Brown Rice Vinegar as an Olfactory Field Attractant for Drosophila suzukii (Matsumura) and Zaprionus indianus Gupta (Diptera: Drosophilidae) in Cherimoya in Maui, Hawaii, with Implications for Attractant Specificity between Species and Estimation of Relative Abundance

Drosophila suzukii (Matsumura) is an agricultural pest that has been observed co-infesting soft-skinned fruits with Zaprionus indianus Gupta. The characterization of olfactory preferences by species is a necessary step towards the development of species-specific attractants. Five olfactory attractants were used to survey the populations of two invasive drosophilids in cherimoya in Maui, Hawaii. The attractants used were apple cider vinegar (ACV), brown rice vinegar (BRV), red wine (RW), apple cider vinegar and red wine (ACV+RW; 60/40), and brown rice vinegar and red wine (BRV+RW; 60/40). For D. suzukii, BRV+RW resulted in more captures than BRV, ACV, and RW, while ACV+RW resulted in more captures than ACV. No differences were observed between BRV+RW and ACV+RW. BRV had greater specificity in attracting D. suzukii compared to ACV, ACV+RW, and RW. For Z. indianus, no significant differences were observed in either the mean captures or specificity for any attractant used. Collectively, these findings demonstrate that (1) BRV and BRV+RW are effective field attractants and (2) D. suzukii has unique olfactory preferences compared to non-target drosophilids, while (3) Z. indianus’ preferences do not appear to vary from non-target drosophilids, and (4) the accuracy of relative abundance is impacted by the specificity of the attractants.


Introduction
International trade and travel has supported the dispersal of many insect species outside of their native range including Drosophila suzukii (Matsumura) and Zaprionus indianus Gupta. Native to Southeast Asia, D. suzukii is now commonly found throughout North America [1,2], South America [3], and Europe [4][5][6]. The first adult specimens in the United States were captured on Oahu and Kauai in 1983 [7], and records have since been published across the U.S. mainland [8]. Similarly, Z. indianus, native to the Afrotropics, is now reported throughout North America, South America, Europe, and Asia [9,10]. Like many dipterans, D. suzukii and Z. indianus play a vital ecological role in nutrient cycling through catalyzing decomposition processes. However, when introduced to a non-native range, temperate climates and the abundance of host plants paired with the lack of natural predators can contribute to unregulated population expansion [11]. Furthermore, these vinegar fly species possess high adaptability [12,13] and fecundity [14,15] that facilitate successful colonization. Not surprisingly,

Trap Assembly
The traps used for the following three experiments were a novel modification of the traditional deli cup design ( Figure 1). Assembled using a Mason jar (236 mL) and a red Solo cup (473 mL), the jar served to house the attractant-drowning solution, while the Solo cup provided the headspace that enabled passive diffusion of the attractant and fly entry into the trap. To fabricate the trap, the interior of the Solo cup was scoured with sandpaper, rinsed with 70% ethanol, and then coated with epoxy. The silver band from the Mason jar lid was secured inside the Solo cup, as deep as the diameter of the silver band permitted, and was then left to cure for 24 h. A soldering tool (Cold-Heat ® ) with a 5-mm diameter tip was used to gently puncture eight entry holes into the Solo cup-two on each of the four sides. Two of the holes were used in trap installation with nylon rope. The removable trap base permitted simple trap maintenance as the attractant solutions were aliquoted into clean jars weekly which then replaced the spent jars on-site. After the lids were affixed to the spent jars, they were transported off-site for specimen preservation in alcohol and quantification.
Insects 2019, 10, x FOR PEER REVIEW 3 of 18 indianus. Rice vinegar, with and without red wine, has also been characterized by several researchers [30,32,46]. Recently, Akasaka et al. [52] demonstrated that brown rice vinegar was more attractive to D. suzukii than apple cider vinegar in laboratory choice assays, substantiating the need for the characterization of attractant efficacy in field settings. The primary objective of this experiment was to examine attractant efficacy and specificity by attractant type (apple cider vinegar, brown rice vinegar, red wine, apple cider vinegar blended with red wine, brown rice vinegar blended with red wine) and species (Drosophila Suzukii (Matsumura), Zaprionus Indianus Gupta, non-target drosophilids) in cherimoya (Annona cherimola Miller). The secondary objective was to examine whether attractant efficacy varied over time in the field capture of D. suzukii, Z. indianus, and non-target drosophilids. The tertiary objective was to characterize the relative abundance of two invasive drosophilids, D. suzukii and Z. indianus. This report provides the first published characterization of the relative abundance of Z. indianus in Maui, Hawaii, since the reported introduction in 2017 [53]. Furthermore, at the time of writing, this report provides the first published quantitative comparison of brown rice vinegar as an attractant solution in a field setting in the United States.

Trap Assembly
The traps used for the following three experiments were a novel modification of the traditional deli cup design ( Figure 1). Assembled using a Mason jar (236 mL) and a red Solo cup (473 mL), the jar served to house the attractant-drowning solution, while the Solo cup provided the headspace that enabled passive diffusion of the attractant and fly entry into the trap. To fabricate the trap, the interior of the Solo cup was scoured with sandpaper, rinsed with 70% ethanol, and then coated with epoxy. The silver band from the Mason jar lid was secured inside the Solo cup, as deep as the diameter of the silver band permitted, and was then left to cure for 24 h. A soldering tool (Cold-Heat ® ) with a 5mm diameter tip was used to gently puncture eight entry holes into the Solo cup-two on each of the four sides. Two of the holes were used in trap installation with nylon rope. The removable trap base permitted simple trap maintenance as the attractant solutions were aliquoted into clean jars weekly which then replaced the spent jars on-site. After the lids were affixed to the spent jars, they were transported off-site for specimen preservation in alcohol and quantification.

Attractant Solutions
Five attractants were used in this experiment: apple cider vinegar (ACV, supplied by Marukan Vinegar Co. Ltd., Kobe, Japan), brown rice vinegar (BRV, supplied by Marukan Vinegar Co. Ltd.), red wine (RW, Oak Leaf Vineyards, Merlot), ACV+RW (blended; 60/40%), and BRV+RW (blended; 60/40%). ACV, RW, and ACV+RW were chosen due to the historical use of vinegars and wine as olfactory baits in field traps [54]. BRV was selected due to recent evidence suggesting BRV is more attractive to D. suzukii than ACV in laboratory choice-assays [52]. BRV+RW was prepared to assess the dilution of BRV and synergy with red wine volatiles that had been previously observed with ACV+RW blends [30,32]. Attractants were aliquoted (40 mL per trap) into Mason jars on the morning of trap installation. A drop of unscented dishwashing liquid (Dawn ® Ultra Free and Gentle, Procter & Gamble, Cincinnati, OH, USA) was added to each jar using a 3-mL transfer pipette (Karter Scientific, Lake Charles, LA, USA). The addition of a surfactant was necessary to break the surface tension, transforming the attractant into a drowning solution. After the surfactants were added, the Mason jar lids were affixed and ready for transport.

Trap Installation and Maintenance
The traps were installed at fruiting height and the attractant solution was replaced weekly. The trap height corresponded with low hanging fruit, as research in foraging ecology has shown higher D. suzukii capture rates in traps positioned on the lower canopy [55]. After trap installation, the trap bases were replaced weekly to collect captured specimens and to refresh the attractant solution. Once the specimens were collected, they were separated from the attractant solution using a 1-mm mesh filter and were stored in 70% isopropyl alcohol prior to identification.

Species Identification
The specimens were classified as D. suzukii (male or female), Z. indianus (male or female), non-target drosophilid, or other non-target captures. Standard keys were used in the examination of external morphology for D. suzukii [56] and Z. indianus [57,58]. Taxonomic confirmation for D. suzukii was confirmed by Dr. Douglas Pfeiffer while Z. indianus confirmation was provided by Dr. Luc Leblanc, University of Idaho, and Dr. Amir Yassin, Muséum d'Histoire Naturelle de Paris. The specimens were classified as non-target drosophilids if the individuals possessed all of the following attributes: (1) a body length of 3-4 mm, (2) a golden-yellow-brownish body color, (3) red eyes, (4) antennae with branched arista, and (5) wings that have three breaks in the costa vein [59]. The specimens that did not meet the criteria for D. suzukii, Z. indianus, or non-target drosophilid, were classified as non-target other captures. Note that this included the captures of any true fruit fly from Tephritidae, such as Ceratitis capitata (Wiedemann), the Mediterranean fruit fly, or Bactrocera cucurbitae (Coquillett), the melon fly.

Experimental Design
The survey location that was selected was a cherimoya grove at the University of Hawaii, at Mānoa: Kula Research Station (20.7 • N; −156.3 • W, 980 m) ( Figure 2). There were 24 traps installed, which included four replicates of the six attractant treatments (five attractants and one control). Distilled water was used to control for a confounding variable-drosophilid attraction to visual stimuli [31], i.e., the red Solo cup. Once the traps were delineated and installed, attractant treatment was randomized. Each week, the captured specimens were collected, trap bases were replaced, and attractant treatment was re-randomized. Thus, each week corresponded to one run or replicate of the experiment. For each run, the four repeats of each attractant level were treated as individual data points and were not pooled or averaged for analysis, as the experimental unit was defined as an individual trap. traps were installed and were maintained weekly until 11 December 2017 (N = 16). Four rows of trees were selected, and six traps were installed within each row. The traps were placed in a single tree with at least 5 m between each trap.

Statistical Analysis
Statistical analysis was conducted with SAS ® (SAS Institute Inc., Cary, NC, USA), with the statistical model as a generalized linear mixed model (GLMM) (PROC GLIMMIX procedure). A GLMM was selected as the data analyzed were non-normal and not transformed, and the model was subjected to both fixed and random effects. The significance threshold for mean separation was defined as p < 0.05 for post-hoc analysis. Tukey's Honestly Significant Difference (HSD) test was used for mean separation (within the LSMEANS statement). Several resources were utilized in the modeling of count data with a Poisson distribution [60][61][62]. In order to determine whether there were significant differences in the mean captures between the sexes, the total number of D. suzukii males was divided by the total number of male and female D. suzukii captures.
The sex specificity value results in a single response variable with a value between 0 and 1 for each datapoint (row of the spreadsheet). A value of 1.0 implies that only males were captured, while a value of 0.5 implies that there was equal male to female captures, and a value of 0.0 indicates that only females were captured.
To examine whether there were sex differences in mean capture between attractants, a generalized linear mixed model (GLMM) was used (PROC GLIMMIX) with the attractant as a fixed effect and with the week and replicate as random effects. The method was set as Laplace, and a beta distribution was specified as the sex specificity calculation was expressed as a decimal fraction. If there were no significant differences in sex specificity between either male and female D. suzukii or between male and female Z. indianus, further analysis was performed with the sex counts pooled.
To examine differences in mean capture between attractants, a GLMM was used (PROC GLIMMIX), with the attractant as a fixed effect and with the week and replicate as random effects. The method was set as Laplace, and a Poisson distribution was specified for the raw count measurements for each specimen category (D. suzukii female, D. suzukii male, Z. indianus female, Z. indianus male, non-target drosophilid, and other non-targets). The cherimoya grove at the University of Hawaii, at Mānoa: Kula Research Station, contained fruit that was ripening or ripe. The grove floor was shady with small amounts of plant debris in the early stages of decomposition and weeds that were routinely treated with herbicides. The perimeter of the grove contained approximately 2.5 m of well-managed turfgrass. On 13 November 2017, 24 traps were installed and were maintained weekly until 11 December 2017 (N = 16). Four rows of trees were selected, and six traps were installed within each row. The traps were placed in a single tree with at least 5 m between each trap.

Statistical Analysis
Statistical analysis was conducted with SAS ® (SAS Institute Inc., Cary, NC, USA), with the statistical model as a generalized linear mixed model (GLMM) (PROC GLIMMIX procedure). A GLMM was selected as the data analyzed were non-normal and not transformed, and the model was subjected to both fixed and random effects. The significance threshold for mean separation was defined as p < 0.05 for post-hoc analysis. Tukey's Honestly Significant Difference (HSD) test was used for mean separation (within the LSMEANS statement). Several resources were utilized in the modeling of count data with a Poisson distribution [60][61][62]. In order to determine whether there were significant differences in the mean captures between the sexes, the total number of D. suzukii males was divided by the total number of male and female D. suzukii captures.
The sex specificity value results in a single response variable with a value between 0 and 1 for each datapoint (row of the spreadsheet). A value of 1.0 implies that only males were captured, while a value of 0.5 implies that there was equal male to female captures, and a value of 0.0 indicates that only females were captured.
To examine whether there were sex differences in mean capture between attractants, a generalized linear mixed model (GLMM) was used (PROC GLIMMIX) with the attractant as a fixed effect and with the week and replicate as random effects. The method was set as Laplace, and a beta distribution was specified as the sex specificity calculation was expressed as a decimal fraction. If there were no Insects 2019, 10, 80 6 of 18 significant differences in sex specificity between either male and female D. suzukii or between male and female Z. indianus, further analysis was performed with the sex counts pooled.
To examine differences in mean capture between attractants, a GLMM was used (PROC GLIMMIX), with the attractant as a fixed effect and with the week and replicate as random effects. The method was set as Laplace, and a Poisson distribution was specified for the raw count measurements for each specimen category (D. suzukii female, D. suzukii male, Z. indianus female, Z. indianus male, non-target drosophilid, and other non-targets).
To examine the specificity between D. suzukii vs. drosophilid captures and Z. indianus vs. drosophilid captures, the raw data (capture counts) were manipulated to result in a single response variable (specificity ratios) on a per trap basis. For each trap, the attractant specificity for D. suzukii was calculated by dividing D. suzukii captures by drosophilid captures. Similarly, to calculate the attractant specificity for Z. indianus, Z. indianus captures were divided by drosophilid captures, resulting in a specificity value for each trap. The species specificity calculation results in a value between 0 and 1 for each datapoint (row of the spreadsheet) with the interpretation of values similar to as previously described for sex specificity. Using this calculation, mean specificity by attractant type could be evaluated.
To examine whether there were differences in the specificity of captures between attractants, a generalized linear mixed model (GLMM) was used (PROC GLIMMIX) with attractant as a fixed effect and with week and replicate as random effects. The method was set as Laplace, and a beta distribution was specified as the specificity results were expressed as decimal fractions. To support the secondary objective, a GLMM was used (PROC GLIMMIX) to examine whether the mean captures or specificity varied by week (fixed effects are attractant, week, and attractant*week; random effect is replicate). Similar to the previous analysis, the method was set as Laplace, and the distribution was set as Poisson and beta for count data and decimal fractions, respectively. To support the tertiary objective, relative abundance ratios were generated by dividing the total quantity of target specimens captured (D. suzukii or Z. indianus) by the total quantity of vinegar flies captured (D. suzukii, Z. indianus, and non-target drosophilid). Though similar to the calculation of specificity, the estimation of relative abundance produces a single value which includes all attractant solutions, while specificity is calculated on the level of each subject resulting in 80 values for statistical analysis.

Zaprionus indianus: Mean Captures and Attractant Specificity
Since there were no significant differences between male and female Zaprionus indianus captures by attractant (f = 1.43, df = 4, 14, p = 0.2762), the captures were pooled for both sexes for analysis. No significant differences were observed for Zaprionus indianus captures by attractant type (f = 1.41, df =

Estimated Relative Abundance of D. suzukii, Z. indianus, and Non-Target Drosophilids in Cherimoya
To characterize the relative abundance of exotic drosophilids, total captures by species and attractant type were reported in a contingency table (Table 1), then the relative abundance of D. suzukii, Z. indianus, and non-target drosophilids was calculated ( Figure 12).

Estimated Relative Abundance of D. suzukii, Z. indianus, and Non-Target Drosophilids in Cherimoya
To characterize the relative abundance of exotic drosophilids, total captures by species and attractant type were reported in a contingency table (Table 1), then the relative abundance of D. suzukii, Z. indianus, and non-target drosophilids was calculated ( Figure 12). All experimental data, including raw counts and calculated specificity ratios, can be found in the supplementary workbook (S1: Data).

Discussion
The results of this experiment for the field capture of drosophilids in cherimoya in Maui suggest that (1) BRV and BRV+RW are effective attractants for D. suzukii and Z. indianus, (2) D. suzukii has unique olfactory preferences compared to all non-target drosophilids, while (3) there are no differences in olfactory preferences between Z. indianus and non-target drosophilids, (4) attractant preferences vary over time for D. suzukii but not for Z. indianus, and (5) the accuracy of relative abundance is impacted by the specificity of the attractants used. At the time of writing, this is the first reported use of BRV and BRV+RW as olfactory attractants in a quantitative field trapping study in the United States. Furthermore, this report provides the first characterization of the relative abundance of Z. indianus since its reported introduction in Maui [53]. Since there were relatively few captures in the control treatment (distilled water), which controlled for drosophilid attraction towards visual stimuli (to the red solo cup) [31], it can be concluded that the captures in this experiment were due to olfactory attraction or the combination of olfactory and visual stimuli-not to visual stimuli alone. The limitations of this experiment include limited duration (4 weeks) and relatively low repeats (N = 16), suggesting that further research is needed for greater confidence in these trends. Nonetheless, this experiment provides insight into attractant efficacy by species and emphasizes the need to assess the synergy between attractant solutions and other compounds within the host plant environment, such as the outgassing of volatiles emitted from plants and fruit, which has been a topic of recent investigation [63,64]. Furthermore, since the experiment was limited to one host plant on one island, additional experiments in other host plants and localities are needed to have confidence in extrapolating these trends across multiple environments.
Five attractants were used (ACV, BRV, RW, ACV+RW, and BRV+RW) to evaluate attractant effectiveness for D. suzukii and Z. indianus in cherimoya. There were significant differences in the mean captures of D. suzukii by attractant type, but no significant differences were observed for Z. indianus. The use of BRV+RW resulted in greater D. suzukii captures than BRV, RW, and ACV. There were no significant differences between BRV+RW and ACV+RW, suggesting that BRV+RW may be equivalent to ACV+RW as an olfactory attractant for D. suzukii. Similar to the results of previous experiments, the use of ACV+RW resulted in greater D. suzukii captures than ACV alone [32,46]. Researchers conducting field experiments in berry crops found that wine-vinegar blends were more attractive to D. suzukii than wine and vinegar alone [46] or aqueous red wine [32]. Cha et al. [46], reported that blended RW and rice vinegar (60/40%) was more attractive to D. suzukii than rice vinegar or red wine alone, while Landolt et al. [32] found that RW blended with rice vinegar (60/40%) was more attractive to D. suzukii than RW with ACV (60/40%) or RW with white wine vinegar (60/40%) or aqueous RW (60%). In contrast to the previous experiments described, there was no significant difference between ACV+RW and RW, which could be attributed to other factors present between the different host plant environments examined-cherimoya in this experiment vs. blackberry [32] and berry crops [46]. For Z. indianus, no significant differences were observed in the mean captures for any attractant used, which suggests that, in practice, any of the attractants assessed could be used to identify whether Z. indianus adults are present in a field. For non-target drosophilids, greater captures were observed with RW, ACV+RW, and BRV+RW than with ACV or BRV alone. This suggests that for non-target drosophilids and Z. indianus, wine volatiles alone performed similarly to the synergy of wine and vinegar volatiles, while the latter was a stronger preference for D. suzukii.
There were significant differences in the mean captures by week for D. suzukii and non-target drosophilids, but not for Z. indianus. After analysis of D. suzukii attractant preferences by week, there appeared to be slight differences in the distribution of the captures. For non-target drosophilids, even though the captures varied by week, attractant preferences were similar throughout the duration of the experiment. Inclusive of all attractant types examined, 65% of total D. suzukii captures were male and 35% were female. Since there were no significant differences in the proportion of male to female captures by attractant type, this suggests that the trend was consistent regardless of which attractant was used with 60-67% male captures to 33-40% female captures. Though it is not possible to know the actual sex ratio for the population and since the visitor rate was not assessed (e.g., pesticide treatment of traps coupled with a bucket), it is not possible to determine whether there were more males present in the population or simply more male captures.
In order to develop species-specific attractants, the characterization of attractant specificity by species is needed. BRV had greater specificity in attracting D. suzukii compared to ACV, ACV+RW, or RW. On average,~50% of the drosophilid captures in the BRV traps were D. suzukii, which was reduced to~40% with BRV+RW and ACV,~30% with ACV+RW, and~25% with RW. The results of this field experiment support previous research where, in a laboratory trapping assay, BRV was found to be significantly more attractive to D. suzukii than ACV [52]. Though there were no significant differences in specificity observed between BRV and BRV+RW, there were also no differences between BRV+RW and ACV+RW, suggesting that the volatile compound present in BRV, once diluted with red wine, becomes less specific to attracting D. suzukii. Since the results suggest that BRV has greater D. suzukii specificity, it follows that using BRV as a field attractant for surveying could reduce non-target impacts and the time necessary to identify specimens. However, further research in additional host plants and localities is needed to better understand this trend and identify any potential ecological and economic impacts. The specificity reported in this experiment was calculated from observed captures and did not take into account the true population, as this was an open-field experiment. Future research could examine attractant specificity between D. suzukii and Z. indianus in a controlled environment, which may better isolate the variables in question and provide valuable data towards the development of a species-specific attractant. The preferential attraction observed in this experiment is likely associated with evolutionary adaptations that enable co-habitation of D. suzukii and Z. indianus with other drosophilids over finite resources in a given environment [23,41,42]. These unique olfactory preferences have implications for both farmers and researchers. In order to mitigate revenue loss, farmers must detect the presence of pests as early in the season as possible to start implementing pest reduction strategies. These findings suggest that for field surveying drosophilid pests, multiple attractants should be used to cover a range of attractive compounds.
The results of this experiment beg the question-if specificity varies depending on attractant, how can an accurate representation of relative abundance be depicted from the field captures of wild flies? In this experiment, the relative abundance of D. suzukii and Z. indianus ranged from 22.3-46.4% and 1.1-5.8%, respectively, depending on the attractant used. The low relative abundance of Z. indianus observed could be the result of low attractant efficacy in the host plant environment, or that cherimoya was not a preferred host-both of which may have underestimated the actual population. Alternatively, the low relative abundance could correspond with the recent introduction of this species to the region [53] coupled with interspecies competition for resources that limited population expansion. In central Brazil, Tidon et al. [65] reported that Z. indianus had an initially low relative abundance followed by a population explosion the following year when the populations were re-surveyed. Since the actual relative abundance cannot be known in wild populations, and there is attractant bias, the interpretation of field data on relative abundance is challenging and may overestimate or underestimate the actual population. This experiment emphasizes the challenges and limitations associated with the interpretation of relative abundance calculations from field-generated data.
Previous research has demonstrated the potential for co-infestation of D. suzukii and Z. indianus in a variety of crops including strawberry [24], guava [2] and grape [66]. As a secondary pest with opportunistic oviposition tendencies, Z. indianus can increase damage to soft-skinned fruits with oviposition injury from D. suzukii [66] or mechanical damage from other insects [24]. In Veracruz, Mexico, D. suzukii and Z. indianus co-infested guava orchards comprising more than 80% of the total drosophilids captured (55.0 and 26.3% relative abundance, respectively) [2], which suggests that there may be deleterious interspecies competition between endemic and invasive drosophilids. There is also evidence of interspecies competition between D. suzukii and Z. indianus. For example, larval competition between Z. indianus and D. suzukii increased development time and mortality for D. suzukii in Virginia [66]. Even though co-infestation was observed in the cherimoya grove, no attempt was made to examine interspecies competition or to quantify the losses from direct and indirect injury. Furthermore, research is needed regarding D. suzukii and Z. indianus co-infestation to identify economic thresholds and provide practical pest management strategies for farmers.
Several limitations were identified in this experiment including the short duration of the study. The short-duration of the experiment (4 weeks) limited the characterization of olfactory preferences by fruiting stage over time. Over the four-week experiment, there were few total captures for traps treated with distilled water-three D. suzukii specimens were captured (one male, two female), zero Z. indianus specimens were captured, five non-target drosophilids were captured, and nine non-target other specimens were captured. Due to the low quantity of the control captures, there is confidence that the dataset analyzed contained values that were the result of olfactory attraction and not confounded by visual stimuli. Additional research is necessary to assess whether an attractant is more or less attractive in different visual environments. Further characterization of attractant specificity by species is needed as fruit ripeness, population dispersal, and interspecies competition alter drosophilid species composition throughout the growing season. Future research could assess attractant efficacy by host plant fruiting stage by collecting fruit samples weekly and identifying the outgassing fruit volatiles. If novel compounds or the synergistic effects of multiple compounds were identified, this would aid in the development of a species-specific attractant. In the meantime, this may also provide farmers with the most effective attractant to use while surveying depending on host plant fruiting stage.

Conclusions
The results of this study support the notion that D. suzukii displays unique olfactory preferences compared to non-target drosophilids, whereas no attractant preferences were observed between Z. indianus and non-target drosophilids. The results suggest that BRV, or BRV+RW, is an effective alternative to ACV or ACV+RW for the field survey of drosophilid pests. Furthermore, BRV appears to have high specificity for D. suzukii as compared to non-target drosophilids, which warrants further research towards the development of a species-specific attractant. This study provides an initial characterization of the relative abundance of Z. indianus (inclusive of all attractants, 2.9%) in a cherimoya grove in Kula, Maui, Hawaii, since the reported introduction in 2017. However, the estimation of relative abundance varied by attractant type (between 1.1-5.8%) and may have been underestimated as none of the attractants used in the experiment had high specificity for Z. indianus field captures. Alternatively, the low relative abundance observed could be indicative of a low population level. The actual drosophilid distribution can only be estimated, as the true population is unknown.