Conference Agenda

Assessing biodiversity is vital for conservation and ecosystem management. Traditional methods, like field surveys, can be labor-intensive, hindering precise data acquisition. In recent decades, diverse methodologies utilizing remote sensing have emerged, comprehensively assessing structural and species diversity. The "Height Variation Hypothesis" (HVH) is an indirect approach tailored for this purpose. It posits that greater vegetation height heterogeneity (HH), measured through LiDAR or photogrammetry, corresponds to increased complexity in ecosystem structure and heightened species diversity. Tested in various ecosystems, the HVH has gained attention in recent literature. In forests, it's been applied with airborne laser scanning LiDAR and GEDI LiDAR data, showcasing space-borne LiDAR's capability to discern biodiversity patterns. The HVH's application extends to forest areas affected by extreme weather events, like the Vaia windstorm, aiming to understand the impact of HH on on forest biodiversity and stability. Beyond forests, it proves valuable in agricultural landscapes, integrating LiDAR data with ecological modeling for avian biodiversity conservation. Additionally, it assesses flower and bee biodiversity in grasslands using photogrammetric data from UAVs.

In conclusion, the HVH presents a promising approach for estimating biodiversity across ecosystems through LiDAR and photogrammetric data, offering insights from recent literature to advance understanding and support effective conservation strategies.

Each day, trillions of animals take to the skies in pursuit of increased survival or reproductive output. In the current biodiversity crisis, continuous large-scale monitoring of these aerial activities would provide key information on (changes in) this aerial biodiversity, enabling targeted conservation actions for a large part of the terrestrial biodiversity. Radars are powerful remote-sensing tools that can provide such information in detail by quantifying the intensity, timing, and spatial distribution of aerial animal abundance and movements. Small-scale biological radars simultaneously track and classify individual birds, bats, and insects from 3 m up to several hundred meters above ground. Various features measured by these radars such as the size or shape facilitate differentiating into several species-groups. We present a summary on the current biodiversity monitoring possibilities with such biological radar, the open-source tools to process the data, and showcase the potential of a network of such radars for large-scale, detailed biodiversity monitoring by demonstrating their use in quantifying the abundance and temporal movement patterns of insects across Europe. If networks of biological radars would be established through national and international governmental programs, their detailed, temporally-continuous, and large-scale monitoring of aerial biodiversity could provide key input on global Essential Biodiversity Variables.

Europe's Alpine landscape has undergone significant modifications in the last decades due to socio economic changes. The Trentino region, Italy, has shown the same trend with a significant increase in forest coverage as a result of the abandonment of marginal agricultural sites. Due to the natural forest recolonization, open areas like meadows and pastures have shrunk in many mountain rural areas changing the ecological landscape mosaic. Priority habitats where a unique botanical biodiversity is preserved are among the open spaces that are endangered, together with the fauna that depends on open areas (capercaillie, black grouse...). Remote sensed data are crucial to detect, georeference and quantify the spatial environmental changes with the aid of GIS and machine learning algorithms. Combining remote sensed data with historical sources like cartography and forest management plans allows to rebuild past landscape and model future scenarios that can help to guide environmental management choices, for example highlighting the most endangered priority habitats or the effect of climate change Different examples of application of these techniques (developed using FOSS4G) applied at different spatial scales are presented, discussing how the changes affect the distribution of biodiversity, social and cultural dynamics, perception of the landscape, and ecosystem services.

Bats are facing a global decline primarily due to habitat loss and agricultural intensification. In Europe, bats are protected by laws, and the monitoring of their populations is required by the European Union. However, studying bat species is challenging due to their nocturnal, elusive habits, and high mobility. Direct sampling methods for their monitoring can disturb the animals, are often costly, usually limited to small geographical areas and require specialised personnel. Nowadays, automated and miniaturised bat detectors allow long-term and large-scale monitoring studies on bat communities.

Here we present a broad-scale case study using automatic bat detectors to assess the effects of agricultural intensification on bat diversity in mountain environments (South Tyrol, Italy). We selected 47 sites in open agricultural areas, considering pastures, hay meadows, dry grasslands, and annual crops. We used generalised linear mixed models to analyse the activity of bat foraging guilds in relation to agricultural practices, natural structural elements, and landscape variables. Overall, woody features, hedgerows, and wetlands positively influence especially endangered bat species and play a key role in agricultural areas. Bat detectors contribute to efficient large-scale survey and monitoring systems, advancing our understanding on bat distribution even in remote areas and facilitating targeted conservation actions.

1. Camera trapping, a popular non-invasive wildlife monitoring tool, produces numerous images. The overlooked variation in classification performance, which affects ecological inference, arises from using systems like Citizen Science (CS) and Convolutional Neural Networks (CNNs), class types, and camera/environment factors.

2. We evaluated a Citizen Science project and two CNN architectures on a dataset of 100,059 expert-classified images, identifying diurnal and nocturnal images across six classes: humans, cervids, leporids, wild boars, red foxes, and empty images.

3. Citizen Science showed high precision, while CNNs had a superior recall, particularly for leporids. Both struggled with nocturnal photos, especially in precisely identifying empty images, indicating possible under-detection at night. CNNs processed up to 1,150,000 images daily, far outpacing Citizen Science, which annotated only 25% of the same volume. Filtering out low-confidence classifications leads to an impractical increase in the manual workload, lowering the overall detection rate due to fewer images for reliable identification.

4. These findings highlight a key distinction: CNNs effectively maximize species detection with speed and high recall, whereas Citizen Science is better for minimizing classification errors. The study underscores the need to evaluate misclassification variations across systems, classes, and environmental factors for accurate ecological inference.