Mitigating arthropod mortality during roadside mowing: the potential of the EcoCut insect flushing bar

<b>Abstract</b>Due to their size and corridor function, roadside verges are of concern for the conservation and protection of insects and other arthropods, and any opportunity to manage them in an insect-friendly way, e.g. insect-friendly mowing technology, should be seen as an important contribution. Here we investigated the effectiveness of the EcoCut flushing bar (air rail blower, model F360, Fischer Maschinenbau GmbH &amp; Co. KG, Gemmrigheim, Germany), a blowing device to chase away insects, on nine grassland arthropod groups. We compared (1) a section treated with the EcoCut flushing bar mounted on a switched off flail mower head and (2) an untreated control section on two experimental sites (i.e. ‘ecological compensation area’ and ‘extensive orchard meadow’) that differed in vegetation composition, height and density. Our results demonstrate the effectiveness of the EcoCut flushing bar but also suggest that its effectiveness depends on the vegetation characteristics of the area to be treated: In high and dense vegetation (i.e. ‘ecological compensation area’) only Diptera were effectively deterred. In less high and dense vegetation (i.e. ‘extensive orchard meadow’), such as roadside verges, seven of the nine groups studied were successfully “scared off” by 46 to 64% by the blowing device. Overall, the study demonstrates the potential of the EcoCut flushing bar in significantly mitigating arthropod losses during mowing of roadside verges.<b>Material and Methods</b>The experiment took place on June 7, 2024, on two experimental sites (‘ecological compensation area’ (specific area for the promotion of landscape diversity in agroecosystems) and an ‘extensive orchard meadow’) near Kleinaspach, Baden-Württemberg, Germany at sunny and dry weather conditions between 10:30 a.m. and 5 p.m.. Both sites had not been mown at the time of the experiment in 2024 and showed relatively high (especially the ecological comparison site) and dense vegetation.To study the effect of the EcoCut insect flushing bar on the arthropod fauna, we compared (1) a section treated with the EcoCut flushing bar (air rail blower, model F360; capacity 80 m³/min, Fischer Maschinenbau GmbH &amp; Co. KG) mounted on a switched off flail mower head (MK3-124, working with 1.25 m, Fischer Maschinenbau GmbH &amp; Co. KG) and (2) an untreated control section. The flail mower head incl. flushing bar (air rail) was mounted to a mowing arm (DZ3; Fischer Maschinenbau GmbH &amp; Co. KG) attached to a tractor. The driving speed was 2 – 3 km/h.Each study site (‘ecological compensation area’ and ‘extensive orchard meadow’) consisted of two parallel tracks with five sections each. The flushing bar treatment and the control were randomly assigned to the tracks for each section.Each track and section were treated and sampled separately, with the treatments and sampling per section being carried out at similar time intervals. To determine the direct effects of the flushing bar, three isolation squares (1×1×0.8 m, ‘biocenometer’ according to Mühlenberg (1989) were placed on each section. The isolation squares for the control were placed on the corresponding track prior to the application of the flushing bar treatment to prevent migration of startled insects into the controls. The isolation squares of the respective flushing bar section were placed immediately after the treatment. The actual sampling was carried out using a motorized insect vacuum with sample bag (EcoVac Insect vacuum, ecoTech Umwelt-Messsysteme GmbH, Bonn, Germany) after a 5-minute waiting period to allow the captured insects to settle. Each isolation square was completely vacuumed for 5 min. The entire content of each insect vacuum sample bag was transferred into zip-lock plastic bags, stored on ice and later frozen at – 20°C until further processing. In total, we sampled five sections (15 samples) treated with the flushing bar, and five control sections (15 samples) at each experimental site, resulting in 30 samples per treatment. The following taxonomic groups were sorted and counted from all 60 samples using a stereomicroscope: Aranea, Orthoptera, Cicadina, Heteroptera, Hymenoptera (without Formicidae), Coleoptera, Diptera and holometabolous larvae. All non-target taxa (e.g. Thysanoptera, Dermaptera) were classified as ‘other arthropods’. All sorted individuals are stored as voucher specimens in 80% denatured ethanol (Evolutionary Biology of Invertebrates, University of Tübingen). Sorting was done anonymously, i.e. at the time of sample processing it was not known which treatment the sample had come from, to avoid inadvertently influencing the results.Statistical analysis was performed using RStudio (Posit team, 2023) with R 4.3.2 (R Core Team, 2023) following Santon et al. (2023) to run generalized linear mixed models (glmmTMB) (Brooks et al., 2017) using Template Model Builder (TMB) (Kristensen et al., 2016). Track was used as a factor predictor and section was included as a random factor. Due to the large differences between the two experimental sites (vegetation composition, height, and density), the analysis was performed separately for each experimental site, resulting in n = 15 for both treatments on the ‘ecological compensation area’ and the ‘extensive orchard meadow’, respectively. Pairwise comparisons (mean and 95% compatibility intervals), taking into account the predictions, were used to determine the percentage differences between the flushing bar treatment and the control (Santon et al., 2023). No statistical analysis could be performed for Orthoptera for the ‘ecological compensation area’ as only two individuals were recorded.References:Brooks, M. E., Kristensen, K., Benthem, K. J. van, Magnusson, A., Berg, C. W., Nielsen, A., Skaug, H. J., Mächler, M., &amp; Bolker, B. M. (2017). glmmTMB Balances Speed and Flexibility Among Packages for Zero-inflated Generalized Linear Mixed Modeling. <i>The R Journal</i>, <i>9</i>(2), 378–400.Kristensen, K., Nielsen, A., Berg, C. W., Skaug, H., &amp; Bell, B. M. (2016). TMB: Automatic Differentiation and Laplace Approximation. <i>Journal of Statistical Software</i>, <i>70</i>, 1–21. https://doi.org/10.18637/jss.v070.i05Mühlenberg, M. (1989). <i>Freilandökologie</i> (2nd ed.). Quelle &amp; Meyer Verlag.Posit team. (2023). <i>RStudio: Integrated Development Environment for R</i> (Version 2023.9.0.463) [Computer software]. Posit Software, PBC. http://www.posit.co/R Core Team. (2023). <i>R: A Language and Environment for Statistical Computing</i> (Version 4.3.1) [Computer software]. R Foundation for Statistical Computing. https://www.R-project.org/Santon, M., Korner-Nievergelt, F., Michiels, N. K., &amp; Anthes, N. (2023). A versatile workflow for linear modelling in R. <i>Frontiers in Ecology and Evolution</i>, <i>11</i>. https://www.frontiersin.org/articles/10.3389/fevo.2023.1065273<b>Description of the dataset</b>Numbers of individuals per m² for the nine assessed arthropod groups (Araneae, Orthoptera, Cicadina, Heteroptera, Hymenoptera, Coleoptera, Diptera, holometabolous larvae and 'other Arthropods‘) recorded for the different treatments: control (untreated) and EcoCut (air rail blower, mounted on switched off mowing head) on June 07 2024. Sampling was carried out using 1-m² isolation squares and an insect vacuum on the two study sites 'ecolocigal compensation area' and 'extensive orchard meadow'. Each study site was divided into two parallel tracks (T1,T2) with five sections (S1-S5; S6-S10) each. Per section and treatment three samples were taken at the same time (daytime).<br>