed the high-quality leaves from two of the three sites. The results indicate that the larvae of O. vitiosa are restricted to feeding on flush foliage with low toughness. Additionally, variations in foliar percentage dry mass and nitrogen influence larval survival and performance. This knowledge benefited the development of mass-production nursery sites and the selection of suitable release sites, which facilitated the establishment of this biological control agent.     

Circadian and seasonal study of the cinnamate chemotype from Lippia origanoides Kunth

The leaves of Lippia origanoides Kunth are used in culinary as flavoring regional dishes and remedy for gastrointestinal and respiratory diseases in the Amazon region. The circadian and seasonal study of its essential oil was characterized by GC and GC–MS analysis. The oil components were grouped into monoterpenes, sesquiterpenes and phenylpropanoids, during the dry and rainy season. The main constituents were (E)-methyl cinnamate, (E)-nerolidol, p-cymene, 1,8-cineole, carvacrol, α-pinene, (E)-caryophyllene and γ-terpinene, with great variation throughout the year. In this work, we are reporting the occurrence of a new chemotype for L. origanoides, characterized by an essential oil rich in (E)-methyl cinnamate and (E)-nerolidol, with fruity-woody odor, reminiscent of cinnamon, strawberry and wood. The oil yield varied from 1.7% to 4.6%, which is considered a significant value for the production of essential oils on an industrial scale. This new chemotype may have ecological, chemosystematics and taxonomic significance in the management and economic utilization of the species.     

Essential oil polymorphism of Thymus praecox subsp. arcticus on the British Isles

The essential oils of 732 individual plants of Thymus praecox Opiz subsp. arcticus (E. Durand) Jalas (syn. T. drucei Ronn.) collected in Scotland, Ireland, and in the south of England have been analysed by gas chromatography (GC) and mass spectrometry (GC–MS) in order to elucidate the chemical character of this subspecies on the British Isles. In total, 69 components were identified, most of them monoterpenoids and sesquiterpenoids with hedycaryol, linalyl acetate, linalool, the germacradienols, trans-nerolidol, T-cadinol, and β-caryophyllene being the most important compounds. The analysis of the quantitative essential oil data by means of neural networks revealed that T. praecox subsp. arcticus growing in Britain is highly polymorphous. There were 17 chemotypes with the hedycaryol chemotype as the most frequent (24% of the plants), followed by the linalool/linalyl acetate chemotype (22% of the plants) and germacra-1(10),4-dien-6-ol chemotype (18% of the plants). It seems that each part of the British Isles has its special chemotype pattern with 13 chemotypes in Scotland, 11 in Ireland, and 17 in the south of England. An overview of the North Atlantic region of Europe revealed that the polymorphism of T. praecox subsp. arcticus in the essential oil is more distinctive in the southern than in the northern regions, with only 2, 5, and 1 chemotypes in Greenland, Iceland, and Norway, respectively.     

Resource regulation of an invasive tree by a classical biological control agent

The invasive tree Melaleuca quinquenervia experienced substantial declines in growth and reproduction in response to chronic herbivory by the defoliating weevil Oxyops vitiosa. Plants subjected to unrestricted defoliation replaced leaves that were more suitable for feeding by the next generation, a process envisioned by the Resource Regulation Hypothesis which posits that attack by one generation increases the amount of the preferred host resources for the next, resulting in a positive feedback loop for the herbivore. The production of juvenile replacement leaves stimulated additional bouts of oviposition and feeding by O. vitiosa, which ultimately produced positive effects for the herbivore with negative consequences for the plant. The addition of water resources to the plant prolonged the positive feedback loop such that more than twice as many insects were produced on irrigated versus non-irrigated trees. In a more simple, reassembled food web on M. quinquenervia, the lack of biotic constraints like parasitoids may have prevented the earlier termination of the feedback loop and thus increased the impact of the biological control agent on the target. The overall effectiveness of this classical biological control program can be attributed, in part, to the phenomenon of the target plant’s induced susceptible response to a herbivore.     

Herbivory alters resource allocation and compensation in the invasive tree Melaleuca quinquenervia

Abstract. 1. Plants may compensate for the effects of herbivory, especially under favourable growing conditions, limited competition, and minimal top‐down regulation. These conditions characterise many disturbed wetlands dominated by introduced plants, implying that exotic, invasive weeds in these systems should exhibit strong compensatory responses. 2. The Australian native Melaleuca quinquenervia is highly invasive in the Florida Everglades, U.S.A., where it experiences limited competition or herbivory from native species, making it a likely candidate for compensation. The introduced biological control agent Oxyops vitiosa feeds exclusively on the seasonal flushes of developing foliage at branch apices, which represents ≈15% of the total foliar biomass. 3. The hypothesis that M. quinquenervia compensates for folivory by O. vitiosa was tested in a series of field‐based experiments. Trees experiencing folivory over four consecutive years maintained similar levels of foliar biomass after attack yet possessed twice the number of leaf‐bearing terminal stems as undamaged trees. The biomass of these stems was similar among attacked and unattacked trees, indicating that herbivore‐damaged trees produce greater quantities of smaller terminal branches. However, undamaged trees were 36 times more likely to reproduce than herbivore‐damaged trees. 4. In a separate herbivore exclusion study, a single bout of herbivory on previously undamaged M. quinquenervia trees caused an 80% reduction in reproductive structures the following season. Herbivore‐damaged trees also possessed 54% fewer fruits than undamaged trees. An increase in the herbivory frequency (two bouts per year) or magnitude (100% simulated herbivory) did not result in a further reduction in fitness. 5. It has been concluded that M. quinquenervia partially compensates for herbivory by producing new stems and replacing foliage, but this compensation results in a substantial reduction in reproduction.     

Intraspecific volatile oil variation in Myrceugenia cucullata (Myrtaceae)

The analysis of volatile oil content of Myrceugenia cucullata, a treelet from southern Brazilian montane forests, was investigated. Two markedly distinct chemotypes were identified. One of them is dominated by (E)-nerolidol and the other by (−)-α- and (−)-β-pinenes. Field observations showed a distinct ecological behaviour of plants with distinct chemical compositions, and a closer morphological analysis hints to a possible separation of them. The diverse oil composition is discussed and the possibility of specific segregation of the chemotypes is commented.     

The influence of plant size on attack rates of Oxyops vitiosa,an introduced herbivore of Melaleuca quinquenervia

Oxyops vitiosa Pascoe (Coleoptera: Curculionidae) was released in Florida in 1997 as a biological control agent of the invasive tree Melaleuca quinquenervia (Cav.) S.T. Blake. Three years after release, the weevil’s dispersal into three replicated M. quinquenervia stands was monitored to determine if O. vitiosa preferentially selects larger trees during patch colonisation. Data indicated that tree size did not influence the likelihood of an attack by O. vitiosa and smaller individuals within host patches did not escape herbivory through enemy-free space.     

Demonstration and characterization of (E)-nerolidol synthase from maize: a herbivore-inducible terpene synthase participating in (3E)-4,8-dimethyl-1,3,7-nonatriene biosynthesis

Upon herbivore attack, maize (Zea mays L.) emits a mixture of volatile compounds that attracts herbivore enemies to the plant. One of the major components of this mixture is an unusual acyclic C₁₁ homoterpene, (3E )-4,8-dimethyl-1,3,7-nonatriene (DMNT), which is also emitted by many other species following herbivore damage. Biosynthesis of DMNT has been previously shown to proceed via the sesquiterpene alcohol, (E )-nerolidol. Here we demonstrate an enzyme activity that converts farnesyl diphosphate, the universal precursor of sesquiterpenes, to (3S)-(E )-nerolidol in cell-free extracts of maize leaves that had been fed upon by Spodoptera littoralis. The properties of this (E )-nerolidol synthase resemble those of other terpene synthases. Evidence for its participation in DMNT biosynthesis includes the direct incorporation of deuterium-labeled (E )-nerolidol into DMNT and the close correlation between increases in (E )-nerolidol synthase activity and DMNT emission after herbivore damage. Since farnesyl diphosphate has many other metabolic fates, (E )-nerolidol synthase may represent the first committed step of DMNT biosynthesis in maize. However, the formation of this unusual acyclic terpenoid appears to be regulated at both the level of (E )-nerolidol synthase and at later steps in the pathway. Received: 20 August 1999 / Accepted: 27 October 1999     

A comparative study of the essential oils of Cistus salviifolius in several populations of Crete (Greece)

The composition of the essential oils of fifteen populations of Cistus salviifolius L. (Cistaceae) from Crete (Greece) and their interpopulation variability was investigated by GC-MS. 167 compounds were identified representing an average of 96–100% of the oil composition. Labdane diterpenes were detected and identified in the essential oils of Cistus salviifolius L., for the first time. The results obtained from GC-MS analysis of the oils were submitted to a Principal Component Analysis (PCA). Three main chemotypes (clusters) were differentiated. All the essential oils were tested for their antimicrobial activity against Gram-positive and Gram-negative bacteria. Only chemotype III having camphor as a major constituent was active against Gram-positive bacteria. Camphor, viridiflorol, longiborneol, phyllocladene, abietatriene and cis-feruginol are the main constituents, while the group of oxygenated sesquiterpenes has the highest percentage composition.     

Antipredatory Activity of the Weevil Oxyops vitiosa: A Biological Control Agent of Melaleuca quinquenervia

The larvae of the leaf-feeding weevil Oxyops vitiosa, a biological control agent of Melaleuca quinquenervia, are covered with a viscous orange coating that is thought to protect against generalist predators. This coating is gradually lost as the larvae drop to the ground and pupate in subterranean pupal cells. To test the antipredator activity of this species, four immature life stages (early instars, late instars, prepupae, pupae) were exposed to a common generalist predator, the red imported fire ant Solenopsis invicta. Choice tests were conducted by placing an O. vitiosa individual and a control larva of the weevil Neochetina eichhorniae into an arena containing a S. invicta colony and observing subsequent ant behaviors. S. invicta workers contacted O. vitiosa early instars, late instars, and prepupae less frequently than control N. eichhorniae larvae, and upon contact S. invicta was less likely to behave aggressively toward these O. vitiosa life stages than toward N. eichhorniae larvae. However, S. invicta contacted, attacked, and consumed naked (nonencased) O. vitiosa pupae and N. eichhorniae larvae with equal frequency. Encased O. vitiosa pupae buried in sand were not attacked compared to susceptible encased pupae on the sand surface. By shifting from a chemical defense during the larval stages to a physical defense during the pupal stage, O. vitiosa reduces the risk of attack by this generalist predator.     

Modeling the Influence of Resource Availability on Population Densities of Oxyops vitiosa (Coleoptera: Curculionidae), a Biological Control Agent of the Invasive Tree Melaleuca quinquenervia

The Australian snout beetle, Oxyops vitiosa, was introduced to south Florida, USA, as a biological control agent of the invasive tree Melaleuca quinquenervia during the spring of 1997. As part of ongoing post-release evaluations of this weevil, we sought to quantify the population density that can be maintained by O. vitiosa larvae when fully exploiting the available melaleuca foliage. Seasonal population densities were modeled as a function of tree size distribution and density, plant phenology, quantity of acceptable foliage (suitable for larval development) and larval consumption values. Larval densities varied dramatically due to fluctuating resource availability and ranged from 830 946 larvae/ha in July–August to >4.5 million larvae/ha in January–February. The expected population density for a generalized (mixed sized) tree distribution was estimated to be 13.9 million larvae/ha. Larval densities increased to 18.8 million larvae/ha per year when large trees (>20 cm diameter at breast height) represented the dominant size class, whereas habitats dominated by medium (10–19 cm) and small (1.3–9 cm) trees were predicted to have larval densities of 11.3 and 6.3 million/year, respectively. Validation data obtained from smaller, more accessible plants suggested that the model overestimated realized yearly larval densities by 9% or 15 416 larvae/ha per year.     

Biocontrol without borders: the unintended spread of introduced weed biological control agents

An underlying assumption of classical biological control implies that intentionally introduced natural enemies will remain within the boundaries that delineate the program’s area of implementation. A weed biological control program targeting Melaleuca quinquenervia in Florida, USA has resulted in the release and establishment of Oxyops vitiosa and Boreioglycaspis melaleucae. An international survey of M. quinquenervia populations in 13 other states or countries where the insects have not been intentionally introduced was initiated to monitor the long range dispersal of O. vitiosa and B. melaleucae beyond the herbivores’ intended geographic range (Florida). Surveys in 2006 resulted in the discovery of B. melaleucae within the canopies of several M. quinquenervia trees near San Juan, Puerto Rico. In 2007, O. vitiosa was observed on the island of New Providence in the Bahamas but neither herbivore was detected on nearby Grand Bahama or Andros islands. In 2009, B. melaleucae was observed attacking M. quinquenervia trees in Los Angeles, California (USA). The herbivores have not been detected on other surveyed M. quinquenervia populations in Cuba, Jamaica, Texas (USA), Costa Rica, Brazil, Hawaii (USA) or South Africa. There is no evidence to suggest that herbivore colonization of New Providence, Puerto Rico, or California was influenced by linear distance between Florida and the recipient M. quinquenervia stand. While the dispersal pathway(s) remains unknown, biological control agents were detected from 200 to >3500 km from their original release location (Florida) and at locations that have strong links via tourism and trade as indicated by the number of airline flights connecting south Florida with colonized tree populations. Implications of this unintended spread are discussed in relation to permeability of biogeographical barriers and risk assessment of biological control agents.     

Intra-population terpene polymorphism of Thymus pulegioides L.: Evidence for seven chemotypes in a German limestone grassland

For Thymus pulegioides L. (Lamiaceae), occurring in almost entire Europe, about 20 different essential oil chemotypes are described in approximately 25 studies. However, only few studies mention chemotype diversity on population level, describing up to five chemotypes growing together. The aim of the study was to investigate the chemotype diversity within one T. pulegioides population of a limestone grassland in Germany. Essential oil compounds from samples of 93 thyme cushions were extracted with solid phase extraction (SPE) and analysed by GC-MS. Cluster analysis and nonmetric multi-dimensional scaling (NMDS) of the samples’ essential oil composition revealed the existence of seven different chemotypes: linalyl acetate-chemotype (56 individuals), geraniol-citral-linalyl acetate-chemotype (5), geraniol-citral-chemotype (7), thymol-chemotype (8), carvacrol-chemotype (2), linalool-chemotype (1), and β-caryophyllene-germacrene D-β-bisabolene-chemotype (14). The pattern of major monoterpenes from all chemotypes could be explained by the inhibition of specific steps in monoterpene biosynthesis. The sesquiterpene-dominant chemotype might be caused by a block in a very early step of monoterpene production or a regulatory/channeling mechanism. While the geraniol-citral-linalyl acetate-chemotype has been found for the first time, the other chemotypes have been found in other regions and the β-caryophyllene-germacrene D-β-bisabolene-type is similar to other sesquiterpene chemotypes rarely found in other areas. The large intra-population chemical polymorphism in this study raises questions about the chemotype diversity of thyme in other regions of Germany and potential correlations between essential oil composition and abiotic factors or biotic interactions.     

Comparative toxicity,growth inhibitory and biochemical effects of terpenes and phenylpropenes on Spodoptera littoralis (Boisd.)

Eleven monoterpenes, phenylpropenes and sesquiterpenes were evaluated for their insecticidal and growth inhibitory activities against the second and fourth larval instars of Spodoptera littoralis. Among the tested compounds, 1,8-cineole revealed the highest fumigant toxicity against the 2nd and 4th larval instars with LC₅₀ values of 2.32 and 3.13 mg/L air, respectively. The monoterpenes, p-cymene, α-terpinene, (−)α-pinene and (−)-carvone were highly toxic to both larval stages as their LC₅₀ values ranged between 7.35 and 13.79 mg/L air against 2nd larval instar and between 14.66 and 32.02 mg/L air against 4th larval instar. In topical application assay against the 4th larval instar, (−)-carvone (LD₅₀ = 0.15 mg/larva) and cuminaldehyde (LD₅₀ = 0.27 mg/larva) were the most potent contact toxicants. In residual film assay, trans-cinnamaldehyde, (−)-citronellal and p-cymene showed the highest insecticidal activity against the 2nd larval instar, while α-terpinene and (−)-carvone were most effective compounds against the 4th larval instar. Moreover, the tested compounds caused strong growth reduction of both larval stages with growth inhibition higher than 80% in the 2nd larval instar and higher than 70% in the 4th larval instar. On the other hand, (−)-carvone, cuminaldehyde and (Z,E)-nerolidol showed pronounced inhibitory effects on acetylcholinesterase (AChE) and adenosine triphosphatases (ATPases) activity of S. littoralis larvae. Cuminaldehyde (IC₅₀ = 1.04 mM) and (Z,E)-nerolidol (IC₅₀ = 0.02 mM) caused the highest inhibition of AChE and ATPases, respectively. Taken together, the results indicate that monoterpenes, phenylpropenes and phenylpropenes could be used to develop new botanical insecticides for S. littoralis management.     

Genetically-controlled leaf traits in two chemotypes of Salix sachalinensis Fr. Schm (Salicaceae)

Salix sachalinensis has two chemotypes: one biosynthesises ampelopsin as a major component of low molecular weight phenolics in their leaves (A-type), and the other biosynthesises β-d-glucopyranose-1-trans-p-coumarate (PG1) and β-d-glucopyranose-1-trans-cinnamate (PG2) in addition to ampelopsin (AP-type). We investigated phenotypic and genetic variations and clonal repeatabilities of the pubescence density, leaf mass per area (LMA), and concentrations of total phenolics, condensed tannin, ampelopsin, PG1 and PG2. Leaves of wild A-type trees contained significantly higher concentrations of total phenolics and ampelopsin, and lower concentration of condensed tannin than those of wild AP-type trees. In the greenhouse experiment that compared leaf traits between cloned trees obtained from wild chemotypes, there were significant between-type variations in the leaf phenolic concentrations, pubescence density, and LMA. Since chemotypes of cloned trees in the greenhouse were the same as those of wild parent trees, chemotype can be considered as a genetically controlled property. There were also significant within-chemotype variations in the pubescence density, LMA, total phenolics, ampelopsin, PG1, and PG2 concentrations, but not in concentration of condensed tannin for either chemotypes. Genetic variation of leaf traits except for LMA in AP-type was significant. PG1 and PG2 exhibited the highest clonal repeatabilities (0.73 and 0.78, respectively). Thus, the ability to produce and the amount of production of PG1 and PG2 are genetically controlled.