Multi-Dimensional Environmental Monitoring Implications for Sustainable Land Management

FACTA UNIVERSITATIS Series: Economics and Organization Vol. 22, No 3, 2025, pp. 215 - 228 https://doi.org/10.22190/FUEO250904014L © 2025 by University of Niš, Serbia | Creative Commons Licence: CC BY-NC-ND Original Scientific Paper MULTI-DIMENSIONAL ENVIRONMENTAL MONITORING IMPLICATIONS FOR SUSTAINABLE LAND MANAGEMENT 1 UDC 502.521 502.174:502.173 Polina Lemenkova Alma Mater Studiorum - Università di Bologna, Dipartimento di Scienze Biologiche, Geologiche ed Ambientali, Bologna, Italy ORCID iD: Polina Lemenkova https://orcid.org/0000-0002-5759-1089 Abstract. Mountain areas of Italy experienced substantial socio-ecological changes over the last four decades. Partly due to outmigration, traditional agro-forestry activities and cultural landscapes were progressively abandoned and forest expanded substantially. Biodiversity trends observed during the last 30 years of rewilding are unlikely to continue under climate change. Understanding the decadal impacts of rewilding remains challenging. Biodiversity monitoring data are scant, scarcely integrated and scattered. Some vegetation plot time-series do exist, but they are essentially point observations in space. Achieving a complete picture of biodiversity change requires integrating different data, expertise, and mostly, viewpoints. Integrating biodiversity surveys with data that is available with complete spatial coverage, such as remote sensing, is a necessary first step. But interpreting biodiversity changes also requires considering people’s perceptions and knowledge. Only a qualitative-quantitative approach allows us to understand not only the patterns and trends of biodiversity change, but also the reasons why it changed. This study proposes an interdisciplinary, data-driven approach to understand historical and future trends of plant biodiversity in Italian rewilded mountain areas. We produced spatially explicit assessments of vegetation change between 1990 and 2020 in three mountain protected areas spanning the Italian Peninsula: 1) Eastern Alps (Lagorai mountain range); 2) Northern Apennines (Foreste Casentinesi Monte Falterona e Campigna NP), and 3) Central Apennines (Velino Massif). To do so, we used generalized dissimilarity modelling (GDMs) to model the dissimilarity of plant assemblages as a function of their geographical distances, ecological dissimilarities, and differences in land cover derived by remote sensing. After calibrating the GDMs with historical and newly collected vegetation plot data, the maps of predicted species composition at different times were created, whose comparison allowed highlighting areas where change in land cover induced the largest changes in species composition, and quantified the consequences for regional plant diversity. Key words: environment, risk assessment, ecology, mapping, sustainability JEL Classification: Q51, Q52, Q54, Q56 Received September 04, 2025 / Accepted November 10, 2025 Corresponding author: Polina Lemenkova Alma Mater Studiorum - Università di Bologna, Dipartimento di Scienze Biologiche, Geologiche ed Ambientali, Via Zamboni, 33 - 40126 Bologna, Italy | E-mail: [email protected] 216 P. LEMENKOVA 1. INTRODUCTION 1.1. Summary This project focuses on multi-hazard geological risk assessment in the 3 pilot regions of Italy. Nonlinear nature of environmental-geological hazards, such as climate change, temperature rise, earthquakes and landslides, causes challenges in modelling the occurrence and forecasting of risks. Accurate monitoring of natural hazards and risks becomes only possible using holistic integration of complex methods: cartographic, statistical and narrative ones. In this study, we employ such approach to multi-hazard environmental assessment of pilot Italian landscapes through automated algorithms of GIS, statistical analysis and information extraction. We analyse and employ novel approaches of computational GIS to environmental data analysis for processing datasets available from environmental monitoring centers of National Parks and National observation centers of Italy. The estimated hazards include extreme weather (strong winds), natural disasters (earthquakes and related landslides in the mountains), and environmental issues (land cover change). This project performs environmental risk assessment and thematic mapping, with the aim to contribute to the development of ecological early warning (EEW) for mitigating natural vulnerability in Italy. 1.2. Background During recent decades, the landscapes of the Earth have been affected significantly due to the cumulative effects from the geological and climate hazards. Among them, natural hazards remain some of the most devastating and powerful with destructive consequences for society. In this regard, environmental risk assessment methods in the Mediterranean region are crucial for the safety and mitigation of possible threats caused by climate change and environmental instability. One of the important characteristics of climate activity is landscape dynamics, because vegetation responds to the fluctuations in temperature and precipitation. Since climate variables can be recorded and evaluated, their measurements are important characteristics for evaluating environmental sustainability. From the middle of the 20th century, environmental monitoring in Italy has undergone active development. This included observation and mapping of active geologic structures and assessing potential risk of seismic hazards and frequency of earthquakes which has become possible on the regular basis for protection of the society. Recent advancements in technological development, including machine learning (ML), enabled notable positive developments in seismic monitoring. This becomes possible also due to the increased precision of the seismic instruments and availability of data collection for retrospective monitoring and prognostic forecasting of land cover changes using datasets. The development of GIS and natural risk assessment methods has been reported previously. Italy is one of the countries most vulnerable to such forces, especially in the mountains, which explains the importance of Earth observation monitoring for environmental risk assessment. 1.3. Research formulation 1.3.1. Research objective The objective of this research is a GIS-based analysis of plant biodiversity and ecosystem change in protected areas of Italy. To this end, the goal is to evaluate the land Multi-Dimensional Environmental Monitoring Implications for Sustainable Land Management 217 use impacts and scenarios using integrated methods of cartographic analysis and fieldwork. The overall objective of this study is to understand historical trends and future outlooks of plant diversity and ecosystems in Natura 2000 mountain areas undergoing spontaneous rewilding. The specific objectives are the following: i. to assess how temporal changes in land cover affected the composition and spatial distribution of plant biodiversity, highlight areas of maximum compositional change and quantify the proportion of species for which rewilding substantially improved or worsened the conservation outlook; ii. to disentangle and characterize the multiple pressures driving habitats and biodiversity changes and develop narratives for interpreting these changes; iii. to predict the impacts of different climate change scenarios on the future distribution of vegetation, combining vegetation vulnerability and exposure to future climate change to identify the areas with the higher risk of vegetation change. 1.3.2. Research questions Rewilding has had far-reaching effects on all the domains of biodiversity, especially vegetation, but many questions remain unanswered. In this study, we aim at answering the following research questions: ▪ Did rewilding increase overall plant and habitat diversity in Natura 2000 areas? ▪ How did these changes affect landscape mosaics? ▪ Did semi-natural ecosystems gain in resilience to climate change? Answering these questions is crucial for understanding the applicability of rewilding to other mountain areas of Europe. 1.3.3. Research goals This project conducts the multi-hazard environmental risk assessment in Italy. It is an interdisciplinary data-driven research project aiming to understand the geological hazard risks in selected 3 areas of Italy (north, central and southern regions). The target areas span the Alpine-Apennines region: 1. The Lagorai mountain range, 2. Foreste Casentinesi and 3. Velino Massif. We evaluated the threat of multiple geo-environmental hazards in Italy, including landscape dynamics, biodiversity, land cover change and climate-related effects. We consider the potential interactions of geologic and climate risks and their impacts on the society and built environment (structures and roads). The complex approach is essential for disaster preparedness and efficient urban planning of Italy, because the region of Mediterranean has high vulnerability of landscapes to environmental risks. Vulnerable landscapes create high exposure at environmental-climate hazards due to susceptibility of ecosystem components (vegetation, soil, topography) to effects from climate and changing environmental settings. 218 P. LEMENKOVA 2. METHODOLOGY 2.1. Work packages The study project is composed of five Work Packages (Wps), Figure 1. Fig. 1 Study workflow Source: author. Gathering predictors of biodiversity change and reconstructing land cover history of the study areas is performed using remote sensing data analysis. For each study area, we gathered a set of potential predictors of biodiversity change, making use of available data such as SoilGrid and CHELSA, but also available historical maps and aerial photography. Decadal collections of the RS imagery was obtained by the Copernicus and Landsat 8-9 OLI-TIRS archives through Google Earth Engine and reported in the existing papers. All vegetation plots were matched to their corresponding pixel or set of pixels in the satellite images of the year these plots were sampled, as a proxy for land cover. 2.2. Study Area The study region is located in three case studies, Figure 1. In this project, we focused on 3 different geographical regions of Italy. Collectively, the data were gathered from the sources on ecology, plant taxonomy, geoinformatics, habitat interpretation, big data analysis and management, biodiversity modelling, land use, remote sensing, and geography of Italy. This mix of multi-source data is central to our interdisciplinary research ethos, and we strive to ensure the multidisciplinary nature of this approach. Multi-Dimensional Environmental Monitoring Implications for Sustainable Land Management 219 Fig. 2 Three study areas in Italy Source: author. Mapping: QGIS. As the overall project goal, we employed the expertise in environmental and social science, biodiversity monitoring, ecosystem modelling, climate change adaptation and mitigation, and qualitative and quantitative sociological research. This, combined with the finely-tuned capabilities to conduct large-scale interdisciplinary projects, ensures a broad and in-depth oversight of all deliverables. The data on forest ecology and vegetation were used for biodiversity modelling, analysis of land use, conservation biology and macroecology. The protection regime guarantees a limited anthropogenic disturbance since agro-pastoral activities are declining year after year, and the only human activities in the area at the moment are horse grazing and hiking. The vegetation of the area is represented by a set of communities, which in some cases constitute serial relationships representing stages with different evolution or characterized by a different disturbance regime. In this study area we aimed to resurvey 84 vegetation plots of 2x2 m established in 2006 and monitored also in 2011 and 2016. During the first field season, in late spring of 2024, the working group responsible for the Research Unit focused on the lower half of the study area, namely from 1100 and 1554 m asl, and resurveyed plots. 2.3. Workflow To achieve research objectives, the data management and integrity included logical steps related to database preparation and implementation along with the ecological metadata infrastructure. Collection of new vegetation plot data in the field and identifying 220 P. LEMENKOVA plant specimen was performed during the fieldwork campaign. All were implemented logically during the phases of data collection, analysis, manuscript preparation and dissemination. As a result, we set up the following workflow: (a) Model biodiversity changes using satellite images; (b) Highlight areas of maximum compositional change; (c) Disentangle the relative contribution of different drivers of biodiversity change; (d) Combine qualitative and quantitative data in multidisciplinary project; (e) Develop narratives for interpreting these changes. 2.4. Approaches The sampling campaigns were conducted in summer period to maximize the output of vegetation monitoring. This study reflects the cutting-edge research project that integrates the vegetation science and biodiversity conservation, but is also complementary in terms of GIS expertise and mapping in Italy. To achieve a deeper understanding of the multiple pressures driving environmental and biodiversity change in the study areas, we integrated quantitative estimation to stimulate stakeholder participation and capture qualitative expert opinions on the accuracy and interpretation of our results. To do this, we implemented the innovative go-along walking methods, an increasingly used research tool for capturing data relating to people's experience, knowledge and attitudes to surrounding landscapes. The qualitative-quantitative results were used to prepare the analysis of the impacts of rewilding-induced land cover change over the last thirty years on the study areas in the Foreste Casentinesi NP, in the Lagorai range, and one summarizing the lessons learnt and ways forwards across all areas. The study work is focused in three study areas: 1. The Lagorai mountain range is placed in eastern Trentino and represents the southwestern-most outpost of the Dolomites. 2. Foreste Casentinesi, Monte Falterona and Campigna National Park (Foreste Casentinesi NP, hereafter for brevity) is located in the Northern Apennines. 3. The Velino Massif is located within the Sirente Velino Regional Park, one of the largest regional parks in the Central Apennines. Gathering future climatic predictors under alternative scenarios We integrated gathered set of bioclimatic variables and calculated climate change velocity. To obtain alternative future predictions, two distinct socio-economic pathways were considered. For each one, we used three different general circulation models (the GFDLESM4, the IPSL-CM6A-LR and the MRI-ESM2-0) to account for predictions’ uncertainties across GCMs. These data were used to create thematic maps of climate change. Spatial downscaling of current and projected bioclimatic variables Bioclimatic variables are only available at a relatively coarse spatial resolution (2501000 m), much coarser than the resolution of typical vegetation plots (4-100 m). To avoid this scale mismatch, we will downscale all the spatial predictors to a 20 m of resolution using geographically weighted regression models. Multi-Dimensional Environmental Monitoring Implications for Sustainable Land Management 221 Fitting statistical models to identify how climate contributes to the observed temporal trends We tested whether the climatic conditions of a vegetation plot, when originally sampled, are useful for predicting observed variables using established statistical models (GLMs, GAM). The relative contribution of single climate predictors will be assessed using variation partitioning, and model accuracy quantified through root mean square error (RMSE). In this way, the areas where the highest vegetation change can be expected were highlighted. Creating maps of climate change related risk for vegetation models The maps were used to predict the near future using current bioclimatic variables as predictors, and to provide a spatially explicit estimation of vegetation change over time. By highlighting areas where vegetation is predicted to change the most under climate change, we will create maps of vegetation vulnerability. These are combined with maps of exposure to climate change and maps of climate change velocity. Final task includes the producing of maps showing the risk of vegetation changes under alternative climate change scenarios, as the combination of vulnerability and exposure. 3. RESULTS AND DISCUSSION By comparing the relative importance of land cover and other predictors in 3 case studies, we have disentangled the drivers of historical vegetation change and understood the consequences of landscape abandonment on plant biodiversity, including climate change. 3.1. Case study 1 - Foreste Casentines National Park The selection of the areas where the data were collected for vegetation plot dataset in the field was based on the data retrieved in environmental monitoring, and later consolidated in a database, Figure 3. The updated vegetation-plot database was built for Fig. 3 Location of sample points in the 1st study area in Italy Source: author. Mapping: QGIS. 222 P. LEMENKOVA the study area of Foreste Casentines National Park. The vegetation plot data from the relevant sources were retrieved, aggregated and georeferenced. A workflow for taxonomic standardization was created to identify the spatial and temporal gaps in the data gathered in previous steps (Bartolucci & Conti, 2013). A survey scheme was defined for collecting additional data. The results included the data collected in the field to increase the spatial and temporal coverage of existing vegetation plot data and information on flora distribution (Viciani et al., 2010). The spatio-temporal gaps were identified in the vegetation data to check whether the vegetation data aggregated in the dataset are representative of the three study areas, both in terms of geographic distribution and habitats sampled. Maps of land cover (CORINE Land Cover) or habitat distribution (EUNIS) were used as a benchmark, following the successful use of CORINE in previous studies (Lemenkova, 2025). The maps were created to illustrate sampled areas in each study region. The resampling of a representative subset of historical vegetation plots sampled in the 90s or earlier was prioritized among those having sufficient positional accuracy. We defined survey protocols and a sampling design for the study area. After the identification of the areas to sample, and historical plots to resurvey, we defined a standard field survey protocol to be employed in future research. A statistically sound sampling design was developed to maximize the compositional and environmental variability of the area to sample. The sampling efforts were defined as a function of the following three criteria: a) the size of the study areas, b) the number of vegetation plots available, and c) the spatial and habitat representativeness of existing data. This was used to perform a gap analysis aimed at identifying those geographical areas in the three study sites having climatic conditions dissimilar from those vegetation plots where the data were located. The gap analysis, together with in-depth discussions with some of the engaged stakeholders, was used to prioritize the areas where to conduct fieldwork. Fieldwork started in the month of May 2024 in the mountainous area of Foreste Casentines National Park. In the boundaries of the Foreste Casentines National Park, two priority areas were identified for data collection: 1. Biotopo di Capria - on the northern side of the National Park. This is a small valley, which was proposed for strict protection as early as 1972, 20 years before the establishment of the National Park. Here, some old vegetation plot data were retrieved for the areas and in contact with the National Park authority, we deemed relevant to resample these areas to gain insight into the evolution of vegetation in the last two decades. 2. Monte Falterona area - in the SW corner of the National Park. An agricultural area owned by the "'ISMEA (Istituto di Servizi per il Mercato Agricolo Alimentare)", which for the last three decades has been rented out to a small agricultural enterprise ("Azienda Agricola Falterona"). This enterprise used the land extensively as a mosaic of pastures, forest plantations, and (to a smaller extent) cropland. The area has been inaccessible to botanists for the last three decades and very little vegetation data is available. The next part of the results included the collection of new vegetation plot data in the field. Based on the sampling design and survey protocols, new vegetation plots were collected in the field. The compatibly with the available resources was achieved through sampling a minimum of 40 new vegetation plots and resampling a minimum of 20 historical plots across all areas. All vegetation plots were sampled at the peak of the vegetative season (spring-summer). Multi-Dimensional Environmental Monitoring Implications for Sustainable Land Management 223 Since 2023 the area is under the custody of the National Park, while ISMEA finalizes an auction to sell off the land. We therefore chose the area as a priority for additional data collection, especially since it contains several abandoned pastures being encroached by shrubland and young, developing forests, all vegetation types severely underrepresented in our database. The vegetation plot data for the study area was retrieved from the existing databases. As a first step for building an electronic database of georeferenced vegetation-plot data for the study areas, a fine-tuned search for already-digitized data was performed by querying the main Italian vegetation-plot repositories (VegItaly, AMS-VegBank, VPD-Sapienza, CircumMed), and negotiating with database holders the conditions for accessing the data. 3.2. Case study 2 – Regional Park Sirente Velino The study area is located in the Abruzzo region (Central Italy) in Velino, Figure 4. Fig. 4 Location of sample points in the 2nd study area in Italy: Regional Park Sirente Velino, red dots indicate the plot surveyed during May-July 2024 Source: author. Mapping: QGIS. The Velino Massif partially overlaps with a Natura 2000 site; it extends over around 450 ha, mainly inside the National Natural Reserve of Mount Velino which is included inside the Regional Natural Park of Sirente-Velino. More specifically, the area subject to resurvey activities is located on the south-western slope of Mount Velino, between Vallone di Sevice and Rave della Chiave, and covers an altitudinal range from 1100 to 2400 m asl (Anzalone & Veri, 1975). The massif encompasses an elevational range from ca. 1,000 m to 2,487 m asl, and has a complex orographic structure partly determined by quaternary glaciations. This complexity offers a wide range of microclimatic conditions, resulting in very high plant and habitat diversity: the massif hosts more than 650 plant species and 20+ Natura 2000 Habitats. The area was intensely grazed for centuries and is still almost totally treeless. Analysis of temporal changes of plant communities through vegetation plot time series. The Velino study area was used as a benchmark for studying the species that increased in occurrence or abundance over time, and the most vulnerable plant species were identified. For all vegetation plots where more than one temporal survey is available, the change in species richness was calculated, gained and lost species identified and temporal changes estimated. We used the existing approaches of Shannon’s index of diversity (Ercanli, 2018; Haller, & Bender, 2018), Pielou’s index of evenness (Jászayová et al., 2024; Blanco et al.,