Introduction
Herbicide applications using remotely piloted aerial application systems (RPAASs) are gaining popularity in several cropping systems worldwide (Martin et al. Reference Martin, Singh, Latheef and Bagavathiannan2020). However, their current use for managing weeds in turfgrass is limited due to the lack of research on techniques developed specifically for this purpose. Benefits of using RPAAS for site-specific herbicide applications on turfgrass areas (e.g., golf courses, sports fields, parks) include but are not limited to reduced product amounts and environmental impacts, the ability to spray difficult-to-reach areas, flexible operation scheduling, and reduced damage to turf that otherwise arises from foot and vehicle traffic associated with ground applications.
Site-specific herbicides can be applied when sensors detect weeds, allowing the RPAAS to apply an herbicide specifically on those weeds or infested patches (Czarnecki et al. Reference Czarnecki, Samiappan, Wasson, McCurdy, Reynolds, Williams and Moorhead2017). This scenario is especially applicable in cases of low weed densities. Site-specific herbicide application can also be a tool for managing herbicide-resistant weeds through spot applications of alternative herbicides (Hunter et al. Reference Hunter, Gannon, Richardson, Yelverton and Leon2019a; Zanin et al. Reference Zanin, Neves, Teodoro, da Silva Júnior, da Silva and Teodoro2022).
Nozzle selection is an important factor for RPAAS-based herbicide applications. Herbicides are typically applied with an RPAAS using very low volume centrifugal or hydraulic nozzles. However, this can lead to inadequate weed control with increased drift, resulting in higher herbicide deposition on nontarget areas (Kim et al. Reference Kim, Yuan, Kim, Kyung and Noh2023; Xiongkui Reference Xiongkui2018). Hoffmann and Kirk (Reference Hoffmann and Kirk2005) showed that nozzles producing higher percentages of smaller droplets (<200 µm) increased spray drift in aerial applications. Matthews et al. (Reference Matthews, Bateman, Miller, GA, R and P2016) indicated that larger droplets (>200 µm) will be deposited rapidly, with reduced potential for spray drift.
Sarghini et al. (Reference Sarghini, Visacki, Sedlar, Crimaldi, Cristiano and Vivo2019) observed that when using an RPAAS with XR11001 flat-fan nozzles delivering 10 L ha−1 of spray volume at 300 kPa pressure, only 16% of the total amount applied reached the target, suggesting that nozzles with a coarser droplet spectrum are needed to reduce spray drift. Shan et al. (Reference Shan, Wang, Wang, Xie, Wang, Wang, Chen and Lan2021), investigating RPAAS herbicide applications with halauxifen-methyl and florasulam on wheat using centrifugal atomization nozzles (spray volumes of 7.5 to 30 L ha−1 and droplet sizes of 150 to 300 µm), found that the greatest droplet density and coverage occurred with nozzle parameters providing a spray volume of 15 L ha−1 and a droplet size of 150 µm.
Spray volume is another important parameter with aerial herbicide applications. Low spray volumes are fundamentally essential to the efficient use of RPAAS for herbicide applications. In general, reductions in spray volume are desirable for improved operational capacity, which is dependent on low-flow-rate nozzles and smaller droplets (Carvalho et al. Reference Carvalho, Chechetto, Mota and Antuniassi2017; Costa and Sofiatti Reference Costa, Sofiatti, AGF and V2015). Herbicides applied to field crops with an RPAAS typically occur via spray volumes that range from 12 to 15 L ha−1. The coverage area for each flight is governed primarily by battery and spray tank capacities (Cao et al. Reference Cao, Yu, Xu, Du, Guo and Zhang2021; Kharim et al. Reference Kharim, Wayayok, Shariff, Abdullah and Husin2019); therefore, reducing the frequency of spray tank filling through the use of lower spray volumes can help overcome one of the major limiting factors.
Reducing spray volumes for RPAAS-applied herbicide treatments can increase the risk of poor applications due to extremely small droplet sizes and increased application speeds (Gibbs et al. Reference Gibbs, Peters and Heck2021). Typical speeds for ground applications range from 1 to 4 m s−1, while an RPAAS typically operates at speeds between 5 and 7 m s−1. Additionally, nozzle heights above the target typically range from 0.4 to 1.0 m for ground applications and 2 to 5 m for RPAAS applications. Thus, aerial herbicide applications can reduce herbicide deposition on the target and increase the potential for spray drift losses (Costa and Sofiatti Reference Costa, Sofiatti, AGF and V2015), eventually leading to reduced weed control. Ahmad et al. (Reference Ahmad, Qiu, Dong, Ma, Huang, Ahmed and Chandio2020) evaluated herbicide spray deposits from an RPAAS equipped with TeeJet TT110015 flat-fan nozzles on artificial targets (water-sensitive cards and glass samplers). They found that the greatest spray deposit was achieved at the lowest operational speed of 2 m s−1 (corresponding to the highest spray volume) and the lowest application height of 2 m. In this context, spray drift, deposition, coverage, and efficacy are influenced by the spectrum of droplet sizes (Costa et al. Reference Costa, Velini, Negrisoli, Carbonari, Rossi, Corrêa and Silva2007; Hunter et al. Reference Hunter, Gannon, Richardson, Yelverton and Leon2019b). In addition, spray drift and droplet distribution can also be influenced by the airflow (i.e., downwash) generated by the RPAAS rotors (Chen et al. Reference Chen, Lan, Zhou, Deng and Wang2021; Sarghini et al. Reference Sarghini, Visacki, Sedlar, Crimaldi, Cristiano and Vivo2019; Xiongkui et al. Reference Xiongkui, Bonds, Herbst and Langenakens2017).
Although RPAAS-based herbicide applications are becoming increasingly important, limited research has been conducted on the effects of nozzle selection and spray volume for site-specific herbicide applications to turf. In this research, these two RPAAS parameters were tested for the control of large crabgrass, a major weed of turf (Areces-Berazain Reference Areces-Berazain2022; UMETP 2011), with quinclorac.
Materials and Methods
Study Location
Experiments to evaluate the effects of nozzle type and spray volume on site-specific herbicide applications to turfgrass using RPAAS were conducted on June 8, 2022, at the Turfgrass Field Laboratory (30.616°N, 96.366°W), and on June 9, 2022, at the Penberthy Recreation Sports Complex (30.583°N, 96.333°W), both situated in College Station, TX. These sites are characterized by a humid subtropical climate, with a mean annual precipitation of 1080 mm and an average ambient temperature of 20.5 C (Beck et al. Reference Beck, Zimmermann, McVicar, Vergopolan, Berg and Wood2018; Anonymous 2023). Each site area was characterized by sandy loam soils and common bermudagrass [Cynodon dactylon (L.) Pers] as the primary turfgrass species. The turfgrass at each site was mown using a rotary mower set to a 5-cm cutting height at weekly intervals throughout the growing season. The Field Lab area did not have supplemental irrigation but the Penberthy area did, and neither site received routine fertilizer or pesticide applications during the trial period.
Experimental Design and Setup
The treatments were arranged in a 3 × 3 factorial design. The first factor was nozzle type, which included three flat-fan nozzles (TeeJet Technologies, Glendale Heights, IL) with different droplet classes: XR80015 conventional nozzle (fine droplet), DG80015 drift guard nozzle (medium droplet), and AI80-015 air induction nozzle (ultra-coarse droplet). The second factor was spray volume, with 10 and 15 L ha−1 applied using an RPAAS, and 102 L ha−1 applied with a CO2-pressurized, four-nozzle boom backpack sprayer. The RPAAS spray volumes were selected for higher operational capacity, based on low values commonly used for this sprayer type (Virk Reference Virk2022; Xiongkui Reference Xiongkui2018; Zhang et al. Reference Zhang, Zhang, Sun, He, Fu, Qi, Yu, Jin, Zhang and Liu2021). Two independent treatments were included: a pure herbicide (quinclorac without dilution in water) applied via the RPAAS with an XR 80005 flat-fan nozzle (very fine droplet) and an untreated control. To obtain a spray volume corresponding to the dosage for the pure herbicide application, it was necessary to use a lower-rate nozzle (4.6 L ha−1) with finer droplets than the other treatments. Nine replicates were used to evaluate droplet coverage (%) on an artificial target (Kromekote cards). Using pot-grown crabgrass plants (Figure 1), six replicates were used to assess herbicide spray solution deposition on weed plants, while another six replicates were used to evaluate weed control through visible control ratings and dry biomass production.
Figure 1. Field strips used for individual treatment applications at the Field Lab and Penberthy locations, College Station, TX, June 2022. A, B and C indicate placement of large crabgrass plants and Kromekote cards in the treatment area.
Individual treatment application areas at both locations were 1.8-m × 7-m strips that were spaced at least 2 m apart. The strips were georeferenced using an Emlid Reach real-time kinematic –enabled GPS/GNSS unit mounted on a 2-m surveying pole (Emlid Inc, Hongkong, China). Three randomly selected 50- × 50-cm areas were used in each strip. Four 0.35-L circular pots (71 cm2 surface area) filled with a horticultural substrate (Pro-Line C/20 Growing Mix; Jolly Gardner Products Inc., Atlanta, GA) containing roughly 40 greenhouse-grown large crabgrass seedlings, were placed in three areas within each strip, resulting in a total of 12 pots. Large crabgrass plants were at the vegetative stage, with 1 tiller and approximately 8 cm in height. This plant stage allowed for more efficient monitoring of the spray droplet deposition on the target weeds in the turf. Additionally, three Kromekote cards (5.0 × 11.5 cm) were fixed horizontally on 2-cm-tall wooden blocks in these three areas (Figure 1).
Herbicide Applications
The aerial treatments were applied using an RPAAS model PV35X (Leading Edge Aerial Technologies, New Smyrna Beach, FL) equipped with a single-nozzle mount on the underside of the drone body (Figure 2A). The georeferenced strip was used for the automatic navigation of RPAAS, operating at 1.1 m above the ground. The ground applications used a CO2-pressurized backpack sprayer and boom containing four nozzles spaced 50 cm apart and held 50 cm above the crabgrass plant canopy level (Figure 2B). The walking speed for the ground sprayer applications was 1.6 m s−1. The spray pressure was 207 kPa for all treatments. All spray volumes evaluated were obtained by adjusting flight or ground speed to the nozzle flow rates and spray swath of each sprayer. Spray swaths were previously confirmed with the spray height tested through water-sensitive papers under and along the spray plume of each sprayer. The speed was set through the RPAAS control software for the flights, ranging from 3 to 4.5 m s−1. The direction of displacement for both sprayers was longitudinal to each application strip (Figure 1). The spray application setup for each treatment is presented in Table 1.
Figure 2. A) The remotely piloted aerial application system (RPAAS) with a single nozzle, and B) the backpack sprayer with a boom containing four nozzles, used for the treatment applications at the Field Lab and Penberthy locations, College Station, TX, June 2022.
Table 1. Spray application setup for RPAAS and backpack sprayer treatments at the Field Lab and Penberthy locations, June 2022.
a TeeJet Technologies, Glendale Heights, IL.
b Droplet size class at 30 PSI: fine, medium, and ultra-coarse approximately ranging from 106 to 235 µm, 236 to 340 µm, and higher than 665 µm, respectively (ASABE 2009; SPRAYING SYSTEMS CO 2014).
c Spray volume corresponded to the quinclorac dose recommended for crabgrass control from a commercial product label.
The herbicide quinclorac (Drive® XLR8; BASF Corporation, Research Triangle Park, NC) was applied once at 827 g ai ha−1 for all treatments. With the exception of the pure herbicide treatment, a modified vegetable oil adjuvant (Super Spread® MSO; Wilbur-Ellis, Aurora, CO) consisting of methyl soyate and nonylphenol ethoxylate was added to all spray solutions at a rate of 1.7 L ha−1. The herbicide and adjuvant rates used were based on herbicide label recommendations (Anonymous 2019).
Fluorescent dye tracers (1,3,6,8-pyrenetetrasulfonic acid tetrasodium salt [PTSA] at 12.4 g ha−1) and rhodamine (20 ml L−1, Vision Pink; Garr Co, Converse, IN) were added to all spray mixes for spray deposition quantification on plants and Kromekote cards, respectively (Alves et al. Reference Alves, Cunha and Palladini2014; Hoffmann et al. Reference Hoffmann, Fritz and Ledebuhr2014). The herbicide, adjuvant, and PTSA dye concentrations were adjusted for each spray volume to maintain the same dose for all treatments.
Data Collection
Wind speed, temperature, and relative humidity were measured during treatment applications with a portable weather station (Vantage Pro2; Davis Instruments, Hayward, CA) set up 10 m from the experimental areas. The applications were performed from 0900 to 1230 hours at the field laboratory, and from 0715 to 1000 hours at Penberthy on June 8th and 9th 2022, respectively.
After treatment application and allowing the herbicide solution to dry on large crabgrass plant surfaces, the total shoot biomass of six crabgrass plants was randomly collected from a group of six pots per treatment, regardless of pot position. The six plant samples were pooled into a single sample to represent a replication (n = 6). Samples were placed in plastic bags and stored in insulated thermal boxes to protect them from outside temperature and light during transport to and storage in the laboratory. The plant samples were washed within 24 and 48 h after applications at the Field Lab and Penberthy locations, respectively. To remove the dye tracer deposits, 40 ml of distilled water was mixed with 91% isopropyl alcohol (9:1). The shoot samples were placed into a forced ventilation oven at 60 C until they were completely dry (72 h), and then weighed to determine the dry mass (in grams). The fluorescence in washed solutions was determined in a spectrofluorophotometer (RF-5301 PC; Shimadzu, Kyoto, Japan) through a computer interface and Hyper FR (v. 1.57, Shimadzu). The optical density was measured with emission wavelengths from 375 to 402 nm for PTSA dye quantification.
To develop a calibration curve for the PTSA recovery test, a set of six large crabgrass plants were treated with 1, 5, 10, 25, and 125 µL of spray solution, with four replications for each prepared solution. To simulate the spray drop deposits on plants, chromatography microsyringes of 10 and 200 µL were used. The spray solutions were randomly allocated at different points on the plant surfaces by droplets formed through the microsyringe needles. These samples were washed 24 and 48 h after application, and the PTSA concentration was determined following the same procedure used for the field samples.
Based on readings from a calibration curve with known concentrations of PTSA in distilled water containing 91% isopropyl alcohol (9:1), the optical density from the spray solutions used in each treatment and from the plant washing samples were converted to dye concentrations (mg L−1). From these data, the volume of spray solution deposited on the plant shoots was calculated using the equation 1:
$${{C1V1}} = {{C2V2}}$$
where C1 = PTSA concentration in the spray solution (mg L−1), V1 = volume of spray solution deposited on plants (µL), C2 = PTSA concentration in the washing solution (mg L−1), and V2 = volume used for washing (40,000 µL).
The amount of herbicide (in milligrams of active ingredient, mg ai) deposited on each plant sample was calculated based on the volume of spray solution deposited on the plants (in microliters, µL) and the quinclorac concentrations in the spray solutions (mg ai/µL) for each spray volume. Both the deposited solution volume and herbicide amount were individually divided by the plant biomass, yielding spray solution and herbicide deposition data expressed in µL g−1; and mg ai g−1, respectively. After the spray deposits dried, Kromekote cards were placed in paper bags and transported to a laboratory for analysis. Digital images were taken with a desktop scanner (Epson Perfection V39) set at 600 dpi resolution. Spray droplet coverage (%) on each card image was then measured using DepositScan (USDA-ARS, Wooster, OH).
Six pots of large crabgrass (two from each of the three pot placement areas) of each strip, which were maintained in the greenhouse and irrigated daily, were used to assess weed visible control at 21d after application (DAA). Herbicide injury was estimated on a scale of 0% (no injury) to 100% (plant death) (Frans et al. Reference Frans, Talbert, Marx, Crowley and Camper1986). The large crabgrass plants in all six pots (replications) in the potted areas of each strip were clipped to the soil surface using scissors. The plants were placed in a forced ventilation oven set at 60 C until they were completely dry (72 h) to determine the dry matter (in milligrams, mg) using an analytical balance. The values were used to calculate dry biomass reduction (%) in relation to the dry biomass of the untreated control.
Statistical Analysis
The effects of nozzle type and spray volume treatment on Kromekote card coverage, spray solution, and herbicide deposition on large crabgrass shoots, as well as dry matter reduction from treatment applications, were subjected to ANOVA using Assistat (Statistical Assistance 7.7; UFCG, Campina Grande, PB, Brazil). Treatments (nozzles and spray volumes) and their interactions were treated as fixed effects, while the locations and replications were treated as random effects. Interactions between treatments were evaluated, and when data were significant, nozzles and spray volume data were separated for analysis. Significant treatment means for nozzles and spray volumes were separated using Tukey’s test with a significance level of α = 0.05. The average values pooled across nozzles and spray volumes were used for comparisons with the additional treatments (pure herbicide and untreated control) using Dunnett’s test at α = 0.05. The Dunnett test is appropriate for comparing the means of multiple experimental groups to a single control group.
Results and Discussion
In general, the weather conditions at both trial locations were within the acceptable limits of relative humidity (>50%), wind speed (<2.8 m s−1), and air temperature (from 18 to 30 C) for herbicide application (Table 2; Antuniassi et al Reference Antuniassi, Carvalho, Mota and Chechetto2017; Carvalho et al. Reference Carvalho, Chechetto, Mota and Antuniassi2021; Crop Life International 2022; NDSU 2023). Some exceptions were observed at the Field Lab, where the maximum wind speed and air temperature values were 5.4 m s−1 and 33.9 C, while at Penberthy, these values were 4.5 m s−1 and 31.7 C, respectively (Table 2). Due to these specific weather conditions, the results are presented and discussed separately for each experimental area. Statistical significance was observed for the majority of evaluated characteristics referring to nozzles, spray volume, and their interactions. The findings and analysis are presented sequentially, examining the effects of nozzle type and spray volume on first the droplet coverage on Kromekote cards, then the amount of deposition of the spray solution and the herbicide, and finally the control efficacy and dry mass reduction of large crabgrass.
Table 2. Weather conditions during the herbicide treatment applications at each experimental location, June 2022.
Droplet Coverage on the Kromekote Cards
Overall, comparing spray volumes by nozzle type at both locations determined that coverage was greater with 102 L ha−1 than 15 or 10 L ha−1 (Table 3). The DG nozzle provided the greatest Kromekote card coverage at the Penberthy location with spray volumes of 15 and 102 L ha−1, and at the Field Lab location with a spray volume of 102 L ha−1. For the 10 L ha−1 spray volume at the Field Lab, the AI nozzle provided the highest coverage (7.8%) compared to the XR (4.7%) or DG (4.6%) nozzle types. The AI nozzle produced the lowest coverage at the Field Lab with spray volumes of 15 and 102 L ha−1 (5.7% and 27.8%, respectively) and at Penberthy with 102 L ha−1 (28.6%), compared to the DG and XR nozzles. The greater spray coverage observed for the XR and DG nozzles at both locations with 15 and 102 L ha−1 can be partially attributed to the droplet size generated by the nozzles at 207 kPa. The XR, DG, and AI nozzles produce fine, medium, and ultra-coarse droplets, respectively, with approximate volume mean diameter values ranging from 106 to 235 µm, 236 to 340 µm, and higher than 665 µm, respectively (ASABE 2009; SPRAYING SYSTEMS CO 2014). Generally, smaller droplet sizes tend to provide greater coverage of the sprayed surface (Shan et al. Reference Shan, Wang, Wang, Xie, Wang, Wang, Chen and Lan2021). The pure herbicide treatment applications using the RPAAS provided the lowest coverage at both sites, registering 1.7% at the Field Lab and 0.6% at Penberthy. This was likely because of the minimal spray volume (4.6 L ha−1) applied in this treatment. The results from the current study further validate that finer droplets and greater spray volumes contribute to greater spray coverages. Furthermore, the absence of an adjuvant in the undiluted herbicide treatment may also have diminished droplet spreading on the Kromekote surface, in turn contributing to the lowest spray coverages observed.
Table 3. Nozzle type and spray volume effects on spray droplet coverage (%) on Kromekote cards at the Field Lab and Penberthy locations, June 2022.h
a For each location, means followed by the same upper case letter in each row and lower case letter in each column do not differ based on Tukey’s HSD test at α = 0.05.
b Nozzle types used in the study (all from TeeJet Technologies, Glendale Heights, IL) included XR-80015 (conventional, fine droplet), DG-80015 (drift guard, medium droplet), and AI-80015 (air induction, ultra-coarse droplet).
c The spray volumes of 10 and 15 L ha−1 corresponded to the RPAAS applications, and 102 L ha−1 to the backpack sprayer application.
d For each location, means followed by the same letter in each column do not differ based on the Dunnett’s test at α = 0.05.
e Averages pooled across nozzle types and spray volumes.
f Spray volume corresponded to commercial herbicide dosage: 4.6 L ha−1.
g Significance is noted as follows: ns, nonsignificant; **, significant at α = 0.01; *, significant at α = 0.05.
h Abbreviations: C, control; F, calculated F-value; N, nozzle type; NxSV, interaction between nozzle type and spray volume; N×SV × PH+C, nozzle type and spray volume contrasted to pure herbicide and control; PH × C, pure herbicide contrasted to control; PH, pure herbicide; RPAAS, remotely piloted aerial application system; SV, spray volume.
Amount of Spray Solution and Herbicide Deposited on Large Crabgrass
The highest spray application volume (102 L ha−1) resulted in the largest spray solution deposits on large crabgrass from each nozzle at the Field Lab, and regardless of the nozzle type at Penberthy (Table 4). However, considering the nozzle effect, the DG nozzle at 15 L ha−1 provided the greatest herbicide deposition at the Penberthy location (Table 5). Given that wind is an important factor that affects spray drift and, consequently, droplet deposition (Baio et al. Reference Baio, Antuniassi, Castilho, Teodoro and Silva2019; Lan et al. Reference Lan, Qian, Shengde, Zhao, Deng, Wang, Zang, Wang and Qiu2021; Shengde et al. Reference Shengde, Lan, Jiyu, Zhiyan, Aimin and Yuedong2017; Wang et al. Reference Wang, Xu, Li and Cao2022), the lowest spray solution and herbicide deposits obtained with the AI nozzle compared to the other nozzles at 102 L ha−1 at the Field Lab (Tables 4 and 5, respectively) can also be related to strong wind gusts during that treatment at that location. However, based on results from RPAAS applications using different nozzle types in a wind tunnel system, under distinct simulated flight speeds and weather conditions, Wang et al. (Reference Wang, Zeng, He, Song, Herbst and Gao2020a) recommended substituting conventional flat-fan or hollow-cone nozzles with air-induction nozzles to reduce spray drift.
Table 4. Nozzle type and spray volume effects on the amount of spray solution deposited on large crabgrass at the Field Lab and Penberthy locations, June 2022.h
a For each location, means followed by the same upper case letter in each row and lower case letter in each column do not differ based on Tukey’s HSD test at α = 0.05.
b Nozzle types used in the study (all from TeeJet Technologies, Glendale Heights, IL) included XR-80015 (conventional, fine droplet), DG-80015 (drift guard, medium droplet), and AI-80015 (air induction, ultra-coarse droplet).
c The spray volumes of 10 and 15 L ha−1 corresponded to the RPAAS applications, and 102 L ha−1 to the backpack sprayer application.
d For each location, means followed by the same letter in each column do not differ based on the Dunnett’s test at α = 0.05.
e Averages were pooled across nozzle types and spray volumes.
f Spray volume corresponded to commercial herbicide dosage: 4.6 L ha−1.
g Significance is noted as follows: ns, nonsignificant; **, significant at α = 0.01; *, significant at α = 0.05.
h Abbreviations: C, control; F, calculated F-value; N, nozzle type; N×SV, interaction between nozzle type and spray volume; N×SV × PH+C, nozzle type and spray volume contrasted to pure herbicide and control; PH × C, pure herbicide contrasted to control; PH, pure herbicide; RPAAS, remotely piloted aerial application system; SV, spray volume.
Table 5. Nozzle type and spray volume effects on the amount of herbicide deposited on large crabgrass at Field Lab and Penberthy locations, June 2022.h
a For each location, means followed by the same upper case letter in each row and lower case letter in each column do not differ based on Tukey’s HSD test at α = 0.05
b Nozzle types used in the study (all from TeeJet Technologies, Glendale Heights, IL): XR-80015 (conventional, fine droplet), DG-80015 (drift guard, medium droplet), and AI-80015 (air induction, ultra-coarse droplet).
c The spray volumes of 10 and 15 L ha−1 corresponded to the RPAAS applications, and 102 L ha−1 to the backpack sprayer application.
d For each location, means followed by the same letter in each column do not differ based on the Dunnett’s test at α = 0.05.
e Averages pooled across nozzle types and spray volumes.
f Spray volume corresponded to commercial herbicide dosage: 4.6 L ha−1.
g Significance is noted as follows: ns, nonsignificant; **, significant at α = 0.01; *, significant at α = 0.05.
h Abbreviations: C, control; F, calculated F-value; N, nozzle type; N×SV, interaction between nozzle type and spray volume; N×SV × PH+C, nozzle type and spray volume contrasted to pure herbicide and control; PH × C, pure herbicide contrasted to control; PH, pure herbicide; RPAAS, remotely piloted aerial application system; SV, spray volume.
The spray volume of 102 L ha−1 resulted in the highest spray solution deposition values at both experimental locations (Table 4). Ahmad et al. (Reference Ahmad, Qiu, Dong, Ma, Huang, Ahmed and Chandio2020) tested RPAAS herbicide deposition using a TT110015 nozzle on tall fleabane and documented the greatest spray deposits with the highest spray volume applied at the lowest ground speed tested. Our results also support the findings reported by Ahmad et al. (Reference Ahmad, Qiu, Dong, Ma, Huang, Ahmed and Chandio2020) in that the highest deposition on large crabgrass was obtained with the highest spray volume (102 L ha−1). For the pure herbicide treatment, the spray solution and herbicide deposition observed were lower than those of the other RPAAS and backpack sprayer treatments (Tables 4 and 5, respectively).
The greatest herbicide deposits were observed at 102 L ha−1 using the XR and DG nozzles at the Field Lab (Table 5). In a separate study, RPAAS applications at 18.7 and 37.4 L ha−1 resulted in lower coverage and droplet density on the adaxial leaf surface of the weeds Palmer amaranth (Amaranthus palmeri S. Watson) and ivyleaf morningglory (Ipomoea hederacea Jacq.) compared to backpack sprayer applications at 140 L ha−1 (Martin et al. Reference Martin, Singh, Latheef and Bagavathiannan2020). Therefore, the highest spray volume applied with the backpack sprayer (102 L ha−1) generally resulted in greater herbicide deposits on large crabgrass compared to volume with the RPAAS (10 and 15 L ha−1) (Table 4). It is worth noting that Palmer amaranth and ivyleaf morningglory are not typically recognized as common weed species in turfgrass systems.
Increasing spray deposition through higher spray volumes is a known strategy in agricultural pesticide applications, for both ground and aerial sprays (Costa et al. Reference Costa, Maciel, Rogoski, Lira and Helvig2019; Wang et al. Reference Wang, Lan, Qi, Chen, Hewitt and Han2019, Reference Wang, Zeng, He, Song, Herbst and Gao2020a, Reference Wang, Li, Andaloro, Chen, Song, Shan, Chen and Lan2020b), but RPAAS spray volumes are limited due to spray tank capacity and battery life (Cao et al. Reference Cao, Yu, Xu, Du, Guo and Zhang2021; Kharim et al. Reference Kharim, Wayayok, Shariff, Abdullah and Husin2019). Consequently, despite drift potential, there is a general tendency to use nozzles with finer spray droplet sizes to gain more coverage with RPAAS. The greater potential for spray losses with RPAAS compared to ground sprayers is likely related to the higher application heights (Wang et al. Reference Wang, Zeng, He, Song, Herbst and Gao2020a). The rotor downwash, design, and power ratings of each RPAAS influence the performance and efficiency of these aerial application systems, affecting spray pattern uniformity, effective swath width, canopy penetration, droplet deposition, and overall coverage (Gong et al. Reference Gong, Fan and Peng2019; Martin et al. Reference Martin, Woldt and Latheef2019; Xiongkui et al. Reference Xiongkui, Bonds, Herbst and Langenakens2017).
Control Efficacy and Dry Mass Reduction of Large Crabgrass
The highest levels of visible weed control were obtained with the DG and XR nozzles (70.2% and 71.3%, respectively) for the 102 L ha−1 application rate at the Penberthy location (Table 6). Part of this effect was confirmed for the XR nozzle by the highest reduction in dry mass (81.3%) at this same spray volume (Table 7). However, the greatest control levels (from 62.0% to 65.2%) and biomass reductions (from 58.2% to 68.8%) were observed with RPAAS applications (10 and 15 L ha−1) using DG and AI nozzles. These results were perhaps related to the greatest herbicide deposits observed in the same experimental area, especially with the DG nozzle (Table 5). No statistically significant differences were observed for nozzle type or spray volume in terms of visible control or dry mass reduction of large crabgrass at the Field Lab.
Table 6. Nozzle type and spray volume effects on visible control of large crabgrass at the Field Lab and Penberthy locations, June 2022. g
a For each location, means followed by the same upper case letter in each row and lower case letter in each column do not differ based on Tukey’s HSD test at α = 0.05.
b Nozzle types used in the study (all from TeeJet Technologies, Glendale Heights, IL) included XR-80015 (conventional, fine droplet), DG-80015 (drift guard, medium droplet), and AI-80015 (air induction, ultra-coarse droplet).
c The spray volumes of 10 and 15 L ha−1 corresponded to the RPAAS applications, and 102 L ha−1 to the backpack sprayer application.
d For each location, means followed by the same letter in each column do not differ based on the Dunnett’s test at α = 0.05.
e Averages pooled across nozzle types and spray volumes.
f Spray volume corresponded to commercial herbicide dosage: 4.6 L ha−1.
g Abbreviations: C, control; F, calculated F-value; N, nozzle type; N×SV, interaction between nozzle type and spray volume; N×SV × PH+C, nozzle type and spray volume contrasted to pure herbicide and control; PH × C, pure herbicide contrasted to control; PH, pure herbicide; RPAAS, remotely piloted aerial application system; SV, spray volume.
Table 7. Nozzle type and spray volume effects on dry mass reduction of large crabgrass at Field Lab and Penberthy locations, June 2022. g
a For each location, means followed by the same upper case letter in each row and lower case letter in each column do not differ based on Tukey’s HSD test at α = 0.05.
b Nozzle types used in the study (all from TeeJet Technologies, Glendale Heights, IL) included XR-80015 (conventional, fine droplet), DG-80015 (drift guard, medium droplet), and AI-80015 (air induction, ultra-coarse droplet).
c The spray volumes of 10 and 15 L ha−1 corresponded to the RPAAS applications, and 102 L ha−1 to the backpack sprayer application.
d For each location, means followed by the same letter in each column do not differ based on the Dunnett’s test at α = 0.05.
e Averages pooled across nozzle types and spray volumes.
f Spray volume corresponded to commercial herbicide dosage: 4.6 L ha−1.
g Abbreviations: C, control; F, calculated F-value; N, nozzle type; N×SV, interaction between nozzle type and spray volume; N×SV × PH+C, nozzle type and spray volume contrasted to pure herbicide and control; PH × C, pure herbicide contrasted to control; PH, pure herbicide; RPAAS, remotely piloted aerial application system; SV, spray volume.
Nangle et al. (Reference Nangle, Raudenbush, Morris and Fidanza2021) found that in turfgrass sites with a heavy large crabgrass population (40% of covered area), one postemergence application of quinclorac is best optimized with spray nozzles that produce spray droplets from very coarse (401–500 μm) to medium (226–325 μm) sizes. This information also partially corroborates the visible control and dry mass reductions observed for RPAAS applications (10 and 15 L ha−1) at the Penberthy location, where the highest control and dry biomass reduction percentages were obtained with the DG and AI nozzles (medium and ultra-coarse droplet sizes, respectively).
For the pure herbicide treatment applied by RPAAS at the Penberthy area, the result for visible crabgrass control (12.7%) and dry mass reduction (5.9%) was lower than the other herbicide treatments (54.2% and 53.8%, respectively) (Tables 6 and 7). However, this treatment with pure herbicide was similar to the other treatments at the Field Lab for both evaluations. Therefore, the lowest spray or herbicide deposits verified for pure herbicide treatment (Tables 4 and 5, respectively) did not necessarily result in poor control of large crabgrass. One possible hypothesis to explain the similar biological efficacy observed in this study is the higher concentration of quinclorac in the deposited droplets for the pure herbicide treatment compared to the other treatments.
A dose of 827 g ha−1; across all treatments resulted in quinclorac concentrations of 180, 83, 55, and 8 g L−1; in the spray volumes of 4.6 L ha−1; (pure herbicide treatment), 10, 15, and 102 L ha−1, respectively. Thus, the amount of quinclorac in the pure herbicide treatment was 2.2, 3.3, and 22.2 times more concentrated than it was in the spray solutions for the treatments with 10, 15, and 102 L ha−1, respectively. Others researchers have observed that triclopyr applied at 0.81 kg ha−1 or more provided higher herbicide concentrations to achieve good (75%) control of blue violet (Viola sororia Willd.) in turfgrass (Patton et al. Reference Patton, Weisenberger and Wenmew2020).
Pooled across the nozzle type and spray volumes, the visible control level was 66% at the Field Lab and 54.2% at the Penberthy location, and these values are lower than acceptable levels for a field application scenario (CWSS 2018; EWRC 1964; Nangle et al. Reference Nangle, Raudenbush, Morris and Fidanza2021). Along with targeting the appropriate weed stage, most label recommendations, including that explaining quinclorac dosage and adjuvant use, were followed (Anonymous 2019). The suboptimal control levels generally observed with quinclorac in this study could not be fully explained; however, the control levels achieved were sufficient to compare the treatments and reveal relative differences.
Shan et al. (Reference Shan, Wang, Wang, Xie, Wang, Wang, Chen and Lan2021), who evaluated the use of halauxifen-methyl and florasulam with an RPAAS on wheat, obtained more spray coverage and droplet density when spray volume was increased and droplet size were reduced from 7.5 to 30 L ha−1 and from 150 to 300 µm, respectively. However, droplet size class by nozzle type variabilities can be more robust for systemic herbicides (such as quinclorac) than for contact herbicides, which require good coverage to achieve high weed control efficacy (Costa and Sofiatti Reference Costa, Sofiatti, AGF and V2015; Prokop and Veverka Reference Prokop and Veverka2003).
The differences in deposition of the spray solution and the herbicide for the nozzles and spray volumes tested here did not directly correspond with the observed levels of visible crabgrass control and biomass reduction with quinclorac. According to a report by Martin et al. (Reference Martin, Singh, Latheef and Bagavathiannan2020), downwash and wind turbulence generated by the RPAAS can cause leaf fluttering and improve spray deposition, which may help increase the efficacy of contact herbicides. The research reported by these authors also demonstrated that an RPAAS may be effectively used for herbicide applications in place of conventional ground/backpack sprayers.
Overall, this research demonstrates that RPAAS-based applications of quinclorac using DG and AI nozzles at spray volumes of 10 L ha−1; or higher can achieve crabgrass control comparable to that of backpack/ground applications at 102 L ha−1, highlighting the potential of RPAAS for precision weed control in turfgrass systems. However, several questions and challenges remain to be addressed. Future research should explore new combinations of factors, including herbicide type (contact vs. systemic), different weed species, and varying environmental conditions. For RPAAS settings and characteristics, factors such as spray volume, ground speed, application height, nozzle type, and embedded technologies that can affect herbicide application quality should also be considered in future studies. There are also some indications in this research that certain herbicides can be sprayed using RPAAS without the need for mixing in water, but the response could be influenced by various factors, which require further experimentation.
Practical Implications
Considering the various factors that can influence RPAAS-based herbicide applications, the documentation of comparable efficacy levels between RPAAS and ground applications in this study is significant and offers promise for further development of this technology for site-specific herbicide applications to turf. This is particularly relevant for RPAAS platforms equipped with flat-fan nozzles and spray volumes of 10 L ha−1 or greater. Soon, such application platforms could assist turf managers in reducing herbicide usage in turfgrass systems by allowing applications to be explicitly directed at weed-infested areas.
Acknowledgments
the Texas A&M Weed Science Research and Turfgrass Research teams for their technical and logistical support.
Funding statement
Author A.G.F. Costa was awarded grant No. 2020/03807-6 from the São Paulo Research Foundation (FAPESP) to support his research visit to Dr. Bagavathiannan’s research group at Texas A&M University. This study was also funded in part by PBI-Gordon Corporation.
Competing interests
The authors declare they have no competing interests.
Open access
