Anastrepha grandis (Macquart) (Diptera: Tephritidae), one of the larger tephritid fruit flies, occurs in much of the low elevation Andean region of South America attacking fruits of the Cucurbitaceae (Silva & Malavasi 1993; Bolzan et al. 2015). It is considered a regulated pest of cucurbits by most tropical and subtropical countries where it does not occur. Despite its regulatory significance, no phytosanitary treatments have been developed against this pest. Although it is covered by the 150 Gy generic phytosanitary irradiation dose approved for all Tephritidae, no irradiation research has been published on A. grandis (Hallman 2012).

Placement in temperatures near 0 °C for 2 to 5 wk has been used for a century to disinfest fruits of tephritid fruit flies (Gould 1994; Heather & Hallman 2008). From then until the present, it has been the most commonly applied phytosanitary treatment for various types of citrus fruits exported from several countries, such as Argentina, Australia, Morocco, Spain, and the USA. It is also used for a variety of other fruits such as Physalis peruviana L. (Solanaceae) exported from Colombia to the USA.

To verify the efficacy of a phytosanitary treatment, the most tolerant life stage present in the commodity should be determined, and then large-scale confirmatory testing is usually conducted with that stage to demonstrate that it is efficacious to a high degree of certainty (Heather & Hallman 2008). The objectives of this research were to determine the most cold-tolerant developmental stage of A. grandis and to estimate a time period required for a phytosanitary cold treatment in zucchini squash, Cucurbita pepo L. (Cucurbitaceae). Cucurbita pepo was the best of several hosts of A. grandis as measured by infestation index, fecundity, and fertility (Bolzan et al. 2015).

Materials and Methods

Specimens of A. grandis were obtained from fruits originating from South America in 2012 and reared in the Insect Pest Control Laboratory of the Joint FAO/IAEA Division of Nuclear Techniques in Food and Agriculture in Seibersdorf, Austria, on zucchini squash obtained from a local wholesaler in Vienna. Zucchini weighing 200 to 400 g were placed in cages (44 × 44 × 44 cm) with adult A. grandis for 6 to 8 h for infestation to occur and then held at 26 °C for 0 to 20 d to obtain insects of different stages. To ensure that insects in all sizes and ages within a stage were represented in the analyses, infested fruits were treated at all ages between 0 and 20 d after infestation. The ages for each stage were determined by dissection of infested fruits and established as follows: 0 to 3 d: egg; 4 to 7 d: 1st instar; 8 to 13 d: 2nd instar; and 14 to 20 d: 3rd instar. These time periods do not precisely coincide with their respective stages, as there is overlap at the days adjoining consecutive stages, but for the purposes of robust determination of the most tolerant stage, this scheme is useful. At ~20 d, late 3rd instars are present in the fruit, but adults will soon begin to emerge.

When the desired stage developed in the zucchini, the infested fruit samples were placed in plastic boxes (44 × 33 × 12 cm) in a 2 m3 cold treatment chamber (model SE-2000-4, Thermotron Industries, Holland, Michigan) at a target temperature of 1.1 °C for 8 to 14 d. Temperatures in the chamber and within the fruit were recorded (datalogger model OM-CP-OCTTEMP2000, Omega, Stamford, Connecticut) with Type T, 30-gauge thermocouples that were calibrated to 0.0 °C in a slurry of reverse osmosis-purified ice and water. Upon removal, the fruits were returned to 26 °C and held until A. grandis would have reached the late 3rd instar, which was calculated by subtracting the number of days between exposure to adults and placement in the cold treatment chamber from 21; an additional day was added to allow surviving insects to recover from the cold treatment before examination. This additional day simulates the minimum time required for treated fruit to reach an importing inspection point. Controls for all stages were also held at 26 °C until late 3rd instars developed.

Non-treated but infested fruits (controls) were held at 26 °C; the proportion of fruits held as controls depended on the developmental stage of the insect. Dead eggs, 1st instars, and early 2nd instars could not readily be observed upon visual examination of the fruits. Therefore, the non-treated “controls” in this case were not used to estimate the level of response (mortality) not attributable to the cold treatment. Instead, non-treated fruits were used to estimate the number of insects present in the fruit, i.e., the number treated. Thus, there was no estimate of natural (non-cold treatment induced) mortality for eggs and early instars. A large proportion (30 to 50 %) of infested fruits was not treated to minimize variation in the estimate of number of insects present in the fruits per replicate. As late 2nd instars and 3rd instars could readily be observed upon visual examination of the fruits, the control for these stages was only needed to determine natural mortality, and ~10% of the fruits were held as controls. It is assumed that natural (non-cold treatment induced) mortality was similar in all stages tested.

Because the research depended upon infestation via oviposition to avoid making untested assumptions related to more artificial techniques used in phytosanitary treatment research (Hallman 2014), sometimes not all fruits were infested or they rotted before they could be used. Therefore, the number of replicates varied from 2 to 4 per stage-time combination.

After a total of 21 d at 26 °C pre and post cold treatment, the zucchinis were opened and examined for dead and live larvae. Larvae that were the color of live ones but not moving were kept until they were obviously dead or alive. Larvae that moved upon inspection or pupariated were considered survivors regardless of subsequent fate.

Data were subjected to probit analysis (PoloPlus, Petaluma, California). For each stage, only the lowest dose that resulted in 100% mortality and that was not followed by a higher dose resulting in 1 or more survivors was used in the analysis; i.e., doses ≥⃒11 d for eggs and 1st and 2nd instars all resulted in 100% mortality, thus, only data from 8 to 11 d were used in the analyses for those 3 stages.

Results

The infestation rate of non-treated zucchini fruits varied over the course of the study from 0 to 68 recovered larvae per fruit; more heavily infested fruits as well as some other fruits were disintegrating and not used in the research. The mean ± SE cold-treatment temperature during the study was 1.04 ± 0.02 °C, and the inside of fruits reached that temperature nearly 1 d after placement in the cold treatment chamber.

Mean natural mortality in the controls of 3rd instars was 0.78 ± 0.39%. Although there was no measurement of natural mortality in non-treated earlier instars or eggs, it is assumed that natural mortality in these stages was also low and would not significantly affect the results of the probit analysis.

Table 1.

Total numbers of Anastrepha grandis eggs and instars in zucchini squash, Cucurbita pepo, fruits treated at 1.1 ± 0.1 °C for 8 to 14 d.

Table 2.

Results of probit analysis of Anastrepha grandis eggs and instars in zucchini squash, Cucurbita pepo, fruits treated at 1.1 ± 0.1 °C for 8 to 14 d: χ2, heterogeneity, and lethal dose (LD) estimates (d ) for 90%, 99%, and ≈99.9968% (“probit 9″) mortality (95% fiducial limits).

Mortality of the stages of A. grandis was similar for eggs and 1st and 2nd instars, for which 100% mortality was achieved at 11 d (Table 1). Complete mortality of 3rd instars required 14 d. Estimates of doses to achieve various levels of mortality by probit analysis, log10 of dose, are given in Table 2. At 23.2 d, the estimate for 99.9968% (probit 9) mortality of 3rd instars was approximately twice the estimate for the other stages.

Discussion

Eggs and 1st and 2nd instars were more susceptible to cold treatment than 3rd instars. There was no estimate of natural (non-cold treatment induced) mortality in the eggs and early instars, and it was assumed to be negligible because natural mortality in the controls of late instars was negligible (mean = 0.78%). As natural mortality increases, efficacy of the treatment decreases as some of the mortality that would normally be ascribed to the treatment is in reality due to other causes. However, at the high levels of efficacy achieved with the doses used in this study, corrections for natural mortality would be negligible even if they were in the order of ~20%.

Cold tolerance in A. grandis follows that for most other tephritids (Hallman et al. 2013a,b) in that the 3rd instar was the most tolerant. The estimate to provide a very high level (99.9968%) of mortality of 3rd instars in zucchini fruit is 23.2 d. However, fit of the data to the probit model was not good, as exemplified by the upper 95% fiducial limit being 144 d. Because estimates of very high levels of efficacy are not always accurate (West & Hallman 2013), large-scale confirmatory testing, where at least 10,000 3rd instars are treated with no survivors, would usually be required to substantiate that the treatment was efficacious with a high degree of confidence.