Volume 20, Issue 1 (4-2026)
مرتع 2026, 20(1): 0-0 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Mahmoudzadeh L, Tahmasebi P, Ebrahimi A, Mahzooni Kachapi S S. Correlation between Vegetation Indices and Ecological Diversity Components Using Landsat-8 Imagery. مرتع 2026; 20 (1)
URL: http://rangelandsrm.ir/article-1-1340-en.html
URL: http://rangelandsrm.ir/article-1-1340-en.html
Faculty of Natural Resources and Earth Science, Shahrekord University
Abstract: (35 Views)
Background and Objective: Precise monitoring of biodiversity, particularly plant diversity, is crucial for the sustainable management of rangeland ecosystems. Advances in remote sensing technologies and the availability of satellite imagery with appropriate spectral resolution have provided new opportunities for non-invasive biodiversity assessment. This study aimed to evaluate the effectiveness of vegetation indices derived from Landsat-8 satellite imagery in predicting three key dimensions of plant diversity-alpha, beta, and functional diversity in semi-steppe rangeland ecosystems.
Materials and Methods: The study area encompasses approximately 5118.1 hectares of semi-steppe rangelands in Chaharmahal va Bakhtiari Province, characterized by high biodiversity. Six to eight sampling sites were selected, and vegetation sampling was conducted during peak growing season using a randomized systematic approach. At each site, three 30*30 m2 macroplots were established, and species cover was evaluated along three transects within three 2*2 m2 subplots. Simultaneously, Landsat-8 Operational Land Imager (OLI) satellite images with a 30-meter spatial resolution were utilized to calculate a suite of vegetation indices, categorized into slope-based indices (e.g., NDVI, EVI, SAVI) and distance-based indices (e.g., AVI, MSAVI2, TSAVI2). Field data were aligned with satellite-derived indices to extract index values for each macroplot. Three diversity components-alpha, beta, and functional diversity were quantified: alpha diversity via species richness indices; beta diversity through similarity and dissimilarity metrics (e.g., Bray-Curtis and Cody indices); and functional diversity employing functional indices including Community Weighted Mean (CWM), Rao’s quadratic entropy (RaoQ), Functional Richness (FRic), Functional Evenness (FEve), and Functional Divergence (FDiv) calculated using R and PAST software. Linear regression analyses were performed to assess the relationships between vegetation indices and diversity components, evaluating significance levels and coefficients of determination.
Results: For alpha diversity, the AVI index (R²=0.25), which is highly sensitive to structural changes and vegetation density, exhibited the strongest correlation with species richness, while MSAVI2, EVI, and NDVI also showed significant associations. Beta diversity, influenced by spatial heterogeneity driven by grazing and topographic variation, was well captured by AVI (R²=0.50) and MSAVI2 (R²=0.51). Regarding functional diversity, EVI demonstrated the highest correlation with the FRic index (R²=0.32), indicating its strong capability in detecting vegetation structural attributes and functional composition. NDVI showed moderate explanatory power (R²=0.28) for certain functional diversity components, and MSAVI2 notably contributed to estimating CWM-PG. The saturation effect of NDVI in dense vegetation and soil background variability were identified as limiting factors for its accuracy. Overall, AVI, EVI, and MSAVI2 outperformed other indices, particularly in analyzing spatial differentiation (beta diversity) and functional plant traits, establishing robust statistical relationships. These results suggest that, under specific ecological conditions of semi-steppe rangelands, selected spectral vegetation indices serve as effective tools for spatial-temporal monitoring of biodiversity, accurately reflecting plant diversity patterns.
Conclusion: This study confirms that vegetation indices derived from Landsat-8 data especially AVI, EVI, MSAVI2, and NDVI are capable of accurately estimating and monitoring biodiversity at multiple levels (alpha, beta, and functional diversity) within semi-steppe rangeland ecosystems. Incorporating these remote sensing-based indices into sustainable rangeland management can provide essential information for informed decision-making and ecosystem conservation. Although the accuracy of remote sensing data may be constrained compared to detailed field measurements, this approach offers a cost-effective complementary tool for ecological monitoring and management. Future research is recommended to integrate hyperspectral remote sensing data and more detailed field sampling to enhance prediction accuracy and generalizability.
Materials and Methods: The study area encompasses approximately 5118.1 hectares of semi-steppe rangelands in Chaharmahal va Bakhtiari Province, characterized by high biodiversity. Six to eight sampling sites were selected, and vegetation sampling was conducted during peak growing season using a randomized systematic approach. At each site, three 30*30 m2 macroplots were established, and species cover was evaluated along three transects within three 2*2 m2 subplots. Simultaneously, Landsat-8 Operational Land Imager (OLI) satellite images with a 30-meter spatial resolution were utilized to calculate a suite of vegetation indices, categorized into slope-based indices (e.g., NDVI, EVI, SAVI) and distance-based indices (e.g., AVI, MSAVI2, TSAVI2). Field data were aligned with satellite-derived indices to extract index values for each macroplot. Three diversity components-alpha, beta, and functional diversity were quantified: alpha diversity via species richness indices; beta diversity through similarity and dissimilarity metrics (e.g., Bray-Curtis and Cody indices); and functional diversity employing functional indices including Community Weighted Mean (CWM), Rao’s quadratic entropy (RaoQ), Functional Richness (FRic), Functional Evenness (FEve), and Functional Divergence (FDiv) calculated using R and PAST software. Linear regression analyses were performed to assess the relationships between vegetation indices and diversity components, evaluating significance levels and coefficients of determination.
Results: For alpha diversity, the AVI index (R²=0.25), which is highly sensitive to structural changes and vegetation density, exhibited the strongest correlation with species richness, while MSAVI2, EVI, and NDVI also showed significant associations. Beta diversity, influenced by spatial heterogeneity driven by grazing and topographic variation, was well captured by AVI (R²=0.50) and MSAVI2 (R²=0.51). Regarding functional diversity, EVI demonstrated the highest correlation with the FRic index (R²=0.32), indicating its strong capability in detecting vegetation structural attributes and functional composition. NDVI showed moderate explanatory power (R²=0.28) for certain functional diversity components, and MSAVI2 notably contributed to estimating CWM-PG. The saturation effect of NDVI in dense vegetation and soil background variability were identified as limiting factors for its accuracy. Overall, AVI, EVI, and MSAVI2 outperformed other indices, particularly in analyzing spatial differentiation (beta diversity) and functional plant traits, establishing robust statistical relationships. These results suggest that, under specific ecological conditions of semi-steppe rangelands, selected spectral vegetation indices serve as effective tools for spatial-temporal monitoring of biodiversity, accurately reflecting plant diversity patterns.
Conclusion: This study confirms that vegetation indices derived from Landsat-8 data especially AVI, EVI, MSAVI2, and NDVI are capable of accurately estimating and monitoring biodiversity at multiple levels (alpha, beta, and functional diversity) within semi-steppe rangeland ecosystems. Incorporating these remote sensing-based indices into sustainable rangeland management can provide essential information for informed decision-making and ecosystem conservation. Although the accuracy of remote sensing data may be constrained compared to detailed field measurements, this approach offers a cost-effective complementary tool for ecological monitoring and management. Future research is recommended to integrate hyperspectral remote sensing data and more detailed field sampling to enhance prediction accuracy and generalizability.
Type of Study: Research |
Subject:
Special
Received: 2025/08/19 | Accepted: 2025/12/20 | Published: 2026/04/4
Received: 2025/08/19 | Accepted: 2025/12/20 | Published: 2026/04/4
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
