Detection and monitoring of pink bollworm moths and invasive insects using pheromone traps and encounter rate models

Modelling push-pull management of pest insects using repellents and attractive traps in fruit tree orchards

BACKGROUND

Push-pull with semiochemicals in pest management uses repellents to reduce response of pests to food-mate resources (push) and attractive traps to reduce populations (pull). Simulation models of push-pull can aid understanding of plant-insect interactions in nature and suggest hypotheses for field tests that improve management. A previous model indicated advantages of push-pull for controlling ambrosia beetle, Euwallacea fornicatus, pest of avocado trees. However, the simulated behavior of repellency was inconsistent with field observations.

RESULTS

We simulated individual-based movement of insects in push-pull to reveal relationships between parameters of radii (strength) of attractive traps, pest aggregations, and repellents with densities of each in an avocado orchard to visualize and understand the interactions and significance. Simulations indicated placement of traps along a 1-ha area periphery as a barrier resulted in similar trapping and mating as when traps were in a grid, either when insects originated randomly inside the plot or came from outside the plot. However, when insects originated from outside, both arrangements caught slightly more than when insects originated within the plot.

CONCLUSION

There were no differences in capture rates whether traps were spaced in a barrier along the plot's periphery or in a grid covering the plot. Push-pull was more effective than pull (mass trapping) alone. Repellent behavior of individuals when encountering a repellent radius was modelled by approximate 90° turns (right or left at random) when about to enter an infestation, consistent with earlier observations of effects of repellents on bark beetles orienting to aggregation pheromone. © 2022 Society of Chemical Industry.

Index of host habitat preference explored by movement-based simulations and trap captures

Animal species likely have different strengths of host habitat preference (HHP) that might be characterized by a standardized index ranging from 0 (no preference) to 1 (maximum preference). We hypothesized that in some species, HHP may result from individuals dispersing out of the host habitat having a probability of turning back at the boundary, or after entering host habitat by reducing speed or increasing size of turning angles. Computer simulations of individuals moving between various sized patches of host and nonhost habitat were conducted based on these three behaviours hypothesized to affect HHP. In the rebounding model, simulations resulted in equilibria of animal numbers inside and outside of host habitat that depend on sizes of these areas, initial number and the rebounding probability. Curvilinear regression of simulation results suggested an equation that predicted numbers in the host habitat and was solved for rebounding probability. A modified equation that sampled population densities (e.g., insect pheromone trap catches) inside and outside host habitat areas gave the rebounding probability, an index of HHP, without requiring the sizes of the areas. Simulations with traps and moving animals verified that the modified equation could predict the index correctly. The modified equation also estimates an index of HHP from sampled densities due to speed reductions and a combination of this and rebounding. Changes in angular turning size upon entering host habitat, however, did not affect habitat preference. Using pheromone trap captures, we found that the lesser date moth Batrachedra amydraula has a HHP for date Phoenix dactylifera plantations of 0.96. Host habitat preference indexes also were calculated from sampled insect densities reported in the literature. The new index of HHP is useful to characterize habitat patches of many organisms and aid understanding of animal spatial distributions and speciation processes. In addition, the index can be applied in studies of invasive species, trap crops of pest insects and conservation management.

Inhibitory Effects of Semiochemicals on the Attraction of an Ambrosia Beetle Euwallacea nr. fornicatus to Quercivorol

The Euwallacea fornicatus (Eichhoff) species complex includes the polyphagous shot hole borer (PSHB), an ambrosia beetle infesting avocado limbs, Persea americana Mill. Synthetic quercivorol, a monoterpene alcohol, is known to attract females (males are flightless) over a range of release rates spanning three orders of magnitude. The upper release dose was extended 10-fold using sticky traps baited with quercivorol released at 1× (0.126 mg/day), 10×, and 108× relative rates to obtain a dose-response curve fitting a kinetic formation function. Naturally infested limbs of living avocado trees were wrapped with netting to exclude the possibility of catching emerging beetles on the encircling sticky traps. The results indicate PSHB are significantly attracted to infested limbs. Ethanol released over a 64-fold range (lowest rate of 7.5 mg/day) was moderately inhibitory of PSHB attraction to 1× quercivorol. β-caryophyllene and eucalyptol did not appear to affect attraction at the rates tested. A field test of potential inhibitors of 1× quercivorol was done using ~1 mg/day releases of monoterpene ketones: (-)-(S)-verbenone, (+)-(R)-verbenone, 3-methyl-2-cyclo-hexen-1-one (MCH or seudenone), piperitone, (+)-(S)-carvone, and racemic cryptone. Only piperitone and the two enantiomers of verbenone were strongly inhibitory. A blend of piperitone and verbenone tested together at different distances (0, 0.5, 1, 2, and 4 m) from a 1× quercivorol baited sticky trap became increasingly ineffective in inhibiting the attractant as separation distance increased. Due to the relatively short-range repellency (<1 m), the inhibitors would need to be released from several places on each tree to effectively repel PSHB from avocado trees. Effective attraction radii, EAR, and circular EARc are estimated for the quercivorol baits released at 1×, 10× and 108× rates. Push-pull simulations of moving beetles were performed in 1 ha plots with 2, 4, or 16 traps of 10× EARc and 400 trees (0, 1, or 3 inhibitors per tree) of which ten had an infested limb (EARc = 0.5 m). The simulations indicate that push-pull methods would be more effective in reducing PSHB mating than simply using mass-trapping alone.

Monitoring and mass-trapping methodologies using pheromones: the lesser date moth Batrachedra amydraula

The lesser date moth (LDM) Batrachedra amydraula is a significant pest of date palm fruits. Previously, detection and monitoring of the pest was inaccurate due to high costs of sampling with lifting machines. We report a practical system for detection and monitoring of LDM based on pheromone traps and relevant models. Dose-response experiments with LDM pheromone traps indicated a 1 mg lure is optimal for monitoring. Delta traps with adhesive covering their entire inner surface gave the highest captures while trap colour was unimportant. Sampling pheromone traps throughout the night indicated male flight began at 1:00-2:00 and reached a peak 2 h before sunrise. Monitoring traps exposed all year long in Israel revealed three generations with different abundance. Trapping transects in a date plantation indicated interference from a monitoring trap became minimal at distances >27 m away. Inter-trap distances closer than this may lower efficiency of monitoring and mass trapping in control programs. Our estimate of the circular effective attraction radius (EARc) of a 1 mg delta trap for LDM (3.43 m) shows this bait is among the most attractive compared with baits for other insects. We developed encounter-rate equations with the pheromone trap EARc to model the interplay between population levels, trap density and captures that are useful for detection of invasive LDM and its control by mass trapping. The integrated methodologies are applicable to many pest species.

Attraction of the Euwallacea sp. near fornicatus (Coleoptera: Curculionidae) to Quercivorol and to Infestations in Avocado

The Euwallacea sp. near fornicatus (Euwallacea sp. 1 hereafter) feeds on many woody shrubs and trees and is a pest of avocado, Persea americana Mill., in several countries including Israel and the United States. Quercivorol baits are commercially available for Euwallacea sp. 1 females (males do not fly), but their attractive strength compared to other pheromones and potential for mass trapping are unknown. We used sticky traps baited with quercivorol released at 0.126 mg/d (1×) and at 0.01×, 0.1×, and 10× relative rates to obtain a dose-response curve of Euwallacea sp. 1 attraction. The curve fitted well a kinetic formation function of first order. Naturally infested limbs of living avocado trees had attraction rates equivalent to 1× quercivorol. An effective attraction radius (EAR) was calculated according to previous equations for each of the various baits (1× EAR = 1.18 m; 10× EAR = 2.00 m). A pole with six sticky traps spaced from 0.25-5.75 m in height had captures of Euwallacea sp. 1 yielding a mean flight height of 1.24 m with vertical flight distribution SD of 0.88 m (0.82-0.96 m, 95% CI). The SD with specific EAR was used to calculate EARc, two-dimensional EAR (1× EARc = 0.99 m; 10× EARc = 2.86 m), for comparison with other insect pheromone traps and for use in simulations. The simulation methods described previously were performed with combinations of 1-16 traps with 1-50 aggregations per 9-ha plot. The simulations indicate mass trapping with quercivorol could be effective if begun in spring before Euwallacea sp. 1 establishes competing sources of attraction.

Large-Scale Evaluation of Association Between Pheromone Trap Captures and Cotton Boll Infestation for Pink Bollworm (Lepidoptera: Gelechiidae)

Although transgenic cotton producing insecticidal proteins from Bacillus thuringiensis (Bt) is a cornerstone for pink bollworm control in some countries, integrated pest management remains important for bolstering sustainability of Bt cotton and is critical for controlling pink bollworm where Bt cotton is not available or where this pest has evolved resistance to Bt cotton. Here, we used data on moth captures in gossyplure-baited pheromone traps and boll infestations for 163 Bt and 152 non-Bt cotton fields from Arizona to evaluate accuracy of chemical control decisions relying on moth trapping data and capacity of Bt cotton to suppress survival of offspring produced by moths. Assuming an economic injury level of 12% boll infestation, the accuracy of decisions based on moth captures corresponding to economic thresholds of 6%, 8%, and 10% boll infestation increased from 44.7% to 67.1%. The association between moth captures and boll infestation was positive and significant for non-Bt cotton fields but was not significant for Bt cotton fields. Although chemical control decisions based on trapping data were only moderately accurate, pheromone traps could still be valuable for determining when moth populations are high enough to trigger boll sampling to more rigorously evaluate the need for insecticide sprays.

Field Estimates of Attraction of Ceratitis capitata to Trimedlure and Bactrocera dorsalis (Diptera: Tephritidae) to Methyl Eugenol in Varying Environments

Measuring and modeling the attractiveness of semiochemical-baited traps is of significant importance to detection, delimitation, and control of invasive pests. Here, we describe the results of field mark-release-recapture experiments with Ceratitis capitata (Wiedemann) and Bactrocera dorsalis (Hendel) to estimate the relationship between distance from a trap baited with trimedlure and methyl eugenol, respectively, and probability of capture for a receptive male insect. Experiments were conducted using a grid of traps with a central release point at two sites on Hawaii Island, a Macadamia orchard on the East side of the island and a lava field on the West side. We found that for B. dorsalis and methyl eugenol there is a 65% probability of capture at ∼36 m from a single trap, regardless of habitat. For C. capitata, we found a 65% probability of capture at a distance of ∼14 m from a single trap in the orchard and 7 m in the lava field. We also present results on the spatial and temporal pattern of recaptures. The attraction data are analyzed via a hyperbolic secant-based capture probability model.