A framework for standardizing the processes of eDNA monitoring and an accessible vision of the future
摘要: eDNA (environmental DNA)是指从环境样品（水体、土壤、沉积物、空气、混合物等）中提取的DNA，是各种生物的DNA混合物，区别于传统从单物种样品中提取的DNA。eDNA监测是指从环境样品中提取DNA，用物种特异性引物或特定DNA metabarcoding引物对其进行扩增测序、分类学分析、相对丰度分析、功能预测等，以监测环境中是否存在某特定物种，或者获得环境中物种组成、群落结构、生态功能等相关信息。eDNA监测主要应用在特定物种监测检出、生物多样性调查评估、群落结构和功能分析、生态系统过程分析评估等领域。eDNA监测可以应用在陆地、水体、空气、器物表面、生物（内）表面等任何有可能存在不确定DNA的场景。因为eDNA监测的非侵入性、灵敏检出、大类群检出、易标准化方法与结果、低人力物力时间成本、结果可核查等特点，未来很有希望发展成为一个常规的物种监测、群落功能预测、生态过程分析的方法，对象可涵盖所有生境条件下的所有生物类群。但要实现这个图景，需要从普遍性和特殊性层面完成eDNA监测技术链条中10个关键环节的技术标准化：样品重复数设计、采样时间设计、采样点设计、采样方法设计、样品前处理、样品保存、扩增引物选择、DNA提取-扩增-测序-分析程序、序列比对注释、监测结果后评估。目前来看，10个环节的标准化基本不存在理论上的困难，并且一部分支撑标准化的研究已经启动，甚至一部分标准化的组织工作已在推进，在未来数年内有望完成关键知识、关键数据的积累，推动eDNA监测成为长期性周期性常规生态监测内容，支撑渐行渐近的数据驱动型科学研究和生态管理。
Abstract: Environmental DNA (eDNA) is DNA extracted from any type of environmental sample (e.g. water, soil, sediment, air, mixture, etc.), which is a DNA mixture originated from different species and individuals, distinguish from a pure DNA sample extracted from a particular organism. eDNA monitoring refers the processes that 1) extracting DNA sample from environmental sample, 2) using definite species-specific primers or meta-barcoding primers to amplify and sequence eDNA sample, 3) clustering the operational taxonomic units (OTUs) and identifying their taxa against reference databases, 4) calculating the relative abundance of each OTUs/ species and other biodiversity indexes, 5) analyzing the corresponding ecosystem structure, processes or function. According to eDNA monitoring, a definite species (or other taxonomic units) in the sampling site could be identified, and the biological information about species composition, community structure, ecosystem processes, ecological function of the research area could be collected. eDNA monitoring has been applied in monitoring and early warning definite species, investigating and assessing biodiversity, detecting and analyzing community structure and function, studying and quantifying ecosystem processes and so on. eDNA monitoring could work in any type of environmental scene where there is unidentified DNA trace, such as in terrestrial environment, aquatic environment, air environment, body surface, organism (inner) surface and so on. As an emerging tool for documenting species presence without direct observation, allowing for sensitive and efficient detection, easy-to-standardize sampling and analyzing approach, comprehensive taxonomic groups coverage, less reliant on taxonomic expertise and auditable by third-party researchers, eDNA monitoring would be a prospective general method for species monitoring, community function predicting and ecosystem processes analyzing in future. Moreover, the objective scope of eDNA monitoring covers all environmental conditions and all biological taxonomies. However, to realize the prospective application vision of eDNA monitoring, there are ten crucial links that need to be standardized at both general level and definite level. 1) Design of duplicated samples for a region with definite environment conditions. The number of duplicated samples could be generally identified just using species accumulation curves. 2) Design of sampling time for a region with definite environment conditions. The interval of sampling time could be generally identified by quantifying the degradation ratio or the retention time of the eDNA from different taxonomic organisms in definite environment conditions. 3) Design of sampling sites for a region with definite environment conditions. The distance of sampling sites could be generally identified by quantifying the effective transportation distance or the spatial heterogeneity of the eDNA from different taxonomic organisms in definite environment conditions. 4) Design of sampling method. For different study areas, objects and aims, there are different optimal sample types (water, soil, sediment or other samples). Don’t combine different duplicated samples, or some rare species would be omitted because of their too weak signals. 5) Pretreatment of samples. Pretreatment of samples mainly refers filtration of water samples. It’s suggested that filtering water samples should use finer millipore glass fiber filter. Don’t remove large particles by prefiltering water sample, or some species signals could be removed. 6) Storage of samples. It’s suggested that samples could be stored at -20 or -80 centigrade, except water samples. Water samples should be kept cool in ice bath and be filtered as soon as possible. 7) Choosing of primers. The primers of metabarcoding of the 16S rRNA gene are widely used for detecting bacteria and archaea. The primers of metabarcoding of the ITS and 18S rRNA genes are widely used for detecting fungi. The primers of metabarcoding of the mitochondrial CO1, 12S rRNA and Cyt b genes are widely used for detecting metazoan. Metagenome is another choice for identifying species. 8) Experiment processes of DNA extraction, amplifying, sequencing and analyzing. As the experiment processes are more and more tending to be processed by commercial biolabs, a set of general experimental parameters is needed. 9) Taxonomic identification of OTUs. Good reference databases, either comprehensive reference databases or local customizable reference databases, are required. 10) Post-evaluation of results. Post-evaluation of results mainly pays attention on whether the number of duplicated samples is sufficient, whether the frequency of sampling is suitable, whether the spatial distance between sampling sites is suitable, whether the taxonomic identification of OTUs is accurate. Until now, there is no theoretical difficulty in standardizing these ten crucial links. Now, the mainly work is the accumulation of datasets and knowledge. Some studies on supporting the standardization have been processed. Parts of standardizing works have been organized both at home and abroad. We expect that the accumulation of crucial datasets and knowledge on eDNA monitoring in hot regions could finish in future several years, and then the eDNA monitoring could be a general work, even a long term basic work in hot regions. As the eDNA monitoring could produce comprehensive and standard datasets, along with the long term basic work of eDNA monitoring realizing, the long time series datasets could be used to detect the biodiversity (especially hiddenbiodiversity) variations and study the dynamic and evolution of ecosystem structure, processes, function and health. Moreover, we expect a series of datasets with high quality, rigour, availability and transparency in future to support the open science and the data-intensive scientific discovery and ecosystem management.