In order to understand the state-of-the-art in eDNA to study PAI and enable an effective discussion of this application, we qualitatively reviewed studies incorporating eDNA methods into PAI research (targeted and metabarcoding approaches) by searching the literature published between 2009 to 2021 using Google Scholar (https://scholar.google.com/) with the following search conditions: 1) “[eDNA] AND [plant-animal interactions]”; 2) “[eDNA] AND [pollination]”; 3) “[pollen metabarcoding]”; 4) “[eDNA] AND [herbivory]”; 5) “[herbivory metabarcoding]”; and 6) “[fecal metabarcoding]”. The content of the top 100 publications on Google Scholar were verified manually. We selected papers that worked on non-destructive eDNA-based methods for representation. The selected publications were then evaluated to understand the potential advantages and limitations of eDNA-based methods in the study of PAI.
Why use eDNA-based methods for studying PAI?
Species interactions are dynamic processes and their subsequent observation is difficult using discrete means of data collection (i.e. the conventional methods which are difficult to scale up in space and time). Studying species interactions would therefore require sampling methods that provide broad spatial and temporal inference. DNA-based methods offer a broad output with the capability of identifying multiple PAI simultaneously, and the ease at which DNA is collected and analysed also affords multiple sampling events for an integrative approach. The conventional applications of DNA-based methods (e.g., metabarcoding of gut contents and bulk samples) have already proved useful in elucidating complex species and trophic interactions (García-Robledo et al., 2013). For instance, conventional DNA analysis from gut content or bulk samples have identified different nodes across various food webs, and reconstructed the trophic links in terrestrial (Wirta et al. 2014, 2015a, 2015b, 2016; Gogarten et al., 2020), aquatic (Leray et al., 2012; Leray et al., 2015) and often inaccessible environments, such as deep-sea beds, hydrothermal vents, and cold-seeps (Olsen et al., 2014). Several reviews to date have summarized the history, achievements, and current applications of studying species interaction using conventional DNA-based methods across multiple fields (Symondson, 2002; Valentini et al., 2009; Pompanon et al., 2012; Clare, 2014; Kress et al., 2015; Evans et al., 2016).