Postharvest Biology and Technology 19 (2000) 193 199 www.elsevier.com/locate/postharvbio Short communication Phosphine as a replacement for methyl bromide for postharvest disinfestation of citrus P. Williams *, G. Hepworth, F. Goubran, M. Muhunthan, K. Dunn Agriculture Victoria, Institute for Horticultural De elopment, Pri ate Bag 15, South Eastern Mail Centre, Victoria 3176, Australia Received 17 September 1999; accepted 3 March 2000 Abstract Methyl bromide is the only fumigant registered for disinfestation of citrus in Australia. However, it has been identified as an ozone-depleting chemical and its use is being restricted in accordance with an international agreement. The only alternative to a 2-h methyl bromide fumigation is cold treatment at 1 C for 16 days, and a more rapid alternative is desirable. The phosphine cylinder gas formulation ECO 2 FUME (phosphine with CO 2 as a carrier gas) was recently registered for 15-h fumigation of cut flowers. This treatment was used as a basis for experimental fumigations of oranges infested with larvae of Queensland fruit fly, Bactrocera tryoni in a 900-l chamber. Uninfested oranges were included in all fumigations to check for possible adverse effects. A 16-h fumigation at 20 C with an initial phosphine concentration of 0.98 g m 3 resulted in 96.4% mortality of fly larvae. This was a promising result but the mortality achieved was insufficient to meet mortality requirements for interstate (99.5%) or export trade (99.9%). Exposure times, temperatures and phosphine concentrations were all increased in subsequent fumigations. In the final series of fumigations with export grade Washington Navel oranges the exposure time was 48 h, thermostat settings for the chamber heater were 23 or 25 C, initial phosphine concentrations were 1.67 g m 3 and the concentrations were topped up to 0.7gm 3 after 24 h. No adverse effects on the oranges were observed, and a mortality of 99.998% was achieved with 48 000 larvae exposed. This would meet requirements for interstate trade and possibly also some international trade, particularly if incorporated as part of a systems approach. 2000 Elsevier Science B.V. All rights reserved. Keywords: Phosphine; Postharvest; Disinfestation; Citrus; Fruit fly 1. Introduction * Corresponding author. Tel.: +61-3-92109222; fax: +61-3-98003521. E-mail address: peter.williams@nre.vic.gov.au (P. Williams) Citrus infested with larvae of Queensland fruit fly (Qfly), Bactrocera tryoni (Froggatt), is unacceptable in the marketplace. Two treatments are currently registered for control of the fruit fly 0925-5214/00/$ - see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0925-5214(00)00093-4
194 P. Williams et al. / Posthar est Biology and Technology 19 (2000) 193 199 when fruit is sent interstate or exported. They are fumigated with methyl bromide for 2 h or cold treatment at 1 C for 16 days. Methyl bromide can damage some citrus, e.g. mandarin oranges, and it has been identified as an ozone-depleting chemical with its use is being restricted in accordance with an international agreement, the Montreal Protocol (Anonymous 1998). Currently there are no restrictions on use of methyl bromide as a commodity treatment for pre-shipment and quarantine purposes, but in the future restrictions may be imposed, with the deadline for phase-out now advanced to 2005. Cold treatment for 16 days places time constraints on handling of fruit, and some types of citrus, e.g. oranges and lemons, may suffer chilling injury (Goubran, 1989). Handling constraints are a particular problem for growers when outbreaks of Qfly occur in areas generally free of the fly. Exclusion zones of 15 km diameter centred on the outbreak site are imposed to make movement of fruit illegal unless it has been subjected to an approved disinfestation treatment. Many growers have limited cool room space and would be severely disadvantaged if all their fruit had to be cold-treated before it could go to market. Consequently development of other alternatives to methyl bromide is considered a high priority research objective by the citrus industry (Anonymous, 1996). Surface treatments with petroleum oils have recently been developed for control of surface pests such as scale insects (Taverner et al., 1998) but they do not provide adequate control of pests that live within the fruit, such as fruit fly larvae. Phosphine is used extensively as a fumigant for stored produce, having the advantage of leaving no detectable toxic residues in produce following ventilation. However, it has the disadvantage that longer fumigation times are required than for methyl bromide (Bond, 1984). Traditionally, phosphine fumigation is carried out using solid formulations of aluminium and magnesium phosphide from which the gas is gradually evolved. Some work has been done with these formulations, placing them in water to accelerate the rate of release of phosphine. However, exposure times were limited to 4 h (to be competitive with methyl bromide), concentrations were similar to those used in grain and the treatments were not sufficiently effective against Qfly (R. Corcoran, personal communication, 1997). Successful fumigation of fruit, including tomatoes and grapefruit infested with Oriental, Mediterranean and Caribbean fruit flies, using solid phosphine generating formulations, has been reported (Seo et al., 1979; Hatton et al., 1982). Exposure periods for these fumigations were much longer (up to 4 days) than for methyl bromide. During the 1980s, a cylinder gas formulation of phosphine, Phosfume (2% phosphine with carbon dioxide as a carrier gas) was developed (Winks and Russell, 1994), eliminating problems of powder residues associated with solid formulations and enabling fumigation times to be reduced. Phosfume, recently renamed ECO 2 FUME, has been used successfully in experiments to control several important pests of Australian wildflowers for export using phosphine concentrations of up to 1 g m 3 and exposure periods of up to 16 h (Muhunthan et al., 1997; Williams and Muhunthan, 1998). The objective of this study was to assess the potential for use of the phosphine formulation ECO 2 FUME for disinfestation of citrus infested with Qfly larvae. 2. Materials and methods 2.1. Infesting and handling oranges Export quality Navel oranges were infested with Qfly eggs at the Horticultural Research and Advisory Station, Gosford NSW. The rind of each orange was pierced with a spiked block which made 20 holes to make it easy for flies to oviposit. Oranges were placed on top of fly mesh cages of adult fruit flies, which laid through the mesh into the oranges. For each series of fumigations, 20 cartons of infested oranges and two of uninfested control oranges were transported to the Keith Turnbull Research Station, Frankston, Victoria (KTRI) where fumigations were carried out. The cartons of infested oranges were enclosed in terylene voile bags to prevent escape of any flies
P. Williams et al. / Posthar est Biology and Technology 19 (2000) 193 199 195 that might have crawled into the boxes. At KTRI the oranges were held in a quarantine insectary at 26 C ( 2 C) and 40 70% RH to provide favourable conditions for the development of Qflies. The control oranges were held in a refrigerator at 11 C. Oranges used were Leng Navels from Mildura, Victoria and Washington Navels from Yanco and Forbes, New South Wales. 2.2. Fumigating oranges Cartons of oranges were fumigated with ECO 2 FUME in a 900-l chamber fitted with a heater and a gas recirculation and exhaust fan. Phosphine concentrations were determined using gas detector tubes (Dräger ) shortly after introduction of ECO 2 FUME, and again just before the end of each fumigation. In fumigations lasting longer than 24 h, phosphine concentrations were often also determined after half the fumigation time had elapsed. On some occasions additional phosphine was introduced at this time and the boosted concentration measured. The first fumigations of Leng Navel oranges were based on a schedule developed for wild flowers (Muhunthan et al., 1997). Initial phosphine concentrations ranging from 0.94 to 1.36 g m 3 were used with exposure times ranging from 16 to 24 h. Average temperatures during fumigation ranged between 20 and 23 C. The heater thermostat in the fumigation chamber was set at 20 C, but often the temperature in the chamber was above 20 C, because of the higher initial temperature of the oranges and higher ambient temperatures. Temperatures were recorded every 2 min by two Tinytag loggers with probes placed amongst infested oranges. In each fumigation, three or four cartons each containing 80 96 infested oranges were used together with a carton of 20 uninfested oranges. In fumigations of Washington Navel oranges, exposure time was increased up to 48 h and initial phosphine concentrations of up to 1.8 g m 3 were used. The thermostat was set to give 20, 23 or 25 C and temperatures within an infested orange were measured by a probe. 2.3. Assessing effecti eness of fumigations Four groups of ten infested oranges were sampled from the cartons. Each group was placed in a plastic box 580 mm 380 mm 165 mm with a perforated base, which fitted on top of a box of similar dimensions with a solid base, covered with fine grade vermiculite. Mature larvae left the fruit, passed through the perforations and pupated in the vermiculite. Each pair of boxes was enclosed in a terylene voile bag to prevent escapes. About 10 days later the vermiculite was sieved for puparia. If larvae were found, sieving was repeated a few days later. The number of puparia formed was used to estimate the number of larvae infesting oranges that were fumigated. Following each fumigation oranges were taken from the cartons and placed in similar plastic boxes to those used for the unfumigated, infested oranges. After the last puparia from the boxes of unfumigated infested oranges were collected, the boxes of fumigated oranges were sieved for puparia. Then the oranges, vermiculite and puparia were autoclaved. A few puparia were kept to see if they would complete development to the adult stage, before being killed. Effectiveness of the fumigations was assessed by estimating percentage mortality using the number of puparia produced compared with the estimated number of larvae exposed. The estimate of larval numbers was based on the average number of pupae per orange produced from batches of unfumigated infested oranges. Lower 95% confidence limits for mortality were found using exact calculations based on the binomial distribution. This is the method of calculation used by Couey and Chew (1986) to relate the number of insects exposed to mortality and probit values. 2.4. Assessing effects of fumigations on oranges Twenty uninfested oranges from each consignment were included in each fumigation and subsequently examined by a panel of three for appearance in comparison with control batches. Five of the treated oranges from each fumigation were also cut open and assessed for rind condition, juiciness and taste after a day airing and again up to 4 weeks later.
196 P. Williams et al. / Posthar est Biology and Technology 19 (2000) 193 199 3. Results and discussion 3.1. Effecti eness of fumigations on fruit fly lar ae In fumigations with Leng Navel oranges the thermostat on the fumigation chamber was set to give a temperature of 20 C, initial phosphine concentrations ranged from 0.94 to 1.36 g m 3 and exposure times ranged from 16 to 24 h (Table 1). When the oranges were transferred from the cartons in which they were fumigated to the incubation boxes, dead larvae were observed on many of the oranges. The number of larvae that emerged from these oranges to pupate was far less than the estimated number of larvae exposed (based on pupae produced from the infested unfumigated oranges) (Table 1). While the furnigations showed that phosphine had promise as a control measure, the number of pupae produced was considered commercially unacceptable for export consignments of citrus, for which a minimum mortality level of 99.99% is generally required. For interstate trade the minimum level of mortality accepted is 99.5%. Although this level was exceeded in the best fumigation (Table 1) it was considered that using an initial phosphine concentration of 1.3gm 3 or more and an exposure time of 24 h would give greater reliability to the treatment. Increasing exposure time is known to be very effective in improving efficiency and reliability of phosphine fumigations (Winks, 1987). Washington Navel oranges were used in two subsequent series of fumigations. In the first of these the thermostat was set on 20 C, the maximum initial dosage of phosphine was 1.46 g m 3 and exposure times were 40 and 48 h. In two of the 48-h fumigations phosphine concentrations were measured after 24 h and then topped up to 0.7gm 3. There appeared to be advantages in using a 48-h rather than a 40-h exposure period (Table 2). Results obtained in the 48-h fumigations were better than for any previous fumigations and would certainly be acceptable for interstate trade and export to some countries, as all these fumigations gave mortality levels exceeding 99.9%. Further improvement was possible by increasing either the exposure time or the temperature of the fumigation. It was decided to increase the temperature rather than the exposure time, because it was considered likely to be more acceptable to industry. Thermostat settings were increased to 23 and 25 C. Two fumigations were carried out with the thermostat set at 23 C and two at 25 C. In the first fumigation, phosphine concentrations were measured at the beginning and end of the exposure time (Table 2). The final concentration of phosphine (0.14 g m 3 ) was considered lower than desirable, particularly because it could have been low for many hours. To overcome this problem in subsequent fumigations, concentrations were checked after 24 h and topped up to 0.58 or 0.70 g m 3. Only one pupa, from which the adult fly did not emerge, was produced from all the fumigations with thermostat settings of 23 or 25 C in which initial phosphine concentrations were 1.67 g m 3 and the concentrations were topped up after 24 h. The percentage mortalities achieved and lower confidence limits shown in Table 1 Fumigations of Leng Navel oranges infested with Queensland fruit fly larvae Temperature Exposure time Phosphine concentration Estimated no. of No. of pupae Mortality (%) ( C) (h) (g m 3 ) fly larvae produced Initial Final 20 16 0.98 0.55 11 688 421 96.39 21 20 0.94 0.40 8460 53 99.37 20 20 1.08 0.81 12 375 488 96.06 23 20 1.36 0.58 12 874 201 98.44 21 24 1.26 0.49 9248 16 99.83
Table 2 Fumigations of Washington navel oranges infested with Queensland fruit fly larvae Temperature ( C) Exposure time (h) Phosphine concentration (g m 3 ) Estimated no. of No. of pupae Mortality (%) 95% lower confidence fly larvae produced limit Initial Middle Final 20 40 1.13 0.35 12 112 23 99.810 20 48 1.25 0.31 0.70 0.36 12 112 2 99.983 20 48 1.46 0.54 0.70 0.35 14 896 8 99.946 20 48 1.39 0.14 14 144 0 100 23 48 1.46 0.14 19 552 5 99.974 99.909 23 48 1.67 0.29 0.58 0.10 19 552 0 100 99.985 25 48 1.67 0.21 0.70 0.14 16 256 1 99.994 99.971 25 48 1.67 0.35 0.70 0.14 12 960 0 100 99.977 Both 25 48 1.67 0.70 0.14 29 216 1 99.997 99.984 23 25 48 1.67 0.58 0.70 0.10 0.14 48 768 1 99.998 99.990 P. Williams et al. / Posthar est Biology and Technology 19 (2000) 193 199 197
198 P. Williams et al. / Posthar est Biology and Technology 19 (2000) 193 199 Table 2 indicate that the schedule used for these fumigations would be reliable for interstate trade and possibly also for some international trade. particularly if incorporated as part of a systems approach integrating use of control measures, such as baiting, with monitoring and timing of harvest to minimise the chance of citrus being infested. Temperatures within oranges and cartons were between 23 and 27 C, except in one fumigation where the carton temperature rose to above 30 C, probably because of moulding and rotting oranges. 3.2. Effects of fumigations on oranges The Leng Navel oranges maintained their good condition following fumigation. There was no evidence of damage to the rind, and taste and juiciness were unaffected. A similar result was obtained with the Washington Navel oranges, with the exception of one fumigation. Fumigated oranges were similar to the unfumigated ones in appearance, juiciness and taste, both on day 1 after fumigation and after storage for a further 4 weeks, when they were all considered commercially acceptable. The exception was the final 48-h fumigation in the first series of fumigations with Washington Navels. There was a peculiar smell emanating from the carton of uninfested oranges, and as well the oranges themselves had a noticeably bitter taste. The cartons of infested oranges in this fumigation contained many oranges with extensive moulding, mainly from green mould (Penicillium digitatum). These cartons had a similar smell to that detected in the carton of uninfested oranges, so the offodour and bitter taste probably resulted from contamination from the cartons of infested oranges rather than from the fumigant. 4. Conclusions This study has demonstrated that phosphine has potential for use as a replacement for methyl bromide for disinfestation of citrus from Qfly. Further work needs to be done to develop means of utilising the fumigation technique by the citrus industry. It is anticipated that, as with grain fumigations, there will be no problems with residues, but this needs to be demonstrated. Acknowledgements This work was conducted with support from the Horticultural Research and Development Corporation (HRDC), the Australian Citrus Industry, Agriculture Victoria, NSW Agriculture, Gosford and BOC Gases. References Anonymous, 1996. The Australian Citrus R&D Plan 1996 2000. Horticultural Research and Development Corporation, Gordon NSW, 25 pp. Anonymous, 1998. 1997 Update of the handbook for the international treaties for protection of the ozone layer. Ozone Secretariat United Nations Environmental Programme, Nairobi, Kenya 233 pp. Bond, E.J., 1984. Manual of fumigation for insect control. FAO Plant Production and Protection Paper 54, 432 pp. Couey, H.M., Chew, V., 1986. Confidence limits and sample size in quarantine research. J. Econ. Entomol. 79, 887 890. Goubran, F.H., 1989. A cold treatment for lemons. Rural Q. 2, 23. Hatton, T.T., Cubbedge, R.H., von Windeguth, D.L., Spalding, D.H., 1982. Phosphine as a fumigant for grapefruit infested by Caribbean fruit fly larvae. Proc. Fla. State Hortic. Soc. 95, 221 224. Muhunthan, M., Williams, P., Thorpe, G.R., 1997. Phosphine an alternative to methyl bromide for postharvest disinfestation of wildflowers in containers. Agric. Eng. Aust. 26, 29 33. Seo, S.T., Akamine, E.K., Goo, T.T.S., Harris, E.J., Lee, C.Y.L., 1979. Oriental and Mediterranean fruit flies: fumigation of papaya, avacado, tomato, bell pepper, eggplant, and banana with phosphine. J. Econ. Entomol. 72, 354 359. Taverner, P., Bailey, P., Roush, R., 1998. Old myths and new oils: insecticidal oils in action. In: Zalucki, M.P. Drew, R.A.I., White, G.G. (Eds.), Proceedings of the 6th Australian Applied Entomological Research Conference, vol. 2, Brisbane, Australia, pp. 185 193. Williams, P., Muhunthan, M., 1998. Fumigants for postharvest control of insect pests of cut flowers. Acta Hortic. 464, 291 296.
P. Williams et al. / Posthar est Biology and Technology 19 (2000) 193 199 199 Winks, R.G., 1987. Strategies for effective use of phosphine as a grain fumigant and the implications of resistance. In: Donahaye, E., Navarro, S. (Eds.), Proceedings of the 4th International Working Conference on Stored-Product Protection, Tel Aviv, Israel, pp. 335 344. Winks, R.G., Russell, G.F., 1994. Effectiveness of SIROFLO in vertical silos. In: Highley, E., Wright, E.J., Banks, H.J., Champ, B.R. (Eds.), Proceedings of the 6th International Working Conference on Stored-Product Protection, Canberra, Australia, pp. 1245 1249..