Name: Assessing the Repellency of an Insecticide-Fungicide Combination to Bees Using Bioacoustics (Video) | JoVE Journal
Uploaded: 2026-07-03T13:00:00.000+00:00
Duration: 4 min 57 s
Description: Explore bioacoustic monitoring of bee activity in soybean fields using machine learning analysis. This method helps researchers assess pesticide impacts on pollinators and inform sustainable agricultural practices for crop protection.

ai_agent_IP

ai_agent

maintenance

instantTrialDurationMinutes

isBiopharma

enableNewMyJoveSectionDesign

educationFeatures

seoMetaApi

ai_copilot_institution_blacklist

ai_chat_institution_blacklist

freeAccessWindow

ssoRoleCapture

scheduledMaintenanceBanner

joveLabsFeatures

disable_autoplay_for_bots

extended_trial_duration

5 minutes

playerTrialDesktop

The authors thank Allen Geyer from Ohio State University for field planting, management, and harvesting&#160;and Arthur Rodrigues de Souza, James Underwood, and Allison Davis for assistance with field work.&#160; This work was funded by the Agriculture and Food Research Initiative of the US Department of Agriculture&#8217;s National Institute of Food and Agriculture under grant 2022-67019-36437 and The Ohio State University College of Food, Agricultural, and Environmental Science&#8217;s Internal Grant Program (Project # 2023-013). Research support was provided by state and federal funds appropriated to the Ohio State University, College of Food, Agricultural, and Environmental Sciences, Ohio Experiment Station (OHO01277; OHO01355-MRF).

Experimental design and plot preparation
The experiment was conducted in July and early August of 2023 and 2024. Typically, the soybean growing season in Ohio roughly takes place from April to November. The sites used during the 2023 growing season were located in Clinton County (39.488, -83.769) and Union County (40.296, -83.283), Ohio. The sites used during the 2024 growing season were located in Clinton County (39.489, -83.762), Preble County (39.893, -84.611), and Licking County (39.948, -82.501), Ohio.
The experiment employed a randomized complete block design to control for spatial effects within the soybean field. Within each block, 5 plots were designated for a specific treatment. The treatments included two growth stages (R3 and R5) and two application timings (mid-morning and mid-day), and an untreated control. This resulted in the following 4 pesticide applications made on plots within each block: R3 mid-morning, R3 mid-day, R5 mid-morning, R5 mid-day. Applications during the mid-morning or mid-day occurred within a ~2 h window, with mid-morning applications approximately centered around 10:00 AM and mid-day applications approximately centered around 12:30 PM. In addition to the 5 treatment plots within each block, 3 plots were designated as spares that could be switched into the treatment plots in case of poor soybean establishment or growth. Soybean plant stages were determined with the help of staging guides such as the one published by the University of Minnesota Extension39.
Blocks of soybean plots were spaced about 1.016 m (40 in) apart to allow researchers to walk easily between them. Soybean plots were 1.9 m by 8.5 m (0.0016 ha; 6.25 ft by 28 ft), and the soybean variety used for all sites was Pioneer P35T155E (Corteva, Indianapolis, IN). 6 rows were planted in each plot and were spaced 38 cm (15 in) apart. A seeding rate of 370,658 seeds per ha (150,000 seeds per ac) was used. Bees involved in the experiment were all from local populations. No supplemental bees were added by researchers. The authors had no knowledge of either stationary or migratory apiaries near the experiment sites. All materials used for the experiment are listed in the Table of Materials.
Pesticide preparation
The pesticides used were Fitness fungicide (41.8% propiconazole active ingredient) and Fastac EC insecticide (10.9% alpha-cypermethrin active ingredient). Pesticide application rates were 6 fl oz/ac (438.47 mL/ha) for the SBI fungicide and 3.8 fl oz/ac (277.70 mL/ha) for the pyrethroid insecticide. To prepare the pesticides for application, 9.375 mL of SBI fungicide and 5.937 mL of pyrethroid insecticide were mixed with water to achieve a 3 L final volume as the tank mixture. Chemical-resistant gloves were worn while preparing the mixture. The final concentration of SBI fungicide in the tank was 3.125 mL per L. The final concentration of pyrethroid insecticide in the tank was 1.979 mL per L.
Pesticide application
This mixture was applied to individual soybean plots within each block using a custom-made handheld backpack sprayer. The pesticide mixture was applied at the top of the soybean canopy, at a speed of ~5 km per hour. After starting the pesticide application process for a given treatment, it took ~10 min for all plots to receive the application. For each 3 L tank, an area of ~0.027 ha can be covered with the pesticide mixture. The spray volume consumed per hectare is approximately 113 L. This means that ~0.18 L of the tank mixture was applied per plot. To avoid pesticide drift to other plots due to wind, applications were avoided on windy days. Full personal protective equipment (long-sleeved shirt and long pants, chemical-resistant gloves, shoes plus socks, and protective eyewear) was worn while applying the pesticide. If there was excess pesticide mixture after the conclusion of a pesticide application, it was disposed of according to recommendations from the state environmental control agency.
Recorder setup/use
The general setup and use of audio recorders is similar to a protocol described by other authors40. Handheld audio recorders (Sony ICD-PX370) were fastened to plastic step-in electric fence stakes with hook-and-loop tape, and open-cell foam was placed over the microphone as a windscreen. A cone-shaped 3-D printed rain cover was affixed to the stake ~3 cm above the recorder to protect it from weather.
Audio recorders were then placed in the center of each plot ~24 h before the first pesticide application and left running for at least 24 h after each application. Twelve recorders were deployed for each application day, four for each time-of-day group (mid-morning, mid-day, control) in both R3 and R5.
Audio data processing
All audio recordings were processed through the machine learning tool “buzzdetect”41 which can be found at: https://github.com/OSU-Bee-Lab/buzzdetect. This tool has been trained to distinguish insect buzzing noises from other sounds that frequently occur in the experimental environment, such as cars, combines, or airplanes. The program is not capable of distinguishing Apis mellifera buzzing from other bee species. However, buzzDetect is capable of distinguishing buzzing noises from some other non-buzzing insect noises, though specific noises (such as cricket calls, mostly during nighttime) are still hard to distinguish. More specific instructions for how to run buzzDetect can be found at: https://buzzdetect.readthedocs.io/en/latest/gui/. After audio analysis, R42 was used to calculate the detection rate within 1-hour and 8-hour time intervals.
Defoliation data
Defoliation assessments were performed by visually assessing leaves from the upper soybean canopy on 10 plants in each plot involved in the experiment, including all treatments and untreated control plots as recommended by the University of Nebraska Extension43. These assessments were done 24 h after the pesticide application during the R5 growth phase.
Yield data
Soybeans were harvested, and yield was measured in bushels per acre using a plot combine and a grain gauge weighing system. Yield measurements were adjusted to 13% moisture. Plots had 6 rows each; only the 4 center rows were harvested for yield measurements.
Statistical analysis
All statistics were performed in R42. The data were analyzed with either mixed effects linear models via the lme4 package44 or beta regression via the glmmTMB package45. The raw data from BuzzDetect are available in the supplementary materials. Three main types of data were analyzed for this study. Average bee activity data over the course of 1 h and 8 h following a pesticide application, the yield of each plot of soybeans, and the defoliation of soybeans observed at the R5 growth stage.
Average bee activity was obtained by calculating the proportion of time positive for bee buzzing in either an 8-hour or 1-hour window. The 1-hour window is intended to estimate the acute effects of pesticide application on bee activity, while the 8-hour window is intended to estimate the acute and residual effects. Residual toxicity data are not available for alpha-cypermethrin, but residual toxicity for other pyrethroids has been reported to be ~8 h after initial application46. Because the calculation for bee activity is a proportion, differences in activity were analyzed using beta regression conducted with the R package “glmmTMB”. The models examined average detection rates for either 1 h or 8 h following a pesticide application and how they varied when comparing treated plots to control plots (the “sprayed” condition), soybean growth stage, time of day the pesticide was applied, or a growth stage and time of day interaction term. The comparison between treated and control plots is referenced by the variable “sprayed” in both the 1-hour and 8-hour model formulas.
In the 1-hour window model, the fixed effects were the “sprayed” condition, growth stage, time of day, and a growth stage and time of day interaction term. The model had both a year and a year by site interaction term fitted as random effects. The complete model formula was:
<img alt="Statistical model equation for data analysis using glmmTMB in ecological research." src="/files/ftp_upload/70440/70440eq1v4.jpg" style="font-size: 14px;" data-altupdatedat="1791864094000" data-alt="Equation 1">  (1)
The variable detection rate reflects the average bee detections that were observed relative to the total observation window—in this case, 1 h. The variable sprayed corresponds to whether a plot received a pesticide application or if it was an untreated control plot. The variable time_of_day corresponds to whether the detections were captured after the mid-morning or the mid-day pesticide applications for the desired time frame. The variable stage corresponds to the growth stage in which the detections were captured.
In the 8-hour window model, the fixed effects were the growth stage and the “sprayed” condition. The 8-hour model includes nighttime hours for the mid-day spray time, which makes for an unfair comparison between the mid-morning and mid-day spray times. This is because the mid-day application times do not test for the full 8 h of residual toxicity due to bees being inactive during the evenings, a large component of these 8-hour windows. Therefore, mid-day applications were excluded from the model, as were the “time of day” and “time of day:stage”. terms. The complete formula for this model was:
<img alt="Generalized linear mixed model equation, glmmTMB, showcasing detection rate factors." src="/files/ftp_upload/70440/70440eq2v2.jpg" data-altupdatedat="1791864094000" data-alt="Equation 2">     (2)
The variables detection rate, sprayed, and stage have the same meanings in this formula as they do in the formula used for the 1-hour windows. The only difference is that the variable detection rate is calculated for an 8-hour window. Yield data were analyzed with a mixed-effects linear model. The model looked at bushels per acre for each plot and how the yield varied by treatment. The fixed effect in the model was the treatment, and site and year were fitted as random effects. The complete model formula is written as:
<img alt="Linear mixed-effects model formula: yield ~ treatment + (1|site) + (1|year of spray) equation." src="/files/ftp_upload/70440/70440eq3v2.jpg" data-altupdatedat="1791864094000" data-alt="Equation 3">      (3)
Soybean leaf defoliation was measured as the average percentage of missing leaf surface area per plot. A mixed effects linear model was used to compare the effect of time of day of the pesticide application on the defoliation percentage. The fixed effect in the model was the time of day that the plots were sprayed. Site and year were fitted as random effects. The complete model formula is:
 <img alt="Linear regression equation; defoliation vs. time, site, spray; statistical analysis." src="/files/ftp_upload/70440/70440eq4v2.jpg" data-altupdatedat="1791864094000" data-alt="Equation 4">    (4)
All the protocols used in this study are compiled in Supplementary File 1.

<ol><li>Raine, N. E., Rundlöf, M. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Pesticide+Exposure+and+Effects+on+Non-Apis+Bees%5BTitle%5D)+AND+%22Annual+Review+of+Entomology%22%5BJournal%5D)">Pesticide Exposure and Effects on Non-Apis Bees.</a> Annual Review of Entomology. , (2024).</li>
<li>Vanbergen, A. J. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Initiative+the+IP.+Threats+to+an+ecosystem+service%3A+pressures+on+pollinators%5BTitle%5D)+AND+%22Frontiers+in+Ecology+and+the+Environment%22%5BJournal%5D)">Initiative the IP. Threats to an ecosystem service: pressures on pollinators.</a> Frontiers in Ecology and the Environment. , (2013).</li>
<li>Zhao, G. P., Yang, F. W., Li, J. W., Xing, H. Z., Ren, F. Z., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Toxicities+of+Neonicotinoid-Containing+Pesticide+Mixtures+on+Nontarget+Organisms%5BTitle%5D)+AND+%22Environmental+Toxicology+and+Chemistry%22%5BJournal%5D)">Toxicities of Neonicotinoid-Containing Pesticide Mixtures on Nontarget Organisms.</a> Environmental Toxicology and Chemistry. , (2020).</li>
<li>Andrews, H., Rose, M. A. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Ohio+State+University+Extension%5BTitle%5D)+AND+%22Protecting+Pollinators+While+Using+Pesticides%22%5BJournal%5D)">Ohio State University Extension.</a> Protecting Pollinators While Using Pesticides. , (2018).</li>
<li><a target="_blank" href="http://scholar.google.com/scholar?hl=en&safe=off&q=%22U.S.+Environmental+Protection+Agency%E2%80%99s+Policy+to+Mitigate+the+Acute+Risk+to+Bees+from+Pesticide+Products%22">U.S. Environmental Protection Agency’s Policy to Mitigate the Acute Risk to Bees from Pesticide Products</a>. , United States Environmental Protection Agency. (2017).</li>
<li>Lydy, M., Belden, J., Wheelock, C., Hammock, B., Denton, D. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Challenges+in+Regulating+Pesticide+Mixtures%5BTitle%5D)+AND+%22Ecology+and+Society%22%5BJournal%5D)">Challenges in Regulating Pesticide Mixtures.</a> Ecology and Society. , (2004).</li>
<li>Gandini, E. M. M., Costa, E. S. P., dos Santos, J. B., Soares, M. A., Barroso, G. M., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Compatibility+of+pesticides+and%2For+fertilizers+in+tank+mixtures%5BTitle%5D)+AND+%22Journal+of+Cleaner+Production%22%5BJournal%5D)">Compatibility of pesticides and/or fertilizers in tank mixtures.</a> Journal of Cleaner Production. , (2020).</li>
<li>Fishel, F. M. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Tank-Mixing+Pesticides+without+Disasters%5BTitle%5D)+AND+%22EDIS%22%5BJournal%5D)">Tank-Mixing Pesticides without Disasters.</a> EDIS. , (2020).</li>
<li>Tosi, S., Sfeir, C., Carnesecchi, E., vanEngelsdorp, D., Chauzat, M. P. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Lethal%2C+sublethal%2C+and+combined+effects+of+pesticides+on+bees%3A+a+meta-analysis+and+new+risk+assessment+tools%5BTitle%5D)+AND+%22Science+of+The+Total+Environment%22%5BJournal%5D)">Lethal, sublethal, and combined effects of pesticides on bees: a meta-analysis and new risk assessment tools.</a> Science of The Total Environment. , (2022).</li>
<li>Pal, V., Singh, G., Dhaliwal, S. S. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Effects+of+Sequential+and+Tank+Mix+Applications+of+Zinc%2C+Iron+and+Nitrogen+on+Symbiotic+Parameters%2C+Productivity+and+Economics+of+Chickpea+%28Cicer+arietinum+L.%29+Under+Field+Conditions%5BTitle%5D)+AND+%22J+Soil+Sci+Plant+Nutr%22%5BJournal%5D)">Effects of Sequential and Tank Mix Applications of Zinc, Iron and Nitrogen on Symbiotic Parameters, Productivity and Economics of Chickpea (Cicer arietinum L.) Under Field Conditions.</a> J Soil Sci Plant Nutr. , (2023).</li>
<li>Kammler, K. J., Walters, S. A., Young, B. G. <a target="_blank" href="http://scholar.google.com/scholar?hl=en&safe=off&q=author%3aKammler%2C+KJ+%22Halosulfuron+Tank+Mixtures+and+Adjuvants+for+Weed+Control+in+Pumpkin+Production%22">Halosulfuron Tank Mixtures and Adjuvants for Weed Control in Pumpkin Production</a>. , (2008).</li>
<li>Shannon, B., Walker, E., Johnson, R. M. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Toxicity+of+spray+adjuvants+and+tank+mix+combinations+used+in+almond+orchards+to+adult+honey+bees+%28Apis+mellifera%29%5BTitle%5D)+AND+%22Journal+of+Economic+Entomology%22%5BJournal%5D)">Toxicity of spray adjuvants and tank mix combinations used in almond orchards to adult honey bees (Apis mellifera).</a> Journal of Economic Entomology. , (2023).</li>
<li>Younesabadi, M., Das, T. K., Sharma, A. R. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Effect+of+tillage+and+tank-mix+herbicide+application+on+weed+management+in+soybean+%28Glycine+max.%29%5BTitle%5D)+AND+%22IJA%22%5BJournal%5D)">Effect of tillage and tank-mix herbicide application on weed management in soybean (Glycine max.).</a> IJA. , (2013).</li>
<li>Bruff, S. A., Shaw, D. R. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Tank-mix+Combinations+for+Weed+Control+in+Stale+Seedbed+Soybean+%28Glycine+max.%29%5BTitle%5D)+AND+%22Weed+Technology%22%5BJournal%5D)">Tank-mix Combinations for Weed Control in Stale Seedbed Soybean (Glycine max.).</a> Weed Technology. , (2017).</li>
<li>Gill, K. A., O’Neal, M. E. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Survey+of+Soybean+Insect+Pollinators%3A+Community+Identification+and+Sampling+Method+Analysis%5BTitle%5D)+AND+%22Environ+Entomol%22%5BJournal%5D)">Survey of Soybean Insect Pollinators: Community Identification and Sampling Method Analysis.</a> Environ Entomol. , (2015).</li>
<li>Lin, C. H., Suresh, S., Matcham, E., Monagan, P., Curtis, H., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Soybean+is+a+Common+Nectar+Source+for+Honey+Bees+%28Hymenoptera%3A+Apidae%29+in+a+Midwestern+Agricultural+Landscape%5BTitle%5D)+AND+%22Journal+of+Economic+Entomology%22%5BJournal%5D)">Soybean is a Common Nectar Source for Honey Bees (Hymenoptera: Apidae) in a Midwestern Agricultural Landscape.</a> Journal of Economic Entomology. , (2022).</li>
<li>Rondeau, S., Bosanac, L. F., Raine, N. E., Forrest, J. R. K. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Wild+bees+and+wasps+%28Hymenoptera%29+visiting+soybean+%28Fabaceae%29+flowers+in+eastern+Canada%5BTitle%5D)+AND+%22The+Canadian+Entomologist%22%5BJournal%5D)">Wild bees and wasps (Hymenoptera) visiting soybean (Fabaceae) flowers in eastern Canada.</a> The Canadian Entomologist. , (2025).</li>
<li>Robacker, D. C., Flottum, P. K., Sammataro, D., Erickson, E. H. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Effects+of+climatic+and+edaphic+factors+on+soybean+flowers+and+on+the+subsequent+attractiveness+of+the+plants+to+honey+bees%5BTitle%5D)+AND+%22Field+Crops+Research%22%5BJournal%5D)">Effects of climatic and edaphic factors on soybean flowers and on the subsequent attractiveness of the plants to honey bees.</a> Field Crops Research. , (1983).</li>
<li>Blettler, D. C., Fagúndez, G. A., Caviglia, O. P. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Contribution+of+honeybees+to+soybean+yield%5BTitle%5D)+AND+%22Apidologie%22%5BJournal%5D)">Contribution of honeybees to soybean yield.</a> Apidologie. , (2018).</li>
<li>Kengni, S. B., Tchuenguem Fohouo, F. N., Ngakou, A. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Impact+of+the+foraging+activity+of+Apis+mellifera+adansonii+Latreille+%28Hymenoptera%3A+Apidae%29+and+Bradyrhizobium+fertilizer+on+pollination+and+yield+components+of+Glycine+max+L.+%28Fabaceae%29+in+the+field%5BTitle%5D)+AND+%22IJBR%22%5BJournal%5D)">Impact of the foraging activity of Apis mellifera adansonii Latreille (Hymenoptera: Apidae) and Bradyrhizobium fertilizer on pollination and yield components of Glycine max L. (Fabaceae) in the field.</a> IJBR. , (2015).</li>
<li>De O. Milfont, M., Rocha, E. E. M., Lima, A. O. N., Freitas, B. M. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Higher+soybean+production+using+honeybee+and+wild+pollinators%2C+a+sustainable+alternative+to+pesticides+and+autopollination%5BTitle%5D)+AND+%22Environ+Chem+Lett%22%5BJournal%5D)">Higher soybean production using honeybee and wild pollinators, a sustainable alternative to pesticides and autopollination.</a> Environ Chem Lett. , (2013).</li>
<li>Ng, S. J., Lindsey, L. E., Michel, A. P., Dorrance, A. E. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Effect+of+Mid-Season+Foliar+Fungicide+and+Foliar+Insecticide+Applied+Alone+and+In-Combination+on+Soybean+Yield%5BTitle%5D)+AND+%22Crop%2C+Forage+%26+Turfgrass+Management%22%5BJournal%5D)">Effect of Mid-Season Foliar Fungicide and Foliar Insecticide Applied Alone and In-Combination on Soybean Yield.</a> Crop, Forage & Turfgrass Management. , (2018).</li>
<li>Hershman, D. E., Sikora, E. J., Giesler, L. J. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Soybean+Rust+PIPE%3A+Past%2C+Present%2C+and+Future%5BTitle%5D)+AND+%22Journal+of+Integrated+Pest+Management%22%5BJournal%5D)">Soybean Rust PIPE: Past, Present, and Future.</a> Journal of Integrated Pest Management. , (2011).</li>
<li>Swoboda, C., Pedersen, P. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Effect+of+Fungicide+on+Soybean+Growth+and+Yield%5BTitle%5D)+AND+%22Agronomy+Journal%22%5BJournal%5D)">Effect of Fungicide on Soybean Growth and Yield.</a> Agronomy Journal. , (2009).</li>
<li>Akem, C. N. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((The+effect+of+timing+of+fungicide+applications+on+control+of+frogeye+leaf+spot+and+grain+yield+of+soybeans%5BTitle%5D)+AND+%22Eur+J+Plant+Pathol%22%5BJournal%5D)">The effect of timing of fungicide applications on control of frogeye leaf spot and grain yield of soybeans.</a> Eur J Plant Pathol. , (1995).</li>
<li>Bueno, A. F., Panizzi, A. R., Hunt, T. E., Dourado, P. M., Pitta, R. M., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Challenges+for+Adoption+of+Integrated+Pest+Management+%28IPM%29%3A+the+Soybean+Example%5BTitle%5D)+AND+%22Neotrop+Entomol%22%5BJournal%5D)">Challenges for Adoption of Integrated Pest Management (IPM): the Soybean Example.</a> Neotrop Entomol. , (2021).</li>
<li>St Clair, A. L., Dolezal, A. G., Cass, R. P., Hendriksma, H. P., Stein, D. S., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Insecticide+application+prevents+honey+bees+from+realizing+benefits+of+native+forage+in+an+agricultural+landscape%5BTitle%5D)+AND+%22Sci+Total+Environ%22%5BJournal%5D)">Insecticide application prevents honey bees from realizing benefits of native forage in an agricultural landscape.</a> Sci Total Environ. , (2025).</li>
<li><a target="_blank" href="http://scholar.google.com/scholar?hl=en&safe=off&q=%22Mode+of+Action+Classification%22">Mode of Action Classification</a>. , Insecticide Resistance Action Committee. (2024).</li>
<li>Kumar, G., Singh, S., Nagarajaiah, R. P. K. Detailed Review on Pesticidal Toxicity to Honey Bees and Its Management. <a target="_blank" href="http://scholar.google.com/scholar?hl=en&safe=off&q=author%3aKumar%2C+G+%22Modern+Beekeeping+-+Bases+for+Sustainable+Production%22">Modern Beekeeping - Bases for Sustainable Production</a>. , (2020).</li>
<li>Liu, Q., He, Q., Zhang, S., Chai, Y., Gao, Q., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Toxic+effects+of+detected+pyrethroid+pesticides+on+honeybee+%28Apis+mellifera+ligustica%29%5BTitle%5D)+AND+%22Sci+Rep%22%5BJournal%5D)">Toxic effects of detected pyrethroid pesticides on honeybee (Apis mellifera ligustica).</a> Sci Rep. , (2022).</li>
<li>Haas, J., Nauen, R. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Pesticide+risk+assessment+at+the+molecular+level+using+honey+bee+cytochrome+P450+enzymes%3A+a+complementary+approach%5BTitle%5D)+AND+%22Environ+Int%22%5BJournal%5D)">Pesticide risk assessment at the molecular level using honey bee cytochrome P450 enzymes: a complementary approach.</a> Environ Int. , (2021).</li>
<li>Iverson, A., Hale, C., Richardson, L., Miller, O., McArt, S. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Synergistic+effects+of+three+sterol+biosynthesis+inhibiting+fungicides+on+the+toxicity+of+a+pyrethroid+and+neonicotinoid+insecticide+to+bumble+bees%5BTitle%5D)+AND+%22Apidologie%22%5BJournal%5D)">Synergistic effects of three sterol biosynthesis inhibiting fungicides on the toxicity of a pyrethroid and neonicotinoid insecticide to bumble bees.</a> Apidologie. , (2019).</li>
<li>Schuhmann, A., Schmid, A. P., Manzer, S., Schulte, J., Scheiner, R. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Interaction+of+Insecticides+and+Fungicides+in+Bees%5BTitle%5D)+AND+%22Frontiers+in+Insect+Science%22%5BJournal%5D)">Interaction of Insecticides and Fungicides in Bees.</a> Frontiers in Insect Science. , (2022).</li>
<li>Rieth, J. P., Levin, M. D. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((The+repellent+effect+of+two+pyrethroid+insecticides+on+the+honey+bee%5BTitle%5D)+AND+%22Physiological+Entomology%22%5BJournal%5D)">The repellent effect of two pyrethroid insecticides on the honey bee.</a> Physiological Entomology. , (1988).</li>
<li>Rieth, J. P., Levin, M. D. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Repellency+of+Two+Phenylacetate-Ester+Pyrethroids+to+the+Honeybee%5BTitle%5D)+AND+%22Journal+of+Apicultural+Research%22%5BJournal%5D)">Repellency of Two Phenylacetate-Ester Pyrethroids to the Honeybee.</a> Journal of Apicultural Research. , (1989).</li>
<li>Thompson, H., Wilkins, S. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Assessment+of+the+synergy+and+repellency+of+pyrethroid%2Ffungicide+mixtures%5BTitle%5D)+AND+%22Bulletin+of+Insectology%22%5BJournal%5D)">Assessment of the synergy and repellency of pyrethroid/fungicide mixtures.</a> Bulletin of Insectology. , (2003).</li>
<li>Ekabote, S., Raj, P., Divyajyothi, U. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((In+Vivo+Bio-efficacy+and+Phytotoxicity+of+Propiconazole+%2B+Difenoconazole+on+Soybean%5BTitle%5D)+AND+%22Chemical+Science+Review+and+Letters%22%5BJournal%5D)">In Vivo Bio-efficacy and Phytotoxicity of Propiconazole + Difenoconazole on Soybean.</a> Chemical Science Review and Letters. , (2020).</li>
<li>Hodgson, E. W., McCornack, B. P., Tilmon, K., Knodel, J. J. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Management+Recommendations+for+Soybean+Aphid%5BTitle%5D)+AND+%22J+Integr+Pest+Manag%22%5BJournal%5D)">Management Recommendations for Soybean Aphid.</a> J Integr Pest Manag. , (2012).</li>
<li>Naeve, S. L. <a target="_blank" href="http://scholar.google.com/scholar?hl=en&safe=off&q=author%3aNaeve%2C+SL+%22Soybean+growth+stage%22">Soybean growth stage</a>. , (2018).</li>
<li>Forrester, K. C., Lin, C. H., Johnson, R. M. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Measuring+factors+affecting+honey+bee+attraction+to+soybeans+using+bioacoustics+monitoring%5BTitle%5D)+AND+%22Journal+of+Insect+Science%22%5BJournal%5D)">Measuring factors affecting honey bee attraction to soybeans using bioacoustics monitoring.</a> Journal of Insect Science. , (2024).</li>
<li>Hearon, L. E., Johnson, L. H. P., Underwood, J., Lin, C. H., Johnson, R. M. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((buzzdetect%3A+an+open-source+deep+learning+tool+for+automated+bioacoustic+pollinator+monitoring%5BTitle%5D)+AND+%22J+Insect+Sci%22%5BJournal%5D)">buzzdetect: an open-source deep learning tool for automated bioacoustic pollinator monitoring.</a> J Insect Sci. , (2025).</li>
<li>R Core Team. <a target="_blank" href="http://scholar.google.com/scholar?hl=en&safe=off&q=%22R%3A+A+Language+and+Environment+for+Statistical+Computing%22">R: A Language and Environment for Statistical Computing</a>. , (2024).</li>
<li>Wright, R., McMechan, J. <a target="_blank" href="http://scholar.google.com/scholar?hl=en&safe=off&q=author%3aWright%2C+R+%22Estimating+Soybean+Leaf+Defoliation%22">Estimating Soybean Leaf Defoliation</a>. , (2017).</li>
<li>Bates, D., Mächler, M., Bolker, B., Walker, S. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Fitting+Linear+Mixed-Effects+Models+Using+lme4%5BTitle%5D)+AND+%22Journal+of+Statistical+Software%22%5BJournal%5D)">Fitting Linear Mixed-Effects Models Using lme4.</a> Journal of Statistical Software. , (2015).</li>
<li>Brooks, M. E., Kristensen, K., Benthem, K. J. van, Magnusson, A., Berg, C. W., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((glmmTMB+Balances+Speed+and+Flexibility+Among+Packages%5BTitle%5D)+AND+%22The+R+Journal%22%5BJournal%5D)">glmmTMB Balances Speed and Flexibility Among Packages.</a> The R Journal. , (2017).</li>
<li>US EPA. <a target="_blank" href="http://scholar.google.com/scholar?hl=en&safe=off&q=%22Residual+Time+to+25%25+Bee+Mortality+%28RT25%29+Data%22">Residual Time to 25% Bee Mortality (RT25) Data</a>. , (2014).</li>
<li>Hayashida, R., Hoback, W. W., de Freitas Bueno, A. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((A+test+of+economic+thresholds+for+soybeans+exposed+to+stink+bugs+and+defoliation%5BTitle%5D)+AND+%22Crop+Protection%22%5BJournal%5D)">A test of economic thresholds for soybeans exposed to stink bugs and defoliation.</a> Crop Protection. , (2023).</li>
<li>Zelaya Mejia, W. D., Tilmon, K., Ortez, O., Lindsey, L. E. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Soybean+yield+was+not+influenced+by+foliar+insecticide+application%5BTitle%5D)+AND+%22Crop%2C+Forage+%26+Turfgrass+Management%22%5BJournal%5D)">Soybean yield was not influenced by foliar insecticide application.</a> Crop, Forage & Turfgrass Management. , (2024).</li>
<li>Wernecke, A., Frommberger, M., Forster, R., Pistorius, J. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/pubmed?term=((Lethal+effects+of+various+tank+mixtures+on+honey+bees%5BTitle%5D)+AND+%22J+Consum+Prot+Food+Saf%22%5BJournal%5D)">Lethal effects of various tank mixtures on honey bees.</a> J Consum Prot Food Saf. , (2019).</li>
</ol>

All authors declare no conflicts of interest.

While scouting for pests and making pesticide applications according to an IPM framework is recommended, the pesticide applications in the experiment were prophylactic, made at specific times during the growing season according to a predetermined schedule rather than on the basis of pest pressure. It is possible that yield and defoliation were unaffected in this study following pesticide application because pest pressure was below economic threshold levels. Defoliation was &#60;5% in most cases, which is below the 15% economic damage threshold for reproductive phases of soybean47. There was no apparent effect of the prophylactic insecticide-fungicide application on soybean yield or on the amount of defoliation that was observed during the R5 growth stage. This suggests that the pesticide application did not preserve soybean yield or deter insect herbivory. However, there was no pest scouting done, and the defoliation results suggest there were few insect pests to control in these fields, which is consistent with the findings of other authors48. Similarly, other authors22 saw similar results where tank-mix pesticide applications did not increase soybean yield when there was little to no insect pressure. These results suggest that there are limited benefits to adding an insecticide to a fungicide application when pest pressure is low. Therefore, adding insecticide may be an unnecessary expense, especially for prophylactic applications.
Instances of bee buzzing, which were used as a proxy for bee activity in soybean, were not affected by application of the insecticide-fungicide tank mix at either mid-morning or mid-day applications, or during the R3 or R5 stages. Additionally, a reduction in bee activity was not observed for either the 1-hour or 8-hour observation windows after pesticide application. This is surprising, as previous work has documented a repellent effect of pyrethroid insecticides34. This may be because the pyrethroid insecticide was combined with a fungicide. Combining these two pesticide classes has previously been shown to decrease repellency36.
It is worth noting that this experiment did not measure the potential off-site effects of the pesticide application on honey bees (Apis mellifera) that encountered it in the field. The lack of effect of the pesticide on bee activity does not necessarily indicate that the honey bees foraging in these plots were unaffected by the pesticide exposure. It is possible that honey bees foraging in the field encountered the pesticide, continued to forage, and later experienced negative effects. Studies have documented negative effects of tank mix applications49 and pyrethroid insecticides30 on honey bees. It is possible that the honey bees were negatively affected, but this effect did not translate into decreased activity in the field. This suggests that honey bees may not display avoidance behavior for some pesticides/combinations they encounter in the field. Other bees may be found in soybean fields, such as bumble bees or solitary bees15,17. These bees are likely to be disproportionately impacted by pesticide applications that happen when they are foraging on soybean flowers1.
This experiment did shed light on when bees are active in soybean fields. Bees were more active during peak soybean bloom at the R3 stage than at the R5 stage, consistent with results from other authors40. Additionally, the data suggest that bees are more active in soybeans during the middle of the day than during the morning. Through this experiment, a high-resolution picture of bee activity over a day and throughout the growing season has been generated. This reveals important patterns of bee activity in soybean fields, which will be important for crafting pesticide label and application recommendations. These results suggest that pesticide applicators should avoid spraying during R3 and in the mid-day to reduce pesticide exposure to bees. However, soybean farmers should apply foliar insecticides only when necessary, based on pest insect activity and economic thresholds.

Pesticide applications can affect pollinators that forage in agricultural systems1,2,3. It is possible to mitigate the effect of pesticides, at least in part, by timing applications to minimize pollinator exposure. Pesticides that are known to be toxic to pollinators carry labels that prohibit application when bees are active or when flowers are in bloom. This guidance is consistent with United States EPA guidelines and University factsheets4,5. If followed, this label guidance can mitigate the risk of pesticide application to bees, but does not provide more specific advice on which plant growth stages and particular times of day have the greatest bee activity. Additionally, the regulatory testing that informs label language was performed with singular pesticides and does not consider synergistic interactions between pesticides6.
A &#8220;tank mix&#8221; is the combination of two or more pesticides in a single application7 and is widely practiced to reduce the cost of pesticide applications and to reduce soil compaction by minimizing the travel of heavy application equipment across the field8. However, this is potentially a risky practice for bees, as most pesticide combinations have not been evaluated for their potential effects on non-target organisms9. Despite this, tank mixing is widely used in many cropping systems, including chickpea10, pumpkins11, almonds12, and soybean13,14.
Soybeans are one crop where tank-mixed pesticide applications may be particularly important for bee health. Bees frequently forage on soybean flowers for nectar, and soybeans can be a major nectar source for honey production15,16,17,18. Despite soybeans being self-pollinating, insect pollination has been repeatedly shown to increase yield19,20,21. A survey of soybean growers found that approximately one-third of foliar fungicide applications were tank-mixed with foliar insecticides22. It is recommended that foliar fungicides be applied around the time when soybeans are fully flowering to control common diseases such as anthracnose, Cercospora&#160;leaf blight, frogeye leaf spot, and soybean rust23,24,25. The addition of an insecticide to a fungicide application is viewed by many farmers as a cost-effective way to protect their crop by reducing the risk of arthropod pest issues later in the season26. However, a diversity of bees and other nontarget organisms are at risk when an insecticide is included in fungicide applications made during soybean bloom15,17,27.
One fungicide-insecticide mixture of particular concern is the combination of sterol biosynthesis-inhibiting (SBI) fungicides and pyrethroid insecticides. Pyrethroid insecticides are sodium channel modulators that disrupt nerve impulses in insects28. Many pyrethroids are highly toxic to bees29,30. SBI fungicides inhibit cytochrome P450 monooxygenase enzymes (P450s) that are important for the formation of fungal cell membranes, but can also inhibit detoxicative P450s that metabolize pyrethroids and other xenobiotics in bees31. The combination of SBI fungicides and pyrethroid insecticides has been shown to be more toxic to honey bees and bumble bees than either pesticide alone32,33. For this reason, applications involving tank-mixing pyrethroid insecticides and SBI fungicides may pose a risk to bees and other pollinators.
Previous work has reported that bees can be repelled by certain pyrethroid insecticides34 and that exposure to some pyrethroids can decrease bee activity35. However, it has also been documented that fungicides, when combined with insecticides, may mask the repellency effect36.
Bioacoustics, the study of organisms through the sounds that they make, offers a promising approach to evaluate changes in bee activity over the course of a day or growing season in response to pesticide applications. Using the distinct&#160;noise that bees make while in flight&#8212;buzzing&#8212;as a proxy for bee activity, bioacoustics can be effectively combined with machine learning to assess bee activity in a given area over a desired time period. The audio data collected offers enhanced resolution of bee activity and allows for daily bee foraging patterns to be examined. Additionally, this audio data can be examined over longer periods of time to find trends in activity through a growing season. Decisions on when to apply pesticides occur on both of these scales. Bioacoustics offers an information-rich way to monitor bee activity and answer questions about bee activity in narrow time windows. Compared to traditional methods like sweep-netting and pan-trapping, a bioacoustics approach requires less labor, allows continuous sampling over longer periods of time, and is highly scalable to larger study sites. One such application of bioacoustics is in the investigation of pesticide effects on bee activity in agricultural fields.
Here, work is done to determine the effects that a &#8220;worst-case scenario&#8221; application of a fungicide-insecticide tank mix, applied to soybeans during bloom, may have on bee activity and soybean yield. The pesticides used were formulations containing propiconazole, an SBI fungicide, and alpha-cypermethrin, a pyrethroid insecticide. Both of these pesticides may be applied to soybeans during bloom to control fungal diseases and soybean aphids (Aphis glycines)37,38. A bioacoustics approach was used to determine whether applying this pesticide combination at different times of day (mid-morning or mid-day) and during different soybean growth stages (peak flowering [R3] or late flowering [R5]) affected bee activity. Bees were expected to forage in high numbers and to be most susceptible to negative effects from pesticide application during the middle of the day and during the R3 growth stage.

<table><tbody><tr><td>&amp;frac12; in. (1 cm) soft packing foam</td><td>Uline</td><td>S-8350</td><td>Microphone cover to act as wind screen, cut to 8 x 5 cm</td></tr><tr><td>1.75 mm PETG+ 3D Printer Filament - True Red</td><td>Inland (Microcenter)</td><td>PETG+175FR1</td><td>Filament for making recorder cover, printed in vase mode, using https://www.thingiverse.com/thing:7319146/files</td></tr><tr><td>4 ft. (1.2 m) step-in fence post for fence wire</td><td>Fi-Shock (Tractor Supply Co.)</td><td>3600956</td><td>Stake for holding audio recorder</td></tr><tr><td>6 in. (15 cm) reusable cable ties</td><td>VELCRO</td><td>NA</td><td>For attaching recorder to stake</td></tr><tr><td>AAA batteries</td><td>Procell</td><td>NA</td><td>For powering handheld audio recorders</td></tr><tr><td>Fastac EC insecticide</td><td>BASF Ag Products</td><td>EPA: 7969-298</td><td>Pyrethroid insecticide used throughout the experiment.</td></tr><tr><td>Fitness Fungicide</td><td>Loveland products, Inc.</td><td>EPA: 34704-1031</td><td>Sterol Biosynthesis Inhibiting (SBI) fungicide used throughout the experiment.</td></tr><tr><td>Grain Gauge H2 weighing system</td><td>HarvestMaster</td><td>NA</td><td>For gathering yield data from research plots</td></tr><tr><td>Handheld audio recorder</td><td>Sony</td><td>ICD-PX370</td><td>Audio recorders used throughout the experiment. Used specifically to capture audio data in soybean fields.</td></tr><tr><td>Pioneer P35T155E</td><td>Corteva</td><td>P35T155E</td><td>Soybean variety used throughout the experiment.</td></tr><tr><td>Quantum Core plot combine</td><td>Wintersteiger</td><td>NA</td><td>Combine used to harvest soybeans during the experiment.</td></tr><tr><td>Rain cover</td><td>N/A</td><td>NA</td><td>For protecting audio recorders from rain. Can be found at: https://www.thingiverse.com/thing:7319146</td></tr><tr><td>Rubber Bands - 2 x 1&amp;frasl;16 in. (127 x 1.6 mm), #14</td><td>Uline</td><td>S-14723</td><td>For attaching foam wind screen to recorder</td></tr><tr><td>SeedPro 360</td><td>ALMACO</td><td>NA</td><td>For planting soybean seeds in research plots</td></tr></tbody></table>

assessing the repellency of an insecticide-fungicide combination to bees using bioacoustics

Assessing the Repellency of an Insecticide-Fungicide Combination to Bees Using Bioacoustics

Soybeans (Glycine max (L.) Merr.) and bees have a mutually beneficial relationship where nectar is exchanged for pollination service, which may enhance seed yield. However, pesticide use on soybeans during flowering could disrupt this relationship by harming bees and therefore reducing the yield benefits from pollinatiors. This study examined how pesticide applications during flowering affect bee foraging and soybean yield, focusing on application timing (mid-morning vs. mid-day) and flowering stage (R3&#8212;full bloom vs. R5&#8212;late/end of bloom), compared to an untreated control. To do this, an experiment was conducted in central Ohio in 2023 and 2024, where soybean plots were sprayed with a bee-toxic mixture of a pyrethroid insecticide (0.9% alpha-cypermethrin) and a sterol biosynthesis-inhibiting fungicide (41.8% propiconazole) using a backpack sprayer. Bee activity was monitored using audio recorders placed in each plot and analyzed with the machine learning tool buzzDetect. Bioacoustic analysis showed no effect of pesticide application on bee activity, regardless of timing. Bee activity was significantly higher during the R3 growth stage, when soybeans were at peak bloom, compared to the R5 growth stage, when flowers were less abundant. Bee activity was significantly higher in mid-day than in mid-morning. No significant differences in soybean defoliation or yield were detected across pesticide treatments. These results provide insight into how bee foraging varies across the soybean growing season and throughout the day, which can inform pesticide application recommendations to minimize bee exposure.

Activity
The 8-hour window model (Figures 1A–B) revealed significant differences between bee activity depending on the growth stage of the soybeans (Chi square = 74.35, P &lt; 0.001, df = 1), with significantly higher average detection rates in the R3 stage compared to the R5 stage for the 8 h following pesticide application. No significant differences in bee activity were observed between control plots and those receiving the tank mix (Chi square = 0.053, P = 0.82, df = 1). For this analysis, the mean squared error was 0.00094. The variances for the parameters involved in the experiment were: sprayed = 0.021 and stage = 0.041.
Analysis of audio data for 1 h following pesticide application (Figures 2A–B) revealed significant differences between bee activity depending on the stage of the soybeans (Chi square = 12.72, P &lt; 0.001, df = 1) and the time of day (Chi square = 9.97, P = 0.0016, df = 1). The average detection rate was higher during the R3 stage than in the R5 growth stage. The average detection rate was higher during the hour following the mid-day spray time than during the hour following the mid-morning spray time. No significant differences in bee activity were seen when the pesticide tank mix was applied (Chi square = 0.62, P = 0.43, df = 1). No significant differences were seen with the time of day and growth stage interaction term (Chi square = 2.52, P = 0.11, df = 1). For this analysis, the mean squared error was 0.00094. The estimates of variance for parameters involved in the analysis were: sprayed = 0.027, time of day = 0.040, stage = 0.090, time of day:stage = 0.14.
When averaged over the course of all days in the experiment, the audio detections reveal a daily trend in detection rates that has differences when separated by soybean growth stage (Figure 3). For both stages, detection rates start to increase rapidly at ~10:00 and decrease rapidly at ~17:00. However, this trend is much more pronounced in the R3 growth stage. Additionally, detection rates reach a much higher peak at ~14:00 in R3 than they do in R5.
Yield data
Analysis of the yield data (Figure 4) did not indicate that there were differences in soybean yield across treatments. No significant differences were found for soybean yield on any treatment type (F = 0.16, df = 4, 90.86; P = 0.96).
Defoliation
Analysis of defoliation data (Figure 5) did not indicate significant differences in defoliation based on the time of day of pesticide applications (F = 3.046, df = 2, 56; P = 0.055). Defoliation was generally low (&lt;5% in most cases).
DATA AVAILABILITY: 
All raw data used to create graphs and conduct statistical analyses are included in the supplementary materials section.
<img alt="Detection rate analysis, control vs. treatment, 8-hour post-spray, box plots, time impact study." class="xfigimg" src="/files/ftp_upload/70440/70440fig1v2.jpg" data-altupdatedat="1791864094000" data-alt="Figure 1"> 
Figure 1: Distribution of average detection rates for the 8 h following pesticide application. (A) Distribution of average detection rates (binned by 10 min) for the 8 h following the pesticide application during the R3 growth stage. (B) Distribution of average detection rates (binned by 10 min) for the 8 h following the pesticide application during the R5 growth stage. The control groups received no pesticide application. There are no statistically significant differences between any groups. All boxplots display a median line, upper and lower quartiles (top and bottom of boxes), and whiskers (minimum and maximum). <a href="https://www.jove.com/files/ftp_upload/70440/70440fig1largev2.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
<img alt="Detection rate comparison, R3 and R5 spray, control vs treatment, boxplot, statistical analysis." class="xfigimg" src="/files/ftp_upload/70440/70440fig2v3.jpg" data-altupdatedat="1791864094000" data-alt="Figure 2"> 
Figure 2: Distribution of average detection rates for the 1 h following pesticide application. (A) Distribution of average bee detection rates (binned by 10 min) for 1 h following the pesticide application during the R3 growth stage. The control groups received no pesticide application. (B) Distribution of average bee detection rates (binned by 10 min) for 1 h following the pesticide application during the R5 growth stage. The control groups received no pesticide application. For both R3 and R5, there was a statistically significant difference in average detection rates between mid-morning and mid-day groups. All boxplots display a median line, upper and lower quartiles (top and bottom of boxes), and whiskers (minimum and maximum). <a href="https://www.jove.com/files/ftp_upload/70440/70440fig2largev3.jpg" style="font-size: 14px;" target="_blank">Please click here to view a larger version of this figure.</a>
<img alt="Detection rate graph; time of day vs. detection rate; comparison of R3 and R5 datasets." class="xfigimg" src="/files/ftp_upload/70440/70440fig3v2.jpg" data-altupdatedat="1791864094000" data-alt="Figure 3"> 
Figure 3: Average detection rates of all audio recorders in either the R3 or R5 growth stages. Detection rates (binned by 10 min) averaged across all audio recorders used in the experiment and separated by growth stage. False positives in buzz detections due to cricket calls were observed during nighttime hours, especially ~4:30 a.m.The curves for each growth stage are not stacked on top of each other. <a href="https://www.jove.com/files/ftp_upload/70440/70440fig3v2large.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
<img alt="Yield comparison by pesticide application timing; box plot chart; agriculture experiment results." class="xfigimg" src="/files/ftp_upload/70440/70440fig4v2.jpg" data-altupdatedat="1791864094000" data-alt="Figure 4"> 
Figure 4: Yield in bushels per acre of soybean for each treatment. The control group received no pesticide application. There are no statistically significant differences between any groups. All boxplots display a mean line (black bars), median line, upper and lower quartiles (top and bottom of boxes), and whiskers (minimum and maximum). <a href="https://www.jove.com/files/ftp_upload/70440/70440fig4v2large.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
<img alt="Defoliation impact on plants; box plot; pesticide timing: control, midmorning, midday analysis." class="xfigimg" src="/files/ftp_upload/70440/70440fig5v2.jpg" data-altupdatedat="1791864094000" data-alt="Figure 5"> 
Figure 5: Defoliation levels assessed at the R5 growth stage. Defoliation, expressed as the percentage of leaf area missing for soybean leaves, was grouped by timing of pesticide application during the R5 growth stage. The control group received no pesticide application. There are no statistically significant differences between any groups. All boxplots display a mean line (black bars), median line, upper and lower quartiles (top and bottom of boxes), and whiskers (minimum and maximum). <a href="https://www.jove.com/files/ftp_upload/70440/70440fig5v2large.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
Supplementary File 1: The protocols needed to replicate the experiments are listed in this document. <a href="https://www.jove.com/files/ftp_upload/70440/Supplementary%20file%201.docx" target="_blank">Please click here to download this file.</a>

This study aimed to investigate the relationship between the timing of pesticide application and bee activity, measured using bioacoustics, in soybeans. Analysis showed that pesticide application, regardless of timing, had no effect on bee activity, defoliation, or yield. However, bee activity was higher later in the day and during peak soybean bloom.

Soybeans (Glycine max (L.) Merr.) and bees have a mutually beneficial relationship where nectar is exchanged for pollination service, which may enhance ...

This research examined the visitation patterns of bees on soybean flowers, and investigated whether a tank-mixed pesticide application could deter them. Existing bee monitoring methods lack temporal precision and scalability. This bioacoustics protocol enables continuous, high-resolution monitoring across large geographic areas.To begin, prepare the field for soybean cultivation according to the treatment plan. To set up the handheld audio recorders, place an open cell foam over the microphone to serve as a windscreen, then fasten the recorder to plastic step-in electric fence stakes with hook-and-loop tape. Place the audio recorders at the center of the field approximately 24 hours before the first pesticide application, and keep them running for at least 24 hours. Use 12 audio recorders on each application day, with four recorders assigned to each time-of-day group across both R3 and R5 growth stages.Wear full personal protective equipment. Using a custom-made hand-held backpack sprayer, apply the previously prepared pesticide mixture to individual soybean fields within each block. Apply the pesticide mixture at the top of the soybean canopy at approximately 3 miles per hour. After starting the pesticide application process for a given treatment, allow approximately 10 minutes for all plots to receive the application.To process all audio recordings, use the machine learning tool buzzdetect. Use the program to distinguish insect buzzing noises from other environmental sounds, including cars, combines, and airplanes. Next, distinguish buzzing noises from some non-buzzing insect noises while noting that certain sounds remain difficult to distinguish. After audio analysis, use R to calculate the detection rate within one-hour and eight-hour time intervals. Finally, 24 hours after pesticide application during the R5 growth phase, perform defoliation assessments by visually assessing leaves from the upper soybean canopy on 10 plants from each treatment and untreated control plot.Average detection rates were significantly higher in the R3 stage than in the R5 stage for the eight hours following pesticide application. However, bee activity did not differ significantly when the pesticide tank mix was applied. In the hour following pesticide application, average detection rates were significantly higher during the R3 growth stage compared to the R5 growth stage. Additionally, detection rates were notably greater in the hour after the midday spray than in the hour after the mid-morning spray. However, bee activity was not significantly different when the pesticide tank mix was applied. Detection rates increased rapidly around 10:00 AM and declined sharply around 5:00 PM in both stages. Peak detection rates at approximately 2:00 PM were substantially higher in R3 than in R5. Soybean yield was not significantly different across treatments. Defoliation did not differ significantly based on the time of day of pesticide application. Overall, defoliation levels were low, remaining below 5% in most cases.The protocol allows researchers to measure bee activity more passively over time at specific points of interest. A major challenge of this protocol is managing the large volume of audio files, requiring efficient data storage and organization systems. Future studies can use this approach to investigate bee foraging preferences, flowering resource use, and temporal patterns of bee activity.

assessing-repellency-an-insecticide-fungicide-combination-to-bees

Research

JoVE Journal

Environment

This manuscript demonstrates and discusses techniques used to survey pesticide susceptibility and detect resistance to contact and systemic pesticides in arthropod pests.

Bioassays for Monitoring Insecticide Resistance

A method to rapidly screen host plant volatiles by measurement of the electrophysiological response of adult navel orangeworm (Amyelois transitella) antennae to single components and blends via electroantennographic analysis is demonstrated.

Electroantennographic Bioassay as a Screening Tool for Host Plant Volatiles

A novel behavioral assay is described for investigating the short term gustatory responses of the mouthparts of freely-moving bumble bees (Bombus terrestris) toward nutrients and toxins in solution.

A Novel Behavioral Assay to Investigate Gustatory Responses of Individual, Freely-moving Bumble Bees (Bombus terrestris)

Microbial consortia within bumble bee hives enrich and preserve pollen for bee larvae. Using next generation sequencing, along with laboratory and field-based experiments, this manuscript describes protocols used to test the hypothesis that fungicide residues alter the pollen microbiome, and colony demographics, ultimately leading to colony loss.

Empirical, Metagenomic, and Computational Techniques Illuminate the Mechanisms by which Fungicides Compromise Bee Health

The loss of honey bee colonies presents a challenge to crop pollination services. Current pollinator protection practices warrant an alternative approach to minimize the contact of honey bees to harmful pesticides using repellent chemistries. Here, we provide detailed methods for a visual tracking protocol to screen deterrents for bees.

Video Tracking Protocol to Screen Deterrent Chemistries for Honey Bees

Sublethal doses of pesticides may affect colonies in ways that are difficult to detect using only periodic or visual methods. Methods to measure total adult bee mass, brood, and food resources by weighing hives and hive parts, photographing frames, and installing sensors, are provided. Data analysis is also addressed.

Monitoring Colony-level Effects of Sublethal Pesticide Exposure on Honey Bees

Photo-attraction bioassay arenas are used to determine the optimal light color(s) to maximize insect attraction; however bioassays and methods are specific to target insect behaviors and habitats. Customizable equipment and modifications are explained for nocturnal or diurnal and terrestrial or aerial insects.

Key Elements of Photo Attraction Bioassay for Insect Studies or Monitoring Programs

Fungicide sprays on flowering plants may expose solitary bees to high concentrations of pollen-borne fungicide residues. Using laboratory-based experiments involving in vitro-reared bee larvae, this study investigates the interactive effects of consuming fungicide-treated pollen derived from host and non-host plants.

In Vitro Rearing of Solitary Bees: A Tool for Assessing Larval Risk Factors

This protocol was developed to enhance the understanding of how agrochemicals affect honey bee (Apis mellifera) reproduction by establishing methods to expose honey bee queens and their worker caretakers to agrochemicals in a controlled, laboratory setting and carefully monitoring their relevant responses.

Assessing Agrochemical Risk to Mated Honey Bee Queens

The present protocol explains a method to feed pesticide-contaminated provisions to the larvae of the solitary bees, Osmia excavata. The procedure examines the ecotoxicity of the pesticide to the larvae of the solitary bees.

Evaluating the Effect of Pesticides on the Larvae of the Solitary Bees

We describe the methodology and importance of the topical application bioassay to measure insecticide susceptibility in mosquitoes and fruit flies. The presented assay is high-throughput, utilizes insect mass-thus allowing for calculating a mass-relativized lethal dose instead of concentration-and likely has lower variability than other similar methods.

Topical Application Bioassay to Quantify Insecticide Toxicity for Mosquitoes and Fruit Flies

Explore bioacoustic monitoring of bee activity in soybean fields using machine learning analysis. This method helps researchers assess pesticide impacts on pollinators and inform sustainable agricultural practices for crop protection.

Assessing the Repellency of an Insecticide-Fungicide Combination to Bees Using Bioacoustics

Explore More Videos