1 | Introduction
Ctenophora is a phylum of gelatinous marine zooplankton with nearly 200 named species (Mills 2017). Their unique body plan separates them from other gelatinous plankton and from Cnidarians with which they are sometimes associated (Dunn et al. 2015). They are found throughout the world ocean, from both poles to the equator and from the surface to the deep-sea (Harbison et al. 1978), with the deepest ctenophore observed at over 7,000 meters depth (Lindsay & Miyake 2007). Many species are common and well-studied, particularly coastal species like Mnemiopsis leidyi , which is noted for having been introduced to habitats around the world. However, most deep-living ctenophores remain undescribed because specimens are delicate, difficult to access, and often damaged during collection (Haddock 2004). The use of remotely operated vehicles (ROVs) and specialized sampling equipment have expanded our ability to observe and collect ctenophores in the deep-sea during the last 30 years of research (Haddock et al. 2017). Morphological investigations of specimens collected from the deep sea suggest that we have only begun to reveal the remarkable diversity within this phylum. However, there are few taxonomic experts who work on ctenophores, and morphological identifications often are stymied by damaged specimens, cryptic morphology, and poor preservation in all fixatives. Molecular identifications can provide relatively quick identifications, especially for taxa like ctenophores that have few taxonomic experts, although polymorphic loci and good reference libraries are critical to achieve this goal.
Sequence data from the nuclear 18S ribosomal gene provide a molecular phylogenetic framework for the broad relationships within ctenophores (Podar et al. 2001; Simion et al.2015), and transcriptomes from a handful of species allow for more in-depth studies of some representative diversity (Simion et al.2017). However, the 18S rDNA gene fragment is highly conserved, and the phylogenies often do not effectively discriminate between many species and closely related genera particularly in groups such as Lobata. For example, some genera, such as Bathocyroe, Eurhamphaea , Deiopea and Kiyohimea , have nearly identical18S rDNA sequences, showing the limitations of the utility of the18S fragment with respect to species delineation (Haddock et al. in prep). The “barcoding” mitochondrial cytochrome-c-oxidase subunit-I (COI ) sequence fragment is typically useful for species identification and delimitation (DeSalle & Goldstein 2019). Many degenerate PCR primer sets are available to amplify a broad swath of taxa, from bacteria to humans (Folmer et al. 1994; Gelleret al. 2013; Leray et al. 2013; Siddall et al.2009). While the barcode locus is successfully amplified for many taxa, amplification and/or its utility is problematic for many others, often for non-model organisms (Vrijenhoek & Waples 2012). One such example is Ctenophora, since many species have extremely high rates of mitochondrial evolution, are rich in adenine (A) and thymine (T) residues, and have variable gene order within the mitochondrial genome, even within a genus (Arafat et al. 2018; Kohn et al. 2012; Pett et al. 2011; Schultz et al. 2020). A consequence of high mitochondrial variability is that common primers are often poorly suited to amplify the COI fragment. Unsuccessful amplification of ctenophores by commonly used barcoding primers has a number of important ramifications including: a lack of quick and easy molecular identifications that results in difficulties for revealing diversity, few ctenophore sequences in public databases, and a deficiency of easily amplified phylogenetic and population-genetic markers.
The lack of robust COI primers and the resulting paucity of ctenophore sequences available in public repositories also hampers our understanding of the role of ctenophores in ocean ecology. Although they are delicate animals, ctenophores are carnivorous and play a critical role as predators in food webs. Despite their seemingly low nutrient content, they are also prey for a large range of animals (Choy et al. 2017; Thiebot & McInnes 2020; Yeh et al.
2020). Metabarcoding and eDNA studies are powerful tools to assess community diversity, ecosystem monitoring, and function (Eble et al. 2020). Many manuscripts have highlighted the presence and abundance of ctenophores in zooplankton metabarcoding analyses based on 18SrDNA fragments (Günther et al. 2018; López-Escardó et al.2018; Preston et al. 2020; Schroeder et al. 2020; Yehet al. 2020); However, since the fragment of 18S is often used in metabarcoding studies but lacks resolution to discriminate between most species of ctenophores, often all members of the entire phylum are lumped together (Günther et al. 2018; López-Escardóet al. 2018; Preston et al. 2020; Schroeder et al.2020; Yeh et al. 2020). In manuscripts that used multiple loci in a metabarcoding framework, a significant proportion of 18Ssequences were from ctenophores, yet COI often failed to detect any (Djurhuus et al. 2018; Günther et al. 2018), or the ctenophore sequences had poor taxonomic assignments so the results were not addressed (Pitz et al. 2020). For other environmental studies of metabarcoding and eDNA, the entire phylum of Ctenophora is often missing from diversity estimates (Lacoursière-Roussel et al.2018) or they are lumped in an ‘unassigned taxa’ category (Leray & Knowlton 2015).
For this study, we designed multiple primers to amplify COI from all major clades of ctenophores, including many deep-living, undescribed species. We applied those primers across the phylum and tested taxonomic assignments for various groups. Finally, as a case study, we used our newly generated sequences as a library for an eDNA study along the eastern Pacific coast, and provided species-level resolution for taxonomic assignments of ctenophores.