A methodology to assess environmental pesticide pollution during vine spraying Carole Sinfort* — Eric Cotteux ** — Bernadette Ruelle** — Vincent De Rudnicki** — Bernard Bonicelli ** * Montpellier SupAgro - UMR ITAP 2, pl. Viala F-34060 Montpellier [email protected] ** Cemagref - UMR ITAP BP 5095 F-34033 Montpellier {eric.cotteux, bernadette.ruelle, vincent.derudncki, bernard.bonicelli}@cemagref.fr Vine spraying induces pesticide losses to the environment (ground, air) that are responsible of background pollutions. Three test campaigns (one per year) were set up to : i) quantify pesticide losses, ii) compare the effects of sprayer technologies and iii) evaluate the importance of the machine set-up. Measurements were based on the use of a fluorescent dye (Brilliant Sulfo Flavine) trapped on several kind of collectors fitted to each evaluated compartment (ground, plant, soil). The amount of deposit was then obtained through spectro-fluorimetry analysis. Efficiency of collectors was previously tested in laboratory conditions. The implementation of the collectors within the test plots was also optimised to ensure a good estimation. The tested sprayers belong to the farmers participating to the project or to the research institute (Cemagref). They were adjusted by a specialised technician from the local agricultural advice organisation. The results show that the measurement protocol gives a good estimation of the sprayed pesticide fate as roundly 100% of the sprayed product was found from the collectors of each compartment. The amounts of lost products are considerable : from 10 to 45% on the ground, 15 to 40% in the air. Losses on the ground are most important when spraying at an early vegetative stage and are difficult to reduce in this case, even when improving the machine setup. When the canopy is fully developed the orientation of the sprays is shown to have a significant effect : a bad orientation setup can multiply ground losses by three. ABSTRACT. Les traitements phytosanitaires des vignes entraînent des pertes vers l’environnement (sol, air) qui sont la cause de pollutions diffuses. Trois campagnes d’essais (une par an) ont été mises en place pour : i) quantifier les pertes de pesticide, ii) comparer les effets des technologies RÉSUMÉ. des pulvérisateurs et iii) évaluer l’importance des réglages des machines. Une méthodologie a été développée pour évaluer les pertes de pesticides. Elle est basée sur l’utilisation d’un traceur fluorescent (de la Brillant Sulfo Flavine) piégé sur des collecteurs adaptés à chaque compartiment évalué (sol, plante, air). Les quantités de dépot sont ensuite obtenues par une analyse spectro-fluorimétrique. L’efficience des collecteurs a été préalablement testée en laboratoire. Le positionnement des collecteurs à l’intérieur des parcelles d’essai ont également été optimisées pour assurer une bonne estimation des dépôts réels. Les pulvérisateurs testés appartiennent aux viticulteurs qui participaient au projet ou à l’institut de recherche (Cemagref). Ils avaient été préalablement réglés par un technicien de la Chambre d’Agriculture. Les résultats montrent que le protocole de mesure donne une bonne estimation du devenir des pesticides puisque environ 100% du produit pulvérisé est retrouvé à partir des collecteurs dans les différents compartiments. Les quantités de produit perdu sont considérables : entre 10 et 45% sur le sol et entre 15 et 40% dans l’air. Les pertes au sol sont plus importantes lors des traitements en stade végétatif précoce et sont difficiles à réduire dans ce cas, même en optimisant les réglages. Quand la végétation est totalement développée, l’orientation des jets a un effet significatif sur les pertes : un mauvais réglage de cette orientation peut multiplier les pertes au sol par trois. KEYWORDS: pesticide, sprayer, spraying, pollution, ground, air, field measurements MOTS-CLÉS : terrain produit phyosanitaire, pulvérisateur, pulvérisation, pollution, sol, air, mesures de 1. Introduction France is the world’s 3rd biggest pesticide user after the United States and Japan. 77 000 tonnes of pesticides were consumed nationwide for agricultural purposes in 2007. Although only representing 4% of the total cultivated land, viticulture consumes 20% of these plant protection products. Only a part of the total pesticides sprayed actually reaches the target (weeds, crop, insects, fungi, etc). Over the last years, the role of pesticides in environmental contamination, has led authorities, and society as a whole, to raise questions on the potential risks that the general public are running, primarily with regards to ground water pollution, and more recently in relation to air pollution. Viticulture is particularly concerned because of the quantities of products sprayed and the technologies used. Indeed, to facilitate pesticide penetration of vegetation, vineyard sprayers are equipped with turbines creating a turbulent air flow that assists drop transport and which shake the leaves. Some of this air is deviated by the vegetation barrier, either towards the ground or the atmosphere, taking with it a non-negligible amount of pesticide. [BER 99] estimated that between 30 and 50% of spread masses could be lost towards the atmosphere. During tests conducted on apple orchards [CRO 01], ground losses were estimated to be between 24 and 97% and atmospheric losses ranged between 5 and 51%. These environmental losses are difficult to estimated, due notably to the fact that many molecules have to be detected (several hundred [L’H 09]). The quantitative estimation of losses is made even more complex as it depends on the equipment used, the way the tractor is driven, settings (operator dependent) as well as terrain and weather conditions. Since the 1960ies, much research has described methodologies to measure drift, by assessing ground deposits in close proximity to the treated area and/or deposit estimation on vertical targets: the American consortium “Spray Drif Task Force”has identified 2500 scientific articles on drift [SDT 97]. However, a thorough quantification of contamination of air and ground compartments has not been undertaken. Moreover, a pesticide loss balance for the different compartments was not found in the literature. The study, on which this article is based, was conducted over 3 consecutive years on a small vineyard catchment basin in the Herault “department”– the Valeille catchment basin located in the Nefes “commune”– with the involvement of vine-growers. The objective of the study was to ascertain the practices and equipment used in Languedoc viticulture, and to estimate pesticide losses in the environment resulting from the different practices used on the catchment basin. This article aims to present the methodologies used to: i) quantify pesticide losses in the environment, ii) compare impacts of different sprayer technologies and iii) assess the importance of equipment set-up. After a short summary of methods available to measure pesticide losses to the environment, we describe the methodology developed as well as the protocols implemented, and to conclude the main results obtained over the 3 years of tests are presented and discussed. 2. Methods to estimate pesticide losses to the environment The most practical method to detect drop impacts is based on the use of hydrosensitive paper [DEM 00, LAN 04, FAR 04, PAN 04]. They can be placed directly on vegetation to assess treatment efficiency, on the ground to evaluate losses or also on mountings to assess atmospheric pesticide losses. Although excellent for qualitative assessments, they are not suitable for quantification of pesticide mass losses. Lazer velocimeters or PDAs (Phase Doppler Analyser) have been used since the eighties to evaluate spray drop transport and spray cloud dispersion (cf. [ZAL 80] for airassisted spraying). Later, LIDAR (Light Detection and Ranging) systems were made available with the same objectives in mind; however these systems cannot be used for a quantitative approach [HOF 89, STO 97]. To obtain quantitative evaluations, the methodologies described in scientific literature use alternatively dyes trapped in collectors, which are usually passive, or analytical chemistry techniques. Among the dyes described, there are radioactive tracers [DOB 83], which are very accurate but not readily available, metal-based dyes, which are quantified using atomic absorption spectrometry [YAT 76, CRO 01, DAS 05], by conductimetry or using techniques associated with electro scanning microscopy and image analysis (Electron Beam Analysis [KRA 04]), colourings (especcially tartrazine, food colouring E102), analyzed using a spectrophotometer [PER 01, ADE 07], and fluorescent dyes. The use of the latter has increased significantly these last few years (e.g., [MIL 89, BAY 05, GIL 07]) because spectrofluorometry measurements are both fast and very accurate (µg/L detection capacity). The most commonly used fluorescent dye is Brillant Sulfo Flavine, which offers the advantage of a low light-related breakdown rate [CAI 97]. Other tests mention the use of another dye: Caracid Brilliant Flavine FFN [KIR 04]. The fluorescent dyes are collected on passive media, which are then rinsed using di-ionized water and swabbed, the concentration of solution obtained being then measured. The utilisation of these dyes (fluorescent or otherwise) does not, unfortunately, allow for quantification during the vapour stage [SOL 96]; thus they should only be used in close proximity to the point of emission. Analytical methods remain the most reliable way of estimating pesticide concentrations in the air or on the ground, or on any other media. In the air, samplers that suck air through a filter or a resin, are the most commonly used techniques depending on whether liquid or gas phase is looked for. The type of sampler plays an important role in quantities measured, depending on the characteristics of the product trapped [BUI 98, BRI 2a]. The most commonly used analytical methods used to assess environmental pesticide presence are chromatography for the liquid phase (LPC) [CAR 06] or chromatography in the gas phase (GC), which is complemented by mass spectrometry (MS) [MIL 00, WIT 02, RAV 05]. Although there are solutions that can be used for multi-residue analysis [PIC 95, BRI 2b], these methods are especially used when looking for a given active substance, that has been chosen beforehand, as they rely heavily on the physicochemical properties of the sought-after molecule [GIL 05]. 3. Materials and Methods Taking into account the specificities of the different methods mentioned above and the significant number of tests to be performed, we chose to use a flourescent dye called BSF, and, with regards to out specific objectives, to optimize the nature and positioning of the air, ground and vegetation tracer-collectors as well as test protocols. The fluorescent concentration of samples is determined with a LS45 Perkin Elmer spectrofluorimeter from disposable polystyrene tanks located 1Cm from optical trajectory. The excitation wavelength is 455 nm and the signal emitted is read at 500 nm. 3.1. Choice and Positioning of Collectors 3.1.1. Ground Deposit Collectors The ground collectors are cut strips of 5cm x 1,5m, fitted within flexible irrigation sheaths. The strips are rinsed in 500mL of deionised water and then swabbed. The recovery rate of this media was tested in laboratory conditions by placing a known quantity of the media using a micropipette: 100To limit uncertainties linked to sampling, we chose to measure ground losses along unbroken lines of collectors placed perpendicularly to sprayer trajectory advance. The repetition of several lines allowed us to take into account variability in: vegetation density, ground surface (ruts) and advance speeds along vine rows. Preliminary tests were conducted to determine statistically the number of collectors required to obtain a real value for results with a maximum standard deviation of 5%. The real value was estimated using 30 collectors with 10cm spacings the number of collectors was set at 10 sampling lines of 5 collectors at .5m intervals and covering a width of 7.5 m (3 inter-rows) including the tractor passage. In addition, for the balance measurements, the system was extended using a set of collectors parallel to tractor advancement in order to widen the measurement zone to include 2 further rows – one on each side of the sprayer passage – ( 7 inter row areas are thus covered). 3.1.2. Atmospheric Deposit Collectors To measure air losses, the collectors used consist of PVC wires of 2 mm in outer diameter (1 mm inner) with hardness exceeding 60 Shore A. The capacity of this system to trap the dye and to sample air losses was the subject of several research efforts Cemagref including, a theoretical study, wind tunnel experiments and tests to quantify the recovery rates after rinsing [GIL 07]. The trapping capacity of these collectors was estimated to be 80%. A corrective factor was integrated into the results to take into account this characteristic. During tests 5 wires measuring 10m in length were extended above the plot and above the 5 adjacent inter-rows at a height of 2m50. This height was chosen to reduce evaporation-related losses [GIL 07]. After spraying the wires were collected and individually rinsed in 250 mL of water. Figure 1. Tested sprayer diagrams: T, Tangential ; A1, airblast sprayer ; A2, airblast equipped with deflectors; A3, airblast for narrow vine rows; P1, simple pneumatic sprayer; P2, pneumatic sprayer with width adjustable diffusers. 3.1.3. Vegetation Deposit Collectors On vegetation the leaves and fruit act as collectors. To create a measurement zone, all the leaves and fruit located along 15cms of row length, are collected. The method was used during previous studies [DAS 02] and demonstrated an excellent capacity to curb variability between repetitions in relation to tests where only a few leaves were sampled or when collectors were placed here and there in the vegetation. Collection efficiency of 90% was obtained by proceeding in the way manner as with the ground collectors. This efficiency rate was taken into account in the results. 3.2. Experimental Protocols The machines tested are presented in a diagram form in figure 1. Four sprayers were chosen from among those used by the winegrowers. These sprayers are suited to vine conditions in the region, which are characterized by small plots of gobelet or cordon de Royat wire trained vines: two air assisted sprayers (A1 et A2) and pneumatic sprayers (P1 et P2). Two other sprayers, although not used in this particular catchment basin are used in the region (A3 and T) were also tested in specific configurations. All sprayers are towed, except machine A3, which is mounted. The sprayers were all calibrated before tests by a technician from the local Chamber of Agriculture. Before each test, we calibrated jet orientation, and when possible deflector orientation with regards to the position of the vines. Sprayers were towed/driven at a speed of 5 km/h in compliance with volume per hectare practices used by winegrowers. The flow rates were measured before each series of tests. Weather conditions were recorded throughout the duration of the tests using an ultra sound sensor (Young Model 81000 V) positioned a height of 3m50 within the plot and with temperature and hygrometry sensors (FH A646) positioned at vegetation level. Each test was repeated 3 times. 3.2.1. Technology Impacts The different machines were tested on artificial vines, described in [GIL 07], to mitigate vegetation variability. The artificial vineyard tested was made up of 4 rows of 10m vine lengths with 2m spacings. The structure was composed of metal components 50cm wide and 1m50 high covered with 1mm by 7mm mesh net acting as wind breakers. Tests conducted in a wind tunnel showed that the artificial vines had the same drag coefficient as real vines of the same size at the mature vegetation stage. The sprayers made 2 return passages over the central row. The deposits were measure on the ground and in the air (1 wire above the central inner row and 2 wires suspended on either side of it). 3.2.2. Impacts of Calibration and the Vegetation Growth Stage The tests were conducted on a plot of Marselan vines along “Guyot”tressled trained rows. The vines are made up of 10 rows with 2.5m row spacings, and the vines were planted at 1.75 m intervals along the longest length. The machine treats every second row and is driven along 3 rows. Only ground deposits are measured. Sprayers T, P1 and P2 were used for these tests. The deposits were measured over a distance of 10m for a width of 7m50, in a zone located in the middle of the plot. Spraying is begun 10m before and 10m after the designated zone. The vegetation growth stages at which the tests were conducted are: in May, when flower buds appear, in June, at the full blossom stage, and in July, when the mature vegetation stage is reached. In July, the tests are performed once the vine training wires have been removed and after leaf cutting. At each of these vine-growth stages, sprayer calibration effects were studied. The number of open outlets, in May (all open or only those required) tractor/sprayer speed (5km or 7km/h), in June and nozzle orientation during July. The conditions for this last series of tests concerned the adaption of the sprayer to the distance between rows. The vine growers often have several plots with different row spacings but do not recalibrate sprayer orientation from one plot to another. Consequently we tested 2 different sprayer set ups on the plot with 2m50 row spacings; one optimized for 2m row spacings and the other optimized for 2m5 row spacings (the most commonly used spacing in the region). For sprayer T, for which nozzle orientation cannot be modified we lowered the sprayer modules by 10cm. 3.2.3. Pesticide Balance after spraying The tests were conduced on the same plot, and at the same 3 vegetation growth stages. Tests were only conducted for Sprayer P1. The sprayer made 2 return trips along the same row. An area 20m long and 17.50m wide was marked out in the centre of the plot. The first section (the first 10 meters) was used to take vegetation deposit measurements: the leaves of 4 four plants are collected, 2 are from the row directly treated and the other 2 plants are located on the adjacent rows. Ground and air deposits are estimated over the second section of the designated zone (the following 10 metres). After each test, leaves are collected at 2 specific points to check the initial florescence level. 4. Results 4.1. Technology Impacts Table 1 presents the percentages of atmospheric and ground losses measured for the six sprayers chosen, taking into account spray flow rates. The losses were measured for 3 repetitions on average. Atmospheric losses are much heavier than those recorded at ground level: they are estimated at between 15 and 37%. The variation coefficients computed over the 3 repetitions are quite low for the ground measurements (all below 1%). For atmospheric losses, however, the variation coefficients are all higher. A non parametric test conducted by Kruskal-Wallis on the other sprayers reveals statistically distinct groups: T and A1 in the first group, and A2 and A3 in the other. For atmospheric losses, the non parametric test allowed us to disassociate the T result, which was significantly weaker that those obtained with A2 and A3. Ground Air A1 7,7 (0,4) 24,3 (2,5) A2 10,4 (0,4) 37,7 (6,3) A3 10,2(0,4) 37,4(4,2) P1 8,7 (0,7) 23,8 (4,7) P2 9,1 (0,3) 21,2 (7,7) T 7,7 (0,5) 14,5 (3,2) Table 1. Percentage of ground and air losses in relation to spray flow rates measured on artificial vines; the figures in brackets are the standard deviations. 4.2. Impact of calibration and vegetation growth stage There were significant decreases in measured ground losses, as vegetation matured, over the three periods when the set-up of all sprayers was optimized (cf. Tab.2). During the first tests, ground losses were in the region of 40%; at the mature vegetation stage, the losses fell to approximately 10% except for sprayer T for which sprayer orientation could not be optimized. P1 P2 T may 44,5(3,2) 35,5(4,7) 41(2) june 17,4(2) 25,3(3,2) 26,8(2,6) july 10,3(1,5) 13(1,1) 24,1(3,3) Table 2. Percentages of ground losses in relation to spray flow rate as a function of the vegetation growth stage for optimized calibration; the figures in brackets are the standard deviations. Concerning the impact caused by changing the number of open outlets, analyses of the raw data showed no significant difference between when certain outlets are closed and when all outlet are open concerning sprayers P1 and P2. (cf. first columns of Tables 2 and 3). However, deposits measured for sprayer T when all nozzles were open, are significantly lower (by 50%) than those measured using optimized settings (upper nozzles closed). Nevertheless as wind conditions were more severe during these last tests, this result is questionable. Treatments undertaken at 7 km/h instead of at 4 km/h (second column) had no significant effect on ground deposits measured. The differences are not statistically significant, except for sprayer P1 for which a 5% drop in ground deposits was observed. The impact of incorrect spray orientation (third column) is very clear for sprayers P1 and P2, for which a significant rise in ground losses was observed. For sprayer P2, losses are more than tripled. For sprayer T, the difference is not significant. P1 P2 T may Number of open outlets 43,5(2,4) 29,8(4,2) 20(2,8) june speed of tractor advancement 11,4(1,6) 27,9(2,5) 24,9(2,1) july Spray orientation 13,8(2,2) 37,4(5,9) 22,2(1,8) Table 3. Percentages of ground losses in relation to spray flow rate for poor quality settings; the figures in brackets are the standard deviations. 4.3. Pesticide Balance after Spraying Measurements of pesticide quantities after spraying (cf. Tab.4) show that losses to the atmosphere are very significant in all cases (between 25 and 35%), ground losses range between 15 and 25% and quantities traped in the vegetation vary between 20 and 70%. However, it should be noted that experimental protocols implemented in May were not suitable: vegetation samples were taken over only 2 rows and ground deposits were only estimated over 3 rows. Only 73,6% of sprayed product was accounted for. Taking into account, the vegetation growth stage and the results described in the previous paragraph (4.2), we can expect most of the missing product to have been deposited on the ground. The protocol was optimized for June and July tests. In June, 107.5% of the sprayed product was measured and 85.5%, during July. Quantities recovered in the most distant rows were approximately 1/4 of those found ir rows adjacent to treatment. If we consider a standard treatment of one row out of two, the adjacent rows would receive Twice as much product as the intermediary ones. air adjacent rows distant rows ground Mai 37,9(5,3) 19,6(1,6) not measured 16,1(2) Juin 31(3,2) 43,2(2,5) 15,3(6,1) 18(1,9) Juillet 25.5(3.1) 34,2(3,1) 7,5(1) 19,7(1,6) Table 4. Percentages of atmospheric, ground and plant deposits for the three growth stages; the figures in brackets are the standard deviations. 5. Discussion The validity of results obtained relies heavily of the dimensions of the t areas within which the product was collected. The results indicate that for the first tests (artificial vine experiments, balance test conducted in May) there were not enough ground deposit collectors. The definitive protocol implemented enabled satisfactory sampling within the deposit zones. An error assessment was conducted based on the uncertainties noted by the operators. The following uncertainties were recorded before the tests; tractor advancement speed and flow rate, during the test; uncertainty on the concentration of sprayed product and maintenance of a constant advancement speed, after the tests; determining rinsing and dilution volumes, rinsing efficiency, the choice of leaf sampling zone, spectrofluorometry errors. Rinsing efficiency was optimized in the laboratory by determining rinsing volumes and shaking times for each sample, and error is thus supposed to be negligible. The “operator”effect on the choice of leaf sampling zone could however have an important role (a priori, the least dense zones are systematically chosen), but it is difficult to evaluate. Uncertainty linked to concentration was limited as a standard curve was plotted based on the product sprayed. Error linked to spectrofluorometry was assessed by successive measurements of several samples of the same solution, and was deemed to be negligible. To these uncertainties can be added errors linked to: collector surfaces, their trapping capacities, and errors related to sampling. The last error was determined for ground collectors (cf. par.3.1.1). For atmosphere collectors, their trapping capacity was estimated at 80% by [GIL 07], and was integrated into our calculations. Errors linked to sampling were not evaluated however, and could have an important role in uncertainty. For the leaves, there was no sampling within the measurement zone. However error linked to the size of the measurement zone could also be significant (estimated at 15%). Only those errors which were seen to be both significant and quantifiable were taken into account: flow rate, tractor advancement speed, measurement zone dimensions, volume related errors, and for ground collectors, sampling related errors. The errors therefore calculated are: 10% for ground deposit measurements, 5% for atmosphere related measurements and 17,5% for errors related to measurements made in the vegetation. Other aspects are open to discussion on the methodology used, in particular on outdoor conditions. Generally speaking, the tests were performed in moderate tem- peratures ( below 25°C) in light winds and with humidity levels of 40% and above ; However more severe wind conditions could not always be averted, which distorted the results of at least one measurement. Other questions could be raised on the quantities of product caught on the wires used for t atmospheric product loss measurements: it is possible that a certain quantity of deposits drop onto the measurement zone after the cloud of vaporized product has been emitted. Our observations made with the naked eye, led us to believe that this effect is negligible, but, in any case, it would be difficult to quantify. Taking everything into account, the balance measurement results show satisfactory orders of magnitude, as 85% and 107% of sprayed product was accounted for. These results are compatible with the errors estimated in the methodology. The last point which is subject to discussion concerns the representativeness of a solution dye in water as real products are generally suspended solutions and products sprayed have physicochemical properties that can differ significantly from one to the next. 6. Conclusion Over the three years of the study, experiments based on the usage of a fluorescent dye were performed to measure pesticide losses within the environment. The experimental protocols were gradually improved as the study progressed, notably concerning the number and positioning of the ground deposit collectors. Numerous verifications were made in the laboratory and in the field to reduce the number of sources of error, which were finally evaluated at between 5 and 10% for the atmospheric and ground collectors respectively. Balance tests showed that, on the whole, the measurement method could be used to recover the entire product, with most uncertainty linked to the different types of collectors used. In the balance experiments, the capture of product on the vegetation was the greatest source of error. The results revealed the important influence of the type of equipment used and their calibration. During experiments on artificial vines, and based on optimal settings, ground losses were estimated to vary between 7 and 10%, while, depending on the equipment used, atmospheric losses ranged between 20 and 40%. Results obtained on real vines confirmed these orders of magnitude. The most important calibration, with regards to mitigation of pesticide losses to the environment, concerned the orientation of jets towards the vegetation. When deliberately poor settings were used (nozzle orientation set for vines measuring 2m instead of 2m50), we noted that ground losses were tripled. Ground losses could also be significant using standard equipment settings at the early vegetation growth stage: sometimes exceeding 40% of quantities sprayed. The measurements made at different vine-growth stages show that atmospheric losses remain very significant (between 25 and 45%). Overall, the orders of magnitude are approximately 40-20-40,at the early vine-growth stage, and 10-50-40 for mature vines for the ground, plant, atmosphere compartments respectively. Acknowledgements This work was partly funded by the European Life Aware programme and by the ANR Geduque project. The authors would like to extend their gratitude the winegrowers who volunteered to take part in this study and would also like to thank all the members of the team and trainees involved - M. Béranger, D. Gilles, N. Gimenez, J. Lagrevol and O.Liet - without whom these tests could not have been carried out. 7. References [ADE 07] A DE G., RONDELLI V., “Performance of an air-assisted boom sprayer in the control of Colorado beetle infestation in potato crops”, Biosystems engineering, vol. 97, num. 2, 2007, p. 181–187. [BAY 05] BAYAT A., B OZDOGAN N. Y., “An air-assisted spinning disc nozzle and its performance on spray deposition and reduction of drift potential”, Crop Protection, vol. 24, 2005, p. 951–960. [BER 99] VAN DEN B ERG F., K UBIAK R. ., B ENJEY W., “Emission of pesticides into air”, Water , Air, Soil and Pollution, vol. 115, 1999, p. 195-218. [BRI 2a] B RIAND O., B ERTRAND F., S EUX R., M ILLET M., “Comparison of different sampling techniques for the evaluation of pesticide spray drift in apple orchards”, The science of the total environment, vol. 288, 2002a, p. 199-213. [BRI 2b] B RIAND O., M ILLET M., B ERTRAND F., C LÉMENT M., S EUX R., “Assessing the transfer of pesticides to the atmosphere during and after application. Development of a multiresidue method using adsorption on Tenax and thermal desorption GC/MS”, Analytical and Bioanalytical chemistry, vol. 374, 2002b, p. 848–857. [BUI 98] B UI Q., W OMAC A., H OWARD K. D., M ULROONEY J. E., A MIN M. K., “Evaluation of sampler for spray drift”, Transactions of the American Society of Agricultural Engineers, vol. 41, 1998, p. 37–41. [CAI 97] C AI S. S., S TARK J. D., “Evaluation of five fluorescent dyes and triethyl phosphate as atmospheric tracers of agricultural sprays”, Journal of Environmental Science and Health Part B - Pesticides Contaminants and Agricultural Wastes, vol. 32, 1997, p. 969–983. [CAR 06] C ARLSEN S. C. K., S PLIID N. J., S VENSMARK B., “Drift of 10 herbicides after tractor spray application”, Chemosphere, vol. 64, 2006, p. 787–794. [CRO 01] C ROSS J. V., WALKLATE P. J., M URRAY R. A., R ICHARDSON G. M., “Spray deposits and losses in different sized apple trees from an axial fan orchard sprayer: 1. Effects of spray liquid flow rate”, Crop Protection, vol. 20, num. 1, 2001, p. 13-30. [DAS 02] DA S ILVA A., S INFORT C., B ONICELLI B., VOLTZ M., H UBERSON S., “Spray penetration within vine canopies at different vegetative stages”, International advances in pesticide applications; AAB/BCPC Conference, Unversity of Surrey, Guidford, UK, January 5-7, 2002. [DAS 05] DA S ILVA A., S INFORT C., T INET C., P IERRAT D., H UBERSON S., “A Lagrangian model for spray behaviour within vine canopies”, Journal of Aerosol Science, vol. AS3921, 2005, p. 1–17. [DEM 00] D E M OOR A., L ANGENAKENS J., V EREECKE E., “Image analysis of water sensitive paper as a tool for the evaluation of spray distribution of orchard sprayers”, Aspects of Applied Bilogy, vol. 57, 2000. [DOB 83] D OBSON C. M., M INSKI M. J., M ATTHEWS G. A., “Neutron activation analysis using dysprosium as a tracer to measure spray drit”, Crop Protection, vol. 2, 1983, p. 345352. [FAR 04] FAROOQ M., L ANDERS A. J., “Interactive effects of air, liquid and canopies on spray patterns of axial-flow patterns.”, Paper 041001, ASAE/CSAE Annual International Meeting, Fairmont, Chateau Laurier, Ottawa, Ontario, Canada, 1-4 August, 2004, Paper 041001. [GIL 05] G IL Y., S INFORT C., “Emission of pesticides to the air during sprayer application: a bibliographic review”, Atmospheric Environment, vol. 39, 2005, p. 2183-5193. [GIL 07] G IL Y., S INFORT C., B RUNET Y., P OLVECHE V., B ONICELLI B., “Atmospheric loss of pesticides above an artificial vineyard during air-assisted spraying”, Atmospheric Environment, vol. 41, 2007, p. 2945–2957. [HOF 89] H OFF R. M., M ICKLE R. E., F ROUDE F. A., “A rapid acquisition Lidar system for aerial spray diagnostics”, Transactions of the American Society of Agricultural Engineers, vol. 32, 1989, p. 1523-1528. [KIR 04] K IRK I. W., F RITZ B. K., H OFFMANN W. C., “Aerial Methods for increasing spray deposits on wheat heads”, ASAE/CSAE Annual International Meeting, Fairmont, Chateau Laurier, Ottawa, Ontario, Canada, 1-4 August 2004, 2004, Paper 041029. [KRA 04] K RAUSE C. R., Z HU H., F OX R. D., B RAZEE R. D., D EKSEN R. C., H ORST L. E., Z ONDAG R. H., “Detection and quantification of nursery spray penetration and offtarget loss with electron beam and conductivity analysis”, Transactions of the American Society of Agricultural Engineers, vol. 47, num. 2, 2004, p. 375–384. [LAN 04] L ANDERS A., “Prevention is better than cure - Reducing Drit from Vineyard Sprayers”, Invited Presentation Article - International Conference on Pesticide Application for Drift Management. Waikoloa, Hawaï, 27-29 Octobre, 2004. [L’H 09] L’H ERMITE N., G OUZY A., L E G ALL A.-C., FARRET R., B EDOS C., S INFORT C., B ONICELLI B., M ARLIERE F., “Identification and classification of pesticides of concern for human health: Sph’Air, a decision support system for pesticide air monitoring”, Atmospheric Ennvironment, vol. soumis, 2009. [MIL 89] M ILLER P. C. H., H ADFIELD D. J., “A simulation model of the spray drift from hydraulic nozzles”, Journal of Agricultural Engineering Research, vol. 42, num. 2, 1989, p. 135-147. [MIL 00] M ILLER D. L., S TOUGHTON T. E., S TEINKE W. E., H UDDLESTON E. W., B OSS J. B., “Atmospheric stability effects on pesticide drift from an irrigated orchard”, Transactions of the American Society of Agricultural Engineers, vol. 43, 2000, p. 1057–1066. [PAN 04] PANNETON B., L ACASSE B., T HERIAULT R., “Effect of the characteristics of air jets on the performance of a spray recovery sprayer for vineyards”, ASAE/CSAE Annual International Meeting, Fairmont, Chateau Laurier, Ottawa, Ontario, Canada, 1-4 August, 2004. [PER 01] P ERGHER G., “Recovery rate of tracer dyes used for spray deposit assesment”, Transactions of the American Society of Agricultural Engineers, vol. 44, 2001, p. 787– 794. [PIC 95] P ICHON V., G UENU S., D UPAS S., H ENNION M.-C., “Analyse multirésidus des pesticides dans les eaux par extraction liquide-solide et chromatographie en phase liquide : Analyse de l’eau”, Spectra 2000 analyse, vol. 24, num. 185, 1995, p. 25-31. [RAV 05] R AVIER I., H AOUISEE E., C LÉMENT M., S EUX R., B RIAND O., “Field experiments for the evaluation of pesticide spray-drift on arable crops”, Pest management science, vol. 61, 2005, p. 728-736. [SDT 97] SDTF, “A summary of airblast application studies”, report , 1997, Spray Drift Task Force, available on http://www.agdrift.com. [SOL 96] S OLANELLES F., F ILLAT A., P IFARRÉ C., P LANAS S., “A method of drift measurement for spray applications in tree crops.”, Ageng’96, Madrid, Spain, Paper 96A-133, 1996. [STO 97] S TOUGHTON T. E., M ILLER D. R., YANG X., D UCHARME K. M., “A comparison of spray drift predictions to lidar data”, Agricultural and Forest Meteorology, vol. 88, num. 1-4, 1997, p. 15-26. [WIT 02] W ITTICH K. P., S IEBERS J., “Aerial short-range dispersion of volatilised pesticides from an area source”, International Journal of Biometeorology, vol. 46, 2002, p. 126–135. [YAT 76] YATES W. E., A KESSON N. B., BAYER D., “Effects of spray adjuvant on drift hazards”, Transactions of the American Society of Agricultural Engineers, vol. 19, 1976, p. 41–46. [ZAL 80] Z ALAY A. D., B OUSE L. F., C ARLTON J. B., C ROOKSHANK H. R., E BERLE W. R., H OWLE R. E., S HRIDER K. R., “Measurement of airborne spray with a laser Doppler velocimeter”, Transactions of the American Society of Agricultural Engineers, vol. 23, 1980, p. 548-552.
© Copyright 2025 Paperzz