Alexander J. Robillard, Madeline Standen, Noah Giebink, Mark Spangler, Amy C. Collins, Brian Folt, Andrew Maguire, Elissa M. Olimpi, Brett G. Dickson
Driving off‐highway vehicles (OHVs), which contributes to habitat degradation and fragmentation, is a common recreational activity in the United States and other parts of the world, particularly in desert environments with fragile ecosystems. Although habitat degradation and mortality from the expansion of OHV networks are thought to have major impacts on desert species, comprehensive maps of OHV route networks and their changes are poorly understood. To better understand how OHV route networks have evolved in the Mojave Desert ecoregion, we developed a computer vision approach to estimate OHV route location and density across the range of the Mojave desert tortoise (Gopherus agassizii). We defined OHV routes as non‐paved, linear features, including designated routes and washes in the presence of non‐paved routes. Using contemporary (n = 1499) and historical (n = 1148) aerial images, we trained and validated three convolutional neural network (CNN) models. We cross‐examined each model on sets of independently curated data and selected the highest performing model to generate predictions across the tortoise's range. When evaluated against a ‘hybrid’ test set (n = 1807 images), the final hybrid model achieved an accuracy of 77%. We then applied our model to remotely sensed imagery from across the tortoise's range and generated spatial layers of OHV route density for the 1970s, 1980s, 2010s, and 2020s. We examined OHV route density within tortoise conservation areas (TCA) and recovery units (RU) within the range of the species. Results showed an increase in the OHV route density in both TCAs (8.45%) and RUs (7.85%) from 1980 to 2020. Ordinal logistic regression indicated a strong correlation (OR = 1.01, P < 0.001) between model outputs and ground‐truthed OHV maps from the study region. Our computer vision approach and mapped results can inform conservation strategies and management aimed at mitigating the adverse impacts of OHV activity on sensitive ecosystems.
{"title":"Application of computer vision for off‐highway vehicle route detection: A case study in Mojave desert tortoise habitat","authors":"Alexander J. Robillard, Madeline Standen, Noah Giebink, Mark Spangler, Amy C. Collins, Brian Folt, Andrew Maguire, Elissa M. Olimpi, Brett G. Dickson","doi":"10.1002/rse2.70004","DOIUrl":"https://doi.org/10.1002/rse2.70004","url":null,"abstract":"Driving off‐highway vehicles (OHVs), which contributes to habitat degradation and fragmentation, is a common recreational activity in the United States and other parts of the world, particularly in desert environments with fragile ecosystems. Although habitat degradation and mortality from the expansion of OHV networks are thought to have major impacts on desert species, comprehensive maps of OHV route networks and their changes are poorly understood. To better understand how OHV route networks have evolved in the Mojave Desert ecoregion, we developed a computer vision approach to estimate OHV route location and density across the range of the Mojave desert tortoise (<jats:italic>Gopherus agassizii</jats:italic>). We defined OHV routes as non‐paved, linear features, including designated routes and washes in the presence of non‐paved routes. Using contemporary (<jats:italic>n</jats:italic> = 1499) and historical (<jats:italic>n</jats:italic> = 1148) aerial images, we trained and validated three convolutional neural network (CNN) models. We cross‐examined each model on sets of independently curated data and selected the highest performing model to generate predictions across the tortoise's range. When evaluated against a ‘hybrid’ test set (<jats:italic>n</jats:italic> = 1807 images), the final hybrid model achieved an accuracy of 77%. We then applied our model to remotely sensed imagery from across the tortoise's range and generated spatial layers of OHV route density for the 1970s, 1980s, 2010s, and 2020s. We examined OHV route density within tortoise conservation areas (TCA) and recovery units (RU) within the range of the species. Results showed an increase in the OHV route density in both TCAs (8.45%) and RUs (7.85%) from 1980 to 2020. Ordinal logistic regression indicated a strong correlation (OR = 1.01, <jats:italic>P</jats:italic> < 0.001) between model outputs and ground‐truthed OHV maps from the study region. Our computer vision approach and mapped results can inform conservation strategies and management aimed at mitigating the adverse impacts of OHV activity on sensitive ecosystems.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":"89 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2025-04-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143798030","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Liezl Mari Vermeulen, Koenraad Van Meerbeek, Paulo Negri Bernardino, Jasper Slingsby, Bruno Verbist, Ben Somers
Savannas, defined by a balance of woody and herbaceous vegetation, are vital for global biodiversity and carbon sequestration. Yet, their stability is increasingly at risk due to climate change and human impacts. The responses of these ecosystems to extreme drought events remain poorly understood, especially in relation to the regional variations in soil, terrain, climate history and disturbance legacy. This study analysed time series of a vegetation index, derived from remote sensing data, to quantify ecosystem stability metrics, i.e., resistance and resilience, in response to a major drought event in the semi‐arid savanna of the Kruger National Park, South Africa. Using Bayesian Generalized Linear Models, we assessed the influence of ecosystem traits, past extreme climate events, fire history and herbivory on regional patterns of drought resistance and resilience. Our results show that sandier granite soils dominated by trees have higher drought resistance, supported by the ability of deep‐rooted water access. In contrast, grassier savanna landscapes on basalt soils proved more drought resilient, with rapid vegetation recovery post‐drought. The effects of woody cover on ecosystem drought response are mediated by differences in historical fire regimes, elephant presence and climate legacy, underscoring the complex, context‐dependent nature of savanna landscape response to drought. This research deepens our understanding of savanna stability by clarifying the role of regional drivers, like fire and climate, alongside long‐term factors, like soil composition and woody cover. With droughts projected to increase in frequency and severity in arid and semi‐arid savannas, it also highlights remote sensing as a robust tool for regional‐scale analysis of drought responses, offering a valuable complement to field‐based experiments that can guide effective management and adaptive strategies.
{"title":"Woody cover and geology as regional‐scale determinants of semi‐arid savanna stability","authors":"Liezl Mari Vermeulen, Koenraad Van Meerbeek, Paulo Negri Bernardino, Jasper Slingsby, Bruno Verbist, Ben Somers","doi":"10.1002/rse2.70005","DOIUrl":"https://doi.org/10.1002/rse2.70005","url":null,"abstract":"Savannas, defined by a balance of woody and herbaceous vegetation, are vital for global biodiversity and carbon sequestration. Yet, their stability is increasingly at risk due to climate change and human impacts. The responses of these ecosystems to extreme drought events remain poorly understood, especially in relation to the regional variations in soil, terrain, climate history and disturbance legacy. This study analysed time series of a vegetation index, derived from remote sensing data, to quantify ecosystem stability metrics, i.e., resistance and resilience, in response to a major drought event in the semi‐arid savanna of the Kruger National Park, South Africa. Using Bayesian Generalized Linear Models, we assessed the influence of ecosystem traits, past extreme climate events, fire history and herbivory on regional patterns of drought resistance and resilience. Our results show that sandier granite soils dominated by trees have higher drought resistance, supported by the ability of deep‐rooted water access. In contrast, grassier savanna landscapes on basalt soils proved more drought resilient, with rapid vegetation recovery post‐drought. The effects of woody cover on ecosystem drought response are mediated by differences in historical fire regimes, elephant presence and climate legacy, underscoring the complex, context‐dependent nature of savanna landscape response to drought. This research deepens our understanding of savanna stability by clarifying the role of regional drivers, like fire and climate, alongside long‐term factors, like soil composition and woody cover. With droughts projected to increase in frequency and severity in arid and semi‐arid savannas, it also highlights remote sensing as a robust tool for regional‐scale analysis of drought responses, offering a valuable complement to field‐based experiments that can guide effective management and adaptive strategies.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":"18 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2025-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143734262","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With the advancement of artificial intelligence (AI) technologies, vehicle‐mounted mobile monitoring systems have become increasingly integrated into wildlife monitoring practices. However, images captured through these systems often present challenges such as low resolution, small target sizes, and partial occlusions. Consequently, detecting animal targets using conventional deep‐learning networks is challenging. To address these challenges, this paper presents an enhanced YOLOv7 model, referred to as YOLOv7(sr‐sm), which incorporates a super‐resolution (SR) reconstruction module and a small object optimization module. The YOLOv7(sr‐sm) model introduces a super‐resolution reconstruction module that leverages generative adversarial networks (GANs) to reconstruct high‐resolution details from blurry animal images. Additionally, an attention mechanism is integrated into the Neck and Head of YOLOv7 to form a small object optimization module, which enhances the model's ability to detect and locate densely packed small targets. Using a vehicle‐mounted mobile monitoring system, images of four wildlife taxa—sheep, birds, deer, and antelope —were captured on the Tibetan Plateau. These images were combined with publicly available high‐resolution wildlife photographs to create a wildlife test dataset. Experiments were conducted on this dataset, comparing the YOLOv7(sr‐sm) model with eight popular object detection models. The results demonstrate significant improvements in precision, recall, and mean Average Precision (mAP), with YOLOv7(sr‐sm) achieving 93.9%, 92.1%, and 92.3%, respectively. Furthermore, compared to the newly released YOLOv8l model, YOLOv7(sr‐sm) outperforms it by 9.3%, 2.1%, and 4.5% in these three metrics while also exhibiting superior parameter efficiency and higher inference speeds. The YOLOv7(sr‐sm) model architecture can accurately locate and identify blurry animal targets in vehicle‐mounted monitoring images, serving as a reliable tool for animal identification and counting in mobile monitoring systems. These findings provide significant technological support for the application of intelligent monitoring techniques in biodiversity conservation efforts.
{"title":"How to achieve accurate wildlife detection by using vehicle‐mounted mobile monitoring images and deep learning?","authors":"Leilei Shi, Jixi Gao, Fei Cao, Wenming Shen, Yue Wu, Kai Liu, Zheng Zhang","doi":"10.1002/rse2.70003","DOIUrl":"https://doi.org/10.1002/rse2.70003","url":null,"abstract":"With the advancement of artificial intelligence (AI) technologies, vehicle‐mounted mobile monitoring systems have become increasingly integrated into wildlife monitoring practices. However, images captured through these systems often present challenges such as low resolution, small target sizes, and partial occlusions. Consequently, detecting animal targets using conventional deep‐learning networks is challenging. To address these challenges, this paper presents an enhanced YOLOv7 model, referred to as YOLOv7(sr‐sm), which incorporates a super‐resolution (SR) reconstruction module and a small object optimization module. The YOLOv7(sr‐sm) model introduces a super‐resolution reconstruction module that leverages generative adversarial networks (GANs) to reconstruct high‐resolution details from blurry animal images. Additionally, an attention mechanism is integrated into the Neck and Head of YOLOv7 to form a small object optimization module, which enhances the model's ability to detect and locate densely packed small targets. Using a vehicle‐mounted mobile monitoring system, images of four wildlife taxa—sheep, birds, deer, and antelope —were captured on the Tibetan Plateau. These images were combined with publicly available high‐resolution wildlife photographs to create a wildlife test dataset. Experiments were conducted on this dataset, comparing the YOLOv7(sr‐sm) model with eight popular object detection models. The results demonstrate significant improvements in precision, recall, and mean Average Precision (mAP), with YOLOv7(sr‐sm) achieving 93.9%, 92.1%, and 92.3%, respectively. Furthermore, compared to the newly released YOLOv8l model, YOLOv7(sr‐sm) outperforms it by 9.3%, 2.1%, and 4.5% in these three metrics while also exhibiting superior parameter efficiency and higher inference speeds. The YOLOv7(sr‐sm) model architecture can accurately locate and identify blurry animal targets in vehicle‐mounted monitoring images, serving as a reliable tool for animal identification and counting in mobile monitoring systems. These findings provide significant technological support for the application of intelligent monitoring techniques in biodiversity conservation efforts.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":"9 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2025-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143627562","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Ole Johannes Ringnander Sørensen, Itai van Rijn, Shai Einbinder, Hagai Nativ, Aviad Scheinin, Ziv Zemah‐Shamir, Eyal Bigal, Leigh Livne, Anat Tsemel, Or M. Bialik, Gleb Papeer, Dan Tchernov, Yizhaq Makovsky
The United Nations' sustainable development goal to designate 30% of the oceans as marine protected areas by 2030 requires practical management tools, and in turn ecologically meaningful mapping of the seafloor. Particularly challenging is the mesophotic zone, a critical component of the marine system, a biodiversity hotspot, and a potential refuge. Here, we introduce a novel seafloor habitat management workflow, integrating cm‐scale synthetic aperture sonar (SAS) and multibeam bathymetry surveying with efficient ecotope characterization. In merely 6 h, we mapped ~5 km2 of a complex mesophotic reef at sub‐metric resolution. Applying a deep learning classifier on the SAS imagery, we classified four habitats with an accuracy of 84% and defined relevant fine‐scale ecotones. Visual census with precise in situ sampling guided by SAS images for navigation were utilized for ecological characterization of mapped units. Our preliminary fish surveys indicate the ecological importance of highly complex areas and rock/sand ecotones. These less abundant habitats would be largely underrepresented if surveying the area without prior consideration. Thus, our approach is demonstrated to generate scalable habitat maps at resolutions pertinent to relevant biotas, previously inaccessible in the mesophotic, advancing ecological modeling and management of large seascapes.
{"title":"Bridging the gap in deep seafloor management: Ultra fine‐scale ecological habitat characterization of large seascapes","authors":"Ole Johannes Ringnander Sørensen, Itai van Rijn, Shai Einbinder, Hagai Nativ, Aviad Scheinin, Ziv Zemah‐Shamir, Eyal Bigal, Leigh Livne, Anat Tsemel, Or M. Bialik, Gleb Papeer, Dan Tchernov, Yizhaq Makovsky","doi":"10.1002/rse2.70002","DOIUrl":"https://doi.org/10.1002/rse2.70002","url":null,"abstract":"The United Nations' sustainable development goal to designate 30% of the oceans as marine protected areas by 2030 requires practical management tools, and in turn ecologically meaningful mapping of the seafloor. Particularly challenging is the mesophotic zone, a critical component of the marine system, a biodiversity hotspot, and a potential refuge. Here, we introduce a novel seafloor habitat management workflow, integrating cm‐scale synthetic aperture sonar (SAS) and multibeam bathymetry surveying with efficient ecotope characterization. In merely 6 h, we mapped ~5 km<jats:sup>2</jats:sup> of a complex mesophotic reef at sub‐metric resolution. Applying a deep learning classifier on the SAS imagery, we classified four habitats with an accuracy of 84% and defined relevant fine‐scale ecotones. Visual census with precise in situ sampling guided by SAS images for navigation were utilized for ecological characterization of mapped units. Our preliminary fish surveys indicate the ecological importance of highly complex areas and rock/sand ecotones. These less abundant habitats would be largely underrepresented if surveying the area without prior consideration. Thus, our approach is demonstrated to generate scalable habitat maps at resolutions pertinent to relevant biotas, previously inaccessible in the mesophotic, advancing ecological modeling and management of large seascapes.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":"11 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2025-03-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143607921","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Chhandak Bagchi, Josh Medina, Duncan J. Irschick, Subhransu Maji, Fredrik Christiansen
Aerial photogrammetry is a popular non‐invasive tool to measure the size, body morphometrics and body condition of wild animals. While the method can generate large datasets quickly, the lack of efficient processing tools can create bottlenecks that delay management actions. We developed a machine learning algorithm to automatically measure body morphometrics (body length and widths) of southern right whales (Eubalaena australis, SRWs) from aerial photographs (n = 8,958) collected by unmanned aerial vehicles in Australia. Our approach utilizes two Mask R‐CNN detection models to: (i) generate masks for each whale and (ii) estimate points along the whale's axis. We annotated a dataset of 468 images containing 638 whales to train our models. To evaluate the accuracy of our machine learning approach, we compared the model‐generated body morphometrics to manual measurements. The influence of picture quality (whale posture and water clarity) was also assessed. The model‐generated body length estimates were slightly negatively biased (median error of −1.3%), whereas the body volume estimates had a small (median error of 6.5%) positive bias. After correcting both biases, the resulting model‐generated body length and volume estimates had mean absolute errors of 0.85% (SD = 0.75) and 6.88% (SD = 6.57), respectively. The magnitude of the errors decreased as picture quality increased. When using the model‐generated data to quantify intra‐seasonal changes in body condition of SRW females, we obtained a similar slope parameter (−0.001843, SE = 0.000095) as derived from manual measurements (−0.001565, SE = 0.000079). This indicates that our approach was able to accurately capture temporal trends in body condition at a population level.
{"title":"Automated extraction of right whale morphometric data from drone aerial photographs","authors":"Chhandak Bagchi, Josh Medina, Duncan J. Irschick, Subhransu Maji, Fredrik Christiansen","doi":"10.1002/rse2.70001","DOIUrl":"https://doi.org/10.1002/rse2.70001","url":null,"abstract":"Aerial photogrammetry is a popular non‐invasive tool to measure the size, body morphometrics and body condition of wild animals. While the method can generate large datasets quickly, the lack of efficient processing tools can create bottlenecks that delay management actions. We developed a machine learning algorithm to automatically measure body morphometrics (body length and widths) of southern right whales (Eubalaena australis, SRWs) from aerial photographs (<jats:italic>n</jats:italic> = 8,958) collected by unmanned aerial vehicles in Australia. Our approach utilizes two Mask R‐CNN detection models to: (i) generate masks for each whale and (ii) estimate points along the whale's axis. We annotated a dataset of 468 images containing 638 whales to train our models. To evaluate the accuracy of our machine learning approach, we compared the model‐generated body morphometrics to manual measurements. The influence of picture quality (whale posture and water clarity) was also assessed. The model‐generated body length estimates were slightly negatively biased (median error of −1.3%), whereas the body volume estimates had a small (median error of 6.5%) positive bias. After correcting both biases, the resulting model‐generated body length and volume estimates had mean absolute errors of 0.85% (SD = 0.75) and 6.88% (SD = 6.57), respectively. The magnitude of the errors decreased as picture quality increased. When using the model‐generated data to quantify intra‐seasonal changes in body condition of SRW females, we obtained a similar slope parameter (−0.001843, SE = 0.000095) as derived from manual measurements (−0.001565, SE = 0.000079). This indicates that our approach was able to accurately capture temporal trends in body condition at a population level.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":"54 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2025-03-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143607922","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Siméon Béasse, Louis Sallé, Paul Coiffard, Birgen Haest
Acoustic recordings have emerged as a promising tool to monitor nocturnal bird migration, as it can uniquely provide species‐level detection of migratory movements under the darkness of the night sky. This study explores the use of acoustics to quantify nocturnal bird migration across Europe, a region where research on the topic remains relatively sparse. We examine three migration intensity measures derived from acoustic recordings, that is, nocturnal flight call rates, nocturnal flight passage rates and species diversity, in the French Pyrenees in 2021 and 2022. To assess the effectiveness of these acoustic measurements, we compare them with migratory traffic rates estimated by a dedicated bird radar at three taxonomic levels: all birds, passerines and thrushes. We also test if weather conditions influence these relationships and whether combining acoustic data from multiple simultaneous sites improve the predictive performance. Nocturnal flight passage rates, that is, the number of estimated passing birds independent of call abundance, outperformed predictions using species diversity or nocturnal flight call rates. The predictive accuracy of the acoustics data increased with taxonomic detail: predicting thrush migration using acoustics was far more accurate (R2 = 63%) than for passerines (R2 = 29%) or birds in general (R2 = 27%). Prediction using simultaneous acoustics measurements from several sites strongly reduced the uncertainty of the quantification. We did not find any evidence that weather conditions affected the predictive performance of the acoustics data. Accurate, automated monitoring of migratory flows is crucial as many bird species face steep population declines. Acoustic monitoring offers valuable species‐specific insights, making it a powerful tool for nocturnal bird migration studies. This study advances the integration of acoustic methods into bird monitoring by testing their benefits and limitations and provides recommendations and guidelines to enhance the effectiveness of future studies using acoustic data.
{"title":"Quantifying nocturnal bird migration using acoustics: opportunities and challenges","authors":"Siméon Béasse, Louis Sallé, Paul Coiffard, Birgen Haest","doi":"10.1002/rse2.433","DOIUrl":"https://doi.org/10.1002/rse2.433","url":null,"abstract":"Acoustic recordings have emerged as a promising tool to monitor nocturnal bird migration, as it can uniquely provide species‐level detection of migratory movements under the darkness of the night sky. This study explores the use of acoustics to quantify nocturnal bird migration across Europe, a region where research on the topic remains relatively sparse. We examine three migration intensity measures derived from acoustic recordings, that is, nocturnal flight call rates, nocturnal flight passage rates and species diversity, in the French Pyrenees in 2021 and 2022. To assess the effectiveness of these acoustic measurements, we compare them with migratory traffic rates estimated by a dedicated bird radar at three taxonomic levels: all birds, passerines and thrushes. We also test if weather conditions influence these relationships and whether combining acoustic data from multiple simultaneous sites improve the predictive performance. Nocturnal flight passage rates, that is, the number of estimated passing birds independent of call abundance, outperformed predictions using species diversity or nocturnal flight call rates. The predictive accuracy of the acoustics data increased with taxonomic detail: predicting thrush migration using acoustics was far more accurate (<jats:italic>R</jats:italic><jats:sup>2</jats:sup> = 63%) than for passerines (<jats:italic>R</jats:italic><jats:sup>2</jats:sup> = 29%) or birds in general (<jats:italic>R</jats:italic><jats:sup>2</jats:sup> = 27%). Prediction using simultaneous acoustics measurements from several sites strongly reduced the uncertainty of the quantification. We did not find any evidence that weather conditions affected the predictive performance of the acoustics data. Accurate, automated monitoring of migratory flows is crucial as many bird species face steep population declines. Acoustic monitoring offers valuable species‐specific insights, making it a powerful tool for nocturnal bird migration studies. This study advances the integration of acoustic methods into bird monitoring by testing their benefits and limitations and provides recommendations and guidelines to enhance the effectiveness of future studies using acoustic data.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":"13 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2025-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143599604","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Elamurugu Alias Gokul, Dionysios E. Raitsos, Robert J. W. Brewin, Susana Carvalho, Khaled Asfahani, Ibrahim Hoteit
Coral bleaching, often triggered by oceanic warming, has a devastating impact on coral reef systems, resulting in substantial alterations to biodiversity and ecosystem services. For conservation management, an effective technique is needed to not only detect and monitor coral bleaching events but also to predict their severity levels. By combining high‐resolution satellite measurements (Sentinel‐2 Multispectral Instrument) and a bottom reflectance model within a least‐squares approach, we developed a new ocean color remote‐sensing model specifically designed to detect, map, and predict severity levels (low to high) of coral bleaching events at a high spatial resolution of 10 m. The proposed algorithm was implemented and tested within the Red Sea and compared remarkably well with concurrent and independent in situ data. We also applied the algorithm to investigate the response of corals during and after a bleaching event in the Wadi El‐Gemal region (Egypt) from July to December 2020. Our results show that coral bleaching severity levels and sea surface temperature (SST) were unusually high during August–September 2020. After the event, the coral bleaching signal decreased concurrently with SST during October–November 2020, aligned with a recovery of bleached coral reefs by December 2020. The proposed algorithm offers a cost‐effective approach toward developing a near‐real‐time remote‐sensing system for monitoring coral bleaching events and recovery at multi‐reef scales. Such remote‐sensing tools would aid policymakers and managers in developing and implementing integrated management strategies for coral reef conservation, as well as in supporting reactive management plans, including the identification of priority areas for intervention.
通常由海洋变暖引发的珊瑚白化对珊瑚礁系统造成毁灭性影响，导致生物多样性和生态系统服务发生重大变化。在保育管理方面，不仅需要一种有效的技术来检测和监测珊瑚白化事件，而且还需要一种有效的技术来预测其严重程度。通过结合高分辨率卫星测量（Sentinel - 2多光谱仪器）和最小二乘方法中的底部反射率模型，我们开发了一种新的海洋颜色遥感模型，专门用于在10米的高空间分辨率下检测、绘制和预测珊瑚白化事件的严重程度（从低到高）。提出的算法在红海进行了实施和测试，并与并发和独立的原位数据进行了非常好的比较。我们还应用该算法研究了2020年7月至12月埃及Wadi El - Gemal地区白化事件期间和之后珊瑚的反应。我们的研究结果表明，2020年8月至9月期间，珊瑚白化严重程度和海面温度（SST）异常高。事件发生后，2020年10月至11月期间，珊瑚白化信号与海温同时下降，与2020年12月白化珊瑚礁的恢复一致。所提出的算法为开发近实时遥感系统提供了一种经济有效的方法，用于监测多珊瑚礁尺度上的珊瑚白化事件和恢复。这种遥感工具将有助于决策者和管理者制定和实施珊瑚礁保护的综合管理战略，以及支持反应性管理计划，包括确定优先干预领域。
{"title":"Remotely sensing coral bleaching in the Red Sea","authors":"Elamurugu Alias Gokul, Dionysios E. Raitsos, Robert J. W. Brewin, Susana Carvalho, Khaled Asfahani, Ibrahim Hoteit","doi":"10.1002/rse2.70000","DOIUrl":"https://doi.org/10.1002/rse2.70000","url":null,"abstract":"Coral bleaching, often triggered by oceanic warming, has a devastating impact on coral reef systems, resulting in substantial alterations to biodiversity and ecosystem services. For conservation management, an effective technique is needed to not only detect and monitor coral bleaching events but also to predict their severity levels. By combining high‐resolution satellite measurements (Sentinel‐2 Multispectral Instrument) and a bottom reflectance model within a least‐squares approach, we developed a new ocean color remote‐sensing model specifically designed to detect, map, and predict severity levels (low to high) of coral bleaching events at a high spatial resolution of 10 m. The proposed algorithm was implemented and tested within the Red Sea and compared remarkably well with concurrent and independent in situ data. We also applied the algorithm to investigate the response of corals during and after a bleaching event in the Wadi El‐Gemal region (Egypt) from July to December 2020. Our results show that coral bleaching severity levels and sea surface temperature (SST) were unusually high during August–September 2020. After the event, the coral bleaching signal decreased concurrently with SST during October–November 2020, aligned with a recovery of bleached coral reefs by December 2020. The proposed algorithm offers a cost‐effective approach toward developing a near‐real‐time remote‐sensing system for monitoring coral bleaching events and recovery at multi‐reef scales. Such remote‐sensing tools would aid policymakers and managers in developing and implementing integrated management strategies for coral reef conservation, as well as in supporting reactive management plans, including the identification of priority areas for intervention.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":"95 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2025-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143598968","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
J. M. Wolstenholme, F. Cooper, R. E. Thomas, J. Ahmed, K. J. Parsons, D. R. Parsons
Hedgerows are a key component of the UK landscape that form boundaries, borders and limits of land whilst providing vital landscape‐scale ecological connectivity for a range of organisms. They are diverse habitats in the agricultural landscape providing a range of ecosystem services. Poorly managed hedgerows often present with gaps, reducing their ecological connectivity, resulting in fragmented habitats. However, hedgerow gap frequency and spatial distributions are often unquantified at the landscape‐scale. Here we present a novel methodology based on deep learning (DL) that is coupled with high‐resolution aerial imagery. We demonstrate how this provides a route towards a rapid, adaptable, accurate assessment of hedgerow and gap abundance at such scales, with minimal training data. We present the training and development of a DL model using the U‐Net architecture to automatically identify hedgerows across the East Riding of Yorkshire (ERY) in the UK and demonstrate the ability of the model to estimate hedgerow gap types, lengths and their locations. Our method was both time efficient and accurate, processing an area of 2479 km2 in 32 h with an overall accuracy of 92.4%. The substantive results allow us to estimate that in the ERY alone, there were 3982 ± 302 km of hedgerows and 2865 ± 217 km of hedgerow gaps (with 339 km classified as for access). Our approach and study show that hedgerows and gaps can be extracted from true colour aerial imagery without the requirement of elevation data and can produce meaningful results that lead to the identification of prioritisation areas for hedgerow gap infilling, replanting and restoration. Such replanting could significantly contribute towards national tree planting goals and meeting net zero targets in a changing climate.
{"title":"Automated identification of hedgerows and hedgerow gaps using deep learning","authors":"J. M. Wolstenholme, F. Cooper, R. E. Thomas, J. Ahmed, K. J. Parsons, D. R. Parsons","doi":"10.1002/rse2.432","DOIUrl":"https://doi.org/10.1002/rse2.432","url":null,"abstract":"Hedgerows are a key component of the UK landscape that form boundaries, borders and limits of land whilst providing vital landscape‐scale ecological connectivity for a range of organisms. They are diverse habitats in the agricultural landscape providing a range of ecosystem services. Poorly managed hedgerows often present with gaps, reducing their ecological connectivity, resulting in fragmented habitats. However, hedgerow gap frequency and spatial distributions are often unquantified at the landscape‐scale. Here we present a novel methodology based on deep learning (DL) that is coupled with high‐resolution aerial imagery. We demonstrate how this provides a route towards a rapid, adaptable, accurate assessment of hedgerow and gap abundance at such scales, with minimal training data. We present the training and development of a DL model using the U‐Net architecture to automatically identify hedgerows across the East Riding of Yorkshire (ERY) in the UK and demonstrate the ability of the model to estimate hedgerow gap types, lengths and their locations. Our method was both time efficient and accurate, processing an area of 2479 km<jats:sup>2</jats:sup> in 32 h with an overall accuracy of 92.4%. The substantive results allow us to estimate that in the ERY alone, there were 3982 ± 302 km of hedgerows and 2865 ± 217 km of hedgerow gaps (with 339 km classified as for access). Our approach and study show that hedgerows and gaps can be extracted from true colour aerial imagery without the requirement of elevation data and can produce meaningful results that lead to the identification of prioritisation areas for hedgerow gap infilling, replanting and restoration. Such replanting could significantly contribute towards national tree planting goals and meeting net zero targets in a changing climate.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":"51 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2025-02-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143418078","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Laura B. Vedovato, Luiz E. O. C. Aragão, Danilo R. A. Almeida, David C. Bartholomew, Mauro Assis, Ricardo Dalagnol, Eric B. Gorgens, Celso H. L. Silva‐Junior, Jean P. Ometto, Aline Pontes‐Lopes, Carlos A. Silva, Ruben Valbuena, Ted R. Feldpausch
Secondary forests in the Amazon are important carbon sinks, biodiversity reservoirs, and connections between forest fragments. However, their regrowth is highly threatened by fire. Using airborne laser scanning (ALS), surveyed between 2016 and 2018, we analyzed canopy metrics in burned (fires occurred between 2001 and 2018) and unburned secondary forests across different successional stages and their ability to recover after fire. We assessed maximum and mean canopy height, openness at 5 and 10 m, canopy roughness, leaf area index (LAI) and leaf area height volume (LAHV) for 20 sites across South‐East Amazonia (ranging from 375 to 1200 ha). Compared to unburned forests, burned forests had reductions in canopy height, LAI, and LAHV, and increases in openness and roughness. These effects were more pronounced in early successional (ES) than later successional (LS) stages, for example, mean canopy height decreased 33% in ES and 14% in LS and LAI decreased 36% in ES and 18% in LS. Forests in ES stages were less resistant to fire, but more resilient (capable of recovering from a disturbance) in their post‐fire regrowth than LS stage forests. Data extrapolation from our models suggests that canopy structure partially recovers with time since fire for six out of seven canopy metrics; however, LAI and LAHV in LS forests may never fully recover. Our results indicate that successional stage‐specific management and policies that mitigate against fire in early secondary forests should be implemented to increase the success of forest regeneration. Mitigation of fires is critical if secondary forests are to continue to provide their wide array of ecological services.
{"title":"Impacts of fire on canopy structure and its resilience depend on successional stage in Amazonian secondary forests","authors":"Laura B. Vedovato, Luiz E. O. C. Aragão, Danilo R. A. Almeida, David C. Bartholomew, Mauro Assis, Ricardo Dalagnol, Eric B. Gorgens, Celso H. L. Silva‐Junior, Jean P. Ometto, Aline Pontes‐Lopes, Carlos A. Silva, Ruben Valbuena, Ted R. Feldpausch","doi":"10.1002/rse2.431","DOIUrl":"https://doi.org/10.1002/rse2.431","url":null,"abstract":"Secondary forests in the Amazon are important carbon sinks, biodiversity reservoirs, and connections between forest fragments. However, their regrowth is highly threatened by fire. Using airborne laser scanning (ALS), surveyed between 2016 and 2018, we analyzed canopy metrics in burned (fires occurred between 2001 and 2018) and unburned secondary forests across different successional stages and their ability to recover after fire. We assessed maximum and mean canopy height, openness at 5 and 10 m, canopy roughness, leaf area index (LAI) and leaf area height volume (LAHV) for 20 sites across South‐East Amazonia (ranging from 375 to 1200 ha). Compared to unburned forests, burned forests had reductions in canopy height, LAI, and LAHV, and increases in openness and roughness. These effects were more pronounced in early successional (ES) than later successional (LS) stages, for example, mean canopy height decreased 33% in ES and 14% in LS and LAI decreased 36% in ES and 18% in LS. Forests in ES stages were less resistant to fire, but more resilient (capable of recovering from a disturbance) in their post‐fire regrowth than LS stage forests. Data extrapolation from our models suggests that canopy structure partially recovers with time since fire for six out of seven canopy metrics; however, LAI and LAHV in LS forests may never fully recover. Our results indicate that successional stage‐specific management and policies that mitigate against fire in early secondary forests should be implemented to increase the success of forest regeneration. Mitigation of fires is critical if secondary forests are to continue to provide their wide array of ecological services.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":"34 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2025-02-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143418077","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Arthur Bayle, Baptiste Nicoud, Jérôme Mansons, Loïc Francon, Christophe Corona, Philippe Choler
Multidecadal time series of satellite observations, such as those from Landsat, offer the possibility to study trends in vegetation greenness at unprecedented spatial and temporal scales. Alpine ecosystems have exhibited large increases in vegetation greenness as seen from space; nevertheless, the ecological processes underlying alpine greening have rarely been investigated. Here, we used a unique dataset of forest stand and structure characteristics derived from manually orthorectified high‐resolution diachronic images (1983 and 2018), dendrochronology and LiDAR analysis to decipher the ecological processes underlying alpine greening in the southwestern French Alps, formerly identified as a hotspot of greening at the scale of the European Alps by previous studies. We found that most of the alpine greening in this area can be attributed to forest dynamics, including forest ingrowth and treeline upward shift. Furthermore, we showed that the magnitude of the greening was highest in pixels/areas where trees were first established at the beginning of the Landsat time series in the mid‐80s corresponding to a specific forest successional stage. In these pixels, we observe that trees from the first wave of establishment have grown between 1984 and 2023, while over the same period, younger trees established in forest gaps, leading to increases in both vertical and horizontal vegetation cover. This study provides an in‐depth description of the causal relationship between forest dynamics and greening, providing a unique example of how ecological processes translate into radiometric signals, while also paving the way for the study of large‐scale treeline dynamics using satellite remote sensing.
{"title":"Alpine greening deciphered by forest stand and structure dynamics in advancing treelines of the southwestern European Alps","authors":"Arthur Bayle, Baptiste Nicoud, Jérôme Mansons, Loïc Francon, Christophe Corona, Philippe Choler","doi":"10.1002/rse2.430","DOIUrl":"https://doi.org/10.1002/rse2.430","url":null,"abstract":"Multidecadal time series of satellite observations, such as those from Landsat, offer the possibility to study trends in vegetation greenness at unprecedented spatial and temporal scales. Alpine ecosystems have exhibited large increases in vegetation greenness as seen from space; nevertheless, the ecological processes underlying alpine greening have rarely been investigated. Here, we used a unique dataset of forest stand and structure characteristics derived from manually orthorectified high‐resolution diachronic images (1983 and 2018), dendrochronology and LiDAR analysis to decipher the ecological processes underlying alpine greening in the southwestern French Alps, formerly identified as a hotspot of greening at the scale of the European Alps by previous studies. We found that most of the alpine greening in this area can be attributed to forest dynamics, including forest ingrowth and treeline upward shift. Furthermore, we showed that the magnitude of the greening was highest in pixels/areas where trees were first established at the beginning of the Landsat time series in the mid‐80s corresponding to a specific forest successional stage. In these pixels, we observe that trees from the first wave of establishment have grown between 1984 and 2023, while over the same period, younger trees established in forest gaps, leading to increases in both vertical and horizontal vegetation cover. This study provides an in‐depth description of the causal relationship between forest dynamics and greening, providing a unique example of how ecological processes translate into radiometric signals, while also paving the way for the study of large‐scale treeline dynamics using satellite remote sensing.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":"27 1","pages":""},"PeriodicalIF":5.5,"publicationDate":"2025-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142916837","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：