Contrasting performance of Lidar and optical texture models in predicting avian diversity in a tropical mountain forest
Contrasting performance of Lidar and optical texture models in predicting avian diversity in a tropical mountain forest
复制标题
DOI：
10.1016/j.rse.2015.12.019
复制
发表时间：
2016-03-01
影响因子：
13.5
通讯作者：
Bendix, Joerg
中科院分区：
文献类型：
--
作者：
Wallis, Christine I. B.;Paulsch, Detlev;Bendix, Joerg
世界各地的生态系统受到土地利用和气候变化日益增加的影响的威胁。为了保护它们的多样性和功能性，需要空间上明确的监测系统。在偏远地区，监测很困难，经常性的实地调查费用高昂。通过使用激光雷达或更具成本效益和重复性的光学卫星数据，遥感可以为生境结构提供指标，支持保护生物多样性的措施。在这里，我们比较了机载激光雷达和光学卫星数据在模拟厄瓜多尔东南部复杂热带山区森林生态系统中鸟类生物多样性的空间分布方面的解释能力。我们使用了鸟类实地调查的数据，并选择了三个指标作为多样性不同方面的指标：(I)香农多样性作为α-多样性的指标，它还包括物种的相对丰富度；(Ii)系统多样性作为功能多样性的第一个指标；以及(Iii)群落组成作为组合的α-和β-多样性的指标。我们分别使用激光雷达和光学纹理度量的偏最小二乘回归对这些多样性估计进行建模，并使用留一验证的R-2和均方根误差对模型进行比较。两个遥感数据集对鸟类群落信息的预测效果最好，其次是香农多样性和系统多样性。我们的发现表明，光学纹理度量在预测香农多样性和衡量群落组成方面具有很高的潜力，但不能用于模拟系统多样性。从所调查的热带山区生态系统总结，我们得出结论，从业务卫星系统的多光谱数据中提取纹理信息可以取代昂贵的机载激光扫描来模拟生物多样性的某些方面。(C)2015 Elsevier Inc.保留所有权利。
Ecosystems worldwide are threatened by the increasing impact of land use and climate change. To protect their diversity and functionality, spatially explicit monitoring systems are needed. In remote areas, monitoring is difficult and recurrent field surveys are costly. By using Lidar or the more cost-effective and repetitive optical satellite data, remote sensing could provide proxies for habitat structure supporting measures for the conservation of biodiversity. Here we compared the explanatory power of both, airborne Lidar and optical satellite data in modeling the spatial distribution of biodiversity of birds across a complex tropical mountain forest ecosystem in southeastern Ecuador. We used data from field surveys of birds and chose three measures as proxies for different laspects of diversity: (i) Shannon diversity as a measure of alpha-diversity that also includes the relative abundance of species, (ii) phylodiversity as a first proxy for functional diversity, and (iii) community composition as a proxy for combined alpha- and beta-diversity. We modeled these diversity estimates using partial least-square regression of Lidar and optical texture metrics separately and compared the models using a leave-one-out validated R-2 and root mean square error. Bird community information was best predicted by both remote sensing datasets, followed by Shannon diversity and phylodiversity. Our findings reveal a high potential of optical texture metrics for predicting Shannon diversity and a measure of community composition, but not for modeling phylodiversity. Generalizing from the investigated tropical mountain ecosystem, we conclude that texture information retrieved from multispectral data of operational satellite systems could replace costly airborne laser-scanning for modeling certain aspects of biodiversity. (C) 2015 Elsevier Inc. All rights reserved.