Frequently Asked Questions
1. General questions
1.1 What is the Insect Calculator and what is it used for?
With the Insect Calculator you can answer the following questions: How many insects inhabit this meadow? How does mowing influence the inhabitants of meadows? You'll find answers to these and many more questions right here. Our goal is to provide an online tool to predict the number of insects on a meadow based on the latest scientific knowledge. More info: https://insektentaschenrechner.de/en/about/ https://insektentaschenrechner.de/en/about/
1.2 Who is behind the Insect Calculator?
The Insect Calculator was created within the BioDivKultur project funded by the Research Initiative for the Preservation of Biodiversity (FEdA) of the Federal Ministry of Education and Research (BMBF, now BMFTR). Other researchers contributed data to the Insect Calculator, you can find those project partners and related literature under: https://insektentaschenrechner.de/en/how/
Project page: https://biodivkultur.de/?lang=en
1.3 Why an Insect Calculator?
The Insect Calculator is intended to provide citizens and practitioners with insights into the impact of grassland land use on insects and spiders. It is a practical tool that shows how arthropods react to specific mowing strategies, highlighting the negative impacts and potential improvements for conserving biodiversity in grasslands.
The purpose of the Insect Calculator tool was to highlight the huge and often overlooked insect diversity in urban and rural green spaces and the great impacts of any kind of mowing on the density and diversity of insects – in real numbers. It has been developed and designed for the more specific context of mowing, and particularly for practitioners responsible for urban green spaces, lawns, gardens, roadside verges, and also meadows. Even without the agricultural grasslands, these mown green spaces represent a huge, largely overlooked area (more than all nature reserves together) where biodiversity for meadow-living insects can be improved by simple changes of mowing regimes. And the beneficiaries are not necessarily only the bees or other pollinators that receive a lot of attention, and for which other measures are already implemented. With the Insect Calculator one can make a realistic prediction of insect and spider density in a given site based on the current quantitative scientific knowledge and within the limitations of the datasets available (constraints defined below). In addition, like in games designed for educational purposes, the Insect Calculator offers an unconstrained exploration of different parameters or management strategies and their relative impact on insect and spider density and diversity, based on the current knowledge of effect sizes.
2. Functionality & use
2.1 How does the Insect Calculator work?
The Insect Calculator uses negative binomial models to predict the number of meadow arthropods based on specific mowing parameters. These models provide estimates, which are then calculated with the input that the user enters to predict meadow arthropods. You can find a more detailed description under the following link: https://insektentaschenrechner.de/en/how/
2.2 Which taxa does the Insect Calculator predict?
Foremost, the Insect Calculator predicts the density (the number of individuals) of different above-ground living insect groups and spiders of one square meter of grassland vegetation: spiders (Araneae), leafhoppers/planthoppers (Auchenorrhyncha), beetles (Coleoptera), flies/mosquitoes (Diptera), ants (Formicidae), true bugs (Heteroptera), complete metamorphosis larvae (holometabolous larvae), wasps/bees (Hymenoptera, excluding ants), grasshoppers/crickets (Orthoptera), and aphids (Sternorrhyncha).
These are the groups that are most dominant taxa in suction sampling (biocoenometer, aspirating all arthropods from a “cage” covering 1 m2 of a grassland) conducted in grasslands, excluding groups that are predominantly on the soil surface and inside the soil (e.g. collembolans, oribatid mites). Even though these are collected by suction sampling, this method is not suitable for them due to a bias depending on how much open soil the area has (Their total abundance in such datasets is strongly biased to open soil sucked up with the suction sample). But also they are not typically representative for vegetation-dwelling arthropods.
Additionally, we provide species numbers only for some arthropod groups (spiders (Araneae), leafhoppers/planthoppers (Auchenorrhyncha), beetles (Coleoptera), true bugs (Heteroptera), and grasshoppers/crickets (Orthoptera), for the other groups there was not sufficient data available). These calculations are based on the overall relationship between individual density and species richness in the subset of data where species have been identified.
Density and species numbers are additionally summarised across all focal taxa. However, for species richness the summary only counts in the available five groups.
2.3 Can I also use the calculator for other insect species?
For now, the Insect Calculator can only be used with the available groups. As outlined above (2.2.), these groups contain the majority of vegetation-living arthropod individuals when using the suction sampling method per grassland area (density per area), as it was the case for the underlying data. For soil-dwelling arthropods or flower-visiting insects other methods would be recommended for establishing a new calculator, not covered by the datasets included here. For example, flower visiting insects are often recorded by transect methods or larger areas samples, as they are much more conspicuous and mobile and can be surveyed on flowers directly, but their density (individuals per area for a given time) is typically much lower than for the grassland taxa covered by the Insect Calculator. For example, bees and butterflies are only rarely included in the suction samples of a square meter grassland (these mobile species usually escape before the could be sucked in). We encourage you to contact us about implementing more data, such as suction sampling data with sufficient mowing information, or starting similar approaches for taxa (areas and land use impacts).
2.4 For which areas does the Insect Calculator predict?
Scientifically, the Insect Calculator can only predict within the space and time covered sufficiently by the data. The currently used data representing different types of grasslands and green spaces in Germany and Switzerland, including eleven different datasets with 1,686 samples from 20 regions sampled in different years between 2012 and 2023. This was all the data available for this time. For any location, only the geographic coordinates (latitude and longitude) are used here, not the site-specific climate or bioregion, assuming that the general spatial proximity to the locations covered by the dataset best covers the overall arthropod density. Predictions to other grassland types, different locations, different seasons or years, different climates and other sources of variation depend on the similarity of environmental conditions and similarity of arthropod communities to those represented by the data. Moreover, for some possible combinations of different parameters that are poorly or not covered by the data, the predictions may be inaccurate. For example, rotary mowers may usually not be represented in urban areas, while motor scythes may not be represented in agricultural meadows. Or more biodiversity friendly bar mowers will be not found with a high number of cuts, or mulchers won`t usually be found with a high cutting height. However, these combinations appear to be underrepresented in the management of real-world grasslands, making them less important for practitioners and therefore for this calculator.
To simplify the way to enter coordinates, you can select your current location to see how many insects you would find in a nearby meadow. This location helps include longitudinal and latitudinal gradients of insect diversity in Central Europe. However, you can also select a specific meadow to more precisely predict the effects of mowing.
2.5 Why can't I make predictions for areas outside of Germany, Austria, and Switzerland?
In order to predict the number of insects and spiders in an area, we need sufficient data. Since our data only covers German and Swiss meadows and green spaces, we decided to only predict for the three German-speaking countries. Linear extrapolation outside the gradients covered and predictions for larger and/or more distant countries that are not covered by data would be very uncertain, see disclaimer above in 2.4.
However, we encourage you to contact us about implementing more data, such as suction sampling data with sufficient mowing information, or starting similar approaches for other areas, taxa, and land use impacts.
2.6 How were models built to predict the density of insects and spiders?
To create the models, we used all the parameters that can be found as buttons in “Expert mode” as fixed effects: mowed (yes/no), mower type, mowing height, mowing width, number of cuts, meadow size, coordinates, day of the year (date) and proportion of sealed area (isolation). We ran a model with abundance as the response variable for each arthropod group. The glmmTMB package in R was used to run the negative binomial models. These models are suitable for model predictions for all arthropod groups, even if new data is implemented in the web tool. Their purpose is to make predictions based on a wide variety of data, rather than to find statistical significance values.
The “Walking mode” is similar, but without the variables “mowing width” and “mowing height”. Finally, to simplify the “Walking mode” even further, we have defined default settings for the examples of 4 different meadow types. This allows users who are not familiar with the subject to use the calculator initially without any knowledge of the mower type, for example. More detailed descriptions can be found in the associated article.
2.7 How were calculations connected between the statistical programme R and the website?
We used R to manipulate data and build models for predictions. We extracted coefficients of the model outputs to calculate the predictions manually in a simple linear function (prediction = a+ b*x + c*x...) with the values entered by the calculator. We used this approach to be able to work together with a company that helped create this website. As for them it was more difficult to implement R, we found the most parsimonious way to simply provide a prediction function and provide a csv file that contains all model estimates.
2.8 How were models built to predict the richness of insects and spiders?
Species numbers are only available for some arthropod groups (spiders (Araneae), leafhoppers/planthoppers (Auchenorrhyncha), beetles (Coleoptera), true bugs (Heteroptera), and grasshoppers/crickets (Orthoptera), for the other groups there was not sufficient data available. Based on the extensive species-level data from our BioDivKultur project and the Biodiversity Exploratories project, we predicted the richness of each arthropod group based on abundance. We used a nonlinear model (nls) to account for the saturation of species numbers. Using the individual numbers calculated by the Insect Calculator, we can then estimate species numbers.
2.9 What land use information or parameters are required for the calculations?
See below all the parameters, their descriptions and units that will be entered into the Insect Calculator:
|
Parameter |
Description and unit |
Median |
Min |
Max |
|
|
|
Mown |
Was the meadow previously mown or not? yes: 1077, no: 609 |
|||
 |
Mower |
Mower type: Bar mower (n=112), eco mower (62), lawn mower (n=26), motor scythe with string or knife (n=12), mulcher (n=354), rotary mower (n=1061) and sickle mower (n=39). |
|||
 |
Width |
Mowing width in m |
4.87 |
0.1 |
10 |
 |
Height |
Mowing height in cm |
7.3 |
4 |
20 |
 |
Cuts |
Number of cuts/year |
2 |
0 |
25 |
 |
Date |
Day of the year (between 11th May and 3rd September, Median: 6th July) |
187 |
128 |
265 |
 |
Meadow size |
Size of the sampled meadow in ha |
1.27 |
0.0005 |
130 |
 |
Isolation |
Impervious surfaces: Settlements and transportation areas in % 50 m buffer around the meadow |
0.041 |
0 |
0.976 |
|
|
Longitude |
Longitude coordinates |
8.675 |
6.186 |
14.03 |
 |
Latitude |
Latitude coordinates |
48.73 |
46.38 |
53.15 |
3. Data & Accuracy
3.1 Where does the data used for the calculations come from?
Data was summarised from various published and unpublished studies and projects. The common denominator was suction sampling, the correct taxa, and sufficient information about mowing and grassland management. Samples were taken in various regions of Germany and Switzerland.
3.2 How up to date are the data and calculations?
The calculations are based on the best possible knowledge and data available. Data was collected between 2012 and 2023. It will always be possible to incorporate more data into the models in future to increase their accuracy. To this end, the models were designed to be as robust as possible, so that data can be integrated in the future without the need to repeat the data exploration process, as is done in statistical analyses.
3.3 How accurate are the results of the Insect Calculator?
To illustrate uncertainties, we have included the standard errors alongside the estimates and visualisations provided in the associated article to the Insect Calculator.
The following excerpt of the associated article shows the estimates and standard errors for total arthropod density depending on mowing:
"To illustrate the different influences of the Insect Calculator parameters, we use the default settings of the "Expert mode": Mowing itself had the strongest effect (Est ± SE = -9.811 ± 0.401) with 73% more arthropods on unmown meadows (265 individuals) than on mown meadows (122 individuals). The total density of arthropods varied between mower types. Using the rotary mower (Est ± SE = 3.805 ± 1.295) as a reference, the bar mower (6.170 ± 1.388) showed the highest density, with an increase of +21%. This was followed by the eco mower (5.572 ± 1.473), which showed a modest increase of +5%. In contrast, the most harmful mower types significantly reduced arthropod densities: mulcher (1.5 ± 1.292) by -30%, lawn mower (0.212 ± 1.67) by -42%, sickle mower (-1.327 ± 1.571) and motor scythe (0.305 ± 1.939) by -36%. Total arthropod density decreased by -7% from one cut per year to five cuts (-1.062 ± 0.378)."
The associated article should further visualise potential statistical problems by integrating additional statistical models. It further discusses potential uncertainties in greater detail. In summary, the Insect Calculator, while useful for drawing general conclusions, also has limitations in its implementation. Predictive linear models based on synthesised datasets cannot fully account for the variability and complexity found in real-world conditions, particularly for extrapolations beyond the scope of the included studies and for parameter settings poorly covered by data. Some limitations can depend on the ecological characteristics of organisms themselves. For example, the dense clustering of Formicidae and Sternorrhyncha makes it inherently difficult to predict their abundance. However, as users of the web tool will see ants and aphids in the meadow, which can be quite abundant, it is important to include them in the tool, too. Therefore, uncertainties must be communicated clearly, as we did here on the webpage accompanying the Insect Calculator. Other limitations can result from data availability, for example, the lack of data for certain conditions.
These are some conclusions discussed in the associated article. For further information please consult the article or ask the team.
3.4 Are other grassland management techniques accounted for?
Other land use influences, such as fertilization or grazing (Chisté et al. 2016, Ecosph & Chisté et al. 2018, Oecol), can have an impact on the calculated area. Other grassland management, such as the use of conditioners (Hecker et al. 2022, J Insect Conserv) or maintenance measures such as rolling and leveling (Berger et al. 2024, Ecol Appl) might also play a role. Due to missing data, these parameters could not be taken into account here. However, for example, variables such as mowing frequency can account for fertilisation.
An extension to pastures or including more detailed effects of grazing would be the next logical step, including the extensive sweep netting data from the Biodiversity Exploratories (which we could not use for the Insect Calculator, however, where we decided to limit it to the suction sampling method as it provides a density of arthropods per 1 m² and is the most comparable method).
We are currently planning to develop such extensions. Note however, that developing a model framework with variables that accounts in sufficient detail for the variable effects of different grazers (cows, sheep, horses etc.) and different grazing regimes (pulse disturbance by larger herds for short time, press disturbance over long time). The detail we implemented for mowing was a huge task already, with integrating several observational and experimental data and analyses, to do the same for grazing requires a new approach, new project and more data.
Recent studies on detecting mowing using remote sensing highlight significant challenges: grazing events are difficult to distinguish from mowing, and even mob grazing might occasionally be misclassified as mowing (Holgrave et al. 2023, Schwieder et al. 2022). Detecting finer details, such as grazing intensity, type of grazer, or differences between cattle, sheep, or horses, would be even less feasible with current methods. For these reasons, we could not incorporate such data into the Insect Calculator at this time. However, we gladly include references to these studies in the Discussion to acknowledge the potential for future advancements in this area, should more robust data and tools become available.
Literature
- Chisté, M. N., Mody, K., Gossner, M. M., Simons, N. K., Köhler, G., Weisser, W. W., & Blüthgen, N. (2016). Losers, winners, and opportunists: How grassland land‐use intensity affects orthopteran communities. Ecosphere, 7(11), e01545. https://esajournals.onlinelibrary.wiley.com/doi/pdf/10.1002/ecs2.1545
- Chisté, M. N., Mody, K., Kunz, G., Gunczy, J., & Blüthgen, N. (2018). Intensive land use drives small-scale homogenization of plant-and leafhopper communities and promotes generalists. Oecologia, 186, 529-540. https://link.springer.com/article/10.1007/s00442-017-4031-0
- Hecker, L. P., Wätzold, F., Yang, X., & Birkhofer, K. (2022). Squeeze it or leave it? An ecological-economic assessment of the impact of mower conditioners on arthropod populations in grassland. Journal of Insect Conservation, 26(3), 463-475. https://link.springer.com/article/10.1007/s10841-022-00392-5
- Holtgrave, A. K., Lobert, F., Erasmi, S., Röder, N., & Kleinschmit, B. (2023). Grassland mowing event detection using combined optical, SAR, and weather time series. Remote Sensing of Environment, 295, 113680.
- Schwieder, M., Wesemeyer, M., Frantz, D., Pfoch, K., Erasmi, S., Pickert, J., … & Hostert, P. (2022). Mapping grassland mowing events across Germany based on combined Sentinel-2 and Landsat 8 time series. Remote Sensing of Environment, 269, 112795.
3.5 What are the current limitations of the tool?
The tool relies on existing data collected by the BioDivKultur project and requested by other researchers. It only includes suction sampling, which is ideal for comparison because it is an easy-to-implement method with minimal human interference (e.g., compared to sweep netting). Furthermore, this sampling technique allows for the collection of arthropod density rather than activity measures (e.g., intersection traps). However, of course limiting the Insect Calculator to one comparable method also limits the availability of data. Additionally, some datasets were excluded because they lacked the detailed land use information required by the Insect Calculator.
This demonstrates another limitation: even though we have included a large amount of data (eleven datasets with 1,686 samples from 20 regions sampled in different years between 2012 and 2023), it still does not offer all the information that would be interesting to include in greater detail, such as information on grazing (e.g., which grazers, intensity, and duration), fertilization, and other land use management variables (e.g., post-harvesting processes), other regions and countries, sampling outside of the regular sampling only in summer etc.
In the discussion of the associated article, as well as in some disclaimers along with the calculator, we acknowledge these points that could affect arthropod numbers differently. Here, we also mention other ecological factors that could influence predictions, such as weather and the life history traits of arthropods.
Additionally, many different approaches exist for analysing data. We opted for this combination of parameters and negative binomial models because of their robustness and the availability of the data. Since we implemented the predicted model estimates with a company that was paid during the project to build the website, it is more difficult to make changes after the project ends. However, this company provided us with a lot of information and the ability to make many changes ourselves. However, there are limitations after the funding period, so it is impossible to completely change the models and parameters without new funding. Therefore, for future similar developments, it would be advisable to start an interdisciplinary/transdisciplinary project with computer scientists and IT specialists who work closely with ecologists. Shiny apps, which require fewer specialists to build, might also be an option, but they are possibly less professional than a website built by specialists.
A great advantage of working with the company on the website was using WordPress, which made it better usable and adaptable by non-IT specialists.
In summary, there are different ways these different areas of expertise can be connected. However, each approach probably has its own advantages and disadvantages. However, it is always important to bear in mind that research projects are time-limited, and consideration must be given to how the resulting long-term products are to be managed.
3.6 What could future developments look like?
As mentioned, the Insect Calculator allows inputting new data to increase the robustness of the data. Although it is intended as a tool for analysing the impact of mowing on grassland arthropods in great detail, it can also serve as a prototype for similar approaches. The current data availability shows its limitations, but also the opportunity to build similar tools that could be useful for related topics, such as other grassland land uses (e.g., grazing, fertilization, and post-harvesting processes), as well as different topics in different habitats (e.g., forests or aquatic environments) and different impacts (e.g., forestry and other human impacts). It is a great example of how to make the impacts of land use on arthropods more visible to citizens and practitioners. Please feel free to contact us for further development of ideas.
4. Purpose & Application
4.1 For whom is the Insect Calculator particularly useful?
The Insect Calculator is intended to be a science-based tool that anyone can use without any statistical knowledge. It is interesting as an environmental education tool to give an insight into how many insects and spiders can generally be found in a meadow. The strong effects that mowing has on insects and spiders can be visualised more easily by entering the values yourself. In addition, the Insect Calculator is intended to show walkers, gardeners and practitioners in green space management or agriculture which measures have which effect. Similar as in games designed for educational purposes, the Insect Calculator offers an unconstrained exploration of different parameters or management strategies and their relative impact on insect and spider density and diversity, based on the current knowledge of effect sizes.
4.2 Can I use the results for scientific purposes?
It depends on how you intend to use the Insect Calculator. As with the refuge calculation on the website (see https://insektentaschenrechner.de/en/refuge/), you can use the per-square-metre predictions to estimate the number of invertebrates in a meadow. For more in-depth statistical analysis, it would be worthwhile taking a look at the statistical models in the associated paper or requesting the open access data from the data owners, which is supposed to be publicly available after the publication process.
4.3 For which other purposes can I use the Insect Calculator?
As we have shown in the refuge calculations (see https://insektentaschenrechner.de/en/refuge/) or in educational material for German agricultural students (see https://biodivkultur.de/wp-content/uploads/2025/06/Unterrichtsmaterial_Auswirkungen_von_Mahd_auf_Insketen_LPV_ohne_Loesungsblaetter.pdf) the Insect Calculator results can be applied for different questions and easy calculations. I.e. calculating the number of arthropods you can save with different refuge sizes, or the number of arthropods you can save by changing the mowing equipment.
5. Feedback, Collaboration & Support
5.1 Is there a guide or tutorial on how to use it?
The most information you can find are here in the FAQs and the associated workflow diagram, in the webpage and more specifically in the corresponding paper, which is currently under review. For further questions, please contact us.
5.2 Can I also implement data or contribute?
Yes, please! If you have any suction sampling data with grassland management information, please contact us. We would be glad to implement more data and mention your contribution on the webpage. You can find all parameters in more detail here: https://insektentaschenrechner.de/en/how/
Also, for new ideas of more advanced tools or other kinds of insect calculators, we are glad to share our experience or collaborate. See contact information below.
5.3 Any other questions or feedback?
Please contact us:
Johanna L. Berger: johanna.berger(at)csic.es
Nico Blüthgen: bluethgen(at)bio.tu-darmstadt.de
Or see here, the whole BioDivKultur-Team: https://biodivkultur.de/the-team/?lang=en