M. Garbarino, D. Morresi, Nicolò Anselmetto, P. Weisberg
{"title":"Treeline remote sensing: from tracking treeline shifts to multi‐dimensional monitoring of ecotonal change","authors":"M. Garbarino, D. Morresi, Nicolò Anselmetto, P. Weisberg","doi":"10.1002/rse2.351","DOIUrl":"https://doi.org/10.1002/rse2.351","url":null,"abstract":"","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47608945","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}
{"title":"Comparison of bird migration in a radar wind profiler and a dedicated bird radar","authors":"Nadja Weisshaupt, M. Hervo, B. Haest","doi":"10.1002/rse2.350","DOIUrl":"https://doi.org/10.1002/rse2.350","url":null,"abstract":"","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44514665","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}
Hui Zhang, S. Turvey, Shree P. Pandey, Xiqiang Song, Zhong-yu Sun, Nan Wang
{"title":"Commercial drones can provide accurate and effective monitoring of the world's rarest primate","authors":"Hui Zhang, S. Turvey, Shree P. Pandey, Xiqiang Song, Zhong-yu Sun, Nan Wang","doi":"10.1002/rse2.341","DOIUrl":"https://doi.org/10.1002/rse2.341","url":null,"abstract":"","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48749763","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}
Yuval Werber, Gadi Hareli, Omer Yinon, Nir Sapir, Y. Yovel
Wind energy is a major and rapidly expanding renewable energy source. Horizontal‐axis wind turbines, the main tool in this industry, induce mortality in flying animals and consequently bring about conservation concerns and regulatory restrictions. We utilized a unique combination of RADAR, LIDAR and ultrasonic acoustic recorders to test the utility of a novel technology meant to prevent wind turbine‐related mortality in bats. Our drone‐mounted deterrent device produces a pulsating combination of strong auditory and visual signals while moving through the air. LIDAR was used to assess the device's impact below its flight altitude and RADAR to assess its influence above its flight altitude. Continuous acoustic recordings from ground level to ~400 m above‐ground‐level were used to monitor bat activity in the research site. We recorded the nightly altitudinal distributions of multiple bat species throughout the experiment. Analysis revealed a significant change in activity while the deterrent was flying compared to baseline conditions. We also recorded a significant ~40% decrease below and a significant ~50% increase above the deterrent's flight altitude during its operation compared to the post‐flight control. The tested technology is independent of wind farm activities and does not require modifying wind turbine form or operation procedures. The device differs from previously proposed solutions by being dynamic – moving in the airspace and emitting constantly changing signals – thus decreasing the probability of animal habituation. Our findings suggest that the deterrent could dramatically decrease wind turbine‐related mortality by deterring bats from approaching rotor‐swept airspace. Focused implementation in conditions where bat activity and energy production are in conflict may provide a practical, cost‐effective mortality mitigation solution compared to current alternatives. Thus, our results should be considered by the wind‐turbine industry and environmental monitoring and animal conservation organizations, as well as by regulatory agencies, when pursuing alleviation of wind turbine‐related mortality.
{"title":"Drone‐mounted audio‐visual deterrence of bats: implications for reducing aerial wildlife mortality by wind turbines","authors":"Yuval Werber, Gadi Hareli, Omer Yinon, Nir Sapir, Y. Yovel","doi":"10.1002/rse2.316","DOIUrl":"https://doi.org/10.1002/rse2.316","url":null,"abstract":"Wind energy is a major and rapidly expanding renewable energy source. Horizontal‐axis wind turbines, the main tool in this industry, induce mortality in flying animals and consequently bring about conservation concerns and regulatory restrictions. We utilized a unique combination of RADAR, LIDAR and ultrasonic acoustic recorders to test the utility of a novel technology meant to prevent wind turbine‐related mortality in bats. Our drone‐mounted deterrent device produces a pulsating combination of strong auditory and visual signals while moving through the air. LIDAR was used to assess the device's impact below its flight altitude and RADAR to assess its influence above its flight altitude. Continuous acoustic recordings from ground level to ~400 m above‐ground‐level were used to monitor bat activity in the research site. We recorded the nightly altitudinal distributions of multiple bat species throughout the experiment. Analysis revealed a significant change in activity while the deterrent was flying compared to baseline conditions. We also recorded a significant ~40% decrease below and a significant ~50% increase above the deterrent's flight altitude during its operation compared to the post‐flight control. The tested technology is independent of wind farm activities and does not require modifying wind turbine form or operation procedures. The device differs from previously proposed solutions by being dynamic – moving in the airspace and emitting constantly changing signals – thus decreasing the probability of animal habituation. Our findings suggest that the deterrent could dramatically decrease wind turbine‐related mortality by deterring bats from approaching rotor‐swept airspace. Focused implementation in conditions where bat activity and energy production are in conflict may provide a practical, cost‐effective mortality mitigation solution compared to current alternatives. Thus, our results should be considered by the wind‐turbine industry and environmental monitoring and animal conservation organizations, as well as by regulatory agencies, when pursuing alleviation of wind turbine‐related mortality.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47960510","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}
Richard Turner, Ophélie J. D. Lasne, Kara N. Youngentob, S. Shokirov, Helen L. Osmond, L. Kruuk
1 In wild bird populations, the structure of vegetation around nest-sites can influence the risk of predation 2 of dependent young offspring, generating selection for breeding birds to choose nest-sites with 3 vegetation characteristics associated with lower predation rates. However, for researchers, vegetation 4 structure can be difficult to quantify objectively in the field, which might explain why there remains a 5 general lack of understanding of which characteristics are most important in determining rates of 6 predation. Airborne Laser Scanning (ALS) offers a powerful means of measuring vegetation structure 7 at unprecedented resolution across different spatial scales. Here, we combined ALS with 11 years of 8 breeding data from a wild population of superb fairy-wrens Malurus cyaneus in south-east Australia, a 9 species which nests relatively close to the ground and has high rates of nest and fledgling predation. We 10 derived structural measurements of understorey (0-8 m) vegetation from a contiguous grid of 30 x 30 11 m resolution cells across our c. 65 hectare study area. We tested whether: (i) cells with nests differed in 12 their understorey vegetation structure characteristics compared to those without nests; and (ii) the 13 selection of these sites for nesting was adaptive, by assessing the effects of vegetation characteristics on 14 rates of nest success and fledgling survival, and the subsequent probability of a breeding female having 15 any reproductive success. We found that nest-cells differed from unused cells primarily in having denser 16 vegetation in the lowest layer of the understorey (0-2 m; the ‘groundstorey’ layer). Understorey 17 vegetation was also on average lower in height in nest-cells. However, relationships between 18 understorey vegetation structure characteristics and breeding performance were mixed. Nest success 19 rates decreased with higher volumes of groundstorey vegetation; as did fledgling survival rates, though 20 only in nest-cells with lower height vegetation. Reproductive success was not influenced by any of the 21
在野生鸟类种群中，筑巢地周围的植被结构会影响依赖后代的捕食风险2，从而导致繁殖鸟类选择具有较低捕食率的植被特征的筑巢地。然而，对于研究人员来说，在野外很难客观地量化植被结构，这可能解释了为什么人们仍然普遍缺乏对哪些特征在确定捕食率方面最重要的理解。机载激光扫描(ALS)提供了一种强大的手段，可以在不同的空间尺度上以前所未有的分辨率测量植被结构。在这里，我们将ALS与11年的8个繁殖数据结合起来，这些数据来自澳大利亚东南部的一个超级细尾鹩莺的野生种群Malurus cyaneus，这是一个筑巢相对靠近地面的物种，巢和雏鸟的捕食率很高。我们从30 x 30 11 m分辨率单元的连续网格中获得了下层植被(0-8 m)的结构测量数据，这些单元分布在我们的65公顷研究区域内。我们测试了:(i)有巢的细胞与没有巢的细胞相比，其下层植被结构特征是否存在差异;(ii)通过评估植被特征对筑巢成功率和雏鸟成活率的影响，以及随后雌性繁殖成功率的影响，这些筑巢地点的选择是适应性的。我们发现巢巢细胞与未利用的细胞的主要区别在于在下层(0-2 m;“底层”层)。17层植被的巢室平均高度也较低。18种林下植被结构特征与育种性能之间的关系较为复杂。筑巢成功率随着地面植被的增加而降低;雏鸟的存活率也是如此，尽管只有在植被较低的巢室中才有20%。繁殖成功率不受这21种因素的影响
{"title":"Use of Airborne Laser Scanning to assess effects of understorey vegetation structure on nest‐site selection and breeding performance in an Australian passerine bird","authors":"Richard Turner, Ophélie J. D. Lasne, Kara N. Youngentob, S. Shokirov, Helen L. Osmond, L. Kruuk","doi":"10.1002/rse2.342","DOIUrl":"https://doi.org/10.1002/rse2.342","url":null,"abstract":"1 In wild bird populations, the structure of vegetation around nest-sites can influence the risk of predation 2 of dependent young offspring, generating selection for breeding birds to choose nest-sites with 3 vegetation characteristics associated with lower predation rates. However, for researchers, vegetation 4 structure can be difficult to quantify objectively in the field, which might explain why there remains a 5 general lack of understanding of which characteristics are most important in determining rates of 6 predation. Airborne Laser Scanning (ALS) offers a powerful means of measuring vegetation structure 7 at unprecedented resolution across different spatial scales. Here, we combined ALS with 11 years of 8 breeding data from a wild population of superb fairy-wrens Malurus cyaneus in south-east Australia, a 9 species which nests relatively close to the ground and has high rates of nest and fledgling predation. We 10 derived structural measurements of understorey (0-8 m) vegetation from a contiguous grid of 30 x 30 11 m resolution cells across our c. 65 hectare study area. We tested whether: (i) cells with nests differed in 12 their understorey vegetation structure characteristics compared to those without nests; and (ii) the 13 selection of these sites for nesting was adaptive, by assessing the effects of vegetation characteristics on 14 rates of nest success and fledgling survival, and the subsequent probability of a breeding female having 15 any reproductive success. We found that nest-cells differed from unused cells primarily in having denser 16 vegetation in the lowest layer of the understorey (0-2 m; the ‘groundstorey’ layer). Understorey 17 vegetation was also on average lower in height in nest-cells. However, relationships between 18 understorey vegetation structure characteristics and breeding performance were mixed. Nest success 19 rates decreased with higher volumes of groundstorey vegetation; as did fledgling survival rates, though 20 only in nest-cells with lower height vegetation. Reproductive success was not influenced by any of the 21","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44996843","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}
Nicolás Fuentes‐Allende, P. Stephens, Lynne M. MacTavish, Dougal MacTavish, S. G. Willis
Large herbivores in seasonal environments often experience mass variation due to temporal changes in the availability of critical resources like water and forage, as well as due to breeding events. Yet the documentation of mass variation in mammals of highly seasonal savanna habitats, which host the highest densities of grazing ungulates globally, has rarely been explored. Here, we showcase a method to evaluate seasonal mass variation in bovids. Our method used mineral‐baited scales and camera traps to enable us to track the body mass of three species through a period of wet and dry seasons in a South African savanna ecosystem. To illustrate one potential application of the method, we related body mass data to time, weather and resource availability. This showed that individuals altered their body masses markedly between seasons with, for example, female Kudu (Tragelaphus strepsiceros) gaining, on average, >21 kg over the 15‐week wet‐season period in 1 year. These changes were positively related to factors such as vegetation productivity (assessed using NDVI) and the frequency of rains. This method enables easy, non‐lethal and non‐invasive acquisition of mass data. The equipment is easy to deploy concurrently over large areas. Monitoring by this method has a variety of possible applications, potentially providing a useful early‐warning indicator of body condition to inform management, or providing information about ecological states, such as parturition or the reproductive effort of males. Given the longer and harsher dry seasons experienced in many arid systems in recent decades, and projected in future, this method may provide a straightforward means of monitoring long‐term body condition in animals as a result of environmental change.
{"title":"Remote monitoring of short‐term body mass variation in savanna ungulates","authors":"Nicolás Fuentes‐Allende, P. Stephens, Lynne M. MacTavish, Dougal MacTavish, S. G. Willis","doi":"10.1002/rse2.338","DOIUrl":"https://doi.org/10.1002/rse2.338","url":null,"abstract":"Large herbivores in seasonal environments often experience mass variation due to temporal changes in the availability of critical resources like water and forage, as well as due to breeding events. Yet the documentation of mass variation in mammals of highly seasonal savanna habitats, which host the highest densities of grazing ungulates globally, has rarely been explored. Here, we showcase a method to evaluate seasonal mass variation in bovids. Our method used mineral‐baited scales and camera traps to enable us to track the body mass of three species through a period of wet and dry seasons in a South African savanna ecosystem. To illustrate one potential application of the method, we related body mass data to time, weather and resource availability. This showed that individuals altered their body masses markedly between seasons with, for example, female Kudu (Tragelaphus strepsiceros) gaining, on average, >21 kg over the 15‐week wet‐season period in 1 year. These changes were positively related to factors such as vegetation productivity (assessed using NDVI) and the frequency of rains. This method enables easy, non‐lethal and non‐invasive acquisition of mass data. The equipment is easy to deploy concurrently over large areas. Monitoring by this method has a variety of possible applications, potentially providing a useful early‐warning indicator of body condition to inform management, or providing information about ecological states, such as parturition or the reproductive effort of males. Given the longer and harsher dry seasons experienced in many arid systems in recent decades, and projected in future, this method may provide a straightforward means of monitoring long‐term body condition in animals as a result of environmental change.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-05-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"41840743","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}
{"title":"The vulnerability and resilience of seagrass ecosystems to marine heatwaves in New Zealand: a remote sensing analysis of seascape metrics using\u0000 PlanetScope\u0000 imagery","authors":"K. J. Clemente, M. Thomsen, R. Zimmerman","doi":"10.1002/rse2.343","DOIUrl":"https://doi.org/10.1002/rse2.343","url":null,"abstract":"","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-05-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44879370","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}
R. Hernández-Clemente, A. Hornero, V. González-Dugo, M. Berdugo, J. Quero, J. Jiménez, F. Maestre
{"title":"Global monitoring of soil multifunctionality in drylands using satellite imagery and field data","authors":"R. Hernández-Clemente, A. Hornero, V. González-Dugo, M. Berdugo, J. Quero, J. Jiménez, F. Maestre","doi":"10.1002/rse2.340","DOIUrl":"https://doi.org/10.1002/rse2.340","url":null,"abstract":"","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45709191","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}
James G. C. Ball, Sebastian H. M. Hickman, Tobias D. Jackson, Xian Jing Koay, James Hirst, William Jay, Matthew Archer, Mélaine Aubry‐Kientz, Grégoire Vincent, David A. Coomes
Abstract Tropical forests are a major component of the global carbon cycle and home to two‐thirds of terrestrial species. Upper‐canopy trees store the majority of forest carbon and can be vulnerable to drought events and storms. Monitoring their growth and mortality is essential to understanding forest resilience to climate change, but in the context of forest carbon storage, large trees are underrepresented in traditional field surveys, so estimates are poorly constrained. Aerial photographs provide spectral and textural information to discriminate between tree crowns in diverse, complex tropical canopies, potentially opening the door to landscape monitoring of large trees. Here we describe a new deep convolutional neural network method, Detectree2 , which builds on the Mask R‐CNN computer vision framework to recognize the irregular edges of individual tree crowns from airborne RGB imagery. We trained and evaluated this model with 3797 manually delineated tree crowns at three sites in Malaysian Borneo and one site in French Guiana. As an example application, we combined the delineations with repeat lidar surveys (taken between 3 and 6 years apart) of the four sites to estimate the growth and mortality of upper‐canopy trees. Detectree2 delineated 65 000 upper‐canopy trees across 14 km 2 of aerial images. The skill of the automatic method in delineating unseen test trees was good ( F 1 score = 0.64) and for the tallest category of trees was excellent ( F 1 score = 0.74). As predicted from previous field studies, we found that growth rate declined with tree height and tall trees had higher mortality rates than intermediate‐size trees. Our approach demonstrates that deep learning methods can automatically segment trees in widely accessible RGB imagery. This tool (provided as an open‐source Python package) has many potential applications in forest ecology and conservation, from estimating carbon stocks to monitoring forest phenology and restoration. Python package available to install at https://github.com/PatBall1/Detectree2 .
热带森林是全球碳循环的主要组成部分，也是三分之二陆生物种的家园。上冠层树木储存了大部分的森林碳，可能容易受到干旱事件和风暴的影响。监测它们的生长和死亡对于了解森林对气候变化的适应能力至关重要，但在森林碳储量的背景下，传统的实地调查中大树的代表性不足，因此估算结果的约束很差。航空照片提供了光谱和纹理信息，可以区分不同、复杂的热带树冠中的树冠，这可能为大型树木的景观监测打开了大门。在这里，我们描述了一种新的深度卷积神经网络方法Detectree2，它建立在Mask R - CNN计算机视觉框架的基础上，从机载RGB图像中识别单个树冠的不规则边缘。我们在马来西亚婆罗洲的三个地点和法属圭亚那的一个地点对3797个人工绘制的树冠进行了训练和评估。作为一个应用实例，我们结合了四个地点的重复激光雷达调查(间隔3到6年)来估计上冠层树木的生长和死亡率。Detectree2在14公里的航空图像中描绘了65000棵上冠层树木。自动方法对未见测试树的圈定能力较好(f1得分= 0.64)，对最高类别树的圈定能力较好(f1得分= 0.74)。正如以前的野外研究预测的那样，我们发现生长速率随树高而下降，高大树木的死亡率高于中等大小的树木。我们的方法表明，深度学习方法可以在广泛访问的RGB图像中自动分割树。这个工具(作为一个开源的Python包提供)在森林生态和保护中有许多潜在的应用，从估算碳储量到监测森林物候和恢复。Python包可在https://github.com/PatBall1/Detectree2上安装。
{"title":"Accurate delineation of individual tree crowns in tropical forests from aerial <scp>RGB</scp> imagery using Mask <scp>R‐CNN</scp>","authors":"James G. C. Ball, Sebastian H. M. Hickman, Tobias D. Jackson, Xian Jing Koay, James Hirst, William Jay, Matthew Archer, Mélaine Aubry‐Kientz, Grégoire Vincent, David A. Coomes","doi":"10.1002/rse2.332","DOIUrl":"https://doi.org/10.1002/rse2.332","url":null,"abstract":"Abstract Tropical forests are a major component of the global carbon cycle and home to two‐thirds of terrestrial species. Upper‐canopy trees store the majority of forest carbon and can be vulnerable to drought events and storms. Monitoring their growth and mortality is essential to understanding forest resilience to climate change, but in the context of forest carbon storage, large trees are underrepresented in traditional field surveys, so estimates are poorly constrained. Aerial photographs provide spectral and textural information to discriminate between tree crowns in diverse, complex tropical canopies, potentially opening the door to landscape monitoring of large trees. Here we describe a new deep convolutional neural network method, Detectree2 , which builds on the Mask R‐CNN computer vision framework to recognize the irregular edges of individual tree crowns from airborne RGB imagery. We trained and evaluated this model with 3797 manually delineated tree crowns at three sites in Malaysian Borneo and one site in French Guiana. As an example application, we combined the delineations with repeat lidar surveys (taken between 3 and 6 years apart) of the four sites to estimate the growth and mortality of upper‐canopy trees. Detectree2 delineated 65 000 upper‐canopy trees across 14 km 2 of aerial images. The skill of the automatic method in delineating unseen test trees was good ( F 1 score = 0.64) and for the tallest category of trees was excellent ( F 1 score = 0.74). As predicted from previous field studies, we found that growth rate declined with tree height and tall trees had higher mortality rates than intermediate‐size trees. Our approach demonstrates that deep learning methods can automatically segment trees in widely accessible RGB imagery. This tool (provided as an open‐source Python package) has many potential applications in forest ecology and conservation, from estimating carbon stocks to monitoring forest phenology and restoration. Python package available to install at https://github.com/PatBall1/Detectree2 .","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":"222 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135239414","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}
Ben. G. Weinstein, S. Marconi, Sarah J. Graves, Alina Zare, Aditya Singh, Stephanie A. Bohlman, L. Magee, Daniel J. Johnson, P. Townsend, E. White
Measuring forest biodiversity using terrestrial surveys is expensive and can only capture common species abundance in large heterogeneous landscapes. In contrast, combining airborne imagery with computer vision can generate individual tree data at the scales of hundreds of thousands of trees. To train computer vision models, ground‐based species labels are combined with airborne reflectance data. Due to the difficulty of finding rare species in a large landscape, many classification models only include the most abundant species, leading to biased predictions at broad scales. For example, if only common species are used to train the model, this assumes that these samples are representative across the entire landscape. Extending classification models to include rare species requires targeted data collection and algorithmic improvements to overcome large data imbalances between dominant and rare taxa. We use a targeted sampling workflow to the Ordway Swisher Biological Station within the US National Ecological Observatory Network (NEON), where traditional forestry plots had identified six canopy tree species with more than 10 individuals at the site. Combining iterative model development with rare species sampling, we extend a training dataset to include 14 species. Using a multi‐temporal hierarchical model, we demonstrate the ability to include species predicted at <1% frequency in landscape without losing performance on the dominant species. The final model has over 75% accuracy for 14 species with improved rare species classification compared to 61% accuracy of a baseline deep learning model. After filtering out dead trees, we generate landscape species maps of individual crowns for over 670 000 individual trees. We find distinct patches of forest composed of rarer species at the full‐site scale, highlighting the importance of capturing species diversity in training data. We estimate the relative abundance of 14 species within the landscape and provide three measures of uncertainty to generate a range of counts for each species. For example, we estimate that the dominant species, Pinus palustris accounts for c. 28% of predicted stems, with models predicting a range of counts between 160 000 and 210 000 individuals. These maps provide the first estimates of canopy tree diversity within a NEON site to include rare species and provide a blueprint for capturing tree diversity using airborne computer vision at broad scales.
{"title":"Capturing long‐tailed individual tree diversity using an airborne imaging and a multi‐temporal hierarchical model","authors":"Ben. G. Weinstein, S. Marconi, Sarah J. Graves, Alina Zare, Aditya Singh, Stephanie A. Bohlman, L. Magee, Daniel J. Johnson, P. Townsend, E. White","doi":"10.1002/rse2.335","DOIUrl":"https://doi.org/10.1002/rse2.335","url":null,"abstract":"Measuring forest biodiversity using terrestrial surveys is expensive and can only capture common species abundance in large heterogeneous landscapes. In contrast, combining airborne imagery with computer vision can generate individual tree data at the scales of hundreds of thousands of trees. To train computer vision models, ground‐based species labels are combined with airborne reflectance data. Due to the difficulty of finding rare species in a large landscape, many classification models only include the most abundant species, leading to biased predictions at broad scales. For example, if only common species are used to train the model, this assumes that these samples are representative across the entire landscape. Extending classification models to include rare species requires targeted data collection and algorithmic improvements to overcome large data imbalances between dominant and rare taxa. We use a targeted sampling workflow to the Ordway Swisher Biological Station within the US National Ecological Observatory Network (NEON), where traditional forestry plots had identified six canopy tree species with more than 10 individuals at the site. Combining iterative model development with rare species sampling, we extend a training dataset to include 14 species. Using a multi‐temporal hierarchical model, we demonstrate the ability to include species predicted at <1% frequency in landscape without losing performance on the dominant species. The final model has over 75% accuracy for 14 species with improved rare species classification compared to 61% accuracy of a baseline deep learning model. After filtering out dead trees, we generate landscape species maps of individual crowns for over 670 000 individual trees. We find distinct patches of forest composed of rarer species at the full‐site scale, highlighting the importance of capturing species diversity in training data. We estimate the relative abundance of 14 species within the landscape and provide three measures of uncertainty to generate a range of counts for each species. For example, we estimate that the dominant species, Pinus palustris accounts for c. 28% of predicted stems, with models predicting a range of counts between 160 000 and 210 000 individuals. These maps provide the first estimates of canopy tree diversity within a NEON site to include rare species and provide a blueprint for capturing tree diversity using airborne computer vision at broad scales.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42181383","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：