Monitoring regenerative agriculture practices with XNatura platform

The session begins with Virginia Castellucci, Head of Sustainability at XNatura, welcoming attendees to the first webinar in a new series. She explains that the decision to focus specifically on regenerative agriculture and how the XNatura platform supports it stems from a significant volume of requests from existing clients and ongoing projects. The team felt it was essential to share how their platform is uniquely positioned to handle the specific monitoring needs of these types of agricultural projects. Virginia introduces the speakers for the day: herself, Caterina Castaman (who leads the platform demonstration), and two guests from Bayer, Greta Pignata and Silvano Locardi, who join later to share a specific case study regarding their "Rigenerare per Crescere" project.

The agenda is outlined to include an introduction to the benefits and risks of regenerative agriculture, an explanation of the TNFD framework, a live demo of the platform, the Bayer case study, and a final Q&A session. Virginia proceeds to introduce XNatura for those unfamiliar with the organization. XNatura is described as a nature-tech environmental platform that mixes various technologies to monitor and protect biodiversity. The company has received funding from the European Union to develop the first biodiversity credit, with the first standard actually released in September 2025.

XNatura collaborates with significant institutional and scientific partners, most notably the European Space Agency (ESA), whose satellite images are integral to running the platform. The organization boasts a team of 60 people, with a significant portion dedicated to the Research and Development (R&D) team. They currently serve more than 500 customers and run projects across roughly 25 countries. The platform functions by integrating diverse data sources - remote sensing, IoT sensors, lab analysis, and on-field surveys- into a single tool. While their primary client base consists of private companies from various sectors, they also work extensively with public administrations, municipalities, and natural parks.

The presentation shifts to the core topic: monitoring regenerative agriculture. Virginia outlines the platform's specific benefits for this sector, which include establishing benchmarks for regenerative outcomes, monitoring biodiversity for key biological groups (taxa), linking agricultural practices to measurable impacts, and quantifying nature policy progress over the years. Additionally, the platform supports verified reporting and advanced data visualization. To contextualize the need for this approach, Virginia compares different farming methods.

Traditional agriculture is described as degrading the ecosystem because it focuses on resource extraction. Organic farming, while beneficial because it bans chemicals, does not always actively restore ecosystems. In contrast, regenerative agriculture represents a transformative approach that goes beyond sustainability to actively restore habitats and rebuild organic matter. Virginia provides specific examples of how regenerative agriculture mitigates the negative impacts of traditional farming. Regarding soil degradation and erosion, regenerative practices use cover crops and organic inputs to rebuild the soil structure. To combat the loss of soil carbon, the approach increases carbon sequestration through living roots and organic matter that capture and store carbon.

For biodiversity loss, the use of diverse crops creates essential habitats for soil organisms, pollinators, and wildlife. Finally, regarding water pollution and stress, these practices enhance water retention and quality by improving the soil structure. The methodology is underpinned by six core pillars: biodiversity, soil health, carbon and climate resilience, water stewardship, ecosystem functionality, and economic viability and livelihoods. The core principles driving these pillars are minimizing soil disturbance, maintaining soil cover, maximizing biodiversity, maintaining living roots, integrating livestock, and encouraging ecosystem interaction.

Despite the clear benefits, Virginia acknowledges that the transition from traditional to regenerative agriculture is not easy and presents several risks that farmers may initially face. These include transition risks such as a temporary drop in yields, a long transition period, and a heavy dependence on external environmental factors. There are also physical risks to consider, including extreme weather events, soil degradation, water scarcity, and climate variability. Because of these challenges, it is emphasized that farmers must be supported during this transition, as the ultimate result benefits everyone.

To support this shift, the platform aligns with the TNFD (Taskforce on Nature-related Financial Disclosures) framework, which encourages regenerative practices. The platform is built specifically to follow the TNFD’s "LEAP" approach: Locate, Evaluate, Assess, and Prepare. In the "Locate" phase, XNatura uses satellite images to analyze land use, detect biodiversity hotspots, and identify areas at risk. In the "Evaluate" phase, the platform assesses nature-related impacts and dependencies. In the "Assess" phase, it quantifies risks and opportunities by integrating key indicators such as the MSA (Mean Species Abundance) based on land use. Finally, in the "Prepare" phase, the platform supports target setting, benchmarking, and the generation of compliant reports.

For the "Locate" phase, the system uses land cover data, derived indexes, and KPIs from public datasets. For the "Evaluate" phase, the platform mixes different technologies including satellite images, field sensors, surveys, and lab analysis. Among the sensors, the "Spectrum" sensor is highlighted as the most utilized device for regenerative agriculture projects to date. It provides quantitative data on pollinator presence by counting buzzes. It is particularly valuable because it can detect differences in pollinator activity between particles of land that are geographically very close but managed with different practices. Another sensor is "Birdy", which operates on the same bioacoustic principles as Spectrum but is tuned to detect bird species. XNatura integrates other technologies like the "Biodiversa" app, which is used to run in-field surveys to evaluate ecosystem health. This combination of technologies allows for a comprehensive evaluation of the ecosystem.

The platform helps evaluate performance by providing a robust set of Key Performance Indicators (KPIs). While these are explored in depth during the demo, Virginia provides a list of the most critical ones. For land use, KPIs include land use type, impermeable land share, NDVI, and carbon stock. Microclimate KPIs cover recent rain precipitation, historical climate trends, and the aridity index. Biodiversity state is measured via MSA, nesting sites, and flower availability, while risks are tracked via hydrogeological and water risk metrics.

A highly important feature for regenerative agriculture projects is "Scenario Analysis," a tool developed in the last year. This feature helps evaluate the effectiveness of regenerative practices by simulating them before actual implementation. This allows users to understand the potential changes to the environment, ecosystem, and main KPIs in advance. Finally, regarding the "Prepare" phase of LEAP, the platform supports target setting and the preparation of reporting that is aligned with main frameworks such as TNFD, SBTN, and ESRS, making report generation very rapid. The platform can also be used to monitor other practices like carbon farming by measuring carbon stock, optimizing land management, and supporting carbon credits. Several clients are already using the platform specifically for these regenerative purposes.

Caterina Castaman takes over to conduct the live demonstration of the platform. She begins on the main page, which provides a description of the whole project. Navigating to the "Monitoring" section, she displays a map of all fields, using a demo dataset for 2026. This section is designed to help users prioritize actions; the map uses different colors to represent different risk levels, allowing users to see at a glance where to start taking action. Detailed information is available here, such as the total amount of hectares, the number of hectares currently impacted, and how many sites have a direct impact on biodiversity.

Caterina selects a specific site for the demo, located under a river between Rovigo and Ferrara. She demonstrates how the perimeter of the site is customizable in the settings, as is the "control area radius" (the whitish area surrounding the site used for comparison). She then opens the "Land Cover" field for 2025. This view relies on satellite data to detect various species and artificial surfaces present on the site. Caterina explains that users can go into "Action" and "Manual Edit" to change parts of the land cover if needed. For this demo, she has created a completely new scenario for a different year to serve as a comparison. This new scenario features the same species but managed with a regenerative approach, which will later be used in the scenario analysis to visualize changes.

Keeping with the 2025 data, Caterina highlights the NDVI (Normalized Difference Vegetation Index). She explains that this index ranges from minus one to plus one. Minus one indicates water bodies or a complete absence of vegetation, while plus one indicates maximum vegetation. The platform allows users to see the change in vegetation throughout the whole year, which is crucial for agricultural sites. Caterina points out specific data points on the graph: in August, the NDVI was around 0.7, while in December it dropped to 0.24. This trend can be compared against the control area.

Moving to the "Locate" section following the LEAP approach, Caterina discusses the "Impermeable Land Share" KPI. This index goes from zero to 100%, and the demo site is currently at 50.4% based on satellite data. The best-case scenario would be to decrease permeability to below 50%. She then shifts to the Microclimate section, showing the "Climate Change" score and how temperature has changed and is expected to change in future years. She notes that while the "Rain Score" might look good, it is vital to look at precipitation patterns over time. The "Historical Climate Trend" graph allows users to view data over different timeframes (years, months) and compare KPIs. For example, she illustrates that as temperature rises in summer, soil moisture decreases, showing the relationship between these variables. She also mentions the importance of the Aridity Index and Evapotranspiration, which are highly related.

Caterina emphasizes that Biodiversity KPIs are highly important as they change according to the practices applied. She explains the MSA (Mean Species Abundance) index, which ranges from zero to 100% and is related to human activity impacts. For the selected site in its traditional state, the MSA is very low at 9. The dashed line on the graph represents the benchmark of 35.6, and the site is currently well below this. Caterina outlines several objectives: "Match Neighbor" (reaching an MSA of 14), "Match Benchmark" (reaching 35.6 or a 30% regeneration), the "Half-Earth" objective (reaching 50%), and full recovery. Other biodiversity KPIs like pollinator abundance, nesting sites, and floral availability are also shown to be low or equal to the benchmark in the current state.

Regarding risks, the site has a "Hydrogeological Risk" score of D, which is worse than the benchmark. This is attributed to the site's proximity to a river and the high impermeability of the soil. "Water Stress" is also high, defined as the ratio between available water (low) and water exploited by human activity (high). The "River Discharge Trend" shows peaks related to rain events, indicating a high trend. Other sensors mentioned include "Air Quality" (detected by the PollyX sensor) and "Soil Temperature" (where colder is generally better depending on humidity). Finally, Caterina shows the "Anchor" tab, which deals with double materiality. It provides benchmark scores for production processes (e.g., growing rice has a high impact on freshwater). Users can input their own scores for intensity, reach, and remediability to compare their site's condition against the expected benchmarks for that process.

The demo concludes with the "Scenarios" section, where Caterina compares the "Traditional" land cover (current satellite data) on the left with the "Regenerated" scenario on the right. The platform allows views in absolute values, deltas, or evolution graphs. The simulation shows drastic improvements: the MSA increases from 10 to 44, effectively matching the benchmark and approaching the "Half-Earth" goal of 50%. Pollinator abundance, nesting sites, nectar potential, and floral availability all show significant increases. Perhaps most notably, the soil impermeability drops from 50% to just 1%. Caterina explains that this massive reduction in impermeability leads to higher soil moisture and lower soil temperatures, creating a much better condition for the entire ecosystem.

Following the demo, the floor is given to Greta Pignata and Silvano Locardi from Bayer to present their project, "Rigenerare per Crescere" (Regenerate to Grow), focused on the Italian viticulture sector. Greta begins by outlining Bayer's global vision: "Health for all, hunger for none". This vision is supported by three main pillars. The first is to produce 50% more, a necessity to feed a growing population projected to reach 10 billion by 2050, achieved through genetic gains and more efficient crop protection. The second pillar is to restore nature, which involves commitments to reducing greenhouse gas emissions, optimizing crop protection impacts, and improving water resources. The third pillar is to scale regenerative agriculture through partnerships.

Greta details the six key outcomes Bayer aims to achieve with regenerative agriculture: yield increase and improved productivity, enhanced social and economic well-being for farmers, improved soil health, climate change mitigation and adaptation, preservation and restoration of biodiversity, and conservation of water. For Bayer, future farming is defined as regenerative—producing more and restoring more and they aim to empower this transition through innovation in crop protection, genetics, digital solutions, and services. The approach is described as system-solutional, outcome-driven, and multi-seasonal.

Greta contextualizes the project within the Italian wine industry. While Italy is a world-leading wine producer, the sector faces growing risks from climate change, with production potentially falling by 20% to 30% without adaptation. Farmers are facing specific pain points: rising temperatures are affecting grape ripening, shifting suitable growing areas, and causing earlier germination and harvest, all of which impact quantity and quality. Additionally, farmers face pest and disease pressures and stricter European regulations that limit crop protection products, all while consumers demand more sustainable and organic wines. To address these challenges, Bayer decided to engage wineries and partners to map their portfolio to regenerative outcomes and co-create solutions with forward-thinking farmers who were already leading the way in sustainability.

Silvano Locardi details the practical implementation of the project. Bayer started by matching their portfolio to the six outcomes, using products like Downy Mildew control with proper dosage and application. Crucially, the strategy was not based solely on Bayer products but included commercial products available in the market to ensure the growers' profitability. They utilized decision support systems (DSS) to manage buffer zones and dosages. The project started by focusing on the first, second, and third outcomes (Biodiversity), which involved the partnership with XNatura.

Silvano introduces the four specific partners involved in the project, chosen to represent different regions and approaches. In Piedmont, they partnered with "Montaribaldi," a typical farm with 35 hectares of grapes, and "Fontanafredda," a large-scale operation with over 100 hectares utilizing an organic approach. In Veneto, the partner is "Marcello del Maino," a multi-crop farm managing grapes alongside extensive crops. In Tuscany (Chianti Classico region), the partner is "San Felice," an organic-driven estate with more than 100 hectares. The project involved diffusing the mission through meetings and co-creation sessions with these farmers. A primary need identified by all partners, regardless of size or method, was the monitoring of biodiversity in the vineyard.

They decided to use the XNatura platform for this purpose, which was a point of significant interest for the growers as it was their first time applying such monitoring tools. Additionally, they implemented new products like biostimulants to manage water stress, driven by the "Grapevision" DSS. The 2025 season served as the "zero season," establishing a baseline for biodiversity. They shared the status with growers in November and plan to apply new strategies in the next season to measure results, specifically looking for changes in pollinator biodiversity. Future steps include maintaining Bayer’s leadership in this space, enlarging the partner network, expanding to different wine clusters across Italy, and expanding monitoring beyond pollinators to other biodiversity types.