Naididae, Ehrenberg, 1831

Erséus, Christer, Williams, Bronwyn W., Horn, Kevin M., Halanych, Kenneth M., Santos, Scott R., James, Samuel W., Châtelliers, Michel Creuzé des & Anderson, Frank E., 2020, Phylogenomic analyses reveal a Palaeozoic radiation and support a freshwater origin for clitellate annelids, Zoologica Scripta 49 (5), pp. 614-640 : 631-632

4.3.2 | Naididae and Phreodrilidae

Parvidrilidae and Capilloventridae (but not Randiellidae % are the only clitellates with hair chaetae in both dorsal and lateral bundles. Among all other clitellates, hairs only occur in some naidids (including former Tubificidae and Opistocystidae ; Erséus, Envall, et al., 2010% and most phreodrilids, but then only in dorsal bundles. Naididae and Phreodrilidae have, for this reason and similarities in the genital organs, been considered as closely related to each other, and this was supported by early molecular approaches ( Envall et al., 2006; Erséus, 2005; Sjölin, Erséus, & Källersjö, 2005%. We found maximum support for them as sister families, and that they together are the sister to the remaining crown-group taxa ( Propappus through Acanthodrilidae in Figure 1%. Phreodrilidae , with about 50 described species and two subfamilies (Martin, Martinez-Ansemil, Pinder, Timm, & Wetzel, 2007% and largely restricted to freshwaters of the Southern Hemisphere, is monophyletic, which is also supported by some autapomorphic morphological characters: dorsal chaetae absent from segment II, hair chaetae (when present% each accompanied by a pair of support chaetae, atria (male organs% with thick glandular walls and spermathecae located in XIII ( Pinder & Brinkhurst, 1997 %. Moreover, Jamieson (1981 % noted peculiarities in the acrosome, as well as a long spiral midpiece, in the ultrastructure of phreodrilid sperm.

The phylogeny and classification of the cosmopolitan and diverse Naididae sensu lato (i.e. including Tubificidae and Opistocystidae % have been subject to much debate over the years. The story goes back to the review by Brinkhurst (1971%, but cladistics and molecular approaches have enabled more modern analyses concerning general morphology (e.g. Brinkhurst, 1991, 1994; Erséus, 1987, 1990, 1992; Gustavsson & Erséus, 1999; Marotta, Ferraguti, & Erséus, 2003; Marotta et al., 2008 %, spermatozoan ultrastructure ( Erséus & Ferraguti, 1995; Ferraguti & Erseus, 1999; Ferraguti, Erséus, Kaygorodova, & Martin, 1999; Ferraguti, Ruprecht, Erséus, & Giere, 1994; Jamieson, Erséus, & Ferraguti, 1987; Marotta et al., 2003, 2008%, morphogenesis ( Gustavsson, 2002, 2004; Gustavsson & Erséus, 1997, 1999a % and genetic evolution (e.g. Achurra, Elejalde, & Rodriguez, 2011; Envall et al., 2006; Erséus et al., 2017; Erséus, Envall, et al., 2010; Erséus & Källersjö, 2004; Erséus, Källersjö, Ekman, & Hovmöller, 2002; Erséus, Rota, et al., 2010; Liu et al., 2017; Marotta et al., 2008; Mejlon, De Wit, Matamoros, & Erséus, 2015; Rousset et al., 2008; Sjölin et al., 2005 %. The overall results support monophyly of Tubificinae , Limnodriloidinae , Naidinae and Pristininae , but not for Phallodrilinae and Rhyacodrilinae . And yet, the phylogenetic position of many genera, and subfamily Telmatodrilinae , and the naidid tree of life as a whole were not resolved in previous studies.

Naididae today embraces ca. 1,000 species, but only a few of them have been available for DNA studies to date. In this study, we have 15 representatives, representing six of the eight subfamilies in current use by the first author. Other authors (Schmelz & Timm 2008; Timm 2009%, accepting paraphyly in diagnostically/morphologically distinct taxa, advocate familial rank, that is, the names Tubificidae , ‘ Naididae (s. str. %’, Pristinidae and Opistocystidae , for parts together corresponding to ‘ Naididae (s. lat. %’. Nevertheless, there is consensus that this complex is monophyletic, with marine and freshwater species mixed in most major lineages. Our current analyses yielded a fully resolved Naididae , and despite the limited taxon sample, three main conclusions about the subfamilies can be drawn with stronger confidence than before:

1. Limnodriloidinae ( Limnodriloides % and Tubificinae ( Aulodrilus , Troglodrilus , Potamothrix % are both monophyletic and sister taxa (Erséus, Envall, et al., 2010; Liu et al., 2017; Marotta et al., 2008; Sjölin et al., 2005 %.

2. Phallodrilinae requires taxonomic revision. First, a core group ( Phallodrilinae sensu stricto % must contain the type genus ( Phallodrilus , not yet molecularly studied%. As inferred from an analysis of morphology ( Erséus, 1992 %, Phallodrilinae is herein represented by Albanidrilus and Olavius . The placement of the marine genus Bathydrilus (an alleged phallodriline; Erséus, 1979 % in or near this group is neither corroborated here nor in other studies ( Envall et al., 2006; Erséus, Envall, et al., 2010; Martin et al., 2010; Mejlon et al., 2015; Rousset et al., 2008 %. Another marine taxon, Duridrilus Erséus, 1983 , also first classified in Phallodrilinae , may be closely related to Bathydrilus ( Erséus, 1983 %. Molecular studies underway support this (CE, unpublished data%. On the other hand, the notion that Phallodrilinae (i.e. sensu stricto % is sister to the large genus Heterodrilus (e.g. Envall et al., 2006; Erséus et al., 2002; Mejlon et al., 2015 % is again, and now maximally, supported. However, Heterodrilus (placed in Rhyacodrilinae due to its diffuse prostates and conspicuous coelomocytes; Erséus, 1981 % differs morphologically from Phallodrilus s.str. (with solid prostate gland and inconspicuous coelomocytes; Erséus, 1993 % in such a way that treating it as a subfamily of its own could be advantageous from a diagnostic point of view.

3. Rhyacodrilinae also requires revision. A distinct clade with four of our naidid species represents three subfamilies (Figure 1%: Pristininae ( Pristina %, Rhyacodrilinae sensu stricto ( Rhyacodrilus % and Naidinae ( Paranais , Chaetogaster %. This lineage, albeit with a different (and less strongly supported% internal topology, was recovered before ( Envall et al., 2006; Erséus, Envall, et al., 2010; Rousset et al., 2008; Sjölin et al., 2005 %; in the Erséus, Envall et al. study also with a fourth subfamily, Opistocystinae recovered as sister to Pristininae .Therefore, in a recent multilocus analysis of Naidinae , species of Rhyacodrilinae s. str., Pristininae and Opistocystidae were used as outgroups ( Erséus et al., 2017 %. From these earlier studies, we know that other ‘rhyacodriline’ taxa ( Monopylephorus , Epirodrilus and some Ainudrilus % are also part of this lineage, but there is still no strong evidence that they all form a monophyletic group (i.e. a hypothetical Rhyacodrilinae s.str. % with Rhyacodrilus . To make matters worse, two other ‘rhyacodriline’ taxa, Heronidrilus (marine% and Bothrioneurum (freshwater% are rooted in a more basal part of the naidid tree in the present study as well as in all other studies mentioned here. Here, they appear as sisters and together they are the sister to the rest of Naididae (Figure 1%, but in some previous studies, they occur in other positions. Undersampling may be one reason for these incongruencies. To summarize, Rhyacodrilinae ‘ sensu lato’ must be heavily pruned and divided into smaller subfamily-level taxa following extended molecular sampling.