PESTDECIDE is a further development of the Pesticide Index (Penrose et al 1994) which arose from concerns about simplistic methods proposed for measuring pesticide use in the Australian apple and pear industry. This followed the industry's agreement to be a signatory to the Pesticides Charter (Anon 1991), negotiated between 19 organizations including consumer and environmental groups. The Australian Apple and Pear Growers' Association (AAPGA) was the first fruit grower organization to support the Charter as a result of the devastating effect on sales of red apples following media sensationalism surrounding the use of the plant growth regulator daminozide (Alar ®) in 1989.

The Pesticides Charter committed the AAPGA to "explore the possibilities for achieving ..... reductions in overall pesticide usage ......" through a number of agreed steps. The final target was a 75% reduction by the year 2000.

Pesticide Use Reduction
There were no criteria outlined in the Charter for measuring reduction in pesticide use. Reducing the number of sprays applied during the season was one obvious measure. However, this failed to take into account a move from "hard" insecticides to "soft" alternatives nor did it allow for the continued development of integrated pest management (IPM) programs in which some products were specifically registered for use at reduced rates. Both of these developments were seen as beneficial to the consumer and the environment but neither would necessarily reduce the number of sprays.

Likewise, the discovery and development of more selective and biologically active molecules meant less chemical was being used. For example, the broad spectrum fungicide mancozeb (Dithane ® [registered trade name]), discovered in about 1960, is recommended for use at about 2.5 kg active ingredient (AI) per ha for control of apple scab, whereas the more recently developed hexaconazole (Anvil ®) will provide control of the disease at a rate of just 40 g AI/ha. Ignoring the implications of resistance, replacing mancozeb with Anvil will reduce (by about 60 times) the quantity of pesticide needed to control apple scab, but not necessarily the number of sprays.

Penrose et al (1996) showed some apple growers could achieve commercially acceptable control of pests and diseases with less than half the number of sprays as others within the same area. In most cases, the additional pesticides reflected the risk averseness of the higher users. Often the wish to ensure good cosmetic appearance of the fruit influenced growers to apply the extra sprays which made up a very small proportion of total production costs (Penrose 1995).

Measuring Pesticide Use with PESTDECIDE
Based on the Pesticide Index (PI) concept, a data sheet was prepared for every pesticide registered for use on apples in Australia. A PI score was calculated for 177 insecticide and 99 fungicide recommendations listed on 105 labels available up to June 1995. For some characteristics or attributes, the PI score was derived from objective data, for example mammalian toxicity used published oral LD₅₀s for rats. Other characteristics relied on subjective assessment. Perceptions of field efficacy and the impact of the chemical on apple IPM programs are two examples. All characteristics were given a rating from 1 (desirable) to 5 (undesirable). Some examples of the ratings are given in the table below.

Examples Of Objective (O) And Subjective (S) Ratings For Apple Pesticides

Characteristic or attribute	Pest or disease	Pesticide and (Rating1)
		Desirable	Undesirable
Mammalian toxicity (O)	Codling moth	Fenoxycarb (1)	Parathion-methyl (5)
Persistence on crop (O)	Woolly aphid	Pirimicarb (2)	Vamidothion (4)
IPM compatibility (S)	Apple dimpling bug	Endosulfan (2)	Tau-fluvalinate (5)
Environmental impact (O)	Apple scab	Dodine (2)	Mancozeb (5)

¹range 1-5

A weighting (range 1-4) was applied to emphasise that some characteristics were more important than others, for example potential toxicity to the user (4) was considered more important than cost (1).

As an alternative to counting sprays applied as a measure of pesticide use, PESTDECIDE asks the user to plan the forthcoming season's spray program by selecting products from the data sheets, taking into account the PI scores so as to select the lowest scores possible. This needs to be consistent with the growers' experience of how individual products fit in with orchard management. A maximum total PI (TPI) score for the season will have been set (based on historical data for the district) and the grower aims to stay below the maximum in planning the program. Using spray records, the actual TPI for the season can then be calculated, so the grower can see if the score remains below the maximum.

Exceptional factors such as an unusually wet spring (favourable to apple scab) or a 1 in 20 year invasion by plague thrips has been allowed for within PESTDECIDE, as have the choice of 'green' products such as petroleum oil and copper based fungicides.

Evaluation of PESTDECIDE
In 1995/96, 61 grower cooperators in mainland Australia evaluated the use of PESTDECIDE. A maximum TPI target of a 2100 was set; 82% of the cooperators met that target. The levels of pest and disease were not assessed. Following this evaluation and improvements to the decision support system, a PESTDECIDE package for the Australian apple and pear industry could now be used to plan and measure reduction in pesticide use.

PESTDECIDE is not without its critics. Some growers show reluctance to adopt it because they fear that it will be used as a device to promote fruit from districts with fewer pest and disease problems, and therefore a lower TPI, over those where pest and disease pressures are higher with, presumably, a higher maximum TPI. These differences should be able to be overcome within quality assurance (QA) protocols.

The concept has also been criticised as "flawed and illogical" because the rating of the attributes of the various pesticides is "arbitrary and irrational". Another critic states "farmers need simple tools for making objective decisions concerning pesticide applications" but that "the system is based on a series of subjective criteria and weightings". Admittedly there are some "fuzzy" aspects of PESTDECIDE, such as subjective assessments sometimes based on limited information and the inability to obtain objective data for all pesticides within the one attribute. For example, data from Kovach et al (1992) was used to rate the environmental impact, however not all pesticides registered for apples in Australia in 1995 were listed.

Conclusion
Despite the problems, PESDECIDE is simple, robust and easy to apply. Its application highlights the data gaps needed to improve its objectivity. Although the weightings may be varied, such as from crop to crop or practitioner to practitioner, if they are stated explicitly and not varied for a particular purpose, the system can provide a means of assessing the overall impact of changes in pesticide programs.

PESTDECIDE provides decision support and a logical method for measuring the use of pesticides in apple crop protection programs. Now that it has been developed and evaluated in apples, its modification for use in other cropping systems should be possible.

References
Anon (1991). Pesticides Charter. p 21 in Towards a national food policy. The National Working Group on Food Policy. Australian Consumers' Association, Sydney.

Kovach, J; Petzoldt, C; Degni, J and Tette, J (1992). A method to measure the environmental impact of pesticides. New York's Food and Life Sciences Bulletin No. 139. New York State Agricultural Experiment Station, Geneva, NY. 8 pp.

Penrose, LJ (1995). Fungicide use reduction in apple production - potentials or pipe dreams? Agriculture, Ecosystems and Environment 53: 231-242.

Penrose, LJ; Thwaite, WG and Bower, CC (1994). Rating index as a basis for decision making on pesticide use reduction and for accreditation of fruit produced under integrated pest management. Crop Protection 13: 146-152.

Penrose, LJ; Bower, CC and Nicol, HI (1996). Variability in pesticide use as a factor in measuring and bringing about reduction in pesticide usage in apple orchards. Agriculture, Ecosystems and Environment 59: 97-105.

Address:
NSW Agriculture
Orange Agricultural Institute
Forest Road
Orange, NSW 2800
Australia

Phone: +61 (2) 6391 3800
Fax: +61 (2) 6391 3899
E-Mail: graham_thwaite_at_orange__ai@smtpgwy.agric.nsw.gov.au

Mr. Graham Thwaite is an entomologist with the state department of agriculture in New South Wales, Australia. He is based at the department's Agricultural Institute located at Orange, 300 km west of Sydney in the state's largest apple growing area. During his 30 year career with NSW Agriculture, he has worked on the management of the pests of several field and deciduous fruit crops. In recent years his research has concentrated on apple pests, usually involving collaborative work with colleagues in southern Australia. He was one of two NSW Agriculture delegates at the Australian National Food Policy Conference in the national capital, Canberra, in November 1991 when the Australian "Pesticides Charter" was launched. The Charter committed the Australian apple industry to try to reduce its use of pesticides. Graham has represented his department on apple pest management issues at various national and international forums and has published a number of scientific papers including a recent review on apple integrated pest management in Australia.