Environmental DNA time series reveals seasonal community and functional shifts

Environmental DNA (eDNA) extracted from water is routinely used in river biodiversity research, and via metabarcoding eDNA can provide comprehensive taxa lists with little effort and cost. However, eDNA-based species detection in streams and rivers may be influenced by sampling season, location, and other key factors such as water temperature and discharge. Research linking these factors and also informing on the potential of eDNA metabarcoding to detect shifts in ecological signatures, such as species phenology and functional feeding groups across seasons, is missing. To address this gap, we collected 102 water samples every two weeks for 15 months at a long-term ecological research (LTER) site and at three different positions in the river's cross section, specifically the water surface, riverbed, and riverbank. We analyzed macroinvertebrate species and molecular Operational Taxonomic Unit (OTU) richness and temporal community turnover across seasons and sampling positions based on COI metabarcoding data. Using Generalized Additive Models, we found a significant influence of sampling season but not sampling position on community composition. Community turnover followed a cyclic pattern, reflecting the continuous change of the macroinvertebrate community throughout the year ('seasonal clock'). Although water temperature had no influence on the inferred community composition, higher discharge reduced the number of Annelida and Ephemeroptera species detectable with eDNA. Most macroinvertebrate taxa showed the highest detection rates in spring, in particular merolimnic species with univoltine life cycles. Further, we detected an increase in proportion of shredders in winter and of parasites in summer. Our results show the usefulness of highly resolved eDNA metabarcoding time series data for ecological research and biodiversity monitoring in streams and rivers.

从水体中提取的环境DNA（eDNA）已常规应用于河流生物多样性研究，通过元条形码技术，eDNA能够以极低的人力与成本获取全面的类群清单。然而，基于eDNA的溪流与河流物种检测可能会受到采样季节、采样点位，以及水温、流量等其他关键因素的影响。目前仍缺乏能够整合这些影响因素，并阐明eDNA元条形码技术用于检测生态特征变化（如不同季节的物种物候与功能摄食类群变化）的相关研究。为填补这一研究空白，我们在一个长期生态研究（LTER）站点，于河流横断面的三个不同点位（分别为水面、河床与河岸），每两周采集一次水样，整个采样周期持续15个月，共计获取102份水样。我们基于细胞色素C氧化酶亚基I（COI）元条形码测序数据，分析了不同季节与采样点位下的大型底栖无脊椎动物物种与分子操作分类单元（Operational Taxonomic Unit, OTU）丰富度，以及群落的时间周转情况。借助广义加性模型（Generalized Additive Models），我们发现采样季节对群落组成存在显著影响，而采样点位则无此效应。群落周转呈现周期性模式，反映出全年大型底栖无脊椎动物群落的持续动态变化（即‘季节时钟’）。尽管水温对推断得到的群落组成无显著影响，但更高的流量会降低通过eDNA可检测到的环节动物门（Annelida）与蜉蝣目（Ephemeroptera）物种数量。多数大型底栖无脊椎动物类群的检出率在春季达到峰值，其中尤以具有一化性生活周期的近岸底栖（merolimnic）物种最为显著。此外，我们还发现碎食者类群的占比在冬季有所上升，而寄生者类群的占比则在夏季升高。本研究结果证实了高分辨率eDNA元条形码时间序列数据在溪流与河流生态研究及生物多样性监测中的应用价值。