TY - JOUR
AU1 - Tubby, Katherine
AU2 - Forster, Jack
AB - Abstract British forestry is threatened by numerous pests and diseases. This study investigated the potential for re-introduction of aerial pesticide applications for landscape-scale disease management. In North Scotland in 2013 and 2015, copper oxychloride was applied to Pinus sylvestris L. stands infected with Dothistroma septosporum (Dorogin) Morelet. Helicopters distributed ultra-low-volume (ULV) applications of product via Micronair rotary atomisers, following methods used against D. septosporum in P. radiata D. Don stands in New Zealand. Product deposition was quantified on paper catchers and in foliage, soil and water. Catchers 100 m beyond the plot boundaries intercepted 0.5 per cent of within-plot mean deposition. Foliar analysis revealed slightly elevated copper concentrations (+0.07 μg g−1 dw) 250 m outside plot boundaries. Copper in foliage and needle litter remained above background levels for 109 and 157 weeks after application, respectively, longer than recorded during New Zealand operations. Concentrations in the soil increased over 3 years’ monitoring, whilst deposition into water traps resulted in copper concentrations well within limits set by the Scottish Environmental Protection Agency. No deleterious impacts on vascular and non-vascular ground and canopy flora were recorded. Copper fungicide applications significantly reduced foliar infection at both sites but did not affect needle retention. Further ground-based trials will investigate the efficacy of other actives. In Britain, such aerial operations have not occurred for two decades: this study demonstrated aerial and ground teams have the necessary expertise for their re-introduction, whilst highlighting areas needing further optimization. Introduction In recent years, certain pathogens have had a substantial impact on trees in Britain’s forests, woodlands and trees. In 2009, Phytophthora ramorum Werres, De Cock and Man in’t Veld was detected on Japanese larch (Larix kaempferi (Lamb.) Carr.), a significant component of forestry in Britain (Webber et al., 2010; Harris, 2014). This pathogen has since caused widespread mortality which, together with sanitation felling, is contributing to a predicted 50 per cent decline in volume of standing larch by the 2030s (Forestry Commission, 2014). In eastern England in 2012, Hymenoscyphus fraxineus (T. Kowalski) Baral, Queloz and Hosoya was found on ash (Fraxinus excelsior L.), a common component of British woodland ecosystems (Brasier and Webber, 2013; Clark and Webber, 2017). Research suggests fewer than 10 per cent of trees may exhibit some tolerance to the pathogen (McKinney et al., 2014), and with infection now widespread, and an estimated 126 million ash trees in Britain’s woodlands (Forestry Commission, 2013), this disease could have a landscape-scale impact (Clark and Webber, 2017). Dothistroma needle blight (DNB) caused by Dothistroma septosporum (Dorogin) Morelet affects pine species (Pinus spp.) and has had a major impact on managed pine forests since the late 1990s (Mullett, 2014). Infection causes premature defoliation, and severe crown infection results in decreased increment and, in some cases, tree mortality (Whyte, 1969; Van der Pas, 1981; Old and Dudzinski, 1999). The disease can also affect the aesthetic value of specimen trees and Christmas tree plantations and has been detected in the iconic native Caledonian pine forests in Scotland (Mason, 2000). Options for actively managing these and other tree pests and pathogens in Britain are currently limited. However, risks to trees from pests and pathogens remain high due to the extensive global trade in planting material but inadequate phytosanitary legislation (Brasier, 2008; Santini et al., 2013). Combined with the potential for climate change to affect the behaviour and impact of both native and non-native organisms (Desprez-Loustau et al., 2006, 2007; Tubby and Webber, 2010), a wider toolkit of management options, applicable to a range of disorders, is desperately needed. Current management strategies in Britain for a pathogen such as D. septosporum include (1) avoidance—the most susceptible species, Corsican pine (Pinus nigra subsp. laricio (Poir.)) and “Inland” and “Coastal” origins of lodgepole pine (Pinus contorta var. latifolia (Engelm.)), are not currently planted on the Public Forest Estate; (2) prevention—all forestry pine stock undergoes mandatory inspection for DNB and certification of disease-free status before sale and 3) intervention—early and regular thinning is used to increase airflow and decrease infection levels in affected forests, and heavily infected forest stands are felled early where possible to reduce inoculum levels (Bulman et al., 2016). More generally, a move towards uneven aged and mixed species forests (Ostfeld and Keesing, 2012; O’Hara and Ramage, 2013; Wilson, 2016; Pap et al., 2017), provenance/species change, and exploitation and enhancement of naturally occurring tolerance (Bradshaw, 2004; Clark and Webber, 2017) are being used in forestry both to increase resilience against pests and pathogens and adapt to climate change (Savill, 2017; Silvifuture, 2017). Currently, fungicides are not used on the British Public Forest Estate, but are used widely in tree nurseries to manage pathogens including Dothistroma species. Copper-based fungicides are commonly used in agriculture against, for example, potato/tomato blight (P. infestans (Montagne) de Bary), leaf spot on celery (Septoria apiicola (Spegazzini)) and bacterial canker of cherries and plums (Pseudomonas syringae pv. morsprunorum (Wormald) Young, Dye and Wilkie and P. s. pv. syringae van Hall). Copper has been used in other countries against DNB for c. 70 years (Bulman et al., 2016). As copper dissolves in water, the resulting copper ions cover the needle surface and kill Dothistroma spores released during periods of rain, protecting the foliage from new infection. Copper also stops the fruiting bodies producing and releasing spores (Franich, 1988; Bulman et al., 2013). Occasional aerial applications of copper are made in Serbia against Dothistroma spp. (Pap et al., 2017) whilst, in New Zealand, c. 40 tonnes of copper fungicides per annum are aerially applied to manage DNB on P. radiata D. Don, the main commercially grown plantation species (Lindsay Bulman, personal communication, Scion NZ, 21 June, 2017). Application is triggered when average stand infection reaches 15 per cent (Bulman et al., 2013), before DNB begins significantly to affect growth (Whyte, 1969; Van der Pas, 1981). Aerial operations can be the most effective way of accessing large areas of forest and are currently used for forest protection in Europe, the USA, Canada and New Zealand. Helicopters are particularly useful where terrain is difficult and/or stands are isolated, as they can land and refuel without a long runway. Products including the now banned arsenicals and DDT, and more recently Bacillus thuringiensis (Bt)-based biological pesticides have been used against insect defoliators including pine looper (Bupalus piniaria L.; Schimitschek, 1941 in Slovakia; Butovitsch, 1947 in Sweden), spruce budworm (Choristoneura fumiferana Clem.; Balch et al., 1954; van Frankenhuyzen, 1990; van Frankenhuyzen, 2000; Fournier et al., 2010 in Canada), pine processionary moth (Thaumetopoea pityocampa (Den. and Schiff.); Cebeci et al., 2010 in Turkey) and common forest looper (Pseudocoremia suavis Butler; Rawlings, 1955 in New Zealand). In Britain, aerial applications have been used to protect agricultural crops from pests and to control weeds, but usage has declined in recent years (Sly and Neale, 1983). Aerial forestry operations have been limited in scale, partly due to the smaller relative size of the forestry sector in Britain, but include application of granulated fertilisers (Howe and Little, 1969) and herbicides, particularly for bracken (Pteridium aquilinum L. Kuhn) control (Southgate, 1969; Brown and Robinson, 1997; Brown, 2015). Aerially applied pesticides were used to control pine looper moth in the 1950s–1970s (Crooke, 1959; Scott and Brown, 1973) and pine beauty moth (Panolis flammea Denis and Schiffermüller) in the 1980s and early 1990s, (Evans et al., 1991; Heritage, 1997). A small-scale application of Bt was used against oak processionary moth Thaumetopoea processionea (L.) in Berkshire in 2013 and 2014 (Nigel Straw, Forest Research, personal communication 12 May 2018) but overall, as in agriculture, forestry aerial operations have declined in Britain. This is partly due to increasing environmental constraints and certification standards for sustainable woodland management such as the United Kingdom Woodland Assurance Scheme (ukwas.org.uk) and the Forest Stewardship Council (www.fsc-uk.org) encourage a move away from use of chemical products. However, in the light of recent outbreaks of, for example, DNB and ramorum blight there is renewed interest in aerial operations. In 2013 and 2015, DNB was used as a model to re-investigate the operational feasibility of including aerial applications in a forest industry toolkit for use against landscape-scale pest and disease outbreaks (Willoughby et al., 2015). Pesticide usage in forestry and agriculture operates under considerable legislative and environmental restrictions and due to license requirements restricting applications to <20 ha year−1, these trials were run as empirical studies to investigate operational issues and environmental constraints. The specific objectives of this study were to: record operational experiences to inform future operations test parameters under local (British) environmental conditions, inform the extent of appropriate buffer zones determine the persistence of the product within the environment investigate impacts on key elements of local flora and fauna investigate efficacy of a copper-based fungicide against DNB. Methods Legislative requirements and liaison Health and Safety Executive (HSE) approval was obtained prior to operations (including a Permit for Trial Purposes of Plant Protection Products under PPP regulation (EC) No 1107/2009, and an Aerial Application Plan). This restricted applications to a 3-year period, and an area of forest less than 20 ha per annum. Sites had to be situated at least 100 m away from water courses, residential areas and areas of high recreational use. Prior to the operation a liaison exercise was carried out with interested parties including the Scottish Government and its Agencies (Scottish Natural Heritage, the Scottish Environment Protection Agency), the Scottish Tree Health Advisory Group, local Beekeeper Associations, local elected representatives and community groups to inform and obtain feedback. Press releases and social media channels were also used to explain the impacts of the disease and reasons for the operations. Figure 1 Open in new tabDownload slide (a) Location and layout of sprayed (left) and unsprayed (right) plots in Monaughty Forest, 2013. Sprayed plot contains transects S1–S5 and unsprayed plot contains transects US1-US5, each comprising 30 trees. 100 m cardinal transects extend N, S, E, W (downwind) beyond the sprayed plot boundaries. (b) Sprayed plot in Millbuie Forest, treated in 2015, containing transects S1–S6, each comprising 30 trees. Two additional transects, OPEN 1 and OPEN 2, are situated within an open, treeless area. 100 m cardinal transects extend N, S, E (downwind) and W beyond the plot boundaries. (c) Unsprayed plot in Millbuie Forest, established in 2015 1.5 km distant from sprayed plot, containing transects US1–US6, each comprising 30 trees. (Not to scale. See text for details of sampling). Figure 1 Open in new tabDownload slide (a) Location and layout of sprayed (left) and unsprayed (right) plots in Monaughty Forest, 2013. Sprayed plot contains transects S1–S5 and unsprayed plot contains transects US1-US5, each comprising 30 trees. 100 m cardinal transects extend N, S, E, W (downwind) beyond the sprayed plot boundaries. (b) Sprayed plot in Millbuie Forest, treated in 2015, containing transects S1–S6, each comprising 30 trees. Two additional transects, OPEN 1 and OPEN 2, are situated within an open, treeless area. 100 m cardinal transects extend N, S, E (downwind) and W beyond the plot boundaries. (c) Unsprayed plot in Millbuie Forest, established in 2015 1.5 km distant from sprayed plot, containing transects US1–US6, each comprising 30 trees. (Not to scale. See text for details of sampling). Method of application and site details An AS 350 B2 Single Squirrel Helicopter fitted with “Micronair” rotary atomisers (Micronair, 2020) was used to apply copper oxychloride (Cuprokylt™ supplied as a wettable powder, 50 per cent w/w copper as copper oxychloride, Certis Europe). The working solution, applied at a rate of 5 l ha−1, contained 1.66-kg copper oxychloride (equating to 0.83 kg of elemental Cu), 2 l of oil adjuvant (TOIL™, 95 per cent w/w rapeseed fatty acid esters EAC1, Interagro) to reduce evaporation of the small droplets (Bulman et al., 2004), and 0.05-l marker dye (Blue Marker Dye™, containing 250 g of disodium erioglaucine l−1, De Sangosse), and was made up to 5 l with water. The ultra-low-volume (ULV) application delivered product with a volume median diameter (V.D.M) of 65 μm, previous studies having demonstrated droplets of this size are optimal for coniferous canopy capture through direct interception by gravitational settling and inertial impaction (Picot and Kristmanson, 2012). Fungicide applications were made to dry foliage, with no forecast rain within 2 h of application, and a maximum wind speed of 10 knots. At Monaughty Forest near Forres (X (Easting) = 312 450, Y (Northing) = 857 600; Figure 1a) 5 ha of 18-year-old Scots pine (Pinus sylvestris L.) were treated on the morning of the 25 July 2013. At the second site, Millbuie Forest on the Black Isle, Scotland (X = 270 800, Y = 861 670 and X = 269 325, Y = 862 098; Figure 1b,c) 14 ha of 22-year-old Scots pine were sprayed on the afternoon of the 30 June 2015. Each sprayed plot was paired with a nearby Scots pine stand of similar age and condition, acting as an untreated control. The experimental plots at Monaughty consisted of adjacent 5-ha plots, the sprayed plot to the west of the compartment, the unsprayed control plot to the east. A 100-m buffer was established around the sprayed plot, twice the distance required by the experimental licence. This plot was situated over 100 m away from any water courses. Five transects were established in the sprayed (transects “S1” to “S5”) and unsprayed plot (transects “US1” to “US5”), each transect comprising three adjacent rows of 30 trees. Cardinal transects extended 100 m beyond the sprayed plot boundaries to the north, east, south and west, beginning at 0 m at the mid-point along each edge of the plot boundary (Figure 1a). Six within-plot transects were established in the sprayed (transects “S1” to “S6”) and unsprayed “US1” to “US6”) plots at Millbuie (Figure 1b,c), again including three adjacent rows of 30 trees. An additional two transects 140 and 160 m in length, “Open 1” and “Open 2,” respectively, extended across an open, unforested area within the sprayed plot. Cardinal transects were situated to the north, east, south and west of the sprayed plot, beginning at 0 m at the mid-point along each edge of the plot boundary (0 m), extending another 100 m. Underlying soils were largely brown earths and podzols (Kennedy, 2002) at Monaughty, with intrusions of ironpan towards the western end of the unsprayed plot. At Millbuie, there was an ironpan and podzolic ironpan, with some surface water gley in parts of the unsprayed plot. Applications were timed to coat all foliage, including the recently fully extended current year’s growth with product. A planned 2014 trial was postponed due to unsuitable weather conditions during this application period. Assessment of product distribution Horizontally oriented “A5” (148 × 210 mm) paper catchers were used to collect spray deposition within the sprayed plots and to record off-target drift. They were fixed using metal clips to the distal ends of branches at the base of the live foliage, the middle crown and upper crown of trees 1, 15 and 30 on transects S1, S3 and S5 at Monaughty and trees 1, 15 and 30 along transects S1, S3 and S5, and tree 15 only on transect S6 at Millbuie. Catchers were also laid flat on the ground 0, 20, 40, 60, 80, 10, 120, 140 m along transect “Open1” and 0, 20, 40, 60, 80, 10, 120, 140 and 160 m along transect “Open 2” at Millbuie. Catchers were also situated in the base of the live foliage, the middle and the upper crown of trees 0, 5, 10, 20, 50 and 100 m along the cardinal transects outside the sprayed plot boundaries at both sites. All catchers were collected after the application, digitally scanned (Epson Perfection 1660 Photoscanner) and numbers and area of surface occupied by blue droplets quantified using Olympus CellSens Dimension v.1.7.1 software. An error in mixing led to no dye being applied during the Millbuie application. As spray droplets could not therefore be visualized, approximately 50 g of live foliage was cut from each branch, which held a catcher, but sampling was restricted to “upper crown” positions only due to high analysis costs. Examination of helicopter flight path data indicated potential overspray beyond the western plot boundary at Millbuie, so sampling along the western cardinal transect was extended to 250 m. To enable deposition due to drift caused by the prevailing west to east wind, sampling was extended to 150 m along the eastern transect. Concentrations of copper on the foliage were quantified using methods outlined below, and these data used to compare deposition within and outside the sprayed plot and determine the extent of drift. Wind speed at both sites remained within the tolerances appropriate for the equipment and below 10 knots as stipulated by the Civil Aviation Authority (CAA, 2006). To determine product location after application and monitor changes over time, samples of foliage, ground needle litter and soil were collected pre- and post-application. Foliage from the sprayed and unsprayed plots at Monaughty was sampled from transects S2, S4 and US2 and US4, initially selecting the 1st, 15th and 30th tree along each transect, and in subsequent weeks/months selecting adjacent trees along the same transects. Approximately 50 g samples of live foliage were collected from all foliated years of growth from single branches in the base of the live crown, the middle and top of the crown and from the leader, keeping samples from each of these four crown positions separate. Foliar sampling was carried out 1 week prior to application and then 1, 2, 3, 5, 9, 13 and 26 weeks after the application. Foliar samples were also collected from the sprayed and unsprayed plots at Millbuie Forest, from transects S2A, S3A S5A and US2A US3A and US5A using the same sampling protocol as above. Foliage was initially collected from trees 1, 15 and 30, and in subsequent sampling times, from adjacent trees as at Monaughty. Sampling was carried out 1 week before application and then at 1, 2, 4, 6, 10, 15, 27, 40, 53, 109 weeks post-application. Needle litter was also sampled from transects S2 and S4 in the sprayed and unsprayed plots at Monaughty and transects S2A, S3A S5A and US2A US3A and US5A at Millbuie. Approximately, 50 g of needle litter was collected from near the base of trees 1, 15 and 30 along each of the transects, and subsequently from adjacent trees. 1–1.5 kg of soil was collected from immediately below each litter sample, to a depth of 10 cm. Litter and soil sampling was carried pre-application, and 1, 5, 13, 26, 56, 105 and 157 weeks post-application at Monaughty and 1, 4, 7, 15, 27, 52, 106 and 155 weeks at Millbuie. All foliar, litter and soil samples were stored at 4°C prior to copper quantification analysis in the Environmental Research Laboratory at Forest Research, Alice Holt Lodge, Farnham Surrey. Foliar and litter samples were washed with 5 per cent nitric acid, shaken for an hour and filtered with a 0.45 μm nylon syringe filter. Total soil copper was extracted using microwave-assisted acid digestion of sediments, sludge, soils and oils according to a United States Environmental Protection Agency Standard Operating Procedure 3051A (U. S. EPA, 2007). Biologically “available” copper, i.e. that not bound to substrates within the soil, was extracted using a standard ethylene diamine tetra acetic acid-based method used by the former Ministry for Agriculture, Fisheries and Food (MAFF, 1986). Extracted copper was subsequently quantified using U.S. EPA Method 200.7 for the determination of metals and trace metals in water by Inductively Coupled Plasma- Atomic Emission Spectrometry (U.S. EPA, 1994) using an ICP-OES dual view (Thermo ICap 6500). Both sites were situated at least 600 m away from the nearest water course, but water traps were used to record deposition in accidental over-spray scenarios in the sprayed plots at both sites. 1 × 1 × 0.15-m plastic trays containing five litres of commercially sourced bottled water were placed along transects in the sprayed plots at both sites. 50 ml samples of the water were collected pre- and immediately post-application for analysis of copper ion concentration: at Monaughty, the trays were situated adjacent to trees 1, 15 and 30 on transects S2 and S4, and trees 8, 13, 25, 55 and 95 m along the cardinal transects out-with the sprayed plot (Figure 1a); at Millbuie trays were located adjacent to trees 7 and 22 on transects S1, S3 and S5, and by tree 15 on transect S6 (Figure 1b). Trays were also situated at 0, 20, 40, 60, 80, 100, 120 and 140 m on transects Open 1 and Open 2 and adjacent to trees 8, 13, 25, 55 and 95 m along the cardinal transects out-with the sprayed plot. Copper was quantified in all water samples using the ICP-OES method outlined above. Assessment of biological indicators A range of biological indicators were monitored to assess impacts on the local ecosystem. Vascular and non-vascular ground dwelling plants, representing a diversity of growth forms and habitat preferences, were visually assessed in fixed 1-m2 metal-framed quadrats. These were located on the ground adjacent to trees 1, 15 and 30 on transects S1, S3, S5, US1, US3 and US5 in the sprayed and unsprayed plots at Monaughty (Figure 1a) and trees 1, 15 and 30 along transects S1, S3, S4, S6, US1, US3, US4 and US6 at Millbuie (Figure 1b,c). Percent coverage of vascular and non-vascular vegetation was quantified pre-application and at 1, 12, 24 and 52 weeks post-application at Monaughty and Millbuie, and additionally after 105 weeks at Millbuie. The proportion of main tree stems colonized by non-vascular plants was quantified using fixed, vertical 0.2 × 0.2 m metal quadrats fixed to the north and south aspect of trees at 1.3 m above ground level, and non-vascular plant coverage also assessed using fixed 0.1 × 0.1 m quadrats encircling branches in the top, middle and bottom of the live crowns of trees. The quadrats were situated within trees 1, 15 and 30 along the quadrats detailed above for ground vegetation assessments at both sites, and were assessed at the same time as the ground vegetation assessments. The numbers and species of ectomycorrhizal fungi (ECM) as fungal fruit bodies were counted and identified in the Autumn following the Summer applications and comparisons of numbers and species diversity made between the sprayed and sprayed plots to provide a “broad-brush” illustration of possible fungicide impact. The survey was carried out on 8 September 2013 along five 80 × 4 m transects in each plot at Monaughty. These fungal assessment transects followed the existing marked transects in each plot (Figure 1a), but extended approximately twice their length. Fruit bodies were quantified and identified slightly later in the year (6 November 2015) due to logistical constraints at Millbuie. Assessments were carried out on six 100 × 4 m transects following the lines of the existing marked transects in each plot (Figure 1b,c), but again, extending approximately twice their length. Below ground, ECM fungi associated with roots were also examined in 10 × 10 × 10-cm soil samples taken from just below the needle litter layer. At Monaughty samples were collected from four ordinal points 1 m from the base of trees 1, 15 and 30 along transects S1C, S3C and S5C and US1C, US3C and US5C (Figure 1a). Each set of four ordinal samples was pooled prior to analysis. At Millbuie, samples were collected, using the same protocol, from points adjacent to trees 1, 15 and 30 along S1C, S3C, S4C and S6C and US1C, US3C, US4C and US6C. At both sites, sampling was carried out pre-spray and 12, 40 and 52 weeks after application. All samples were frozen at −20°C, then defrosted, the fine roots (<2 mm) extracted, and root fragments randomly sampled from the soil samples. Subsamples of 100 ECM tips associated with the fragments were identified as far as possible by morphotyping, which uses the morphological features of the mycorrhizal tips to identify the fungi (Agerer, 1986–2002), to obtain taxon richness and an estimate of the relative abundance of each taxon. A search for molluscs (slugs and snails) was undertaken within each of the 1-m2 fixed vegetation quadrats detailed above, pre-application and 1, 12, 24 and 52 weeks post-application at Monaughty and Millbuie, and also after 105 weeks at Millbuie. A search was also made for earthworms within the soil samples prior to their processing for total and biologically available copper quantification (see above). A bee activity survey was conducted in the sprayed plots, noting numbers and species of bees seen before the application (noting time and weather conditions), and 2 weeks post-application, at the same time of day, under similar weather conditions. Assessment of efficacy Crown infection and defoliation were visually assessed in 5 per cent increments, and overall number of years of live needles retained assessed in all 30 trees along all transects in both sprayed and unsprayed plots at both sites to assess tree condition and treatment efficacy. Assessments were carried out 1 week before each application of fungicide and 2 years post-application in July 2015 (Monaughty) and July 2017 (Millbuie). Statistical analysis Product drift, changes in concentrations of copper in foliage, needle litter and soil, and efficacy of treatment in reducing defoliation and infection were analyzed using mixed effects models (nlme4 package, Pinheiro et al., 2018) in R (R Core Team, 2017), with a natural log+1 transformation to normalize the data. Drift analysis through analysis of droplet deposition used total droplet area on the catcher as the response, the interaction of canopy position and location (in plot/outside plot boundary) as fixed effects and tree within transect as random effects. For copper deposition data an initial analysis was carried out to detect any pre-existing differences in copper concentrations between treatments (sprayed and unsprayed) in each site before the application took place, using the interaction of position (within the canopy), treatment and site as fixed effects and transect nested within site as a random effect. For the full datasets, the same model structure was used initially, with the addition of “weeks after application” (continuous variable) as a fixed effect. Transect nested within site was again used as a random effect to account for the repeated measures design, allowing this as an intercept, slope or both. Best fit models were determined by Akaike’s Information Criteria (AIC). As the data illustrate a set of time series, a range of corARMA (autoregressive moving average) models were used to consider temporal autocorrelation and the best fit models again determined by AIC. ANOVA was used to determine significant effects and interactions, with P < 0.05 considered significant, and non-significant effects were removed from models before confirming the best fit versus the initial models using AIC. Models were subsequently applied to the raw data, responses predicted at the population level. Post hoc marginal means and pairwise contrasts were calculated to determine significant differences (P < 0.05) across treatments, with Tukey’s Honest Significant Difference used to correct for multiple comparisons. Fungicide efficacy was modelled using transect within plot (sprayed/unsprayed) within site (Millbuie/Monaughty) as a random effect, and interaction of site and plot as fixed effects. Pre-spray percent infection data were included as a predictor within the model, and percent foliation then converted to a binary response (absence/presence of infection, absence/presence of >5 per cent infection, absence/presence of >10 per cent infection). “Impact factor,” which is calculated from percentage infection and defoliation data showed strong correlation between pre-spray and year 2 values (Kendall’s Tau = 0.45, P < 0.001). Therefore, the change in impact factor was calculated between pre-spray and year 2 per tree, and this percentage change used as the response variable in a linear mixed effects model. Needle retention data were not formally analyzed due to a lack of variance in the response but are described in the following text. Analysis of vegetation cover was restricted to plant species where the median percentage cover across all sampling points was at least 1 per cent. Only ground quadrat data were formally analyzed as all other data were too patchy for unbiased analysis. A negative binomial GLiMM was applied to the data (DHARMa package, Hartig, 2019) with tree nested within plot, nested within site included as random effect and interaction of plant species, weeks after application, site and plot included as fixed effects. No formal analysis of ECM fungi was conducted due to the strongly over-dispersed nature of the data. However, the observed fungal diversity across the sites is discussed in the following text. Results Assessment of product distribution The A5 size paper catchers gave visual confirmation of deposition onto the target area at Monaughty. Within the plot boundaries, deposition varied between transects and 21.5 per cent (98 mm2), 32.7 per cent (149 mm2) and 45.7 per cent (208 mm2) of the total area of intercepted droplets were captured in trees on transects 1, 3 and 5, respectively. Overall, deposition was significantly greater within than outside the plot boundaries (F(1,30.8) = 15.97, P < 0.01), with deposition 10, 50 and 100 m outside the plot boundaries only 10.4 (15.9 mm2), 1.3 (2.0 mm2) and 0.5 per cent (0.7 mm2) the mean within-plot deposition (151.6 mm2). Figure 2a illustrates differences in product drift along the four cardinal transects outwith the plot boundaries. Some overspray was apparent along the south, and to a lesser extent north transects at Monaughty, but the greatest amount of off-target deposition occurred along the western (downwind) transect. Figure 2 Open in new tabDownload slide (a) Extent of off-target drift visualized using paper spray catchers situated in trees on cardinal outside-plot transects at Monaughty in 2013 (Data show mean droplet area on catchers located in top, mid and bottom of tree crown. Prevailing wind direction during applications was east to west). (b) Extent of off-target drift detected through foliar analysis of trees along cardinal transects outside plots at Millbuie in 2015 (prevailing wind during applications was south-west to north east. Data were adjusted for pre-application copper concentrations of 1.1 μg g−1 dw). Figure 2 Open in new tabDownload slide (a) Extent of off-target drift visualized using paper spray catchers situated in trees on cardinal outside-plot transects at Monaughty in 2013 (Data show mean droplet area on catchers located in top, mid and bottom of tree crown. Prevailing wind direction during applications was east to west). (b) Extent of off-target drift detected through foliar analysis of trees along cardinal transects outside plots at Millbuie in 2015 (prevailing wind during applications was south-west to north east. Data were adjusted for pre-application copper concentrations of 1.1 μg g−1 dw). Analysis of foliar samples taken pre-application confirmed that copper concentrations were similar in the plots allocated to spray/no-spray (F(1,3.8) = 0.02; P = 0.90 NS), and at the two sites (F(1,3.6) = 0.09; P = 0.78 NS). Concentrations varied slightly but significantly throughout the tree crown (F(3,10.1) = 5.19; P = 0.02). After application, in the absence of catcher data, the foliage deposition analysis from the Millbuie site demonstrated spray drift had occurred predominantly down-wind along the eastern transect (Figure 2b) where sampling was extended to 150 m beyond the plot boundary. Concentrations slightly above (+1.0 μg g−1 dw copper) background levels were detected at this furthest sampling point at 150 m. Along the western transect, where some overspray had been detected from the flight path data, copper was 0.07 μg g−1 above pre-application concentrations at 250 m. Within the plot boundaries, spray treatment, position within the crown, time after application and site, were all significant in explaining copper concentration variation in foliage. CorARMA (1,0) structure adequately accounted for temporal autocorrelation. There were significant three-way interactions between spray treatment, time after application and canopy position (F(3,253) = 6.5; P < 0.001; Figure 3), and spray treatment, time after application and site (F(2,253) = 6.2; P = 0.002), and significant two-way interactions between all other factors and variables. Immediately after application the copper concentration increased from 0.9 to 46.0 and from 0.7 to 13.4 μg g1 dw at Monaughty and Millbuie, respectively, (mean values calculated from samples taken from all canopy points). Figure 3 Open in new tabDownload slide Changes in foliar copper concentrations in tree crowns over time at both sites. (Monitoring at Monaughty ended at 27 weeks. Shaded area shows 95% confidence intervals). Figure 3 Open in new tabDownload slide Changes in foliar copper concentrations in tree crowns over time at both sites. (Monitoring at Monaughty ended at 27 weeks. Shaded area shows 95% confidence intervals). Post hoc tests indicated copper concentrations in the sprayed plot at Monaughty were still significantly higher (P < 0.05) in foliage at all canopy sampling heights after 27 weeks. At this time concentrations were still 10x higher than before the application at both sites, resulting in a decision to extend the sampling period at Millbuie; copper concentrations remained significantly higher throughout the crowns of sprayed compared with unsprayed trees 66 weeks after application (P < 0.05; Figure 3). After 80 weeks, concentrations remained elevated in all treated foliage except the leader tissue (P < 0.05). By 85 weeks significant differences only existed in the mid- to base of the crown (P < 0.05) and after 109 weeks only the base of the crowns of sprayed trees had significantly elevated copper concentrations compared with unsprayed trees (P < 0.05; Figure 3). Before application, copper concentrations in the needle litter layer were similar in spray or no-spray treatments plots (F(1,6) = 1.57; P = 0.26 NS) and at both sites (F(1,6) = 0.05; P = 0.84 NS) with concentrations of 2.4 μg g−1 dw at Monaughty and 2.3 μg g−1 dw at Millbuie. A CorARMA (1,0) structure adequately accounted for temporal autocorrelation in the remaining dataset. Fungicide application elevated copper concentrations in the treated plots (F(1,7) = 162.9; P < 0.00), and higher levels of copper were found in litter collected at Monaughty than Millbuie (F(1,7) = 23.8; P = 0.001). Following that immediate rise in needle litter concentrations just after application, there were no further changes in copper within the litter over the sampling time (F(1,188) = 0.9; P = 0.34; Figure 4). Post hoc tests showed that at 157-week copper remained significantly higher in treated than in untreated plots (P < 0.05), and further sampling is recommended to monitor any changes over a longer time frame. Figure 4 Open in new tabDownload slide Changes in copper within the litter layer of both sites after application over time (shaded area displays 95% confidence intervals). Figure 4 Open in new tabDownload slide Changes in copper within the litter layer of both sites after application over time (shaded area displays 95% confidence intervals). The two sites differed in soil type and although total soil copper concentrations were similar before application (mean 2.2 and 1.4 μg g−1 dw at Monaughty and Millbuie, respectively; F(1,6) = 2.61; P = 0.16), biologically available copper, i.e. that not chemically bound to organic or inorganic colloids within the soil, was significantly higher at Monaughty (mean = 0.8 μg g−1 dw) compared with Millbuie (0.4 μg g−1 dw; (F(1,6) = 16.93; P = 0.006). Total copper increased in the treated plot soils after application (F(1,10) = 6.4; P = 0.030), but amounts of biologically available copper were not influenced by fungicide treatment (F(1,7) = 2.0; P = 0.20). After application total (F(1,10) = 63.4; P < 0.001), and available (F(1,9) = 58.4; P < 0.001) copper concentrations were higher at Monaughty than Millbuie and both total (F(1,172) = 15.8; P < 0.001; Figure 5) and available (F(1,170) = 31.9; P < 0.001) copper increased slightly over time in both sprayed and unsprayed plots. Figure 5 Open in new tabDownload slide Comparison of total copper concentrations in the soil after application between treatments and over time (shaded areas show 95% confidence intervals). Figure 5 Open in new tabDownload slide Comparison of total copper concentrations in the soil after application between treatments and over time (shaded areas show 95% confidence intervals). The deposition of copper into water traps differed between the sites, with higher concentrations collected under the tree canopy at Millbuie (60.9 μg l−1, SE = 8.8) compared with Monaughty (39.7 μg l−1, SE = 4.6), both values having been corrected for differing initial copper content of the bottled water used. Water traps laid out along two 150 m transects situated in a large open area within the Millbuie plot also captured a mean 59.8 μg l−1 copper indicating overspray of this non-forested part of the plot. The tray area (1 m2) and volume of water (5 l) were used to simulate accidental over-spray of a shallow lake 50 m × 50 m in area, and 5 m deep. The highest recorded in-plot (transect S1 at Millbuie) and outside plot (8 m outside eastern plot boundary) depositions of 88 and 253 μg l−1, respectively, would result in copper concentrations of 0.08 and 0.25 μg l−1 in the water body i.e. lower than the Environmental Quality Standard of 1 μg copper l−1 for freshwater set out by the Scottish Environmental Protection Agency (SEPA, 2020), and over 250× lower than concentrations reported downstream from whisky distilleries in Scotland (Paton et al., 1995). Assessment of biological indicators There was a significant 3-way interaction between site, plot and plant species in determining percentage cover of plant species in ground transects (χ2(4, 42) = 23.2, P < 0.001; Table 1) but no significant changes in vegetation cover, or vegetation type after fungicide application (χ2(1,1156) = 0.9, P = 0.34 NS). The vascular flora at Millbuie was slightly less rich than Monaughty, with 7 and 11 vascular species, respectively, recorded in the ground quadrats. Both sites had high coverage of Calluna vulgaris (L.) Hull (Table 1). Two grasses, Agrostis capillaris L. and Poa pratensis L., were found in Milllbuie but not Monaughty, whilst Dryopteris dilatata (Hoffm.) A. Gray, Galium saxatile L., Luzula multiflora (Ehrh.) Lej., Oxalis acetosella L., Sorbus aucuparia L. and Trientalis europaea L., which formed part of the ground flora at Monaughty, were not found in Millbuie. The sites had a similar diversity of non-vascular species (9 species at Monaughty and 10 at Millbuie) with a preponderance of Dicranum scoparium Hedw. in ground-quadrats and the lichen species, Hypogymnia physodes (L.) Nyl., dominating canopy communities. Table 1 Variation in vegetation cover of 1-m2 quadrats between plots and sites (Data restricted to those species where the median % cover was ≥1% in all sites and plots). Site . Plot . Plant species Mean % quadrat coverage ± SE . Calluna vulgaris (L.) Hull . Deschampsia flexuosa (L.) Trin. . Hylocomium splendens(Hedw.) Schimp. . Pleurozium schreberi (Willd. Ex Brid.) Mitt. . Polytrichum commune Hedw. . Monaughty Control 51.9 ± 12 b* 12.4 ± b 18.9 ± 4 a 1.8 ± 0 a 0 ± 0 a,b,c Sprayed 58.2 ± 13 b 22.1 ± 5 b 14.2 ± 3 a 6.6 ± 2 b 1.4 ± 0 a Millbuie Control 6.8 ± 1 a 14.4 ± 3 b 23.9 ± 4 a 11.7 ± 2 b,c 14.0 ± 2 b Sprayed 11.3 ± 2 a 2.8 ± 1 a 16.8 ± 3 a 16.9 ± 3 c 37.0 ± 7 c Site . Plot . Plant species Mean % quadrat coverage ± SE . Calluna vulgaris (L.) Hull . Deschampsia flexuosa (L.) Trin. . Hylocomium splendens(Hedw.) Schimp. . Pleurozium schreberi (Willd. Ex Brid.) Mitt. . Polytrichum commune Hedw. . Monaughty Control 51.9 ± 12 b* 12.4 ± b 18.9 ± 4 a 1.8 ± 0 a 0 ± 0 a,b,c Sprayed 58.2 ± 13 b 22.1 ± 5 b 14.2 ± 3 a 6.6 ± 2 b 1.4 ± 0 a Millbuie Control 6.8 ± 1 a 14.4 ± 3 b 23.9 ± 4 a 11.7 ± 2 b,c 14.0 ± 2 b Sprayed 11.3 ± 2 a 2.8 ± 1 a 16.8 ± 3 a 16.9 ± 3 c 37.0 ± 7 c *Figures followed by same letter did not differ significantly P < 0.01. Open in new tab Table 1 Variation in vegetation cover of 1-m2 quadrats between plots and sites (Data restricted to those species where the median % cover was ≥1% in all sites and plots). Site . Plot . Plant species Mean % quadrat coverage ± SE . Calluna vulgaris (L.) Hull . Deschampsia flexuosa (L.) Trin. . Hylocomium splendens(Hedw.) Schimp. . Pleurozium schreberi (Willd. Ex Brid.) Mitt. . Polytrichum commune Hedw. . Monaughty Control 51.9 ± 12 b* 12.4 ± b 18.9 ± 4 a 1.8 ± 0 a 0 ± 0 a,b,c Sprayed 58.2 ± 13 b 22.1 ± 5 b 14.2 ± 3 a 6.6 ± 2 b 1.4 ± 0 a Millbuie Control 6.8 ± 1 a 14.4 ± 3 b 23.9 ± 4 a 11.7 ± 2 b,c 14.0 ± 2 b Sprayed 11.3 ± 2 a 2.8 ± 1 a 16.8 ± 3 a 16.9 ± 3 c 37.0 ± 7 c Site . Plot . Plant species Mean % quadrat coverage ± SE . Calluna vulgaris (L.) Hull . Deschampsia flexuosa (L.) Trin. . Hylocomium splendens(Hedw.) Schimp. . Pleurozium schreberi (Willd. Ex Brid.) Mitt. . Polytrichum commune Hedw. . Monaughty Control 51.9 ± 12 b* 12.4 ± b 18.9 ± 4 a 1.8 ± 0 a 0 ± 0 a,b,c Sprayed 58.2 ± 13 b 22.1 ± 5 b 14.2 ± 3 a 6.6 ± 2 b 1.4 ± 0 a Millbuie Control 6.8 ± 1 a 14.4 ± 3 b 23.9 ± 4 a 11.7 ± 2 b,c 14.0 ± 2 b Sprayed 11.3 ± 2 a 2.8 ± 1 a 16.8 ± 3 a 16.9 ± 3 c 37.0 ± 7 c *Figures followed by same letter did not differ significantly P < 0.01. Open in new tab The coverage and composition of non-vascular moss and lichen species present in the ground, stem and branch quadrats at both sites showed no change post-application at either site. No phytotoxic effects (wilting, senescence, foliage discolouration other than that caused by foliar pathogens) were observed in any of the vegetation monitored within the quadrats, or within the treated trees themselves. Above and below ground ectomycorrhizal fungal diversity was relatively low at both sites, typical of P. sylvestris growing on podzolic soils (Jarvis et al., 2013). 2323 fungal fruit bodies were identified at Monaughty comprising 17 EMF species. 81 per cent were Lactarius rufus (Scop, ex Fr.) Fr., and only one other species (Inocybe subcarpta Kühner & Boursier) formed more than 5 per cent of the total (7 per cent). Amanita fulva Fr., Paxillus involutus (Batsch. ex Fr.) Fr., Russula aeruginea Lindblad; -Fr., and Russula cf fragilis (Pers. Fr.) Fr. were present in low numbers and unique to the sprayed block. Gomphideus roseus (Fr.) Karst., Russula nitida (Pers. ex Fr.) Fr. and Suillus bovinus (L. ex Fr.) O. Kuntze were only found in the unsprayed block. More fruit bodies (mean 260) and a slightly greater diversity of species (10) were present in sprayed plot transects than unsprayed transects (mean 200 fruit bodies, 7 species per transect). In total, 102 fruit bodies, comprising 12 EMF were found at Millbuie, again all species typically associated with P. sylvestris (28 per cent Russula emetica (Schaeff. ex Fr.) SF Gray, 24 per cent Hygrophorus hypothejus Fr., 14 per cent L. rufus, with a small number of the remaining 9 species). Similar numbers of fruit bodies were found on sprayed (mean = 8) and unsprayed (10) transects, and the same mean species diversity (mean = 3) was found on all transects. Although pre-application assessments had not been possible, the similar fruit body abundance and species diversity in treated and untreated plots the Autumn following applications at both sites is an indication that the fungal application had had no gross impacts on fungal ground flora. 10 201 and 19 172 sub-surface ectomycorrhizal tips were identified within the soil samples from Monaughty and Millbuie, respectively. In total, 16 taxa were identified in soil at Monaughty, 14 in the sprayed plot and 13 in the unsprayed plot. Seventy-five per cent of all tips examined were L. rufus and only Tomentellopsis submollis (Svrček) Hjortstam formed more than 5 per cent (6.2 per cent) of the remaining tips in both plots. In total, 13 taxa were found at Millbuie, 11 present in the sprayed plot and 10 in the unsprayed plot. L. rufus was again the most common (73 per cent of tips), and only one other taxon, Cenococcum geophilum Fr., formed more than 5 per cent (19.7 per cent) of the remaining samples in both plots. Slight increases in numbers of L. rufus tips were recorded in sprayed plots at both sites over 1 year (Monaughty: increasing from 920 to 972 tips; Millbuie: 1588 to 1760), and an increase in the unsprayed plot at Monaughty (from 794 to 1088) and decrease at Millbuie (from 2281 to 2054) after 1 year, but no marked changes in species diversity or numbers of EMF tips could be attributed to the fungicide applications. No bees were observed during bee activity surveys, no slugs/snails recorded in the vegetation quadrats, and no worms collected during extensive soil sampling. Therefore, in these two trials, it was not possible to gauge the impact of aerial application on these components of the ecosystem. Assessment of efficacy Although DNB survey data (Tubby, unpublished data) indicate Scots pine in this part of Scotland can hold three, sometimes more years of foliated growth within the crown, both sites retained below average foliation before treatment. Changes in needle retention over time could not be formally analyzed due to a lack of variance in the response (91 per cent of trees retained 2 years of needles after application), but needle retention decreased slightly in both plots at Monaughty over the 2 years, and increased slightly in both plots at Millbuie (Table 2), which suggests no influence of fungicide treatment. Table 2 Pinus sylvestris needle retention before and after fungicide applications at Monaughty and Millbuie. Site/treatment . No. years of needles retained in crown (Means ± SE) . Pre-application . 2 years post-application . Monaughty  No treatment 2.4 ± 0.04 2.0 ± 0.03  Copper oxychloride 2.4 ± 0.04 1.9 ± 0.02 Millbuie  No treatment 1.0 ± 0.0 2.0 ± 0.01  Copper oxychloride 1.5 ± 0.05 2.1 ± 0.04 Site/treatment . No. years of needles retained in crown (Means ± SE) . Pre-application . 2 years post-application . Monaughty  No treatment 2.4 ± 0.04 2.0 ± 0.03  Copper oxychloride 2.4 ± 0.04 1.9 ± 0.02 Millbuie  No treatment 1.0 ± 0.0 2.0 ± 0.01  Copper oxychloride 1.5 ± 0.05 2.1 ± 0.04 Open in new tab Table 2 Pinus sylvestris needle retention before and after fungicide applications at Monaughty and Millbuie. Site/treatment . No. years of needles retained in crown (Means ± SE) . Pre-application . 2 years post-application . Monaughty  No treatment 2.4 ± 0.04 2.0 ± 0.03  Copper oxychloride 2.4 ± 0.04 1.9 ± 0.02 Millbuie  No treatment 1.0 ± 0.0 2.0 ± 0.01  Copper oxychloride 1.5 ± 0.05 2.1 ± 0.04 Site/treatment . No. years of needles retained in crown (Means ± SE) . Pre-application . 2 years post-application . Monaughty  No treatment 2.4 ± 0.04 2.0 ± 0.03  Copper oxychloride 2.4 ± 0.04 1.9 ± 0.02 Millbuie  No treatment 1.0 ± 0.0 2.0 ± 0.01  Copper oxychloride 1.5 ± 0.05 2.1 ± 0.04 Open in new tab Percent crown infection was relatively low at both sites, and increased slightly over the 2 year post-application monitoring period. However, the fungicide treatment significantly decreased foliar infection relative to baseline values (χ2(1,18) = 40.6), with a smaller proportion of infected tees in the sprayed plots (Figure 6). Fungicide treatment and site significantly affected the proportion of trees with greater than 5 per cent crown infection (plot—χ2(1,18) = 31.8, P < 0.001; site—χ2(1,18) = 18.3, P < 0.001), and the interaction of the two had a significant effect on infection levels of over 10 per cent (plot × site—χ2(1,18) = 4.5, P = 0.03). However, the differences were less marked at the higher percentage infection cut-offs. The disease “impact factor” only differed significantly between sites (F(1,14) = 135, P < 0.001), and not fungicide treatment (F = 2.3(1,14), P = 0.15). Other potentially defoliating foliar pathogens including Lophodermella sulcigena (Rostr.) v Hohn and L. conjuncta (Darker) Darker were observed during routine Forestry Commission DNB surveys in Millbuie Forest in 2015/2016, but infection by these agents was sporadic, and not quantified during this study. Figure 6 Open in new tabDownload slide Proportion of trees exhibiting any infection, >5% infection or >10% infection, respectively. Data points show estimated marginal means, corrected to a baseline pre-spray infection level of 0%. Error bars illustrate 95% confidence intervals. Lettering shows significant differences within each figure. Figure 6 Open in new tabDownload slide Proportion of trees exhibiting any infection, >5% infection or >10% infection, respectively. Data points show estimated marginal means, corrected to a baseline pre-spray infection level of 0%. Error bars illustrate 95% confidence intervals. Lettering shows significant differences within each figure. Discussion This study was valuable in expanding the evidence base and skill set necessary for conducting aerial pesticide operations in a British forest landscape currently experiencing significant problems from native and non-native pests and diseases (Brasier, 2008; Tubby and Webber, 2010; Santini et al., 2013). Large scale aerial pesticide applications have not been carried out in British forestry for two decades, and prior operations have not been well documented, leading to a loss of vital, practical knowledge. Hence, these trials were a steep learning curve for the various teams involved. They highlighted some limitations in the equipment used and the need for optimization of application technology, the need for clear, prior agreement of operating procedures, and establishment of open communication channels between ground and aerial crews during the operation, as they tend to be physically distant from each other for health and safety reasons. Forestry applications can be technically more difficult than agricultural operations as areas under treatment are often irregular in shape. As in previous aerial studies (Payne et al., 1988; Crabbe et al., 1994; Davis et al., 1994; Damalas, 2015), a degree of off-target drift did occur even though operations were carried out under unstable atmospheric conditions conducive to foliar capture of spray droplets (Davis and Williams, 1990; Crabbe et al., 1994; Picot and Kristmanson, 2012). The majority of droplets were captured within 50 m of the boundaries however, with just 0.5 per cent of the average in-plot deposition recorded at 100 m, the furthest extent of the buffer zone. As foliar analysis showed very slightly elevated copper concentrations as far away as 250 m, it is recommended that in any future trials control and treated sites are at least 250 m distant from each other. Drift from helicopters is generally lower than that experienced with fixed wing aircraft (Salyani and Cromwell, 1992; Payne, 1998) and comparison with New Zealand operations illustrates this extent of drift was typical for the crop and fungicide formulation. Agricultural (Payne, 1992; Davis et al., 1994; Picot and Kristmanson, 2012), and ULV forestry insecticide (Crabbe et al., 1994) applications using fixed wing aircraft, have resulted in drift recorded beyond 400 m. Digital site maps and GPS navigational systems were available to the pilot, but the flight path records illustrated some overspray outside plot boundaries, and across open areas within the Millbuie plot. The “on/off” operation of the rotary atomisers appeared not to be working optimally either, and delayed valve shut-offs led to excess product being delivered at the furthest extents of the flight paths. These issues are currently undergoing further investigation and optimization in field trials, as off-target exposure represents inefficient pest control, an economic loss, and the potential for environmental contamination. Although the same application system was used in both trials, there were differences in product delivery to foliage between sites. Product behaviour depends heavily on droplet size and this is affected by weather conditions, wind-speed and height of spray release relative to the crop canopy (Bird et al., 1996; Felsot et al., 2011). Flights were carried out under similar meteorological conditions and the ULV application should deliver droplets with a median diameter of 65 μm, optimized for capture by fine conifer needles (Picot and Kristmanson, 2012). Although droplet size was not specifically recorded, and catchers were used simply to confirm on-target deposition, the extent of drift, and to establish a baseline for further studies, differences in droplet size populations between sites could explain the different levels of deposition; small droplets with lower settling velocities have a greater propensity to persist in the drift cloud, and larger droplets can bounce off foliage, ending up the ground (Salyani and Cromwell, 1992; Picot and Kristmanson, 2012). Use of electrically driven atomisers in future work is recommended as it would give greater control over droplet generation than the wind-driven atomisers used in this study (Micronair, 2020). Comparing deposition data on foliage, litter, soil and water did not give a consistent picture: lower foliar deposition and higher capture within the ground-based water traps at one site (Millbuie) might imply delivery of larger than optimum droplets, but this was not consistent with the lower copper concentrations found in the ground needle litter samples taken adjacent to the water samples at the same site. Building on these experiences, further aerial operations are ongoing without fungicidal product, to monitor droplet size, drift and deposition in greater detail, under a range of environmental conditions in Britain. In combination with the USDA Forest Service AGDISP™ model (Bilanin et al., 1987), this should enable better prediction and management of drift under a range of flying conditions and better on-target delivery of product. Overall, although differences in delivery to foliage (13 and 46 μg g−1 dw) were recorded between the two operations, this is not unusual for aerial field applications. Post-application depositions ranging between 26 and 50 μg g−1 dw foliage have been recorded in typical New Zealand operations (Lindsay Bulman Scion NZ, pers. Com. 21 June 2017). Comparison of absolute values is not appropriate as these figures originate from P. radiata with different canopy “form”, planting density, and thinning and pruning regimes, all variables likely to affect deposition (Picot and Kristmanson, 2012). Nevertheless, they provide a useful example of variability experienced in routine field applications. Product persistence was higher than expected in this study, as copper was still detectable on foliage after a year, only dropping to background levels after 2 years. In contrast, in New Zealand copper has been found to persist on foliage for only 2–3 months (Bulman et al., 2004). This is likely to be due to the lower annual rainfall in this part of Scotland (547 mm recorded near Monaughty, National Grid Reference NJ 06774 62 804, between July 2013 and end June 2014 and 837 mm near Millbuie Forest at National Grid Reference NH 5819 5851 from July 2015 to end June 2016) compared with e.g. 1400 mm in Rotorua, New Zealand (New Zealand Tourism Guide, 2020). The nature of relationships between effective protection against DNB and decreasing foliar concentrations over time is not known but might give justification to adopting a New Zealand-style policy of typically only spraying every 4–5 years to control the disease (Bulman et al., 2008). Over time the copper could be tracked through the canopy as it was slowly washed off the foliage by rainfall into the lower parts of the crown and onto the ground needle litter layer, and then into the soil. Elevated levels of copper in soil can reduce soil microbial biomass (Dumestre et al., 1999) with subsequent impacts on physical and chemical soil processes, but the actual availability of copper to biological organisms is dictated by soil pH, redox potential, oxide content, and mineral content (Alva et al., 2000). When bound to organic and inorganic colloids, it is relatively unavailable and immobile, reducing its potential toxicity. The differences in initial “total” versus “biologically available” copper indicated a level of variation in these soil components between the sites. However, the organic content of both soils enabled them effectively to “buffer” the copper application, as applications increased total but not available copper. The recorded increases in soil copper over time in all plots, sprayed or not, were more difficult to interpret. Cross-contamination was unlikely as sprayed plots had large buffer zones around them, and at Millbuie the plots were 1.5 km distant, and the untreated site situated upwind (Dalcross Meteorological Station—Meteorological Office, 2012). Improvements in application technology resulting in ten-fold decreases in volumes of copper used against DNB since the 1960s (Bulman et al., 2004) have undoubtedly reduced potential environmental impacts of copper applications. Soil concentrations remain below EU maximum permitted concentrations (140 μg g−1; C.E.C, 1986) even following two to five applications per crop rotation in New Zealand (Bulman, 2008). In contrast, the 7500 tonnes copper fungicides year−1 used on agricultural crops in Australia, can lead to soil concentrations exceeding 340 μg g−1 (Van-Zwieten et al., 2004). Soil copper concentrations in this study consistently remained lower than the EU statutory level, and lower than median copper concentrations in Scottish soils (7.5 μg g−1 mineral soils and 4.5 μg g−1 organic soils; Paterson, 2011). Consequently, impacts on soil microbes are likely to have been be minimal. Certainly, no major changes in the occurrence of soil-dwelling ectomycorrhizal fungi were detected 12 months after application. Extended monitoring of the soil and soil-dwelling biota is recommended, given the new findings that copper is still detectable above natural levels in needle litter after 157 weeks, and therefore potentially still available to contribute to soil copper concentrations. Further monitoring would also be necessary if a programme of multiple aerial operations were carried out on each crop rotation. Earthworms also play a vital role in maintaining soil condition, but are sensitive to copper through ingestion and/or direct absorption through the skin (Streit, 1984). Significant reduction in numbers and biomass has been recorded at concentrations exceeding 180–338 μg g−1 (Streit, 1984; Van-Zwieten et al., 2004). However, these concentrations are far greater than anything recorded within this study. Although impacts on earthworms were not quantified as no worms were found in this study, the soil copper concentrations before and after application fell within the range naturally found in Scotland (Paterson, 2011). Consequently, impacts on earthworms directly attributable to the application would be minimal, although again, longer term monitoring would be necessary in the case of repeated applications. Bees fulfil a crucial role in ecosystem functioning but flying insects are efficient collectors of small spray droplets and they could therefore be particularly vulnerable to aerially dispersed fungicides (Davis and Williams, 1990). The impact of copper fungicides on bees is unclear. Large variations in bee copper concentrations have been reported (Van der Steen et al., 2012), sometimes from exposure to industrial processes (Veleminsky et al., 1990) but also from relatively high natural heavy metal concentrations found in pollen (Lambers et al., 1998). Ecotoxicological studies indicate LD50 values of 18.1 μg bee−1 (oral) and 109.9 μg bee−1 (contact; Turner, 2017), but whereas bees are often used as bioindicators to measure, for example, copper fungicide drift, the actual lethal/sublethal effects on the bees are seldom specifically reported (Fakimzadeh and Lodenius, 2000; Van der Steen et al., 2012). Risks to bees were thought to be minimal in these operations as foraging activity concentrates around flowering plants on the forest floor rather than in the canopies of the trees, which received most deposition, and such ground vegetation was relatively sparse. The spray deposits also dry very rapidly (in c. 1 h), further reducing risks of exposure to bees landing on foliage. Good relationships were established with local bee-keeping organizations during the course of these operations and as a precaution; beekeepers in the vicinity of these trials were recommended to keep their domestic honeybees within the hives during, and for a few hours after the applications. Given the significance of bees as pollinators, and the difficulties experienced in sampling them on site, it would be advisable to focus future sampling on potential bee contamination and differences in behaviour in local hives before and after the operation (Berry et al., 2013). Copper in solution can be highly bioavailable to aquatic communities through diffusionary uptake or via the food chain (Streit, 1984), although this is influenced by the organism and the degree of copper complexation and adsorption (Stauber and Davies, 2000). Elevated concentrations in whisky distillery waste have been shown to have deleterious effects on aquatic microorganisms (Paton et al., 1995. See also Brown et al., 2004; Das and Khangarot, 2011) and the use of copper and other fungicides near water is heavily regulated for these reasons (Payne et al., 1988; Davis and Williams, 1990; Anon, 1992). The 50 m buffer zone originally stipulated by HSE in these British trials had been voluntarily extended to 100 m, but ultimately, very low quantities of copper were detected beyond 50 m and the simulated overspray of water in this study clearly illustrated that applications of copper at rates used for DNB control, even in the absence of an intervening canopy, did not exceed environmental quality guidelines for copper within or outside the target sites (see also Baillie et al., 2017). Elevated copper concentrations can affect terrestrial plants by inhibiting a large number of enzymes involved in processes including photosynthesis, pigment synthesis, and membrane integrity. At the whole plant level, this manifests as reduced growth and general symptoms of senescence (Fernandez and Henrique, 1991; Pätsikkä et al., 2002). Elevated concentrations were recorded within the tree foliage and ground litter after the operation, but no deleterious impacts were observed on the trees or other vegetation. Lichens are often used as bio-indicators due to their sensitivity to air-borne pollution (Van Dobben et al., 2001; Balabanova et al., 2012), but no changes in ectomycorrhizal fungal communities or growth of lichens were observed after the application over 1 or 2 years of monitoring. Visible foliage infection by D. septosporum was low at both sites before the applications, but was reduced significantly by the fungicide treatment. The negative effects of DNB on tree yield and survival are largely due to infected needles falling prematurely, decreasing photosynthetic capacity (Gibson, 1974), but in this study no impacts of treatment on needle retention were detected. Average needle retention in both forests was below the average for Scots pine in good condition in this area of Britain, but the relatively low levels of visible foliar infection suggest the thin crowns were at least partly a result of suboptimal site conditions rather than DNB. It had proved difficult to identify suitable sites due to environmental restrictions and the need to have paired, “control” plots close by. Clearer impacts of treatment might have been obtained at sites with higher levels of prior infection by DNB, and it is recommended that further trials take place to investigate whether treatment can increase needle retention as well as reduce foliar infection. There was a precedent for using copper as the active ingredient in these trials due to its long-term use against DNB in New Zealand. Cuprous oxide (Cu2O) is used more commonly now in New Zealand (Bulman et al., 2004, Bulman, 2008) at least partly because this compound contains proportionally higher amounts of the copper active ingredient than copper oxychloride (H6Cl2Cu4O6). Copper oxychloride was used in these trials as, at the time of the applications, cuprous oxide was not registered for use in Britain. This study showed copper was more persistent on the targeted foliage than has been found in New Zealand’s operations, and applications seem to have a low environmental impact. However, as changes in pesticide legislation may restrict copper’s use in the longer term, other active ingredients are currently being investigated in ground-based field trials (Tubby, unpublished data). If/when consistently effective actives are found, it will then be essential to conduct cost–benefit analyses to examine predicted timber volumes saved, current and future timber prices and all operational costs (Bulman et al., 2008). Conclusion Pathogens such as D. septosporum, P. ramorum and H. fraxineus are forcing change on the forest industry. Practical methods are needed to protect our forests as these three organisms alone could affect 159 million m3 of standing timber in British woodlands (Forestry Commission, 2014—Scots, lodgepole and Corsican pine, larch and ash). This is at a time when there is a strong global timber demand, expanding markets, increased interest in local timber (Forestry Commission, 2017; Grown in Britain, 2020) and environmental incentives for expanding forestry (BEIS, 2017). Awareness of pest and disease threats including DNB is generally high amongst forest managers, but recent studies suggest high levels of inaction reflecting, amongst other things, a lack of evidence from the field, and scepticism over the feasibility and efficacy of proposed management options (Marzano, Fuller and Quine, 2017). This study demonstrated that copper oxychloride has the potential to decrease D. septosporum infection, but further work is needed to see if it can increase needle retention in heavily infected crops. It increased the skill set within British forest industry, provided baseline data on spray behaviour and fungicide persistence under British conditions, addressed potential risks to local ecosystems and highlighted technical issues needing further study. This study was a valuable first step towards the re-inclusion of aerial pesticide application within an integrated, multi-faceted toolkit for future forest protection and management. Acknowledgements I gratefully acknowledge the help and support of the following people: Andy Taylor (James Hutton Institute), Lindsay Bulman Scion NZ, Hugh Clayden (Scottish Forestry), David Henderson (Forestry and Land Scotland), PDG Helicopters, Forest Research Technical Support Unit (Alistair MacLeod, Hazel Andrew, Stephen O’Kane, Fraser McBirnie, Steg McBirnie, Duncan Williams, Colin Smart, Martin MacKinnon, Sandy Bowran, Calum Murray), Forest Research Technical Development Branch (Bill Jones and Michael Wall), Ben Griffin, Peter Walling, Hugh MacKay and local Forestry and Land Scotland District teams, Paul Taylor, François Bochereau and Alberto Morales (FR). I also thank Ian Tubby, and anonymous reviewers for helpful comments on the manuscript. Conflict of interest statement None declared. Funding Forestry Commission Corporate Forestry Services and Forestry Commission Scotland, to whom we extend grateful thanks. Data availability The data underlying this article will be shared on reasonable request to the corresponding author. References Agerer , R. 1986–2002 Colour Atlas of Ectomycorrhizae . Einhorn Verlag . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Alva , A.K. , Huang , B. and Paramasivam , S. 2000 Soil pH affects copper fractionation and phytotoxicity . Soil Sci. Soc. Am. J. 64 ( 3 ), 955 – 962 . doi: 10.2136/sssaj2000.643955 Google Scholar Crossref Search ADS WorldCat Anon 1992 Ontario Ministry of Environment and Energy Buffer Zone Guidelines for Aerial Application of Pesticides in Crown Forests of Ontario , p. 5 . Anon 2012 The Ontario Ministry of Environment and Energy . Baillie , B.R. , Evanson , A.W., Unsworth , D. and Jeram , S. 2017 Aerial application of copper for dothistroma control in New Zealand’s planted forests—effect on stream environments . Environ. Sci. Pollut. Res. 24 ( 3 ) 24494 – 24508 . doi: 10.1007/s11356-017-0020-4 Google Scholar Crossref Search ADS WorldCat Balabanova , B. , Stafilov , T., Šajn , R. and Baèeva , K. 2012 Characterisation of heavy metals in lichen species Hypogymnia physodes and Evernia prunastri due to biomonitoring of air pollution in the vicinity of copper mines . Int. J. Environ. Res. 6 ( 3 ) 779 – 794 . Google Scholar OpenURL Placeholder Text WorldCat Balch , R.E. , Webb , F.E. and Morris , R.F. 1954 Results of spraying against the spruce budworm in New Brunswick . Can. Dep. Agri. Sci. For. Biol. Div. Bi-monthly Prog. Report. 10 ( 1 ) 1 – 2 . Google Scholar OpenURL Placeholder Text WorldCat BEIS. 2017 Clean Growth Strategy: Leading the Way to a Low Carbon Future . HM Government , pp. 1 – 2 . https://www.gov.uk/government/publications/clean-growth-strategy ( accessed on 10 January 2018 ). Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Berry , J.A. , Hood , W.M., Pietravalle , S. and Delaplane , K.S. 2013 Field-level sublethal effects of approved bee hive chemicals on honey bees (Apis mellifera L.) . PLoS One . 8 ( 10 ) e76536 . Google Scholar Crossref Search ADS PubMed WorldCat Bilanin , A.J. , Teske , M.E., Barry , J.W. and Ekblad , R.B. 1987 AGDISP: the aircraft spray dispersion model, code development and experimental validation . Trans. ASAE . 32 ( 1 ) 327 – 334 . doi: 10.13031/2013.31005 . https://www.epa.gov/pesticide-science-and-assessing-pesticide-risks/models-pesticide-risk-assessment ( accessed on 12 December 2017 ). Google Scholar Crossref Search ADS WorldCat Bird , S.L. , Esterly , D.M. and Perry , S.G. 1996 Off target deposition of pesticides for agricultural aerial spray application . J. Environ. Qual. 25 , 1095 – 1104 . Google Scholar Crossref Search ADS WorldCat Bradshaw , R.E. 2004 Dothistroma (red-band) needle blight of pines and the dothistromin toxin: a review . For. Pathol. 34 , 163 – 185 . Google Scholar Crossref Search ADS WorldCat Brasier , C.M. 2008 The biosecurity threat to the UK and global environment from international trade in plants . Plant Pathol. 57 , 792 – 808 . Google Scholar Crossref Search ADS WorldCat Brasier , C. and Webber , J.F. 2013 Vegetative incompatibility in the ash dieback pathogen Hymenoscyphus pseudoalbidus and its ecological implications . Fungal Ecol. 6 ( 6 ) 501 – 512 doi: 10.1016/j.funeco.2013.09.006 Google Scholar Crossref Search ADS WorldCat Brown , R. 2015 A renaissance in the chemical control of bracken? In Proceedings of the 11th National Heathland Conference, 18–20 March 2015 . I. , Alonso, J. , Underhill-Day, S. , Lake (eds.). Sunningdale Park , pp. 60 – 74 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Brown , R. and Robinson , R. 1997 Bracken Management Handbook: A Guide to Best Practice . Essex . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Brown , R.J. , Galloway , T.S., Lowe , D., Browne , M.A., Dissanayake , A., Jones , M.B. et al. 2004 Differential sensitivity of marine invertebrates to copper assessed using multiple biomarkers . Aquat. Toxicol. 66 , 267 – 278 . doi: 10.1016/j.aquatox.2003.10.001 . Google Scholar Crossref Search ADS PubMed WorldCat Bulman , L.S. , Gadgil , P.D., Kershaw , D.J. and Ray , J.W. 2004 Assessment and control of Dothistroma needle blight . Forest Research Bulletin 229 . Forest Research Rotorua , p. 48 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Bulman , L.S. 2008 Determining Copper Content in Soil in Sprayed and Unsprayed Stands . Dothistroma Control Committee Client Report No. 12747 , p. 7 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Bulman , L.S. , Bradshaw , R.E., Fraser , S., Martín-García , J., Barnes , I., Musolin , D.L. et al. 2016 A worldwide perspective on the management and control of Dothistroma needle blight . For. Pathol. 46 , 472 – 488 . doi: 10.1111/efp.12305 . Google Scholar Crossref Search ADS WorldCat Bulman , L.S. , Dick , M.A., Ganley , R.J., McDougal , R.L., Schwelm , A. and Bradshaw , R.E. 2013 Dothistroma needle blight. In Infectious Forest Diseases . P. , Gonthier, G. , Nicolotti (eds.). CABI , pp. 436 – 457 . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Bulman , L.S. , Ganley , R. and Dick , M. 2008 Needle diseases of radiata pine in New Zealand . Forest Biosecurity Research Council , Client Report, 13010 , p. 81 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Bulman , L.S. , Gadgil , P.D., Kershaw , D.J. and Ray , J.W. 2004 Assessment and control of Dothistroma needle blight . FRI Bull. 229 , 48 . Google Scholar OpenURL Placeholder Text WorldCat Butovitsch , V. 1947 Redogörelse för flygbekämpningskampanjen mot tallmätaren under åren 1944–5. Bericht ober die flugzeugbestaubung gegeb den kiefernspanner in den jarren 1944–1945 Medd. Skogsforskningsinst. , 35 ( 9 ) 109 . Google Scholar OpenURL Placeholder Text WorldCat Cebeci , H.H. , Oymen , R.T. and Acer , S. 2010 Control of pine processionary moth, Thaumetopoea pityocampa with Bacillus thuringiensis in Antalya, Turkey . J. Environ. Biol. 31 , 357 – 361 . Google Scholar PubMed OpenURL Placeholder Text WorldCat C.E.C. (Commission of the European Communities) 1986 Council Directive on the protection of the environment, and in particular of the soil, when sewage sludge is used in agriculture . Off. J. Eur. Commun. L181, Annex 1A, 10 p . OpenURL Placeholder Text WorldCat Civil Aviation Authority. 2006 CAP414 The Aerial Application Certificate . www.caa.co.uk/publications ( accessed on 12 December 2017 ). Clark , J. and Webber , J. 2017 The ash resource and the response to ash dieback in Great Britain. In Dieback of European Ash (Fraxinus spp.)—Consequences and Guidelines for Sustainable Management. In The Report on European Cooperation in Science & Technology (COST) Action FP1103 FRAXBACK . R. , Vasaitis, R. , Enderle (eds.). SLU Institutionen för skoglig mykologi och växtpatologi , pp. 228 – 237 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Crabbe , R.S. , McCooeye , M. and Mickle , R.E. 1994 The influence of atmospheric stability on wind drift from Ulta-low-volume forest spray applications . J. Appl. Meteorol. 33 , 500 – 507 . Google Scholar Crossref Search ADS WorldCat Crooke , M. 1959 Insecticidal control of the pine looper in great Britain: I. Aerial spraying . Forestry . 32 ( 2 ) 166 – 196 . doi: 10.1093/forestry/32.2.166 . Google Scholar Crossref Search ADS WorldCat Damalas C.A. 2015 Pesticide drift: seeking reliable environmental indicators of exposure assessment. In Environmental Indicators, R. Armon and O. Hänninen (eds). Springer , pp. 251 – 261 . doi: 10.1007/978-94-017-9499-2_15 ( accessed on 05 November 2018 ). Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Das , S. and Khangarot , B.S. 2011 Bioaccumulation of copper and toxic effects on feeding, growth, fecundity and development of pond snail Lymnaea luteola L . J. Hazard. Mater. 185 ( 1 ) 295 – 305 . doi: 10.1016/j.jhazmat.2010.09.033 . Google Scholar Crossref Search ADS PubMed WorldCat Davis , B.N.K. and Williams , C.T. 1990 Buffer zone widths for honeybees from ground and aerial spraying of insecticides . Environ. Pollut. 63 , 247 – 259 . Google Scholar Crossref Search ADS PubMed WorldCat Davis , B.N.K. , Frost , A.J., Pinder , L.C.V. and Cooke , A.S. 1994 Environmental impact of pesticide drift: aerial spraying . English Nature Research Reports No. 112 , p. 1 – 59 . Desprez-Loustau , M.L. , Robin , C., Reynaud , G., Déqué , M., Badeau , V., Piou , D. et al. 2007 Simulating the effects of a climate-change scenario on the geographical range and activity of forest pathogenic fungi . Can. J. Plant Pathol. 29 , 101 – 120 . Google Scholar Crossref Search ADS WorldCat Desprez-Loustau , M.L. , Marçais , B., Nageleisen , L.M., Piou , D. and Vannini , A. 2006 Interactive effects of drought and pathogens in forest trees . Ann. For. Sci. 63 , 597 – 612 . Google Scholar Crossref Search ADS WorldCat Dumestre , A. , Sauve , S., McBride , M., Baveye , P. and Berthelin , J. 1999 Copper speciation and microbial activity in long-term contaminated soils . Arch. Environ. Contam. Toxicol. 36 , 124 – 131 . Google Scholar Crossref Search ADS PubMed WorldCat Evans , H.F. , Stoakley , J.T., Leather , S.R. and Watt , A.D. 1991 Development of an integrated approach to control of pine beauty moth in Scotland . For. Ecol. Manag. 39 , 19 – 28 . doi: 10.1016/0378-1127(91)90158-R . Google Scholar Crossref Search ADS WorldCat Fakhimzadeh , K. and Lodenius , M. 2000 Honey, pollen and bees as indicator of metal pollution . Acta U. Carol. Environment. 14 , 13 – 20 . Google Scholar OpenURL Placeholder Text WorldCat Felsot , A.S. , Unsworth , J.B., Linders , J.B.H.J., Roberts , G., Rautman , D., Harris , C. et al. 2011 Agrochemical spray drift; assessment and mitigation. A review . J. Environ. Sci. Health B . 46 ( 1 ) 1 – 23 . doi: 10.1080/03601234.2010.515161 Google Scholar Crossref Search ADS PubMed WorldCat Fernandes , J.C. and Henriques , F.S. 1991 Biochemical, physiological, and structural effects of excess copper in plants . Bot. Rev. , 57 ( 3 ), 246 – 273 . Google Scholar Crossref Search ADS WorldCat Forestry Commission 2017 Forestry Statistics and Forestry Facts & Figures. https://www.forestry.gov.uk/forestry/infd-7aqdgc (accessed on 05 November 2018) . Forestry Commission . 2014 50-year Forecast of Softwood Timber NFI Statistical Analysis Report . National Forest Inventory, Forestry Commission . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Forestry Commission 2013 NFI Preliminary Estimates of Quantities of Broadleaved Species in British Woodlands, with Special Focus on Ash . National Forest Inventory, Forestry Commission . https://www.forestry.gov.uk/pdf/NFI_Prelim_BL_Ash_Estimates.pdf/$FILE/NFI_Prelim_BL_Ash_Estimates.pdf ( accessed on 05 November 2018 ). Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Fournier , C. , Bauce , E., Dupont , A. and Berthiaume , R. 2010 Wood losses and economical threshold of Btk aerial spray operation against spruce budworm . Pest Manag. Sci. 66 , 319 – 324 . doi: 10.1002/ps.1878 . Google Scholar Crossref Search ADS PubMed WorldCat Franich , R.A. 1988 Chemistry of weathering and solubilisation of copper fungicide and the effect of copper on germination, growth, metabolism and reproduction of Dothistroma pini . N. Z. J. For. Sci. 18 , 318 – 328 . Google Scholar OpenURL Placeholder Text WorldCat van Frankenhuyzen , K. 2000 Application of Bacillus thuringiensis in forestry. In Entomopathogenic Bacteria: From Laboratory to Field Application . J.F. Charles, A. Delécluse, and C.N.L. Roux (eds). Springer . pp. 371 – 382 . doi: 10.1007/978-94-017-1429-7_20 . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC van Frankenhuyzen , K. 1990 Development and current status of Bacillus thuringiensis for control of defoliating insects . For. Chron. 66 , 498 – 507 . Google Scholar Crossref Search ADS WorldCat Gibson , I.A.S. 1974 Impact and control of dothistroma blight of pines . Eur. J. For. Pathol. 4 , 89 – 100 . Google Scholar Crossref Search ADS WorldCat Grown In Britain. 2020 https://www.growninbritain.org. ( accessed on 08 September 2020 ). Harris , A.R. 2014 The epidemiology of Phytophthora ramorum associated with Larix in the UK . PhD thesis . Dept. Life Sciences, Imperial College , 203 p . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Hartig, F. 2019 R Core Team. DHARMa: Residual Diagnostics for Hierarchical (Multi-Level/Mixed) Regression Models . R package Version 0.2.4. https://CRAN.R-project.org/package=DHARMa. Heritage , S. 1997 Pine beauty moth: its biology, monitoring and control . Forest. Comm. Res. Inform. Note 290 , 7 . Google Scholar OpenURL Placeholder Text WorldCat Howe , R.E.J. and Little , G.B. 1969 Aerial fertilisation at Wark Forest, Northumberland . J. Forest. Comm. 36 ( 1968–69 ), 61 – 65 Google Scholar OpenURL Placeholder Text WorldCat Jarvis , S. , Woodward , S., Alexander , I.J. and Taylor , A.F.S. 2013 Regional scale gradients of climate and nitrogen deposition drive variation in ectomycorrhizal fungal communities associated with native Scots pine . Glob. Chang. Biol. 19 , 1688 – 1696 . Google Scholar Crossref Search ADS PubMed WorldCat Lambers , H. , Chapin , F.S. III and Pons , T.L. 1998 Plant Physiological Ecology . Springer . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Kennedy , F. 2002 The Identification of Soils for Forest Management. Forestry Commission Field Guide . Forestry Commission , pp. 1 – 56 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC McKinney , L.V. , Nielsen , L.R., Collinge , D.B., Thomsen , I.M., Hansena , J.K. and Kjæra , E.D 2014 The ash dieback crisis: genetic variation in resistance can prove a long-term solution . Plant Pathol. 6 ( 3 ), 485 – 499 . doi: 10.1111/ppa.12196 . Google Scholar Crossref Search ADS WorldCat Marzano , M. , Fuller , L. and Quine , C.P. 2017 Barriers to management of tree diseases: framing perspectives of pinewood managers around Dothistroma needle blight . J. Environ. Manag. 188 , 238 – 245 . doi: 10.1016/j.jenvman.2016.12.002 . Google Scholar Crossref Search ADS WorldCat Mason , W.L. 2000 Silviculture and stand dynamics in Scots pine forests in Great Britain; implications for biodiversity . Invest. Agr.: Sist. Recur. For. : Fuera de Serie No. 1 , p. 175 – 197 . Google Scholar OpenURL Placeholder Text WorldCat Meteorological Office. 2012 Meteorological Office Integrated Data Archive System (MIDAS) Land and Marine Surface Stations Data (1853–Current) . NCAS British Atmospheric Data Centre . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Micronair. 2020 Micronair Atomiser Range for Helicopters (AU7000 & AU6539) . http://www.microngroup.com/micronair_atomiser_range_for_helicopters ( accessed on 08 September 2020 ). Ministry of Agriculture, Fisheries and Food. 1986 In Method 10 In Reference Book 427 (3rd Edition). The Analysis of Agricultural Materials: A Manual of the Analytical Methods Used by the Agricultural Development and Advisory Service . E. , Jackson, D.S. , Farrington, K. , Henderson (eds.). Ministry of Agriculture, Fisheries and Food , p. 248 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Mullett , M.S. 2014 The epidemiology of dothistroma needle blight in Britain . PhD thesis . Dept. Life Sciences, Imperial College , 269 p . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC New Zealand Tourism Guide 2020 Rotorua Weather and Climate . http://www.tourism.net.nz/new-zealand/about-new-zealand/rotorua-weather.html ( accessed on 08 September 2020 ). O'Hara , K.L. and Ramage , B.S. 2013 Silviculture in an uncertain world: utilizing multi-aged management systems to integrate disturbance . Forest. Int. J. Forest. Res. , 86 ( 4 ) 401 – 410 . doi: 10.1093/forestry/cpt012 . Google Scholar Crossref Search ADS WorldCat Old , K.M. and Dudzinski , M.J. 1999 Forest pathogen introductions to Australia: experiences, threats and counter measures. Current and potential impacts of pitch canker in radiata pine. In Proceedings of the IMPACT Monterey Workshop . M.E. , Devey, A.C. , Matheson, T.R. , Gordon (eds.). CSIRO , pp. 11 – 20 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Ostfeld , R.S. and Keesing , F. 2012 Effects of host diversity on infectious disease . Annu. Rev. Ecol. Evol. Syst. 43 , 157 – 182 . doi: 10.1146/annurev-ecolsys-102710-145022 . Google Scholar Crossref Search ADS WorldCat Pap , P. , Totić , D., Marković , M. and Galović , V. 2017 Praćenje gljive Dothistroma pini Hulb. na Subotičkoj peščari u godinama posle avio-tretiranja. [Assessment of Dothistroma pini Hulb. In Subotica Sands after aerial application] . Biljni Lekar. 45 ( 3 ), 303 – 310 . Google Scholar OpenURL Placeholder Text WorldCat Paterson , E. 2011 In Geochemical Atlas for Scottish Topsoils . C. , Colin Campbell, M. , Coull, C. , Shand (eds.). The Macaulay Land Use Research Institute . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Paton , G.I. , Palmer , G., Kindness , A., Campbell , C., Glover , L.A., and Killham , K. 1995 Use of luminescent marked bacteria to assess copper bioavailability in whisky distillery effluent . Chemosphere . 31 ( 5 ), 3217 – 3224 . Google Scholar Crossref Search ADS WorldCat Payne , N.J. 1998 Developments in aerial pesticide application methods for forestry . Crop Prot. 17 ( 2 ), 171 – 186 . Google Scholar Crossref Search ADS WorldCat Payne , N.J. 1992 Off-target glyphosate from aerial cilvicultural applications and buffer zones required around sensitive areas . Pestic. Sci. , 34 , 53 – 59 . https://doi.org/10.1002/ps.2780340102 . Google Scholar Crossref Search ADS WorldCat Payne , N.J. , Helson , B.V., Sundaram , K.M.S. and Fleming , R.A. 1988 Estimating buffer zone widths for pesticide applications . Pestic. Sci. 24 , 147 – 161 . doi: 10.1002/ps.2780240206 . Google Scholar Crossref Search ADS WorldCat Pätsikkä , E. , Kairavuo , M., Šeršen , F., Aro , E.M. and Tyystjärvi , E. 2002 Excess copper predisposes photosystem II to photoinhibition in vivo by outcompeting iron and causing decrease in leaf chlorophyll . Plant Physiol. 129 , 1359 – 1367 . Google Scholar Crossref Search ADS PubMed WorldCat Picot , J.J. and Kristmanson , D.D. 2012 Forestry Pesticide Aerial Spraying: Spray Droplet Generation, Dispersion, and Deposition . Environmental Science and Technology Library, Singer-Science + Business Media BV , p. 213 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Pinheiro , J. , Bates , D., DebRoy , S., Sarkar , D. 2018 R Core Team. Linear and Nonlinear Mixed Effects Models . R package Version 3.1-137 . https://CRAN.R-project.org/package=nlme. R Core Team. 2017 R: A Language and Environment for Statistical Computing . R Foundation for Statistical Computing . https://www.R-project.org/. Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Rawlings , G.B. 1955 Epidemics in Pinus radiata forests in New Zealand . N. Z. J. For. 7 , 53 – 55 . Google Scholar OpenURL Placeholder Text WorldCat Salyani , M. and Cromwell , R.P. 1992 Spray drift from ground and aerial applications . Trans. Am. Soc. Agric. Eng. 35 ( 4 ), 1113 – 1120 . Google Scholar Crossref Search ADS WorldCat Santini , A. , Ghelardini , L., De Pace , C., Desprez-Loustau , M.L., Capretti , P., Chandelier , A. et al. 2013 Biogeographical patterns and determinants of invasion by forest pathogens in Europe . New Phytol. 197 , 238 – 250 . doi: 10.1111/j.1469-8137.2012.04364.x . Google Scholar Crossref Search ADS PubMed WorldCat Savill , P. 2017 www.rfs.org.uk/learning/forestry-knowledge-hub/species-profiles-project/ ( accessed on 05 November 2018 ). Schimitschek , E. 1941 Die massenvermehrung des kiefernspanners, Bupalus piniarius L., und seine bekämpfung im jahre 1940 in der Westlowakei. (The mass propagation of the pine looper, Bupalus piniarius L., and its control in 1940 in western Slovakia) Zbl. Ges. Forstw. Lxvii ( 25–46 ), 53 – 59 . Google Scholar OpenURL Placeholder Text WorldCat Scott , T.M. and Brown , R.M. 1973 Insecticidal control of the pine looper moth in Great Britain: III. Aerial spraying with tetrachlorvinphos . Forestry . 46 ( 1 ), 81 – 93 . doi: 10.1093/forestry/46.1.81 . Google Scholar Crossref Search ADS WorldCat SEPA 2020 . Supporting Guidance (WAT-SG-53). Environmental Quality Standards and Standards for Discharges to Surface Waters Version v7.1 April 2020 . Scottish Environment Protection Agency . 34 Silvifuture 2017 www.silvifuture.org.uk ( accessed on 05 November 2018 ). Sly , J.M.A. and Neale , M.C. 1983 Aerial Spraying, Great Britain 1981 . ADAS Survey Report No. 24 . MAFF . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Southgate , G.J. 1969 Helicopter spraying to kill overhead cover at Lavenham Forest . J. Forest. Comm. 36 ( 1968–69 ), pp. 58 – 60 . Google Scholar OpenURL Placeholder Text WorldCat Stauber , J.L. and Davies , C.M. 2000 Use and limitations of microbial bioassays for assessing copper bioavailability in the aquatic environment . Environ. Rev. 8 , 255 – 301 . Google Scholar Crossref Search ADS WorldCat Streit , B. 1984 Effects of high copper concentrationson soil invertebrates (earthworms and oribatid mites): experimental results and a model . Oecologia (Berlin) 64 , 381 – 388 . Google Scholar Crossref Search ADS WorldCat Tubby , K.V. and Webber , J.F. 2010 Pests and diseases threatening urban trees under a changing climate . Forestry . 83 ( 4 ), 451 – 459 . doi: 10.1093/forestry/cpq027 . Google Scholar Crossref Search ADS WorldCat Turner , J.A. 2017 The Pesticide Manual 17th Edition. A World Compendium . BCPC . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC U.S. EPA. 2007 Method 3051A Microwave Assisted Acid Digestion of Sediments, Sludges, Soils and Oils . https://www.epa.gov/sites/production/files/2015-12/documents/3051a.pdf ( accessed on 12 February 2019 ).. U.S. EPA . 1994 “Method 200.7: Determination of Metals and Trace Elements in Water and Wastes by Inductively Coupled Plasma-Atomic Emission Spectrometry”, Revision 4.4, Cincinnati, OH . https://www.epa.gov/homeland-security-research/method-2007-determination-metals-and-trace-elements-water-and-wastes ( accessed on 12 December 2019 ). Van der Pas , J.B. 1981 . Reduced early growth rates of Pinus caused by Dothistroma pini . N. Z. J. For. Sci. 11 ( 3 ), 210 – 220 . Google Scholar OpenURL Placeholder Text WorldCat Van der Steen , J.J.M. , de Kraker , J. and Grotenhuis , T. 2012 Spatial and temporal variation of metal concentrations in adult honeybees (Apis mellifera L.) . Environ. Monit. Assess. 184 ( 7 ), 4119 – 4126 . Google Scholar Crossref Search ADS PubMed WorldCat Van Dobben , H.J. , Wolterbeek , H.T., Wamelink , G.W.W. and Ter Braak , C.J.F. 2001 Relationship between epiphytic lichens, trace elements and gaseous atmospheric pollutants . Environ. Pollut. 112 ( 2 ), 163 – 169 . Google Scholar Crossref Search ADS PubMed WorldCat Van Zwieten , L. , Rust , J., Kingston , T., Merrington , G. and Morris , S. 2004 . Influence of copper fungicide residues on occurrence of earthworms in avocado orchard soils . Sci. Total Environ. 329 ( 1–3 ), 29 – 41 . doi: 10.1016/j.scitotenv.2004.02.014 . Google Scholar Crossref Search ADS PubMed WorldCat Veleminsky , M. , Laznicka , P. and Stary , P. 1990 Honeybees (Apis mellifera) as environmental monitors of heavy metals in Czechoslovakia . Acta Entomol. Bohemoslov. 87 , 37 – 44 . Google Scholar OpenURL Placeholder Text WorldCat Webber , J.F. , Mullett , M. and Brasier , C.M. 2010 Dieback and mortality of plantation Japanese larch (Larix kaempferi) associated with infection by Phytophthora ramorum . New Dis. Rep. 22 , 19 . doi: 10.5197/j.2044-0588.2010.022.019 . Google Scholar Crossref Search ADS WorldCat Whyte , A.G.D. 1969 Tree growth in the presence of Dothistroma pini . Report, Forest Research Institute, New Zealand Forest Service , pp. 51 – 53 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Willoughby , I. , Evans , H., Straw , N. and Tubby , K. 2015 Aerial Application of Pesticides in Forestry . Forest Research , p. 36 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Wilson , S.M.G. Making sense of mixtures . Forestry & Timber News , February 2016, 14 – 15 . OpenURL Placeholder Text WorldCat © Institute of Chartered Foresters, 2020. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com. This article contains public sector information licensed under the Open Government Licence v3.0 (http://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/).
TI - The potential role of aerial pesticide applications to control landscape-scale outbreaks of pests and diseases in British forestry with a focus on dothistroma needle blight
JF - Forestry
DO - 10.1093/forestry/cpaa038
DA - 2020-11-02
UR - https://www.deepdyve.com/lp/oxford-university-press/the-potential-role-of-aerial-pesticide-applications-to-control-Jum0KNA0eu
SP - 1
EP - 1
VL - Advance Article
IS - 
DP - DeepDyve
ER -