Biology subjects

Obrist, M. K.

Publications and source records attributed to Obrist, M. K..

4 recordsLinked to original sources

bioRxiv2026

AWET -- Arthropod Weight Estimation Tool

Arthropods drive essential ecosystem processes such as pollination, decomposition, and nutrient cycling and are widely used as indicators of ecosystem condition and function. Among various arthropod-derived metrics, body weight is a key variable in functional ecology and frequently assessed as dry body weight. However, drying arthropod specimens limits the samples future potential for research, as it prevents further processing such as trait measurements or species identification. Here, we present AWET (Arthropod Weight Estimation Tool), an open-source application for automated estimation of individual fresh body weight and extraction of morphometric measurements from standardized images of pre-sorted arthropod samples. AWET combines automated image analysis with taxon-specific allometric regression models to estimate fresh body weight while simultaneously quantifying body length, width, area, and specimen abundance. The software operates with standard imaging equipment, requires no machine-learning-based classification or segmentation, and allows users to define taxonomic groupings according to their objectives. By preserving specimens for downstream analyses while processing large numbers of individuals within milliseconds, AWET provides an efficient, non-destructive, and cost-effective workflow for high-throughput arthropod phenotyping. The software is a practical and expandable tool for biodiversity monitoring projects investigating changes in arthropod biomass, abundance and individual morphometric measures.

ecology↗

bioRxiv2025

Rewilding beyond the wilderness: Beavers can restore stream biodiversity from urban to agricultural to natural landscapes

1 Beavers have been promoted as a cost-effective nature-based solution to restore degraded stream ecosystems and enhance overall biodiversity. However, the extent to which beaver-engineering boosts biodiversity in human-impacted landscapes remains poorly understood. 2 We assessed the responses of aquatic and terrestrial taxa (amphibians, dragonflies, fish, macrophytes, plankton, terrestrial vegetation, bats, aquatic, flying and terrestrial invertebrates) to beaver-engineering along a gradient of human land-use intensity in Switzerland. In 16 streams, we measured species richness, abundance, unique species richness, evenness, beta and gamma diversity in a beaver-engineered site with beaver dams and a control site without beaver activity. 3 Especially aquatic and semi-aquatic species profited from beaver-engineering. In response to beaver-engineering, amphibians, dragonflies, macrophytes, and plankton increased in species richness, abundance and unique species richness, while fish and terrestrial plants increased in species richness and unique species richness. Bats only increased in unique species richness, while for the other terrestrial taxa, we found no difference between beaver-engineered and control sites. Community evenness remained largely unaffected across all groups. 4 Overall, beaver-engineering increased biodiversity across a land-use gradient. For macrophytes and amphibians (species richness, unique species richness and abundance), as well for dragonflies (species richness and unique species richness), increases were smaller under higher land-use intensity. For the remaining taxa and metrics, we found no relationship between land-use intensity and responses to beaver engineering. 5 However, for beta-diversity, we found that sites with a high land-use intensity had a lower component of nestedness and stronger community turnover in comparison to more natural sites, with a significant effect for macrophytes. Gamma species richness was generally unaffected by land- use intensity, except for amphibians, which were often only found in beaver-engineered sites in high- land-use intensity areas. 6 Synthesis and applications. Our findings demonstrate that beaver-engineering significantly enhances local biodiversity across aquatic and terrestrial ecosystems, even at sites with high land-use intensity. Hence, beavers can effectively restore stream biodiversity from urban to agricultural to natural landscapes. Integrating beaver-engineering into river restoration strategies can substantially advance biodiversity goals under frameworks such as the EU Water Framework Directive and national conservation policies.

ecology↗

bioRxiv2022

Towards a General Approach for Bat Echolocation Detection and Classification

O_LIAcoustic monitoring is a scalable approach for assessing bat populations, yet automating the detection and classification of bat echolocation calls remains challenging, particularly in data-scarce regions. Although deep learning (DL) is increasingly applied to this task, most existing approaches repurpose computer-vision architectures and generate a single prediction per spectrogram clip, offering limited robustness to variable background noise and potentially constraining generality across regions and species assemblages. C_LIO_LIHere, we develop BatDetect2, an open-source DL pipeline for the joint detection and classification of bat echolocation calls. BatDetect2 builds on a 2D convolutional architecture and incorporates two targeted modifications: (i) a temporal self-attention layer designed to capture long-range structure across call sequences, and (ii) convolutional layers augmented with frequency coordinates to explicitly encode frequency information directly. We evaluate model generality using five diverse datasets from four different regions: UK, Mexico, Australia, and Brazil, and conduct ablation analyses using a UK dataset spanning 17 bat species. We further assess whether a trained model can detect echolocation calls from species absent from the training data. C_LIO_LIBatDetect2 consistently outperforms a traditional call-parameter extraction baseline across all datasets and evaluation metrics. Ablation analyses show that the inclusion of temporal self-attention yields a substantial species classification performance gain, increasing mean Average Precision (mAP) from 0.83 to 0.88, while frequency-coordinate augmentation provides no measurable benefit. When applied to novel species assemblages without retraining, model detection performance varies across datasets, with Average Precision ranging from 0.60 to 0.98. C_LIO_LIOverall, BatDetect2 demonstrates strong and transferable performance across acoustically and taxonomically diverse regions. By jointly detecting and classifying all bat calls present in each input clip, the pipeline provides a practical and extensible tool for passive acoustic monitoring. The full training pipeline and a pretrained UK model are released through the open-source Python package batdetect2, enabling practitioners to develop and deploy models using their own data. C_LI

ecology↗

bioRxiv2020

Automated flight interception traps for interval sampling of insects

Recent debates on insect decline require sound assessments on the relative drivers that may negatively impact insect populations. Often, baseline data rely on insect monitorings that integrates catches over long time periods. If, however, effects of time-critical environmental factors (e.g., light pollution) are of interest, higher temporal resolution of insect observations during specific time intervals are required (e.g., between dusk and dawn). Conventional time-critical insect trapping is labour-intensive (manual activation/deactivation) and temporally inaccurate as not all traps can be serviced synchronically at different sites. Also, temporal shifts of environmental conditions (e.g., sunset/sunrise) are not accounted for. We provide a battery-driven automated insect flight-interception trap which samples insects during seven user-defined time intervals. A commercially available flight-interception trap is fitted to a rotating platform containing seven flasks and a passage hole. The passage hole avoids unwanted sampling during time-intervals not of interest. Comparisons between two manual and two automated traps during 71 nights in 2018 showed no difference in caught insects. An experiment using 20 automated traps during 104 nights in 2019 proved that the automated flight interception traps is reliable. The automated trap opens new research possibilities as any insect-sampling interval can be defined. It is efficient and saves manpower and associated costs as activation/deactivation is required only every seven sampling intervals. Also, the traps are accurate because all traps sample exactly the same intervals. The trap is low maintenance and robust due to straightforward technical design. It can be controlled manually or via Bluetooth and smartphone.

ecology↗