O anus, where art thou? An investigation of ctenostome bryozoans

Abstract Ctenostome bryozoans are a small group of approximately 350 currently described species that remain inadequately investigated anatomically. Recently, the importance of soft body morphology of zooids including the digestive tract has become more evident for addressing various biological aspects such as systematic, functional, or phylogenetic analyses. Particularly, the position of the anus shows considerable variation in ctenostomes and in its extreme form can either be at the lophophoral base or at the vestibular wall. However, it has never been analysed in a broader systematic, phylogenetic, or functional context. Hence, the purpose of this study is to assess the distribution of anus position among ctenostomes, analyse whether zooidal or colonial morphology affects anus position, and draw first conclusions on its functional effects. The survey shows that a vestibular anus is ubiquitously present in alcyonidioideans and several, probably closely related, walkerioideans. In other groups such as boring forms, it appears more patchily distributed, or in some currently unassignable genera, such as Monobryozoon, supports a closer relationship to alcyonidioideans. Other zooidal or colonial characters such as tentacle number or zooidal density in the colony do not show a distinct correlation to the position of the anus. It appears that the shift of the anus into a vestibular area occurred once or twice among ctenostomes; the reasons and functional effects remain unknown. Future important aspects of defecation research in bryozoans are discussed.

introversion of the tentacle sheath, which is a thin body wall connecting the lophophoral base with the cystid wall ( Figure 1). The tentacle sheath thus wraps around the lophophore when zooids are retracted. In addition, the vestibular wall, which connects the tentacle sheath to the remaining body wall, can also be highly introvertable (Schwaha, 2019a).
The digestive tract of bryozoans is U-shaped and divided into three distinct areas: foregut, midgut, and hindgut. The latter terminates via the anus in the tentacle sheath (Silén, 1944, Schwaha et al., 2020, see also Figure 1).
Two large clades of bryozoans can be distinguished-Phylactolaemata and Myolaemata: the latter comprising the Stenolaemata and Gymnolaemata (Schwaha et al., 2020).
Phylactolaemates are a small group of freshwater bryozoans.
Stenolaemates are an evolutionarily old taxon with only the Cyclostomata being present in recent times and gymnolaemates are the largest with over 5.000 described species (Taylor & Waeschenbach, 2015). This clade can be divided into the paraphyletic ctenostomes and the monophyletic Cheilostomata, which are calcified and the largest taxon of bryozoans (Taylor & Waeschenbach, 2015;Todd, 2000).
Ctenostome bryozoans show a high diversity of colonial forms that range from tightly encrusting, large erect, to boring, and include monomorphic to polymorphic taxa (Schwaha, 2019b). Recently, in an investigation on the polychaete-tube inhabiting ctenostome Hypophorella expansa, it became evident that the location of the anus on the tentacle sheath was highly unusual and almost at the vestibular wall, close to the cystid wall ( Figure 2a, Pröts et al., 2019), in contrast to other species that have their anus located at the lophophoral base

| MATERIALS AND METHODS
Information on the zooidal position of the anus was taken from numerous ctenostome samples collected or received within the past decade and processed for histological analyses. In addition, data, especially from little-known taxa, were taken from the literature (mainly drawings, illustrations).
For the creation of the comparative table, various zooidal characters were taken into account: anus position, tentacle number, peristome size, and zooidal density/colony form. Tentacle number is categorised as low with 8-10 tentacles, medium 10-20 tentacles, and high with more than 20 tentacles. Peristome size is more difficult to categorise as many zooids in several species essentially consist of a peristome (see Schwaha, 2019b). However, these were ranked from low when there is no or just slight peristomial elevation on the frontal zooidal side, to medium when this is approximately a third of estimated zooidal length, or high when exceeding that. Zooidal density F I G U R E 1 Schematic drawing of the lophophoral (a) and vestibular anus (b) in protruded and contracted condition. The digestive tract is generalised and does not reflect true conditions of the proportions of most bryozoans (mostly due to the lack of comparative data on gut anatomy). Abbreviations: a, anus; cae, caecum; cw, cystid wall; l, lophophore; lb, lophophoral base; ts, tentacle sheath; vw, vestibular wall F I G U R E 2 The position of the anus: (a) vestibular anus as exemplified by H. expansa (modified after Pröts, Wanninger, and Schwaha (2019)). (b) Lophophoral anus in A. uraniae (3D reconstruction, modified from Schwaha, Edgcomb, Bernhard, and Todaro (2019)). Abbreviations: a, anus; ca, cardia; cae, caecum; fg, foregut; int, intestine; l, lophophore; lb, lophophoral base; pm, parietal muscles; rm, retractor muscles; v, vestibulum was assigned as dense for regularly arranged colonies such those in colonies of Alcyonidium or Flustrellidra, to not dense when zooids are more spaced-the latter can be subject to change when the substrate becomes less available and growth can secondarily become very dense. A third category is dense tufts or rows, especially in polymorphic colonies that have stolons from which autozooids branch off.

| RESULTS
In general, two polarised, extreme positions of the anus can be distinguished among ctenostomes: the first is located close to the lophophoral base, closer to the ciliary feeding currents, the second is closer to the vestibular wall and thus further away from the lophophore (Figures 1 and 3). Accordingly, these are termed "lophophoral" and "vestibular" anus, respectively.  (Table 1). Especially, the presence of a distinct vestibular anus in alcyonidioidean and walkerioidean ctenostomes shows opposite zooidal and colonial features: dense versus rather non-dense growth, high versus low tentacle numbers, often very high to low peristomial size.

| Location of the anus in bryozoans
This study shows that there is distinct variation in the position of the anus among ctenostome bryozoans. In comparison, the anus in nonctenostome bryozoans seems much more limited and restricted, showing little to no variation. As potential outgroups of gymnolaemaetes, phylactolaemate, and cyclostome bryozoans show that the anus is always lophophoral (e.g., Boardman, 1998;Mukai et al., 1997;Nielsen & Pedersen, 1979;Ryland, 1970), which indicates that this is the plesiomorphic, original condition. Hence, the vestibular anus among some of the ctenostome taxa is a derived condition, which potentially evolved several times. Little information is available for cheilostome bryozoans, and most descriptions/illustrations generally indicate the anus being located mid-way between lophophoral base and vestibular wall (see, e.g., Calvet, 1900, Harmer, 1902, Marcus, 1937, Lutaud, 1977. However, it generally seems to be associated more with the lophophoral base in protruded zooids (McKinney, 1997). Cases with more a distally located anus have also been reported, however, among cheilostomes (see Lutaud, 1983;Nitsche, 1871).

| A vestibular anus and the fixed anal position of phylactolaemates and cyclostomes
An important restriction in the position of the anus is constructional constraints in the organization of zooids in phylactolaemates and cyclostomes. In phylactolaemates, the position of the anus is fixed and has little possibility to be displaced. This is also connected to the fact that the epistomial coelom originates between the narrow space between the gut shanks and proceeds distally into the epistome above the mouth opening (Gruhl et al. 2009;Schwaha et al., 2019;Schwaha & Wood, 2011). Widening that space would probably affect the functionality and movement of the epistome as it would prevent fluid being easily channelled into the epistome.
Cyclostomes, in general, rarely protrude their lophophore much beyond the orifice or aperture, which has been considered a certain disadvantage concerning feeding competition when compared to the dominant cheilostomes (McKinney, 1988;McKinney & Boardman, 1985). This restricted range of protrusion is related to the morphological design of cyclostomes. They (and probably all stenolaemates) evolved a unique protrusion mechanism by detaching their peritoneal lining from the remaining body wall to form the socalled membranous sac (Borg, 1926;Ernst, 2019;Schwaha et al., 2020). Proximally, the membranous sac is connected to the F I G U R E 3 Schematic drawing of lophophore density and effects of a lophophoral and vestibular anus indicated by faeces in brown for the former and turquoise for the latter. (a) Dense zooidal arrangement. Defecation in the lophophoral anus will mostly likely interfere with feeding currents of neighbouring zooids, whereas the vestibular anus is less likely to do so. (b) "Colonial distancing" with zooids more widely spaced and less inter-zooidal interactions SCHWAHA cystid wall where the retractor muscles attach to the skeleton, whereas distally so-called attachment organs and ligaments are frequently found in the apertural area (Boardman, 1998;Ernst, 2019).
These restrict movement of the polypide in respect to the cystid.
Gymnolaemates, on the other hand, have the possibility to shift their anus in their more flexible and protrusible poylpipes (McKinney 1988;Winston, 1978). However, the mechanism causing the shift in the anus of ctenostomes remains unclear as the current study indicates that there do not seem to be any distinct zooidal or colonial traits correlated with the position of the anus.
The general tendency to a higher polypide protrusion capability is linked to a more efficient feeding mechanisms and flexibility, allowing coordinated colonial integration (e.g., Shunatova & Ostrovsky, 2001, 2002Winston, 2019).

| What we can learn from ctenostome anuses?
Most ctenostomes are only studied as preserved and generally always retracted forms. As previously stated, the introvertable area of  Remane (1938) and Gray (1971) Incertae sedis Panolicellidae Panolicella nutans la Medium High Not dense Jebram (1985) Abbreviations: la, lophophoral anus; ma, mid-positioned anus; va, vestibular anus. a Strictly considered, the entire area containing the polypide is the peristomes, but on the comparison of vestibular wall size it remains low (see Schwaha (2019b)).
individual zooids has a high range with respect to the vestibular wall (Schwaha, 2019a(Schwaha, , 2019b. The vestibular wall can be quite extensive in several species and might even exceed the length of the tentacle sheath if not the polypide itself. This is particularly evident in many alcyonidioidean species that always have a vestibular anus. Consequently, mere introversion of the tentacle sheath has little effect in protruding the tentacle crown. As a consequence, the vestibular wall requires extensive inversion (see also Schwaha, 2019aSchwaha, , 2019b, As already mentioned, the position of the anus does not seem to correlate with colony morphology or zooidal arrangement and hence does not indicate any functional advantages in the feeding process. Instead, the occurrence, in particular, of a vestibular anus appears in closely related taxa, that is, all alcyonidioideans, along with other softtissue characters, aids in characterizing this clade (see Schwaha, 2019aSchwaha, , 2019bSchwaha & Wanninger 2018). Likewise, the walkerioidean genera Triticella, Farrella, and Aeverrillia are often stalked, stolonate forms that also share a vestibular anus, whereas other walkerioideans, such as Mimosella and Walkeria, have a lophophoral anus.
Possibly striking as a clear aid in further addressing its phylogenetic position is the vestibular anus found in Monobryozoon (Table 1, Remane, 1936Remane, , 1938, which supports a closer relationship to the Alcyonidioidea. Likewise, the vestibular anus is another confirmation that Sundanella also belongs to this taxon and is not associated with victorellid ctenostomes (see also Braem, 1939;Schwaha, 2019b).
Other ctenostome taxa show a variety or mosaic concerning their anal position. Boring bryozoans are distributed in four different families which according to their colony morphology and zooidal details probably evolved at least twice independently (Jebram, 1973(Jebram, , 1986Schwaha, 2019b). Their anal distribution currently represents quite a mosaic of lophophoral to vestibular anuses (Table 1).
The small taxon Hislopioidea with fewer than 10 species shows a mid-positioned anus, but similar to the vestibular anus, currently has little functional or evolutionary explanation. Particular lack of data still remains for the "Arachnidioidea," a heterogeneous clade, which almost completely lacks any detailed soft morphological studies (Schwaha, 2019b), although first studies will start to emerge in the near future (Table 1).

| Consequences of the position of the anus: Feeding and defecation
Keeping zooids in a colony in close proximity enhances the capacity of suspension feeding and, with respect to the high competition faced by other benthic suspension feeders, is a vital character for numerous bryozoans. Colonial density also has its drawbacks: feeding currents of neighbouring zooids interact and certain adaptations are necessary for creating exhalant currents for nutrient-depleted water (Shunatova & Ostrovsky, 2001, 2002, Winston, 2019. In a similar manner, defecation interacts between closely spaced zooids (McKinney, 1997, Figure 3A). Colonial distancing has advantages by minimising interactions of feeding or defecation events (Figure3).
However, as previously mentioned, distancing lowers feeding currents with isolated zooids (Winston, 1979) and seems to be a clear selective disadvantage.
Particularly among densely aggregated zooids, faecal pellet disposal is an important aspect of bryozoan coloniality (McKinney, 1997). This is particularly evident in most cheilostomes, where also most observational data are present (McKinney, 1997;Shunatova & Ostrovsky, 2001, 2002Winston, 2019). Numerous colonies create chimneys for excurrent, nutrient depleted water currents that also serve for faecal pellet disposal (see references above). This is present among phylactolaemates (Mukai 1999), recent cyclostomes and cheilostomes (Shunatova & Ostrovsky, 2002) and also is evident among fossil stenolaemates (often by the presence of so-called monticules; Ernst, 2019). Such chimneys usually cover areas devoid of autozooids. In cheilostomes lacking such specific areas, different strategies commonly apply for waste removal such as directional colonial movements effectuated by enlarging, often asymmetrically arranged, lophophores towards the colony margin, or concerted particle removal by "catch and play" behaviour. In the latter, undesirable or unpalatable particles are removed from individual zooids by ciliary reversal of the lophophore. Such particles are thus continuously transported from one zooid to its neighbour until to the colony margin (Shunatova & Ostrovsky, 2001;Winston, 2019).
Most of these are on calcified taxa, whereas the few ctenostome observations were conducted on vesicularioideans. Four different pathways for faecal pellet removal have been recognized among cheilostomes (and a few ctenostomes with a lophophoral anus; McKinney, 1997). Two of these (Pathways 1 and 2) involve faecal pellets entering the lophophore, whereas the other two (Pathways 3 and 4) redirect faecal pellets without entering the circular lophophore. In theory, the displacement of the vestibular anus from the lophophore prevents faecal pellets from entering it (see Figure 3). Also, it would appear improbable for the faecal pellet to be transported on the outer margin of the lophophore (as in Pathway 4, McKinney, 1997). Hence, other pathways probably occur among ctenostomes with vestibular anuses.
In contrast to colonial distancing, a vestibular anus would prevent faecal interference with feeding currents in tight zooidal arrangements ( Figure 3), but implies that faeces would aggregate on the colony surface.
However, faecal accumulations on the colony surface do not seem to be a regular condition among such colonies. Zooids of erect colonies or those that grow on the underside of substrates naturally do not face such a problem. Likewise, frequent water movements (e.g., in epiphytic colonies) also aid in pellet removal, and zooidal intra-colonial interaction also has cleaning purposes of the colony (Shunatova & Ostrovsky, 2001, 2002. In The studied ctenostomes such as Bowerbankia/Amathia(e.g., Winston 1977) that have a typical lophophoral anus were considered to follow defecation Pathway 1 as described for several cheilostomes (McKinney, 1997). Although direct evidence has not been reported for a vestibular anus in the alcyonidioidean Flustrellidra hispida (in contrast to all others of the clade), defecation of individuals occurs in 65% of all cases when polypides protrude or retract (Best & Thorpe, 1987), which indicates that polypide movements might be important if not necessary for such taxa.

| CONCLUSION
This study shows that there is a general variability in the location of the anus among ctenostomes and also underlines how little we still know about many basic features of bryozoans in general. Numerous issues remain open for future studies, including: morphology of the digestive tract and correlation with the position of the anus, general diversity of cheilostome guts, and study of live animals, especially ctenostomes with vestibular anuses. Along with other increasing data on soft tissue morphology (Schwaha, 2019a(Schwaha, , 2019bSchwaha et al., 2020), the position of the anus is an important character for phylogenetic inferences as it does not seem to correlate with zooidal or colonial characters. Molecular trees of ctenostomes just start to appear (e.g., Waeschenbach, Vieira, Reverter-Gil, Souto-Derungs, Nascimento, & Fehlauer-Ale, 2015) and once a new and more complete phylogenetic tree of ctenostome bryozoans is available, it should become clearer how often a vestibular anus has evolved.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.