Karatayev Alexander Y, Mehler Knut, Burlakova Lyubov E, Hinchey Elizabeth K, Warren Glenn J
Great Lakes Center, SUNY Buffalo State, Buffalo, NY, USA.
United States Environmental Protection Agency, Great Lakes National Program Office, Chicago, IL, USA.
J Great Lakes Res. 2018;44(4):629-638. doi: 10.1016/j.jglr.2018.05.003.
In contrast to marine systems where remote sensing methods in studies of benthic organisms have been widely used for decades, these methods have experienced limited use in studies of freshwater benthos due to the general lack of large epifauna. The situation has changed with the introduction of dreissenid bivalves capable of creating visible aggregations on lake bottoms into North American freshwaters in the 1980s and 1990s. The need for assessment of densities prompted exploration of videography as a potentially cost-effective tool. We developed a novel sampling method that analyzes video recorded using a GoPro camera mounted to a benthic sled to estimate coverage, density, and biomass over relatively large areas of the lake bed in the Laurentian Great Lakes compared to traditional sampling methods. Using this method, we compared quagga mussel coverage, density, and biomass estimates based on three replicate Ponar grabs vs. 500 m-long video transects across 43 stations sampled in Lake Michigan in 2015. Our results showed that analysis of images from video transects dramatically increased the bottom area surveyed compared to Ponar grabs and increased the precision of density and biomass estimations at monitoring stations. By substantially increasing the ability to detect relatively small (<20%) changes between years within a particular station, this method could be a useful and cost-effective addition for monitoring populations in the Great Lakes and other freshwater systems where they occur.
与海洋系统不同，在海洋系统中，用于底栖生物研究的遥感方法已经广泛应用了数十年，而由于普遍缺乏大型表栖动物，这些方法在淡水底栖生物研究中的应用有限。随着20世纪80年代和90年代能在湖底形成可见聚集的双壳类斑马贻贝被引入北美淡水水域，情况发生了变化。对密度评估的需求促使人们探索将摄像作为一种潜在的经济高效工具。我们开发了一种新颖的采样方法，该方法分析使用安装在底栖雪橇上的GoPro相机录制的视频，以估计与传统采样方法相比，在 Laurentian 五大湖相对较大的湖床区域上的覆盖率、密度和生物量。使用这种方法，我们比较了基于三个重复的 Ponar 抓斗样本与2015年在密歇根湖43个采样站采集的500米长视频样带得出的斑驴贻贝覆盖率、密度和生物量估计值。我们的结果表明，与 Ponar 抓斗相比，视频样带图像分析极大地增加了调查的底部面积，并提高了监测站密度和生物量估计的精度。通过大幅提高检测特定站点不同年份间相对较小（<20%）变化的能力，这种方法对于监测五大湖及其他出现此类生物的淡水系统中的种群而言，可能是一种有用且经济高效的补充手段。
