Riparian Final Vegclass. Olympic National Park NRCA Westside Rivers Riparian Vegetation Project (2016)

Beschta and Ripple (2012) assert that increased elk populations in the Olympic National Park due to extirpation of wolves in the 1920s has led to a reduction in riparian vegetation. They hypothesize that a decrease in this vegetation has led to an increase in erosion and undercutting of large conifer trees along the river banks, causing woody debris in the river, which in turn impacts channel morphology. Using imagery dating from 1939 and a set of digitized channel margins for each year, we classified vegetation changes that have occurred along the Hoh, Queets, and Quinault Rivers. We focused on identifying large conifers near the river that could impact water flow and channel morphology if undercut and classified all other vegetation as "other". METHODS: Our classification of large conifer trees along the river margins was based on three rules: 1) Minimum Height: we selected only those trees whose height would indicate the tree was at least 50-cm in diameter at the time of the imagery (calculated from LIDAR). Maximum Crown-Width: we used an estimated maximum crown width of 26-m for conifers in the Olympic National Park (Van Pelt et al., 2006) as an indicator of potential root structure size that could be undermined. A 13-m radius buffer was applied to all tree locations that met our minimum height requirement. 2) Distance-to-River: in order to capture potential channel migration that could occur for a given year, we used a 20-m buffer from the channel margin based on Van Pelt's estimated range of 16-26-m per year. We created an ArcGIS ModelBuilder tool that identified the conifers that met our conditions and created polygon outputs that could be used to split the river margins into segments of "conifer" and "other". For those imagery years that occurred prior to the LIDAR dates, we performed a visual inspection of the results and manually classified any areas where large conifer had existed but were undercut by the river historically. For those imagery years that occurred after the LIDAR dates, we manually corrected any locations where large conifer have since been removed by river migration and were misidentified by the tool. REFERENCES: Beschta, R.L., and W.J. Ripple. 2012. The role of large predators in maintaining riparian plant communities and river morphology. Geomorphology 157-158:88-98. Van Pelt, R., T. C. OKeefe, J. J. Latterell, R. J. Naiman. 2006. Riparian Forest Stand Development along the Queets River in Olympic National Park, Washington. Ecological Monographs 76(2):277-298.Beschta与Ripple（2012）指出，奥林匹克国家公园（Olympic National Park）内20世纪20年代灰狼的局部灭绝导致马鹿（elk）种群数量上升，进而造成河岸带植被（riparian vegetation）退化。他们提出假说认为，该类植被的减少会加剧河岸大型针叶树的侵蚀与根蚀掏空现象，使河道内堆积木质碎屑，最终对河道地貌（channel morphology）产生负面影响。 本研究利用1939年的遥感影像以及逐年数字化的河道岸线（channel margin）数据集，对霍河（Hoh River）、奎茨河（Queets River）与奎纳尔特河（Quinault River）沿岸发生的植被变化开展分类工作。研究重点识别河岸附近若发生根蚀掏空便可能干扰水流与河道地貌的大型针叶树，并将其余所有植被归类为"other"（其他植被）。 研究方法：本研究针对河岸大型针叶树的分类基于三项准则： 1. 高度与冠幅约束：仅选取通过激光雷达（LIDAR）数据反推、影像拍摄时胸径至少达50厘米的树木；结合Van Pelt等（2006）对奥林匹克国家公园内针叶树的最大冠幅估算值26米，以此表征潜在根系规模，并为满足高度要求的所有树木点位设置13米半径的缓冲区。 2. 距河距离约束：为涵盖特定年份可能发生的河道迁移范围，本研究参考Van Pelt估算的年河道迁移幅度16~26米，以河道岸线为基准设置20米缓冲区。 本研究构建了ArcGIS ModelBuilder工具，可自动识别符合条件的针叶树并生成多边形输出（polygon outputs）结果，用于将河道岸线划分为"针叶树岸段"与"其他岸段"。对于激光雷达数据采集时间之前的影像年份，研究人员对自动分类结果开展目视校验，并手动标注历史上存在大型针叶树但已被河道掏空的区域；对于激光雷达数据采集之后的影像年份，则手动修正因河道迁移导致大型针叶树被移除、被工具误识别的点位。 参考文献： 1. Beschta, R.L. 与 W.J. Ripple. 2012. 大型捕食者在维持河岸带植物群落与河道地貌中的作用. 《地貌学（Geomorphology）》, 157-158: 88-98. 2. Van Pelt, R., T.C. O'Keefe, J.J. Latterell 与 R.J. Naiman. 2006. 华盛顿州奥林匹克国家公园奎茨河沿岸河岸带林分发育过程. 《生态专著（Ecological Monographs）》, 76(2): 277-298.