Previous Article in Journal
Geographical Storytelling: Towards Digital Landscapes in the Footsteps of Cuchlaine King
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Geographies
Volume 5
Issue 2
10.3390/geographies5020026
geographies-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Influence of Pasture Diversity and NDVI on Sheep Foraging Behavior in Central Italy

by
Sara Moscatelli
1,2,*,
Simone Pesaresi
3,
Martin Wikelski
4,
Federico Maria Tardella
2,
Andrea Catorci
2 and
Giacomo Quattrini
3
1
International School of Advanced Studies, PhD Course “One Health” University of Camerino, Via Madonna delle Carceri 9, 62032 Camerino, Italy
2
School of Biosciences and Veterinary Medicine, University of Camerino, Via Pontoni 5, 62032 Camerino, Italy
3
Department of Agricultural, Food and Environmental Sciences, D3A, Università Politecnica delle Marche, Via Brecce Bianche 12, 60131 Ancona, Italy
4
Department of Migration, Max Planck Institute of Animal Behavior, 78315 Radolfzell, Germany
*
Author to whom correspondence should be addressed.
Geographies 2025, 5(2), 26; https://doi.org/10.3390/geographies5020026
Submission received: 29 March 2025 / Revised: 24 May 2025 / Accepted: 13 June 2025 / Published: 16 June 2025
Browse Figures
Versions Notes

Abstract

:
Pastoral activities are an essential part of the cultural and ecological landscape of Central Italy. This traditional practice supports local economies, maintains biodiversity, and contributes to the sustainable use of natural resources. Understanding livestock behavior in response to environmental variability is essential for improving grazing management and animal welfare and ensuring the sustainability of these systems. This study evaluated the movement patterns of sheep grazing on pastures with differing vegetation indices in the Sibillini Mountains. Twenty lactating ewes foraging on two different pastures were monitored from June to October 2023 using GPS collars and accelerometers. GPS tracks were segmented using the Expectation Maximization Binary Clustering (EmBC) method to characterize movement behaviors, such as foraging, traveling, and resting. The NDVI was used to characterize vegetation dynamics, showing notable differences between the two pastures and across the grazing season. Additive mixed models were used to analyze data, accounting for individual variability and temporal autocorrelation in the sample. The results suggest that variations in the NDVI influence grazing behavior, with sheep in areas of lower vegetation density exhibiting increased movement during foraging. These findings provide valuable insights for optimizing grazing practices and promoting sustainable land use.

Graphical Abstract

1. Introduction

Sustainable pasture management is crucial for maintaining biodiversity, ecological resilience, and economic productivity in mountain regions [1]. Pastures in marginal environments such as the Central Apennines serve as habitats for plant and animal species while sustaining traditional pastoral practices that are deeply embedded in the European culture [2]. However, climate change, the abandonment of agricultural practices, and landscape transformation are profoundly altering these ecosystems, making the need to develop innovative tools for monitoring and managing pastoral resources increasingly urgent [3].
Domestic sheep (Ovis aries) and related species, such as mouflons (Ovis orientalis) and bighorn sheep (Ovis canadensis), exhibit specific movement patterns during foraging that are influenced by resource availability, climatic conditions, and topographic features. Sheep tend to adopt area-restricted search foraging strategies in high-quality foraging patches, reducing movement speed and turning frequency to maximize energy intake [4,5]. On the contrary, in areas with low-quality forage, sheep adopt extensive foraging strategies that involve faster movement and a reduced grazing time in specific locations, allowing them to cover larger areas in search of better resources [4,5]. This adaptive behavior is linked to spatial memory and environmental assessment in optimizing foraging efficiency under varying resource conditions [6].
However, the increasing aridity of pastures due to climate change is a potential threat for the welfare and behavior of grazing animals, particularly sheep and other ovine species. As forage availability and quality decreases in Mediterranean environments, animals must adapt their foraging patterns to meet their nutritional needs, with the risk of increasing energy expenditure and stress [7]. As highlighted in the study by Bailey et al. [4], sheep reduce their grazing time and increase walked distances in degraded pastures in order to optimize energy intake in environments where resources are limited and unevenly distributed. These behavioral adaptations can have negative impacts on animal welfare, as prolonged periods of nutritional stress and increased energy expenditure may result in weight loss, reduced reproductive success, and increased susceptibility to disease [8].
Recent advances in remote sensing technologies, coupled with the increasing availability of open access satellite data (e.g., Landsat, MODIS, and Sentinel) [9], have greatly improved the monitoring of these ecosystems [10,11]. Satellite platforms provide high-resolution, multi-temporal data, allowing for the generation of dense time series of reflectance spectral bands and vegetation indices, which are essential for capturing vegetation dynamics over seasonal and interannual scales [12,13,14,15,16,17]. Among traditional vegetation indices, the Normalized Difference Vegetation Index (NDVI) [18,19] has emerged as a key metric for tracking phenological changes due to its high sensitivity to biomass stress [20,21,22]. For this reason, NDVI time series have been widely used to monitor variation in pasture quality, even in Mediterranean ecosystems [23,24,25,26].
Since grazing animals depend on the availability and spatial distribution of vegetation resources, which change seasonally, understanding these variations is essential for optimizing pasture management. Hurley et al. [27] utilized NDVI time series to examine how spring and autumn phenology influenced the winter survival of mule deer fawns, demonstrating the fundamental role of vegetation seasonality in foraging ecology. Similarly, sheep adjust their movement patterns in response to fluctuations in pasture quality, seasonal dynamics, and environmental conditions [4,28,29,30,31].
The study of animals’ movement through innovative technologies such as satellite radio collars allows for the spatial and temporal patterns of sheep grazing behavior to be tracked and analyzed, providing insights into how environmental factors, such as topography, vegetation, and climate, influence their movement [4,5]. Satellite radio collars have enabled researchers to collect high-resolution data on animal movement, including GPS coordinates, tri-axial acceleration, and environmental metrics [32,33]. This data can be used to analyze not only location data but also energy expenditure and activity patterns [33,34]. Some devices also monitor environmental and physiological metrics, including ambient temperature and body temperature, which can shed light on how sheep adapt to climatic stress or seasonal changes [35,36]. The integration of vegetation and movement satellite data has recently opened new opportunities for the real-time monitoring of animal behavior, enabling the identification of movement patterns and their relation with environmental conditions [37,38], but the application of these techniques to investigate the relationship between resource availability and livestock behavior remains largely unexplored. Integrating satellite-derived vegetation indices with animal movement tracking could open new perspectives for a more comprehensive approach to pasture management. This integration enables not only the quantification of pasture productivity and quality but also the assessment of how animals respond behaviorally to forage availability fluctuations.
This study combines remote sensing and animal behavioral ecology to evaluate differences between two pastures in the Central Apennines. NDVI time series were used to characterize vegetation dynamics, while GPS tracking monitored sheep foraging behavior.
Specifically, the objectives were to
  • Evaluate differences in NDVI values between the two pastures and their seasonal trends;
  • Analyze the influence of pasture conditions on the distance traveled by sheep during foraging.
This multidisciplinary approach provides an innovative perspective on the interaction between vegetation and grazing behavior, offering valuable insights for the sustainable management of mountain pastures.

2. Materials and Methods

2.1. Study Area

The sheep are farmed in Central Italy, in the Sibillini Mountain National Park area. The Monti Sibillini National Park is located between two Italian regions, Marche and Umbria. The extent of the area is approximately 697 km2, which includes numerous mountains above 2000 m, the highest of which is Monte Vettore at 2476 m, and the average altitude of the area is 1173 m a.s.l. Animals were divided into two different areas (Figure 1):
-
“Delizie dei Fratelli Angeli”company
Sheep in this area are farmed by “Delizie dei Fratelli Angeli” company in Pieve Torina. The area consists of hilly grasslands, ranging in elevation from a minimum of 500 to a maximum of 750 m a.s.l. This area was denominated “pasture 1”. At the time of the study the flock consisted of approximately 100 sheep in total, including 50 Comisana sheep and 50 Sopravissana sheep.
-
“Azienda Agricola Piselli” company
Sheep in this area are managed by “Azienda Agricola Piselli” in Monte Cavallo. The area is adjacent to the border of Montagna di Torricchio State Reserve, a protected area managed by the University of Camerino. The area grazed by sheep is a mountain slope that ranges from a minimum of 800 to a maximum of 1400 m a.s.l. This area was denominated “pasture 2”. At the time of the study the flock consisted of approximately 95 animals in total, and the sheep included a mix of productive breeds bred for milk production.
Following Pesaresi et al. [39], the study area belongs to the remperate sub-Mediterranean bioclimatic unit, with a weak sub-Mediterraneity, characterized by the alternation of winter cold stress and summer drought stress [40]. It extends within the lower supra-temperate belt with an upper humid ombrotype and the upper meso-temperate belt with a lower humid ombrotype. The mean annual rainfall is about 1100–1300 mm and the average annual temperature is 9–11 °C in the former unit, while the average annual temperature is 11–13 °C and the mean annual rainfall is 850–1100 mm in the latter unit. The growing season (number of days with minimum temperature over 6 °C) lasts 150–180 days yr-1 and 180–210 days yr-1, respectively [41].
The vegetation is composed of different secondary grasslands, interrupted by small copses and croplands. From a phytosociological viewpoint, the pastoral landscape is referred to Erysimo-Jurineetalia bocconei S. Brullo 1984 order (Festuco hystricis-Ononidetea striatae [41] class sensu Mucina et al., 2016) [42] and Arrhenatheretalia elatioris Tx. 1931 order (Molinio-Arrhenatheretea class). North-facing slopes are covered by semi-mesophylous communities dominated by hemicryptophytes and forbs with a prolonged and late blooming period, whereas south-facing slopes are covered by xerophilous communities rich in annual and chamaephyte species, with an open canopy interrupted by patches of bare soil.

2.2. Study Sample

We applied radio loggers (Table 1) to 20 sheep at the beginning of June 2023 before starting the outdoor season, which ended in October 2023.
Animals were divided into two groups:
  • Ten Comisana lactating sheep on pasture 1;
  • Ten mixed-breed lactating sheep on pasture 2.
All the sheep were adult females, 3 years of age, in lactation. The mixed-breed sheep were compared to the Comisana sheep, both being productive breeds selected for milk production and similar in size. All animals were subjected to semi-extensive pastoral management techniques, kept inside the barn during winter and on extensive pasture from June to November, according to the suitability of the season. During the extensive period the animals were farmed with rotational grazing practices. The sheep spent the night inside small resting paddocks to prevent wolf predation and moved every morning to the foraging area, led by the shepherd, and then came back every afternoon. No paddocks were used during foraging time, but the animals were led to different locations inside the pasture area according to the season progression; in particular, the sheep were moved based on the area’s forage availability and relocated when the forage resource was depleted to ensure the most efficient pasture exploitation.
All the experiments were carried out in accordance with relevant guidelines and regulations. The protocols were in accordance with Legislative Decree No. 146, implementing Directive 98/58/EC of 20 July 1998 concerning the protection of animals kept for farming purposes.

2.3. Animals’ Movement

We collected GPS data every 3 h from the 8 June 2023 to the 9 February 2024. All GPS data used during the study are available on Movebank (movebank.org, study name “ Transhumance and climate Marche 2023”, 2023–2024) [43].
After importing the data into R, they were projected and filtered to include only the period when the sheep were outdoors (June to October). Breed and physiological state information were added for each animal on Movebank before importing data into R, as well as points too far from the study areas, which were marked as outliers. Duplicates were then removed from the dataset. The data from collars 4828 (Comisana sheep on pasture 1) and 4846 (mixed breed on pasture 2) were excluded from subsequent analyses due to functional issues with the collars; in particular, collar 4828 did not record any data, and collar 4846 was found broken on the pasture. We calculated the distance moved, angle turned, and speed for all of the animals to conduct behavioral segmentation using the move2 R package [44] functions mt_distance(), mt_speed(), and mt_turnangle(), and movement metrics were derived from GPS data.

Behavioral Segmentation

We segmented movement data using the Expectation Maximization Binary Clustering (EMbC) algorithm, which classifies GPS locations based on the linear speed and turning angle into four movement categories: high speed–high turning angle (HH), high speed–low turning angle (HL), low speed–high turning angle (LH), and low speed–low turning angle (LL). Given the 3 h GPS sampling interval, we assumed that the turning angles approximate general changes in movement direction rather than capturing detailed, short-term movement dynamics. Nevertheless, the animals followed a highly structured daily routine, moving as a group under the guidance of the shepherd to and from grazing areas. This means it was unlikely that individuals reversed their direction during these phases. Accordingly, we interpreted the HL category (high speed, low turning angle) as indicative of directed movement (hereafter called “traveling”). Although we recorded GPS fixes every 3 h for each individual, sampling times were not synchronized across animals, which provided broader temporal coverage at the group level. This allowed us to identify time windows during which a substantial proportion of individuals were simultaneously classified as traveling (HL), consistent with coordinated group-level displacements. Instead of using EMbC categories to define other behavioral states, we adopted a time-based classification informed by the pastoral context. We classified as resting the periods before and after the main displacement phases, which included inactivity within the night enclosure and low-mobility behaviors outside grazing hours. Based on the animals’ daily routines, we identified the foraging period as the time between the two main traveling phases detected by EMbC, corresponding to the interval from 07:00 to 16:00 when sheep were consistently present in the grazing areas across both sites. Segmentation using Hidden Markov Models (Hmms) was attempted; however, the nature of the data did not allow us to choose the initial parameters; for this reason, it was not possible to ensure reliable results with this segmentation method.

2.4. Environmental Variables

NDVI

The Normalized Difference Vegetation Index (NDVI) was used as a proxy for vegetation quality and availability in the two study areas. NDVI is widely recognized as a predictor of herbivore movement and space use [20,45] and was employed to assess the seasonal dynamics of vegetation across the two pastures. Twenty cloud-free NDVI images from the Sentinel-2 satellite, covering the period from June to October 2023, were analyzed. Sentinel-2, with a spatial resolution of 10 m and a revisit time of five days, is well-suited for capturing fine-scale temporal changes in vegetation [46]. From all the available Sentinel-2 images, to ensure data quality, pixels affected by clouds, cloud shadows, and snow were excluded. For each image, we calculated the mean NDVI value across all valid pixels within each pasture area. These mean values were then aggregated on a weekly basis to generate a time series of average NDVI values for each pasture. The weekly aggregation allowed for a detailed representation of vegetation dynamics, providing insight into short-term fluctuations in forage availability throughout the grazing season.

2.5. Statistical Analysis

2.5.1. Differences Between Pastures

To assess differences in NDVI levels between the two pastures over the season, we fitted a generalized linear model (GLM) with a beta distribution. The beta distribution is appropriate for modeling continuous data constrained between 0 and 1, making it well-suited for NDVI values in the sample. The analysis was conducted using the betareg function (logit link, link.phi = identity) of the betareg package (version 3.2-1) in R [47]. The model included NDVI values as dependent variables, while pasture (pasture 1 and pasture 2), month, and their interactions were included as fixed effects to capture seasonal variations in NDVI between the two pastures. Residuals were visually examined for patterns.

2.5.2. Distance Moved During Foraging

We analyzed segmented tracks (Foraging) to check for the influence of different pastures and NDVI values on the distance moved by animals during foraging. We analyzed the data using generalized additive mixed models (GAMs) to account for the presence of temporal autocorrelation in the data, using the gamm function of the R mgcv package (version 1.9-1) [48]. Log-transformed weekly distance was set as the dependent variable in the model, while pasture ID, NDVI, month, and their pairwise interactions were included as independent variables. Individual ID and farm were included as random factors to account for individual variability and effect of management. Time, particularly day and week, were included as smooth terms. Significance was set as <0.05. We adopted an AIC (Akaike Information Criterion) for the model selection procedure using the AIC function of the stats (version 3.6.2) to compare the full model and reduced models and selected the model with the lowest AIC, which was the reduced model. We included the pasture ID, NDVI, and their interaction as fixed effects.
We visually examined the autocorrelation of the residuals to check for temporal autocorrelation using the acf function of the R-package stats (version 3.6.2). No autocorrelation was detected. Residuals were plotted against fitted values to assess homogeneity and identify potential patterns using the gam.check function of the mgcv R-package [48]. The residuals showed no clear trends or heteroscedasticity. The distribution of residuals was visually inspected using histograms and Q-Q plots.

3. Results

3.1. Differences Between Pastures

The results of the generalized linear model (Table 2) highlight that pasture 2 was associated with significantly lower NDVI values throughout the season compared to pasture 1 (estimate = −0.26544, SE = 0.02358, p < 0.001).
All months included in the model (July, August, September, and October) showed significantly lower NDVI values compared to June (p < 0.001). The largest reduction in the NDVI was observed in August (estimate = −0.72458, SE= 0.02300, p < 0.001), followed by September (−0.54077, SE = 0.02211, p < 0.001), July (−0.41617, SE = 0.02292, p < 0.001), and October (−0.30669, SE = 0.02185, p < 0.001), indicating a seasonal effect on vegetation, as shown in Figure 2.
All interaction terms between pasture 2 and the months were significant and negative (p < 0.001), with estimates ranging from −0.10337 in July to −0.28106 in October.

3.2. Distance Moved During Foraging

The results of the generalized additive mixed model (Table 3) highlight that sheep foraging on pasture 2 moved significantly greater distances compared to sheep foraging on pasture 1 (estimate = 1.45374, SE = 0.29254, p < 0.001). The NDVI was negatively associated with the distance moved by sheep (estimate = −0.90237, SE = 0.18394, p < 0.001), suggesting that sheep moved shorter distances in areas with higher vegetation availability (Figure 3). The interaction between pasture 2 and NDVI_weekly was not significant (estimate = −0.02898, SE = 0.33542, p = 0.931), indicating that the relationship between the NDVI and movement distance did not notably differ between the two pastures.
Among the smooth terms, the week number had a significant non-linear effect on movement distance (edf = 8.852, F = 44.44, p < 0.001), indicating that the distance traveled by sheep varied significantly across weeks in a non-linear pattern.
In contrast, the smooth term for day of year (doy) was not significant (edf = 1.74 × 10−6, F = 0.00, p = 0.686).

4. Discussion

Our results suggest a relationship between forage availability, as reflected by the NDVI values, and livestock foraging patterns. Specifically, lower NDVI values were associated with greater distances walked by animals during foraging. The observed differences between pastures in terms of NDVI values and their seasonal dynamics align with studies using remote sensing to monitor vegetation productivity [24] and its effects on herbivore behavior [49]. All months showed significantly lower NDVI values compared to June. The largest reduction in the NDVI was observed in August, followed by September, July, and October, which reflects the phenological pattern observed in Mediterranean grasslands, with a negative peak observed in August linked to the summer drought [50]. Sheep grazing on pasture 2 moved significantly greater distances compared to sheep grazing on pasture 1, and the increase in the NDVI was associated with significantly smaller distances walked by animals. These findings are broadly consistent with previous research on herbivore foraging strategies, where animals adopt more extensive search patterns when resource availability is limited [51]. Studies have also shown that cattle tend to prefer areas with higher NDVI values, indicative of greater forage availability, thus influencing grazing patterns [4].
Our findings further support the potential value of an interdisciplinary approach: combining vegetation and animal movement data from satellite technologies enhances our ability to better understand animals’ responses to changing environmental conditions [52], while also informing the development of more effective and adaptive grazing management practices [53]. Integrating these methods may help identify critical foraging areas and periods of resource scarcity, enabling targeted interventions to improve animal welfare and optimize pasture utilization [54]. The combined use of the NDVI and movement ecology appears to represent a promising methodological framework for advancing our understanding of herbivore–vegetation interactions and promoting sustainable livestock management [8,25]. Furthermore, Hurley et al. [27] demonstrated that seasonal shifts in the NDVI play a crucial role in the survival of ungulates, emphasizing the need to integrate vegetation dynamics into grazing management strategies.

Limitations and Future Research

Although the results of the present study are promising, several limitations should be acknowledged. GPS locations were recorded every three hours to preserve battery life, which may have limited the resolution of fine-scale movement patterns and behavioral interpretations. Additionally, another limitation of the present study lies in the spatial scale of the investigation, which focused on two specific pastures with different structural characteristics. Notably, pasture 1 is more fragmented, with a greater presence of both natural and anthropogenic discontinuities, which may influence the spatial behavior of the flock. Such heterogeneity could affect sheep movement independently of vegetation productivity and should be accounted for when interpreting the observed foraging patterns. The non-significant interaction between the NDVI and pasture suggests that the relationship between vegetation cover and movement distance was consistent across both sites. However, further research is needed to disentangle the effects of vegetation from other environmental and management-related variables. Moreover, the study covers only a single grazing season. Since grassland productivity and animal behavior can be highly sensitive to interannual climatic variability, the patterns observed may not be fully representative.
Future developments should not rely solely on the NDVI to assess seasonal spectral dynamics. Instead, the full range of spectral bands available from satellite data should be utilized. Different grassland types exhibit distinct phenological responses across different spectral bands, and multivariate approaches such as multivariate functional principal component analysis (MFPCA) can be used to effectively capture joint temporal dynamics [14,55]. Furthermore, texture metrics derived from gray-level co-occurrence matrix (GLCM) techniques offer valuable insights into spatial heterogeneity at ecologically relevant scales, including those that influence foraging behaviors [56]. Combining these spectral and structural descriptors can substantially improve our understanding of landscape functionality and use across spatial and temporal dimensions.
Moreover, future studies could benefit from higher-resolution GPS data (e.g., with fixes every 30 min or less) and finer-scale NDVI products, which would improve both spatial and temporal resolution in analyses of animal–vegetation interactions. Grazing trajectory data contains rich spatiotemporal information, and analyzing the behavior of livestock based on spatiotemporal characteristics can establish a better connection between grassland nutrition and behavior [57]. In addition, detecting the synchrony/asynchrony between peaks of phytomass production and sheep activity may help increase the understanding of how grazing patterns influence pasture resilience and nutrient cycling. For example, the fixed intraseasonal resting periods might be decoupled from the key pulses of nitrogen mineralization and rainfall in the growing season, which can reduce their efficiency in providing a recovery period for grazed grasses [58]. Synchronizing foraging activity with the onset of favorable conditions for high-quality forage may facilitate the intake of nutrients and energy and help to prevent grassland dynamics and forage quality reduction [58].
The transition to high-resolution monitoring in grazing studies presents both unprecedented opportunities and considerable practical challenges. While traditional threshold-based methods, such as Vectorial Dynamic Body Acceleration (VeDBA), remain valuable, continuous monitoring systems now enable more nuanced and detailed behavioral analyses. For instance, accelerometer sampling at 10–15 Hz has been shown to distinguish grazing, resting, and locomotion behaviors with up to 90% accuracy [59,60].
However, this approach generates significant data volumes. Monitoring a single herd at 1 min intervals can produce approximately 1.8 terabytes of data per month, necessitating advanced data processing pipelines to manage, store, and analyze such datasets efficiently [61]. Moreover, the deployment of high-frequency GPS devices (with fixes every 1–5 min) typically demands 3–4 times the battery capacity of standard units, limiting deployment durations and requiring more frequent maintenance [62].
These technical constraints must be weighed against the scientific value of the data. Machine learning approaches, especially Random Forest algorithms applied to high-frequency movement and sensor data, have achieved high classification accuracies without relying on predefined behavioral thresholds [63,64]. These techniques are transforming the field of grazing ecology by offering more flexible and scalable behavior classification models. Nonetheless, they require extensive ground-truthing, usually involving 40–60 h of video-recorded behavioral observations per study site to effectively train and validate the models [61].
Further investigations should also consider forage quality in addition to quantity. As shown by Wang et al. [7], forage quality may be a primary driver of behavioral adjustments, with animals increasing grazing time in areas of lower nutritional value. Integrating measures of forage nutritional composition alongside spectral vegetation indices could thus yield a more comprehensive understanding of foraging strategies. Combining high-frequency movement data with fine-scale environmental information, such as terrain ruggedness, water point distribution, or shade availability, could significantly enhance predictive models of animal behavior under varying ecological conditions. Future research should consider including multiple years, which would allow for an assessment of the stability over time of the relationships between spectral vegetation indices and movement patterns. In addition, the approach presented here could be expanded to a broader range of pasture types and geographical contexts, enabling a more generalizable understanding of how landscape structure and vegetation dynamics influence livestock behavior. Comparative studies across different bioclimatic zones, management regimes, and pasture configurations would further strengthen the applicability of the method. While including more farms would improve statistical power and external validity, it remains a logistical and economic challenge. Nonetheless, this is an important consideration for future research aiming to strengthen inference and broaden the generalizability of the findings.

5. Conclusions

Our results suggest that lower forage availability is associated with increased movement during foraging, as animals adapt their behavior to optimize nutrient intake in resource-limited environments. In addition, the study underscores the value of a multidisciplinary framework that integrates animal movement data with environmental indicators such as the NDVI, offering a more comprehensive understanding of how livestock interact with their ecological context. This holistic perspective enhances our capacity to monitor and predict animal responses to environmental variability, thereby informing more efficient and sustainable grazing systems. By situating animals within dynamic landscapes, we can design management strategies that simultaneously support animal welfare, ecological integrity, and economic resilience. Looking forward, the integration of near-real-time remote sensing products with high-resolution GPS and accelerometry data opens new avenues for adaptive pasture management, allowing for the determination of the grazing intensity of a pasture, grazing areas, and paths and guiding the adjustment of the grazing time and duration. Such a fusion could enable responsive decision-making that aligns forage availability with animal needs in real time, offering a powerful tool for precision livestock systems. Once validated at broader spatial and temporal scales, this approach may significantly advance both academic research in ecological behavior and practical applications in climate-resilient livestock management.

Author Contributions

Conceptualization, S.M., A.C. and M.W.; methodology, S.M., G.Q. and M.W.; software, M.W. and G.Q.; formal analysis, S.M. and F.M.T.; investigation, S.M.; resources, M.W. and A.C.; data curation, S.M. and G.Q.; writing—original draft preparation, S.M. and G.Q.; writing—review and editing, S.M., S.P., A.C., M.W. and G.Q.; visualization, S.P.; supervision, A.C.; project administration, A.C. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Data Availability Statement

The original contributions presented in this study are included in the article. Further inquiries can be directed at the corresponding author.

Acknowledgments

The authors are grateful to the Max Planck Institute of Animal Behavior for the collars’ concession, to George Heine for the collar equipment, and to Kamran Safi for providing technical support in the data analysis.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
NDVINormalized Difference Vegetation Index
EmBCExpectation Maximization Binary Clustering
VeDBAVectorial Dynamic Body Acceleration

References

  1. Catsadorakis, G. The conservation of natural and cultural heritage in Europe and the Mediterranean: A Gordian knot? In Natural Heritage; Routledge: London, UK, 2013; pp. 1–13. [Google Scholar]
  2. Pardini, A.; Nori, M. Agro-silvo-pastoral systems in Italy: Integration and diversification. Pastoralism 2011, 1, 26. [Google Scholar] [CrossRef]
  3. Macdonald, D.; Crabtree, J.R.; Wiesinger, G.; Dax, T.; Stamou, N.; Fleury, P.; Lazpita, J.G.; Gibon, A. Agricultural abandonment in mountain areas of Europe: Environmental consequences and policy response. J. Environ. Manag. 2000, 59, 47–69. [Google Scholar] [CrossRef]
  4. Bailey, D.W.; Gross, J.E.; Laca, E.A.; Rittenhouse, L.R.; Coughenour, M.B.; Swift, D.M.; Sims, P.L. Mechanisms that result in large herbivore grazing distribution patterns. J. Range Manag. 1996, 49, 386–400. [Google Scholar] [CrossRef]
  5. Dumont, B.; Petit, M. Spatial memory of sheep at pasture. Appl. Anim. Behav. Sci. 1998, 60, 43–53. [Google Scholar] [CrossRef]
  6. Edwards, G.R.; Newman, J.A.; Parsons, A.J.; Krebs, J.R. The use of spatial memory by grazing animals to locate food patches in spatially heterogeneous environments: An example with sheep. Appl. Anim. Behav. Sci. 1996, 50, 147–160. [Google Scholar] [CrossRef]
  7. Wang, M.; Alves, J.; Tucker, M.; Yang, W.; Ruckstuhl, K.E.; Chaves, A.V. Effects of intrinsic and extrinsic factors on ruminating, grazing, and bedding time in bighorn sheep (Ovis canadensis). PLoS ONE 2018, 13, e0206664. [Google Scholar] [CrossRef]
  8. Festa-Bianchet, M. Seasonal range selection in bighorn sheep: Conflicts between forage quality, forage quantity, and predator avoidance. Oecologia 1988, 75, 580–586. [Google Scholar] [CrossRef]
  9. Young, N.E.; Anderson, R.S.; Chignell, S.M.; Vorster, A.G.; Lawrence, R.; Evangelista, P.H. A survival guide to Landsat preprocessing. Ecology 2017, 98, 920–932. [Google Scholar] [CrossRef]
  10. Corbane, C.; Lang, S.; Pipkins, K.; Alleaume, S.; Deshayes, M.; Millán, V.E.G.; Strasser, T.; Borre, J.V.; Toon, S.; Michael, F. Remote sensing for mapping natural habitats and their conservation status—New opportunities and challenges. Int. J. Appl. Earth Obs. Geoinf. 2015, 37, 7–16. [Google Scholar] [CrossRef]
  11. Borre, J.V.; Paelinckx, D.; Mucher, C.A.; Kooistra, L.; Haest, B.; De Blust, G.; Schmidt, A.M. Integrating remote sensing in Natura 2000 habitat monitoring: Prospects on the way forward. J. Nat. Conserv. 2011, 19, 116–125. [Google Scholar] [CrossRef]
  12. Schmidt, T.; Schuster, C.; Kleinschmit, B.; Förster, M. Evaluating an intra-annual time series for grassland classification—How many acquisitions and what seasonal origin are optimal? IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2014, 7, 3428–3439. [Google Scholar] [CrossRef]
  13. Pesaresi, S.; Mancini, A.; Quattrini, G.; Casavecchia, S. Mapping Mediterranean forest plant associations and habitats with functional principal component analysis using Landsat 8 NDVI time series. Remote Sens. 2020, 12, 1132. [Google Scholar] [CrossRef]
  14. Pesaresi, S.; Mancini, A.; Quattrini, G.; Casavecchia, S. Functional analysis for habitat mapping in a special area of conservation using Sentinel-2 time-series data. Remote Sens. 2022, 14, 1179. [Google Scholar] [CrossRef]
  15. Stendardi, L.; Karlsen, S.R.; Niedrist, G.; Gerdol, R.; Zebisch, M.; Rossi, M.; Notarnicola, C. Exploiting time series of Sentinel-1 and Sentinel-2 imagery to detect meadow phenology in mountain regions. Remote Sens. 2019, 11, 542. [Google Scholar] [CrossRef]
  16. Féret, J.-B.; Corbane, C.; Alleaume, S. Detecting the phenology and discriminating Mediterranean natural habitats with multispectral sensors—An analysis based on multiseasonal field spectra. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2015, 8, 2294–2305. [Google Scholar] [CrossRef]
  17. Rapinel, S.; Mony, C.; Lecoq, L.; Clément, B.; Thomas, A.; Hubert-Moy, L. Evaluation of Sentinel-2 time-series for mapping floodplain grassland plant communities. Remote Sens. Environ. 2019, 223, 115–129. [Google Scholar] [CrossRef]
  18. Rouse, J.W.; Haas, R.H.; Schell, J.A.; Deering, D.W. Monitoring vegetation systems in the Great Plains with ERTS. In Proceedings of the NASA’s Goddard Space Flight Center, Washington, DC, USA, 10–14 December 1973; Volume 1, pp. 309–317. [Google Scholar]
  19. Huang, S.; Tang, L.; Hupy, J.P.; Wang, Y.; Shao, G. A commentary review on the use of normalized difference vegetation index (NDVI) in the era of popular remote sensing. J. For. Res. 2021, 32, 1–6. [Google Scholar] [CrossRef]
  20. Pettorelli, N.; Ryan, S.; Mueller, T.; Bunnefeld, N.; Jedrzejewska, B.; Lima, M.; Kausrud, K. The normalized difference vegetation index (NDVI): Unforeseen successes in animal ecology. Clim. Res. 2011, 46, 15–27. [Google Scholar] [CrossRef]
  21. White, M.A.; Hoffman, F.; Hargrove, W.W.; Nemani, R.R. A global framework for monitoring phenological responses to climate change. Geophys. Res. Lett. 2005, 32, L04705. [Google Scholar] [CrossRef]
  22. Pasquarella, V.J.; Holden, C.E.; Kaufman, L.; Woodcock, C.E.; Nagendra, H.; He, K. From imagery to ecology: Leveraging time series of all available Landsat observations to map and monitor ecosystem state and dynamics. Remote Sens. Ecol. Conserv. 2016, 2, 152–170. [Google Scholar] [CrossRef]
  23. Serrano, J.; Shahidian, S.; da Silva, J.M. Monitoring seasonal pasture quality degradation in the Mediterranean Montado ecosystem: Proximal versus remote sensing. Water 2018, 10, 1422. [Google Scholar] [CrossRef]
  24. Serrano, J.; Shahidian, S.; Marques da Silva, J. Evaluation of normalized difference water index as a tool for monitoring pasture seasonal and inter-annual variability in a Mediterranean agro-silvo-pastoral system. Water 2019, 11, 62. [Google Scholar] [CrossRef]
  25. Serrano, J.; Roma, L.; Shahidian, S.; Belo, A.D.F.; Carreira, E.; Paniagua, L.L.; Moral, F.; Paixão, L.; da Silva, J.M. A technological approach to support extensive livestock management in the Portuguese Montado ecosystem. Agronomy 2022, 12, 1212. [Google Scholar] [CrossRef]
  26. Vrieling, M.; Meroni, M.; Mude, A. Determining optimal seasonal integration times of NDVI series for index-based livestock insurance in East Africa. In Proceedings of the 2015 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), Milan, Italy, 26–31 July 2015; pp. 169–172. [Google Scholar] [CrossRef]
  27. Hurley, M.A.; Hebblewhite, M.; Gaillard, J.; Dray, S.; Taylor, K.A.; Smith, W.K.; Zager, P.; Bonenfant, C. Functional analysis of normalized difference vegetation index curves reveals overwinter mule deer survival is driven by both spring and autumn phenology. Philos. Trans. R. Soc. B 2014, 369, 20130196. [Google Scholar] [CrossRef]
  28. Balata, D.; Gama, I.; Domingos, T.; Proença, V. Using satellite NDVI time-series to monitor grazing effects on vegetation productivity and phenology in heterogeneous Mediterranean forests. Remote Sens. 2022, 14, 2322. [Google Scholar] [CrossRef]
  29. Dumont, B.; Carrère, P.; D’Hour, P. Foraging in patchy grasslands: Diet selection by sheep and cattle is affected by the abundance and spatial distribution of preferred species. Anim. Res. 2002, 51, 367–381. [Google Scholar] [CrossRef]
  30. Donald, G.; Ahmad, W.; Hulm, E.; Trotter, M.; Lamb, D. Integrating MODIS satellite imagery and proximal vegetation sensors to enable precision livestock management. In Proceedings of the 2012 First International Conference on Agro—Geoinformatics (Agro-Geoinformatics), Shanghai, China, 2–4 August 2012; pp. 1–5. [Google Scholar] [CrossRef]
  31. Fernando, G.; Víctor, C. Pasture monitoring applying normalized difference vegetation index (NDVI) time series with Sentinel-2 and Landsat 8 images, to improve milk production at Santa Mónica Farm, Imbabura, Ecuador. In Computational Science and Its Applications—ICCSA 2020, Proceedings of the 20th International Conference, Cagliari, Italy, 1–4 July 2020; Gervasi, O., Murgante, B., Misra, S., Garau, C., Blečić, I., Taniar, D., Apduhan, B., Rocha, A.M., Tarantino, E., Torre, C., et al., Eds.; Springer: Cham, Switzerland, 2020; pp. 481–496. [Google Scholar] [CrossRef]
  32. Wikelski, M.; Mueller, U.; Scocco, P.; Catorci, A.; Desinov, L.V.; Belyaev, M.Y.; Keim, D.; Pohlmeier, W.; Fechteler, G.; Mai, P.M. Potential short-term earthquake forecasting by farm animal monitoring. Ethology 2020, 126, 931–941. [Google Scholar] [CrossRef]
  33. Wild, M.; Gauly, M.; Zanon, T.; Isselstein, J.; Komainda, M. Tracking free-ranging sheep to evaluate interrelations between selective grazing, movement patterns and the botanical composition of alpine summer pastures in northern Italy. Pastoralism 2023, 13, 25. [Google Scholar] [CrossRef]
  34. Parnell, D.; Kardailsky, I.; Parnell, J.; Badgery, W.B.; Ingram, L. Understanding sheep baa-haviour: Investigating the relationship between pasture and animal grazing patterns. Grassl. Res. 2022, 1, 143–156. [Google Scholar] [CrossRef]
  35. Caldeira, M.C.; Hector, A.; Loreau, M.; Pereira, J.S. Species richness, temporal variability and resistance of biomass production in a Mediterranean grassland. Oikos 2007, 117, 1230–1240. [Google Scholar] [CrossRef]
  36. Mysterud, A.; Langvatn, R.; Yoccoz, N.G.; Stenseth, N.C. Plant phenology, migration and geographical variation in body weight of a large herbivore: The effect of a variable topography. J. Anim. Ecol. 2011, 80, 715–723. [Google Scholar] [CrossRef]
  37. Shepard, E.; Wilson, R.; Quintana, F.; Laich, A.G.; Liebsch, N.; Albareda, D.; Halsey, L.; Gleiss, A.; Morgan, D.; Myers, A.; et al. Identification of animal movement patterns using tri-axial accelerometry. Endanger. Species Res. 2008, 10, 47–60. [Google Scholar] [CrossRef]
  38. Börger, L.; Franconi, N.; Ferretti, F.; Meschi, F.; De Michele, G.; Gantz, A.; Coulson, T. An integrated approach to identify spatiotemporal and individual-level determinants of animal home range size. Am. Nat. 2006, 168, 471–485. [Google Scholar] [CrossRef] [PubMed]
  39. Pesaresi, S.; Biondi, E.; Casavecchia, S. Bioclimates of Italy. J. Maps 2017, 13, 955–960. [Google Scholar] [CrossRef]
  40. Rivas-Martínez, S.; Rivas-Sáenz, S.; Penas, A. Worldwide Bioclimatic Classification System. Glob. Geobot. 2011, 1, 1–634. [Google Scholar] [CrossRef]
  41. Catorci, A.; Cesaretti, S.; Gatti, R. Biodiversity conservation: Geosynphytosociology as a tool of analysis and modelling of grassland systems. Hacquetia 2009, 8, 129–146. [Google Scholar] [CrossRef]
  42. Mucina, L.; Bültmann, H.; Dierßen, K.; Theurillat, J.; Raus, T.; Čarni, A.; Šumberová, K.; Willner, W.; Dengler, J.; García, R.G.; et al. Vegetation of Europe: Hierarchical floristic classification system of vascular plant, bryophyte, lichen, and algal communities. Appl. Veg. Sci. 2016, 19, 3–264. [Google Scholar] [CrossRef]
  43. Wikelski, M.; Davidson, S.C.; Kays, R. Movebank: Archive, Analysis and Sharing of Animal Movement Data. Hosted by the Max Planck Institute of Animal Behavior. Available online: http://www.movebank.org (accessed on 10 September 2023).
  44. Kranstauber, B.; Safi, K.; Scharf, A.K. Move2: R package for processing movement data. Methods Ecol. Evol. 2024, 15, 1561–1567. [Google Scholar] [CrossRef]
  45. Pettorelli, N.; Laurance, W.F.; O’BRien, T.G.; Wegmann, M.; Nagendra, H.; Turner, W.; Milner-Gulland, E. Satellite remote sensing for applied ecologists: Opportunities and challenges. J. Appl. Ecol. 2014, 51, 839–848. [Google Scholar] [CrossRef]
  46. Fernández-Habas, J.; Moreno, A.M.G.; Hidalgo-Fernández, M.T.; Leal-Murillo, J.R.; Oar, B.A.; Gómez-Giráldez, P.J.; González-Dugo, M.P.; Fernández-Rebollo, P. Investigating the potential of Sentinel-2 configuration to predict the quality of Mediterranean permanent grasslands in open woodlands. Sci. Total Environ. 2021, 791, 148101. [Google Scholar] [CrossRef]
  47. Cribari-Neto, F.; Zeileis, A. Beta Regression in R. J. Stat. Softw. 2010, 34, 1–24. [Google Scholar] [CrossRef]
  48. Wood, S.N. Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models. J. R. Stat. Soc. B. 2011, 73, 3–36. [Google Scholar] [CrossRef]
  49. Borowik, T.; Pettorelli, N.; Sönnichsen, L.; Jędrzejewska, B. Normalized Difference Vegetation Index (NDVI) as a predictor of forage availability for ungulates in forest and field habitats. Eur. J. Wildl. Res. 2013, 59, 675–682. [Google Scholar] [CrossRef]
  50. Sanz, E.; Saa-Requejo, A.; Díaz-Ambrona, C.H.; Ruiz-Ramos, M.; Rodríguez, A.; Iglesias, E.; Esteve, P.; Soriano, B.; Tarquis, A.M. Normalized Difference Vegetation Index Temporal Responses to Temperature and Precipitation in Arid. Rangel. Remote Sens. 2021, 13, 840. [Google Scholar] [CrossRef]
  51. Chynoweth, M.W.; Litton, C.M.; Lepczyk, C.A.; Hess, S.C.; Cordell, S. Biology and impacts of Pacific Island invasive species. 9. Capra hircus, the feral goat (Mammalia: Bovidae). Pac. Sci. 2013, 67, 141–156. [Google Scholar] [CrossRef]
  52. Hebblewhite, M.; Haydon, D.T. Distinguishing technology from biology: A critical review of the use of GPS telemetry data in ecology. Philos. Trans. R. Soc. B 2010, 365, 2303–2312. [Google Scholar] [CrossRef]
  53. Augustine, D.J.; Derner, J.D.; Fernández-Giménez, M.E.; Porensky, L.M.; Wilmer, H.; Briske, D.D. Adaptive, multipaddock rotational grazing management: A ranch-scale assessment of effects on vegetation and livestock performance in semiarid rangeland. Rangel. Ecol. Manag. 2020, 73, 796–810. [Google Scholar] [CrossRef]
  54. Berger-Tal, O.; Polak, T.; Oron, A.; Lubin, Y.; Kotler, B.P.; Saltz, D. Integrating animal behavior and conservation biology: A conceptual framework. Behav. Ecol. 2011, 22, 236–239. [Google Scholar] [CrossRef]
  55. Pesaresi, S.; Mancini, A.; Quattrini, G.; Casavecchia, S. Evaluation and selection of multi-spectral indices to classify vegetation using multivariate functional principal component analysis. Remote Sens. 2024, 16, 1224. [Google Scholar] [CrossRef]
  56. Park, Y.; Guldmann, J.-M. Measuring continuous landscape patterns with gray-level co-occurrence matrix (GLCM) indices: An alternative to patch metrics? Ecol. Indic. 2020, 109, 105802. [Google Scholar] [CrossRef]
  57. Gao, F.; Liu, T.; Wang, H.; Shi, H.; Yuan, C.; Song, S.; HaSi, B.; Wu, X. Spatial and temporal variation patterns of summer grazing trajectories of Sunit sheep. Ecol. Inform. 2023, 78, 102322. [Google Scholar] [CrossRef]
  58. Fynn, R.W. Functional resource heterogeneity increases livestock and rangeland productivity. Rangel. Ecol. Manag. 2012, 65, 319–329. [Google Scholar] [CrossRef]
  59. Andriamandroso, A.; Bindelle, J.; Mercatoris, B.; Lebeau, F. A review on the use of sensors to monitor cattle jaw movements and behavior when grazing. Biotechnol. Agron. Soc. Environ. 2016, 20, 427–442. [Google Scholar] [CrossRef]
  60. Debauche, O.; Elmoulat, M.; Mahmoudi, S.; Bindelle, J.; Lebeau, F. Farm animals’ behaviors and welfare analysis with AI algorithms: A review. Rev. Intell. Artif. 2021, 35, 289–299. [Google Scholar] [CrossRef]
  61. Brennan, J.; Johnson, P.; Olson, K. Classifying season long livestock grazing behavior with the use of a low-cost GPS and accelerometer. Comput. Electron. Agric. 2021, 181, 105957. [Google Scholar] [CrossRef]
  62. Perron Chambard, R.; Garel, M.; Marchand, P.; Choler, P. Fine-scale tracking of sheep grazing pressures in mountain pastures: Frugal solution and relevant indicators for improved ecosystem and practice management. SSRN 2024, 4875408. [Google Scholar] [CrossRef]
  63. Muzzo, B.I.; Bladen, K.; Perea, A.; Nyamuryekung’e, S.; Villalba, J.J. Multi-sensor integration and machine learning for high-resolution classification of herbivore foraging behavior. Animals 2025, 15, 913. [Google Scholar] [CrossRef]
  64. Hlimi, A.; El Otmani, S.; Elame, F.; Chentouf, M.; El Halimi, R.; Chebli, Y. Application of precision technologies to characterize animal behavior: A review. Animals 2024, 14, 416. [Google Scholar] [CrossRef]
Geographies 05 00026 g001
Figure 1. The study area. On the left is the location of the study site in the Central Apennines (Marche region, Italy). On the right is a detailed map of the pasture areas: the grazing areas managed by the “Delizie dei Fratelli Angeli” farm are shown in red; those managed by the “Azienda Agricola Piselli” farm are shown in blue.
Figure 1. The study area. On the left is the location of the study site in the Central Apennines (Marche region, Italy). On the right is a detailed map of the pasture areas: the grazing areas managed by the “Delizie dei Fratelli Angeli” farm are shown in red; those managed by the “Azienda Agricola Piselli” farm are shown in blue.
Geographies 05 00026 g001
Geographies 05 00026 g002
Figure 2. Weekly NDVI trends illustrating the seasonal vegetation dynamics of the two monitored pastures between June and October.
Figure 2. Weekly NDVI trends illustrating the seasonal vegetation dynamics of the two monitored pastures between June and October.
Geographies 05 00026 g002
Geographies 05 00026 g003
Figure 3. Average distance moved (m) by sheep on different pastures per month. Red dashed line is median of distance moved by all animals in each pasture in each month.
Figure 3. Average distance moved (m) by sheep on different pastures per month. Red dashed line is median of distance moved by all animals in each pasture in each month.
Geographies 05 00026 g003
Table 1. Loggers’ information.
Table 1. Loggers’ information.
Weight69 grammes
ModelGPS-datalogger series 4000
ManufacturerMax-Planck-Institute Radolfzell
Data connectionsigfox-network, movebank
Power supplyLi-Ion battery 17,500
GPS accuracy±2.5 m
3-axis acceleration sensor
Table 2. The results of the generalized linear model (GLM) for differences in NDVI values between the two pastures and seasonal trends.
Table 2. The results of the generalized linear model (GLM) for differences in NDVI values between the two pastures and seasonal trends.
PredictorEstimateStd. Errorz ValuePr(>|z|)
(Intercept)0.388580.0175222.185<0.001
Pasture (pasture 2)−0.265440.0235811.259<0.001
Month (July)−0.416170.02292−18.156<0.001
Month (August)−0.724580.023−31.501<0.001
Month (September)−0.540770.02211−24.455<0.001
Month (October)−0.306690.02185−14.035<0.001
Pasture2:month (July)−0.103370.03059−3.379<0.001
Pasture2:month (August)−0.183490.03095−5.928<0.001
Pasture2:month (September)−0.256410.03013−8.51<0.001
Pasture2:month (October)−0.281060.02982−9.424<0.001
Table 3. The results of the generalized additive mixed model (GAMs) for the analysis of distance moved by sheep during foraging.
Table 3. The results of the generalized additive mixed model (GAMs) for the analysis of distance moved by sheep during foraging.
PredictorEstimate Std. Errort/F Valuep-Value
Intercept8.310050.1923343.208<0.001
Pasture (pasture 2)1.453740.292544.969<0.001
NDVI−0.902370.18394−4.906<0.001
Pasture 2 * NDVI−0.028980.33542−0.0860.931
Smooth TermsedfRef.dfF valuep-value
s(week(time))8.8528.85244.44<0.001
s(yday(time))1.7418.0000.020.686
Pasture 2 * NDVI = interaction term.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Moscatelli, S.; Pesaresi, S.; Wikelski, M.; Tardella, F.M.; Catorci, A.; Quattrini, G. Influence of Pasture Diversity and NDVI on Sheep Foraging Behavior in Central Italy. Geographies 2025, 5, 26. https://doi.org/10.3390/geographies5020026

AMA Style

Moscatelli S, Pesaresi S, Wikelski M, Tardella FM, Catorci A, Quattrini G. Influence of Pasture Diversity and NDVI on Sheep Foraging Behavior in Central Italy. Geographies. 2025; 5(2):26. https://doi.org/10.3390/geographies5020026

Chicago/Turabian Style

Moscatelli, Sara, Simone Pesaresi, Martin Wikelski, Federico Maria Tardella, Andrea Catorci, and Giacomo Quattrini. 2025. "Influence of Pasture Diversity and NDVI on Sheep Foraging Behavior in Central Italy" Geographies 5, no. 2: 26. https://doi.org/10.3390/geographies5020026

APA Style

Moscatelli, S., Pesaresi, S., Wikelski, M., Tardella, F. M., Catorci, A., & Quattrini, G. (2025). Influence of Pasture Diversity and NDVI on Sheep Foraging Behavior in Central Italy. Geographies, 5(2), 26. https://doi.org/10.3390/geographies5020026

Article Metrics

Back to TopTop