Jackson trap efficiency capturing Bactrocera dorsalis and Zeugodacus cucurbitae with male lures with and without insecticides

Jackson traps baited with male lures with or without insecticides are essential components of surveillance and monitoring programmes against pest tephritid fruit flies. The ability of a trap to capture a fly that enters, sometimes termed ‘trap efficiency’, is dependent on many factors including the trap/lure/toxicant combination. We tested the effects of three important components of Jackson traps on efficiency of capture of two important fruit fly species, using the ‘standard’ (i.e. as they are used in the state‐wide surveillance programme in California) and alternative setups: Insecticide (Naled, DDVP or None), type of adhesive on the sticky panel (Seabright Laboratories Stickem Special Regular or Stickem Special HiTack) and use of a single or combination male lure (Methyl eugenol and/or cuelure). Experiments were conducted in large outdoor carousel olfactometers with known numbers of Bactrocera dorsalis and Zeugodacus cucurbitae and by trapping wild populations of the same two species. Lures were aged out to eight weeks to develop a comprehensive dataset on trap efficiency of the various combinations. Results indicate that the current liquid lure/naled combinations on cotton wicks used in California for surveillance of these flies can be effectively replaced by plastic polymer plugs for the lure and pre‐packaged DDVP strips with no loss of trap efficiency for eight weeks of use or longer. The ‘high tack’ adhesive showed no advantage over the current standard against these flies, and both have low efficiency when used without an insecticide in the trap. Combination lure + DDVP varied when compared to the current standard liquid lure + naled: Olfactometer assays showed similar efficiency between them for B. dorsalis, but higher efficiency for the wafer against Z. cucurbitae. Field result showed similar or slightly higher performance of the wafer compared with the standard for B. dorsalis, but a much lower catch of Z. cucurbitae.


| INTRODUC TI ON
Trapping is likely the most important method of monitoring for the presence and abundance of insect species because of its efficiency.
At its simplest, a trap consists of a method of capturing an insect; an attractant and/or toxicant are often used as well but may not be necessary. Traps are diverse, and examples include pitfall traps (a can or other container in the ground, with or without an attractant, which captures insects walking along the ground), flight intercept traps (a trap consisting of a net that halts an insect in flight and directs it to a vessel for capture) and various 'sticky' traps that may include chemical lures. For tephritid fruit flies, trapping is especially important (FAO/IAEA, 2018).
In areas where they are not established, economically significant fruit flies of the family Tephritidae are often detected with traps (usually a 'sticky'-type trap like a Jackson trap or a wet trap such as a McPhail trap) baited with a male lure (e.g. trimedlure used for Mediterranean fruit fly), or a food-based lure (e.g. Torula yeast solution used for a variety of fruit fly species ;Steyskal, 1977;Tan et al., 2014). Fruit fly trapping is used for regulatory purposes. For example, certain areas may be trapped at a certain density of traps per area. The detection of one or more flies of a given species (and sometimes sex) may trigger eradication or quarantine activities.
Trading partners may require that an area where export crops are grown be trapped to provide assurance that a fruit fly species is not present. A wide variety of traps are available and commonly used to detect and monitor fruit flies (e.g. Katsoyannos, 1994).
The efficiency of trap capture was defined by the proportion of insects approaching a trap that are subsequently caught in it (Muirhead-Thompson, 1991). In an experiment with Lymantria dispar, Cardé et al. (2018) found that higher levels of the standard survey lure attracted more male moths to delta-type traps. The efficiency of capture, however, decreased at higher lure rates. In other words, more moths were attracted to the traps that had more lure, but of the moths that were attracted, fewer were captured by the higher lure traps, possibly because these males had difficulty in localizing the source of the lure. As these authors note, high efficiency of capture is especially important in detection trapping, where a small number of insects may be present, and the probability of capture is ideally maximized.
One of the many trap designs is the 'delta' type, which is a triangular prism in shape and usually constructed of cardboard. This has been found to be generally effective (e.g. Manoukis, 2016). A lure is usually affixed on the inside of the trap, and a 'sticky' insert (again, usually of cardboard) is placed on the inside bottom of the trap to capture insects (Gilbert et al., 2013). The Jackson trap is a simple type of delta trap without invaginated ends. It is used in California for detection and delimitation trapping of certain fruit flies, including oriental fruit fly (Bactrocera dorsalis) and melon fly (Zeugodacus cucurbitae), both of which are federal and state quarantine pests that are not considered established in the United States outside of Hawaii. For purposes of detection and delimitation of these pests in California, Jackson traps with methyl eugenol (in the case of oriental fruit fly) and cuelure (in the case of melon fly) are used to attract and capture male flies. A naled-based toxicant is combined with methyl eugenol and cuelure attractants (Gilbert et al., 2013). The purpose of the toxicant is to provide a rapid kill of the male flies after feeding to increase the likelihood of capture on the 'sticky' insert post-feeding. Shelly et al. (2010) following the approach of Lance and Gates (1994) estimated the capture probability of these two flies in Jackson traps in California and found that the current density used in urban  Currently, in California, 2% liquid naled is combined with methyl eugenol, a powerful male attractant for B. dorsalis (Howlett, 1915), or cuelure, a less attractive but effective male lure for Z. cucurbitae (Manoukis & Gayle, 2015). These mixtures are applied to cotton wicks for traps targeting each of the two species (5 ml for the ME/ naled combination, 5 ml for CL/naled) and hung in plastic baskets within the trap (Figure 1). This presents a significant amount of work per trap. The wicks must be mixed and placed in a trap body by staff with protective gear and cannot be further handled by staff in the field. Staff who work with naled must also undergo medical supervision to ensure their safety. Furthermore, the traps with naled cannot be transported in an enclosed vehicle with people.
While DDVP is in the same class of compounds as naled, it is commercially available as a pre-packaged strip (e.g. Hercon Vaportape II, Hercon Environmental, Emigsville PA). This makes it easier to handle and safer to transport. Furthermore, it can be applied separately from the lure compared with the standard California approach above. This can provide some flexibility, for example if a lure needs to be serviced at a shorter or longer interval than the insecticide, they can be changed independently potentially saving

| Insects, lures and sticky panels
We obtained B. dorsalis and Z. cucurbitae used in olfactometer assays from the colonies maintained at the USDA-ARS Daniel

K. Inouye US Pacific Basin Agricultural Research Center in Hilo,
Hawaii. These lab lines have been maintained for decades (over 400 generations) with occasional infusion of field material using standard rearing diets and protocols (Vargas, 1989). Sexually mature, 10-day-old, male flies were separated in a cold room and allowed to recover for 24 h prior to testing. Flies were held in a 30 × 30 × 30 cm cube cage and given water, protein hydrolysate and sugar. Field experiments relied on captures from the wild population.
Our goal was to test the impact of lure type, toxicant presence and type, and two different sticky panels. Regarding the former, we used five different lure types in experiments. Of these, two were [modification of] the current standard used in California for surveillance trapping: (1) 6 ml Methyl Eugenol (4-allyl-1-2-dimethoxybenze ne-carboxylate) with 1% Naled (dimethyl 1,2-dibromo-2,2-dichloroe thylphosphate) as toxicant, applied to a cotton dental wick approx.
We weathered fruit fly lures and toxicants (Table 1)  All treatments were tested in Jackson traps with standard sticky inserts (Scentry Biologicals) and experimental high tack sticky inserts ( Figure 1).

| Olfactometer assay
We conducted one set of experiments under seminatural conditions in large outdoor rotating-carousel olfactometers ( Figure S1). Treatments of a given age were tested simultaneously in one of the four olfactometers, with traps for each treatment placed in no specific order on the carousel arms. The whole experiment was replicated 4 times for each species, which we ran separately on different dates and with different cohorts of flies. We rotated which age treatment was in a particular cage between replicates, so that each age was tested in each enclosure once per species. We carried out experiments from 24 November 2020 to 24 February 2021. During this time, the average temperature was 21°C (28°C Max , 16°C Min) with an average RH of 83% (93% Max , 61% Min ).
For each of the species, we started experiments at 9:00 a.m. by introducing 1000 male flies into each of the four olfactometers. Flies were allowed to respond and be trapped for 24 h. The following day we collected traps and counted the number of males on the sticky inserts in the laboratory and recorded. We collected lures and toxicants from each trap and wrapped them individually in foil, placed them in a sealed ziplock bag and kept them in a freezer for redeployment in the field portion of the experiment.

| Field experiment
We Trials for each targeted species were conducted separately. We placed Jackson traps with treatments in a single row by age (i.e. each section of the row had traps of a single-aged group, but these varied in lure, toxicant and sticky panel). The order of the traps within age sections was haphazard. Each trap was spaced 30 m from its neighbours, and the rows used ran along the border of the macadamia nut tree farm with neighbouring papaya orchards for treatments targeting Z. cucurbitae and along the side with guava orchards for trials on B. dorsalis. All aged treatments were tested simultaneously for each species targeted, rotated each week by age group and replicated four times.
Traps baited with ME attract B. dorsalis males in great numbers in this field near the guava orchards based on experience, so we left them out for two hours between 9:00 and 12:00. This includes the traps baited with combination wafers that were tested simultaneously with the ME-only treatments-the treatments targeting

| DISCUSS ION
The olfactometer and field experiments we presented here both indicate that the current liquid lure/naled combinations on cotton wicks used in California for B. dorsalis and Z. cucurbitae trapping can be effectively replaced by plastic polymer plugs for the lure plus prepackaged DDVP strips with no loss of trap efficiency for eight weeks or longer. This is consistent with many previous studies both in terms of the solid lures and on the effectiveness of DDVP as a replacement for naled (Jang, 2011;Jang et al., 2013;Shelly, 2013;Vargas et al., 2009Vargas et al., , 2010Vargas et al., , 2015. Relative to liquid systems, solid lures have decreased handling difficulty and increased the flexibility of servicing schedules. As noted in multiple studies, an insecticide is needed for the highest effectiveness against the two species studied here, but packaging DDVP separately is preferable to an impregnated combined lure + toxicant as the latter would require registration with the U.S. EPA (e.x. Vargas et al., 2017). DDVP has been shown to be generally effective against tephritids (Katsoyannos, 1994).
At least out to eight weeks under Hawaii conditions, we saw no systematic differences in the longevity of these two alternatives (liquid lure + naled versus solid lure plus DDVP). This is not particularly surprising given previous studies. For example, Vargas et al. (2015) found effectiveness out to 16 weeks for plastic polymer formulations of cuelure plus DDVP. In some instances, the fresh insecticide reduced catch for the first few weeks (Vargas et al., 2009).
While a repellent effect has been hypothesized (Jang, 2011), a study on Ceratitis capitata suggests that mortality outside the trap is a more likely cause (Manoukis, 2016). Under weather conditions found in California, Arizona and Florida duration of the lures might be shorter, in the range of 6-8 weeks for methyl eugenol and 12 for cuelure (Jang et al., 2013;T. Shelly et al., 2015). Lure and insecticide longevity in the U.S. mainland is affected by season (Vargas et al., 2017).
The 'sticky panel' portion of the Jackson trap is important for keeping dead fruit flies within the trap; without this, flies may be blown out by the wind, for example. For smaller species such as C. capitata, it can be possible to capture flies on sticky panels without the use of an insecticide (Epsky et al., 1999;Gazit et al., 1998 (Hooper, 1978;Ito et al., 1976), but the effect of combination on surveillance systems remains unclear (Stringer et al., 2019). Even for the specific case of B. dorsalis and Z.
cucurbitae and combinations of methyl eugenol and cuelure/raspberry ketone (sometimes with trimedlure, a male attractant for C. capitata, as in this study), results are mixed. While some studies conducted in Hawaii found comparable performance of single and combination lures (T. Shelly et al., 2012;Shelly, 2013;Vargas et al., 2012Vargas et al., , 2015, there is evidence that there may be some inhibitory effect of combining lures  (Shelly et al., 2004;Vargas et al., 2000). The mechanism for any possible inhibition remains unclear, but it seems that the situation also affects other methyl eugenol and cuelure responding species (Stringer et al., 2019). Population density of each species has been suggested to play a role in varying trap catch in combination lures versus singles (Jang et al., 2013). Further study will be needed to fully elucidate how to apply combinations of lures.
In our study, we observed a prominent difference between the olfactometer and field results: A decrease in the catch of melon fly by the combination wafer lure (three male lures + DDVP) in the field experiment. We did not see a similar decrease for oriental fruit fly. This is different than both Vargas et al. (2000) and

ACK N O WLE D G E M ENTS
We would like to thank technical staff who provided significant assistance completing this project, including Steven Souder for lure ageing and Charlotte Aldebron and Jean Auth for help counting trap catch, Mahalo to Island Princess Macadamia Company for allowing us to conduct this study in their fields and University of Hawaii-CTAHR extension field station in Kainaliu HI for allowing us to age our treatments on the West side of the islands. This project was funded by USDA-ARS project 2040-22430-027-00D. Opinions, findings, conclusions and recommendations expressed in this publication are those of the authors and do not necessarily reflect the views of the USDA or CDFA. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture. USDA is an equal opportunity provider and employer.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflicting interest.