Keywords
Plant-animal interactions (PAI), environmental DNA (eDNA), molecular ecology, biodiversity loss, non-destructive, biodiversity sampling, conservation management, ecosystem functioning
1. Introduction
More than one million species are at risk of becoming threatened with extinction (IPBES, 2019), heralding the Anthropocene as the sixth mass extinction event (Myers, 1990; Román-Palacios & Wiens, 2020). The loss of species interactions may occur well before the actual extinction of individual species, thereby initiating deleterious effects on species functionality and its service to the ecosystem (Valiente‐Banuet et al., 2015). This in-turn further accelerates species extinction rates (Simmons et al., 2020), which is especially pertinent for specialist species (Colles et al., 2009). In fact, given that the loss of successive interactions provides an early warning system for the deterioration of ecosystem health (Valiente‐Banuet et al., 2015), documenting, monitoring, and conserving such complex interactions is critical to retain ecosystem functioning.
One of the principal means by which taxa are interconnected in nature is via plant-animal interactions (PAI). These interactions can play pivotal ecological roles and materialize in multiple combinations of positive and antagonistic relationships (e.g. predation; frugivory and herbivory, parasitism, and mutualism). For example, frugivory contributes to propagation and thus facilitates plant restoration (Chama et al., 2013; Monge et al., 2020) and gene flow (Robledo-Arnuncio & Garcia, 2007). Without such mutualistic relationships, some plants may not be able to complete their life cycles, and the animals may starve due to resource deficiency. Herbivory leads to defoliation or root removal, which can regulate or diminish overall phytomass, but can also increase species diversity and influence plant distribution (Milchunas & Lauenroth, 1993; Castagneyrol et al., 2017), thereby regulating ecosystem stability (Wirth et al., 2008; Schallhart et al., 2012; Castagneyrol et al., 2017). In pollinator-plant mutualisms, the former acquires feeding from the latter, and in return serves as an agent of plant propagation and a vector for gene flow (Ellis & Johnson, 2012). Studies documenting the food habits of pollinators and their interactive role in sustaining ecosystems have already shed light on the complex network of species-specificity, habitat preference, and co-evolution between plants and their pollinators (Sargent & Ackerly, 2008). Mutualisms also assist with growth and offer protection from pathogens (e.g., plant- insect associations; Rasmussen et al., 2021). In contrast, antagonistic interactions (e.g., parasites, parasitoids) can affect the growth of plants and result in economical and ecological loss (Derocles et al., 2015). Thus, PAI underpin many of the fundamental processes related to ecosystem structure and functioning (Pacini et al., 2008). However, studying these multifaceted interactions using conventional methods (e.g., field observation, camera, malaise, and pitfall traps, and gut-content analysis), is often difficult and laborious (Thomsen & Sigsgaard, 2019). Alternatively, molecular advancements with the analysis of trace DNA from environmental samples (i.e., environmental DNA or ‘eDNA’) have provided researchers and managers the ability to scale up documentation and monitoring of such relationships, and to do so at increased spatiotemporal frequencies with more cost effectiveness (see Figure 1).
Methodological development for the application of eDNA has rapidly evolved from presence/absence detection of organisms (Ficetola et al., 2008) and abundance quantification of eDNA signals (Taberlet et al., 2012), to the detection of whole communities (Deiner et al., 2021) and even their trophic interactions (Thomsen & Sigsgaard, 2019; D’Alessandro & Mariani 2021). Indeed, eDNA-based methods have experienced a sharp adoption in different fields such as conservation biology (e.g., detection of endangered or invasive species; Piaggio et al., 2014; Stewart et al. 2017), ecological biomonitoring in terrestrial and aquatic ecosystem (e.g., environmental health monitoring; Xie et al., 2017), wildlife forensics (Allwood et al., 2020), wildlife disease monitoring (Barnes et al., 2020), and animal behaviour (Nichols et al., 2015). The application of eDNA methods to investigate a myriad of ecological interactions such as pollination (e.g., plant insects, plant-animal), predation (e.g., herbivory, frugivory), and mutualism (e.g., plant-nematode, plant-insect, plant-animals) (Thomsen & Sigsgaard, 2019; Van Beeck Calkoen et al., 2019; Rasmussen et al., 2021) further demonstrates the application of eDNA as a multidisciplinary approach (Deiner et al., 2021; Veilleux et al., 2021) poised to tackle complex ecological questions regarding inter-taxa relationships. However, as with every newly developed method, the importance of eDNA on PAI studies remains less evaluated.
Here we review the use of eDNA-based methods to study PAI. We discuss the advantages and current limitations of such methods, and propose research priorities that may improve future eDNA-based methods for PAI analysis. Within this context, our goal is to highlight for both researchers and managers, the potential utility of non-invasive/destructive eDNA-based methods, but we also aim to identify and clarify uncertainties and next steps needed to advance these methods for broad application.