Determining terminal velocities of selected insects and substitute materials to develop future test procedures for insect-friendly mowing technology

<b>Abstract</b>Agricultural grasslands are vital habitats for insects that provide key ecosystem services. However, conventional rotary mowers are proposed to cause significant insect mortality due to their high rotational speed of their blades and the resulting airflow. Therefore, innovative insect-friendly mowers with modified airflows are developed to reduce the suction effect on grassland arthropods. However, no standardised and cost-effective testing methods are currently available for these technologies. Here, we present a promising approach for future testing of insect-friendly mowers. We determined the terminal velocities of grassland insects as well as materials that could potentially replace insects in standardised tests. In a comparative analysis of the terminal velocities of 15 insect species and nine substitute materials, we identified several substitutes with terminal velocities closely matching those of the tested insects. These materials are highly suitable for laboratory testing of insect-friendly mowing techniques, particularly for evaluating the airflow and suction effects of the blades. Although substitute materials cannot replicate the natural responses of living insects to airflow, they provide a possible solution for standardised and scalable testing. This approach paves the way for a broader adoption of insect-friendly mowing technologies, ensuring a balance between agricultural productivity and insect conservation.<br><b>Materials and Methods</b>To investigate terminal velocity, a pneumatic conveyor was used. Samples were placed inside the conveyor and separated by increasing air velocities according to their terminal velocities. This resulted in separation curves (Mohsenin, 1986), which depict separated masses at the tested air velocities. These curves serve as effective indicators of particle behaviour in airflow. For the experiments, we selected insects that reflect typical grassland fauna in Germany (Cicadina, Araneae, Hymenoptera, Heteroptera; von Berg et al., 2024). Specifically, we accounted for different sizes, shapes, and lifestyles including inhabitants of the ground and herb layer, typical pollinators as well as parasitoid wasps and predatory insects. When selecting comparable substitutes, we chose inexpensive products that are easy to procure and decompose naturally in the environment.We tested 15 selected insect samples representing a wide range of common grassland insect groups and nine possible substitute particles (Tab. 1, Tab. 2). Insects were either collected from the field, laboratory-bred at the University of Hohenheim (UHOH), or were purchased from specialised insect breeding companies in Germany (DE) (Tab. 1). The insects were killed by freezing shortly before the experiment. Substitute particle samples were commercially available at local grocery stores (Tab. 2). All puffed grain samples were sorted to include only fully puffed grains, non-puffed and partially puffed grains were removed.The test procedure was adapted from Hassebrauck (1966) for tests with non-spherical particles (Hassebrauck, 1966). Frozen insect samples were thawed directly before the experiment. Subsequently, each sample was placed on the mesh inside the air inlet nozzle. The machine was set to separate the sample with increasing increments of air velocity. Each air velocity increment was held for two minutes to ensure consistent separation. After each increment, the machine was stopped and the remaining sample in the air inlet nozzle was weighed. The sample was then returned to the air inlet nozzle and the next air velocity increment was set. The air velocity was adjusted by measuring the pressure difference at the orifice using a Betz manometer. The increments were held constant for all three replicates of a sample but were adjusted for each sample to account for different separation characteristics. The separated particles were carried by the airflow to the cyclone, where they were separated from the air, collected in a container and reused in subsequent replicates. To account for natural variations, three replicates were performed for each sample.Initially, the samples were weighed using a Kern PCB 10000-1 balance (Kern &amp; Sohn GmbH, division: 0.1 g). For samples with a lower total mass, a Precisa PB 520 balance (Precisa Gravimetrics AG, division: 0.0001 g) was used to obtain a higher degree of precision (Tab. 3). The separated masses were converted to percentages of the total sample mass. The mean separated mass fractions were calculated from the three replicates. Air velocity was calculated from the pressure difference at the orifice (DIN Deutsches Institut für Normung e.V., 1982)The results of the pneumatic conveyor tests are presented as separation curves for each sample. These curves plot the fraction of separated mass <i>T</i> against the terminal velocity <i>v</i>_t set in the machine. Each sample is represented by four curves, one for each replicate, and a mean curve calculated from the three replicates. Separation curves are commonly used in agricultural engineering and material science for describing the terminal velocity of natural materials (Hassebrauck, 1966; Karaj &amp; Müller, 2010; Mohsenin, 1986; Persson, 1957; Uhl &amp; Lamp, 1966).<br><b>References</b>Berg, L. von, Sann, M., Steidle, J. L. M., Frank, J., Böttinger, S., &amp; Betz, O. (2024). InsectMow: Development and evaluation of insect- and spider-friendly mowing techniques. In <i>Mitteilungen der DGaaE</i> (Vol. 23, 47-51). Halle (Saale), DE.DIN Deutsches Institut für Normung e.V. (06/1982). <i>Durchflußmessung mit Blenden, Düsen und Venturirohren in voll durchströmten Rohren mit Kreisquerschnitt</i>. (1952). Berlin, DE: Beuth Verlag.Hassebrauck, B. (1964). Das Trennen von Korn-Häcksel-Gemischen in Sichtern mit senkrecht aufsteigendem Luftstrom. <i>Landtechnische Forschung</i>, <i>14</i>(1), 16–20.Hassebrauck, B. (1966). Das Trennen von Korn-Häcksel-Gemischen im Steigsichter. <i>Grundlagen der Landtechnik</i>. (16), 119–122.Karaj, S., &amp; Müller, J. (2010). Determination of physical, mechanical and chemical properties of seeds and kernels of Jatropha curcas L. <i>Industrial Crops and Products</i>, <i>32</i>, 129–138.Mohsenin, N. N. (1986).<i> Physical properties of plant and animal materials: Structure, physical characteristics and mechanical properties</i> (2., rev. and updated ed.). New York, USA: Gordon and Breach.Persson, S. (1957). Eigenschaften des Reinigungsgutes in Mähdreschern. <i>Landtechnische Forschung</i>, <i>7</i>(2), 41–45.Uhl, J. B., &amp; Lamp, B. J. (1966). Pneumatic Separation of Grain and Straw Mixtures. <i>Transactions of ASAE</i>, <i>9</i>(2), 244–246.<br><b>Dataset description</b>Supplementary figures S1 – S3 contain the separation curves with separated mass fraction <i>T</i> (%) and terminal velocity <i>v</i>_t (m/s) for three replicates (black) and calculated mean (red) for all samples. Each diagram contains data for one sample as specified by the figure descriptions.Supplementary tables S4 – S27 contain the values for separated mass fraction <i>T</i> (%) and terminal velocity <i>v</i>_t (m/s) for three replicates and calculated mean for all samples. Each table contains data for one sample and is named accordingly.