Detection of freshwater mussels (Unionidae) using environmental DNA in riverine systems

Abstract

Environmental DNA (eDNA) methods are being developed for use in conservation biology to improve upon conventional species survey techniques. Validation of eDNA methods in different environmental contexts is required if they are to be widely adopted. One potential application of eDNA methods is for the detection of freshwater mussels (Bivalvia: Unionidae), which are among the most imperiled species in North America. Conventional unionid survey methods are highly invasive and can be difficult to conduct due to issues with morphological identification and their cryptic use of habitat. eDNA methods can potentially provide a non-invasive, extremely specific, and highly sensitive alternative. Here, we examine the effectiveness of eDNA methods at detecting an imperiled unionid, the wavy-rayed lampmussel (Lampsilis fasciola), in lotic systems with moderate discharge. We developed a novel qPCR assay for the detection of L. fasciola eDNA, which included a custom internal positive control to check for PCR inhibition. We used different experimental densities of caged L. fasciola specimens as a point source of eDNA within two rivers of the Grand River watershed in Southern Ontario. Sampling occurred at set distances downstream of the cage using purpose-built sampling equipment. Detection was obtained at the cage (i.e., point of eDNA shedding) but not downstream at distances ≥10 m during stream discharges of approximately 1,632–2,332 L/s. The results indicate that eDNA is diluted rapidly in rivers with moderate discharge and that high-resolution spatial sampling efforts may be necessary to obtain meaningful eDNA-based distribution data of unionids, and other sessile organisms, present at low density in lotic systems.

1 INTRODUCTION

North America has the highest level of freshwater mussel (Unionidae) diversity in the world, with 297 native species (Williams, Warren, Cummings, Harris, & Neves, 1993). In addition to having high species diversity, unionids have one of the highest rates of imperilment for organisms in North America, with an estimated 29 species falling to extinction in the last century (Haag & Williams, 2014). Canada is home to 55 species of unionids, 41 of which are found in Ontario (Galbraith, Zanatta, & Wilson, 2015). Of these, 15 are imperiled to some degree. The severe decline of unionid populations is concerning as they contribute to a myriad of ecological processes. Unionids influence bottom-up trophic effects, increase nutrient flux within ecosystems, stabilize substrate, and improve water quality (Allen et al., 2012; Haag & Williams, 2014; Howard & Cuffey, 2006). This general decline may be due to a variety of factors. Anthropogenic effects impact unionids in several ways, including but not limited to: wastewater effluents, siltation, stream impoundment, chemical pollution, agricultural runoff, and the introduction of invasive bivalves (Bogan, 1993; Bringolf et al., 2007; Gillis et al., 2017; Prosser, Rochfort, Mcinnis, Exall, & Gillis, 2017). In particular, the introduction of the highly invasive zebra mussel (Dreissena polymorpha, Pallas 1771) to North America has had significant detrimental effects on unionid populations through fouling and disruption of mussel beds (Bossenbroek et al., 2018; Haag, Berg, Garton, & Farris, 1993).

Another contributing factor impacting the imperilment of unionids is the sensitivity of early life stages to environmental stressors (Galbraith et al., 2015). Unionids release their young as glochidia, which parasitize the gills of fish or amphibians in order to develop into juvenile mussels. Consequently, this relationship makes unionids indirectly sensitive to negative effects on host organisms, which are thought to contribute to lower rates of glochidia recruitment, and in turn higher mortality during early life stages (Bringolf et al., 2007).

Surveys must be conducted to better understand how unionid populations are currently distributed; however, they can be difficult to conduct and require significant expertise (Currier et al., 2018; Mackie, Morris, & Ming, 2008). Unionids are difficult to find in aquatic habitats due to the way they burrow into the benthic substrate, leaving only a portion of their exterior visible. This is further complicated by factors such as water depth and turbidity, often causing visibility of the benthic zone to be non-existent (Mackie et al., 2008; Sansom & Sassoubre, 2017). Traditional surveys attempt to quantitatively identify mussels via random quadrat sampling (RQS), a technique that involves surveying 1 m² sections of substrate and counting the number and abundance of species (Mackie et al., 2008). RQS is not completely effective as it can overlook species present at low density, cause harassment to organisms, and can be very costly to conduct over large, or difficult to survey, areas (Sansom & Sassoubre, 2017). Qualitative surveys, such as timed searches, can be easier to conduct than RQS but have inherent disadvantages for finding cryptic species (Obermeyer, 1998). Alternative survey methods such as adaptive cluster sampling have been proposed for detecting unionids at low density; however, this method can become inefficient when a large search area is required and sample size increases (Smith, Villella, & Lemarié, 2003). Novel survey methods are needed to fully assess unionid populations as current methods are limiting in respect to species occurrence, density estimates, and upon the number of qualified personnel that can conduct them.

Environmental DNA (eDNA) analysis is a rapidly developing environmental survey technique which has the potential to improve many aspects of aquatic species sampling (Goldberg, Strickler, & Pilliod, 2015). eDNA methods have been proven in multiple contexts to be more sensitive, less costly, and less disturbing to the environment than conventional species detection techniques (Goldberg, Strickler, & Fremier, 2018; Hunter et al., 2018; Pilliod, Goldberg, Arkle, Waits, & Richardson, 2013; Simmons, Tucker, Chadderton, Jerde, & Mahon, 2016; Wilcox et al., 2013). However, there are a number of environmental and methodological variables which may injuriously affect the results of any eDNA study if not properly accounted for (Barnes et al., 2014; Jane et al., 2015; Wilcox et al., 2016). One variable of major concern is the presence of PCR inhibitors that prevent DNA amplification and mask eDNA presence (both in qPCR and next-generation sequencing), leading to false-negative results (McKee, Spear, & Pierson, 2015; Wilcox et al., 2018). Another example is the effect of water flow on eDNA detection probability (Deiner & Altermatt, 2014; Jane et al., 2015; Wilcox et al., 2016). Our understanding of how to navigate environmental variables to avoid confounding influences, and to maximize eDNA detection probability, should eventually culminate in a set of eDNA standards for different types of biological systems, pushing eDNA methods toward the forefront of conservation science.

There exists a potential for eDNA methods to be implemented for unionid surveys in conjunction with recovery efforts for imperiled species (e.g., relocation or reintroduction; Fisheries and Oceans Canada, 2018). One such imperiled species is the wavy-rayed lampmussel (Lampsilis fasciola, Rafinesque 1820) classified as “special concern” in 2010 (Fisheries and Oceans Canada, 2018). L. fasciola populations in Canada are limited to four river systems and one delta in southern Ontario. The purpose of this study was to examine eDNA detection rate at set sampling distances, under measured stream discharge, downstream of caged L. fasciola specimens placed in virgin territory, while also controlling for PCR inhibition during analysis. Our results will inform improvements to future eDNA surveys.

请联系上海蔚雨科技获取详细信息