Jump to ContentJump to Main Navigation
The Evolution of Organ Systems$

Andreas Schmidt-Rhaesa

Print publication date: 2007

Print ISBN-13: 9780198566687

Published to Oxford Scholarship Online: September 2007

DOI: 10.1093/acprof:oso/9780198566687.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 04 April 2020

Intestinal systems

Intestinal systems

Chapter:
(p.218) CHAPTER 12 Intestinal systems
Source:
The Evolution of Organ Systems
Author(s):

A. Schmidt-Rhaesa

Publisher:
Oxford University Press
DOI:10.1093/acprof:oso/9780198566687.003.0012

Abstract and Keywords

Animals obtain nutrients using a two-step procedure: food collection and uptake by endocytosis. In eumetazoans, a particular intestinal tract evolved, making it possible to digest food prior to the uptake of the smallest compartments. This chapter presents models for the evolution of this intestinal tract and its further evolution. There is a general trend for evolution from a saclike system to a tubelike one-way gut. In particular, the anterior, ectodermal part of the intestinal tract can be very diverse and contains a pharynx in various forms or it contains cuticular hard structures. Feeding is discussed in the context of larvae evolution and in adults.

Keywords:   endocytosis, sac-shaped intestine, one-way gut, gastrulation, pharynx, sucking pharynx, upstream collecting system, downstream collecting system

The intestinal system is that part of the body into which food is ingested and in which nutrients are absorbed. This is not an exclusive characterization, because nutrients can also be absorbed by other tissues, in particular by the epidermis. The intestinal system is, when it is present, a sack-or tube-like invagination from the body wall. In a sense its lumen can be regarded as a continuation of the external into the body, but as it usually can be sealed off by sphincters or other structures, such characterization is not entirely exact. Typically, the intestinal system originates during development in the gastrulation as the ‘archenteron’. The tissue forming the archenteron is named endoderm, but in adult animals, generally only the central part of the intestinal system is formed by endodermal tissue, whereas the terminal parts are of ectodermal origin. This is of particular importance, because it seems that, for example, cuticular structures can be formed only by tissue of ectodermal origin.

The food of metazoan animals can be very diverse, both in composition and in size. Three processes have to be distinguished: the capture of food, the breakdown into small particles, and the uptake of such particles into the body.

Endocytosis—the central process in intestinal systems

The uptake of the smallest particles into the body is the central process for the function of intestinal systems, and it is basically similar in all metazoan animals. It is performed by endocytosis, the enclosure of fluid or particles in small vesicles, which then detach from the cell membrane and migrate into the cell (see Griffiths 1996 for classification of different modes of endocytosis). When the vesicles fuse with other vesicles containing digestive enzymes (lysosomes), the content is further broken down to the molecular level.

In sponges, which do not have an intestinal system in the sense of a defined organ, endocytosis occurs in pinacocytes and choanocytes (Weissenfels 1976, Willenz & van de Vyver 1984, Langenbruch 1985, Imsiecke 1993). The content of the endocytotic vesicles is then passed to digestive lysosomes (Hahn-Keser & Stockem 1998) and may be transported through the body by archaeocytes (Leys & Reiswig 1998). In animals other than sponges, endocytosis occurs mainly in endodermal cells of the intestinal tract. Principally, other cells are also capable of endocytosis. For example, the uptake of nutrients is partly or completely taken over by epidermal cells in several parasitic animals living in the intestinal tract or in the body cavity of their host (e.g. acanthocephalans, cestodes, nematomorphs; see e.g. Jennings 1989 for the parasitic flatworm Acholades asteris).

Origin of the intestinal tract

From the three processes named in the introduction, sponges use only two: they capture food by creating a water current, which brings in food particles, and then absorb the smallest particles by endocytosis. Any utilization of larger particles requires a breakdown prior to endocytosis. Such breakdown can be performed mechanically or chemically, that is, by the secretion of digestive substances. While mechanical devices are present only in bilaterians, chemical breakdown was the first to evolve. As a secretion of digestive enzymes into the sea water would immediately (p.219) dilute it to insignificant concentrations, it appears clear that such a process only makes sense if secretion is into a defined compartment separated from the external environment. This can in principle be the body of the prey, but there is also a tendency to form such a compartment as an integral part of the body, as the intestinal tract.

It should be noted here, that among sponges there is a fascinating exception to what is written in the paragraph above: sponges from the genus Asbestopluma live in the nutrient-poor deep sea or in caves, and capture larger prey (e.g. small crustaceans) with the aid of hook-like sclerites. This prey is completely overgrown by cells from the vicinity, so that digestion takes place in a temporary internal compartment of the body (Vacelet & Boury-Esnault 1995, Vacelet 2006).

An intestinal tract can be be observed in cnidarians and ctenophores and therefore evolved in the ancestor of Eumetazoa. There are two hypotheses about its origin, both starting from different reconstructions of the eumetazoan ancestor. The most common mode of the ontogenetic development of the intestinal tract is that the hollow blastula invaginates, creating the archenteron and the blastopore. When the eumetazoan ancestor was a blastula-like organism (or when its life cycle included such a stage), the evolutionary origin of the intestinal tract could be similar (Fig. 12.1.). One problem of this hypothesis is that other processes such as ingression of cells, delaminations, or combinations can lead to the same result (the presence of an archenteron). As all such processes are realized in cnidarians (see Chapter 3and Fig. 3.8.), one can not exactly decide which of these is the ancestral mode.

The organization of Trichoplax adhaerens offers another hypothesis for the origin of the intestinal

                   Intestinal systems

Fig. 12.1. Two models for the evolution of the intestinal system, either by invagination (A) or by internalization of the ventral epithelium (B).

(p.220) tract. The ventral epithelium contains gland cells as well as endocytotically active cells and it can be observed that Trichoplax crawls over food particles with a subsequent formation of a ‘feeding pocket’, into which presumably products of the gland cells are secreted (Ruthmann et al. 1986, Grell & Ruthmann 1991). If this process is regarded as the creation of a compartment surrounded by an epithelium containing digestive gland cells, it is basically comparable to digestion in an intestine. If the ventral epithelium in Trichoplax can be homologized with the intestinal epithelium, then the crucial step in Eumetazoa would be to transfer this epithelium completely into the inside of the animal (Grell 1971a, b; Fig. 12.1.). This is probably supported by the expression of a gene, Trox-2, along the outer body margin (Schierwater 2005). Trox-2 resembles antennapedia-like genes which are expressed in the anterior body region. It can be imagined that an internalization of the ventral epithelium would automatically lead to a closure of the margin to form the mouth opening, which would restrict the expression of Trox-2 homologues to the anterior region. This hypothesis is based on the bilaterogastraea-hypothesis (Jägersten 1955), which requires a flat benthic stage in early metazoan evolution. What is problematic in this model is that a stage comparable to Trichoplax, that is, a flat arrangement of two layers, is never seen during development in other animals.

The evolution of an internalized intestinal tract offered the possibility to digest larger food particles and probably stimulated the evolution of a diversity of food capture mechanisms as well as targeted food requirements, which in turn stimulated changes in the sensory and locomotory apparatus.

From a sac-shaped intestine to a one-way gut

Within Bilateria, three principle forms of the intestinal system occur (neglecting its partial or complete reduction): a sac-shaped intestine with one opening to the exterior, a one-way gut with two openings (mouth and anus), and the presence of a digestive syncytium instead of an epithelially bordered intestine.

The last form, the lack of an epithelialized intestine but a digestion by a syncytial tissue, is present in acoels and was once regarded as an important transitory stage in metazoan evolution. Hadzi (1963), Hanson (1963), and Steinböck (1963), especially, advocated a view in which the metazoan ancestor originated from a ciliate-like organism by a multiplication of nuclei with subsequent cellularization. The syncytial nature of the central intestinal tract in acoels was regarded as proof for such a scenario. However, phylogenetic analyses have made this scenario unlikely and made it, in contrast, more likely that this special conformation of the intestinal system evolved within Acoela. The reasons are, on the one hand, that the sister group of Acoela, the Nemertodermatida, has an epithelialized intestine and therefore likely shows the ancestral condition. On the other hand, a basal acoel taxon, the genus Paratomella, shows a cellular organization of two layers around a central lumen. This has been interpreted as the following scenario: the ancestor of Acoelomorpha had a ‘usual’ epithelialized intestine, composed of digestive and gland cells. This condition was preserved in Nemertodermatida, whereas the ancestor of Acoela reduced the gland cells and separated the intestinal epithelial cells into two layers. While this condition was preserved in Paratomella, the proximal cells fused to form a syncytium in the ancestor of the remaining acoels (Euacoela) (Smith & Tyler 1985, Ehlers 1992a, Ax 1996; Fig. 12.2.).

The presence of a sac-shaped intestine in Platyhelminthes and Gnathostomulida has been taken in many cases as an indication of a basal position of these taxa among bilaterians. For example, Ax (1985) regards the Plathelmintho-morpha (Platyhelminthes + Gnathostomulida) as a sister taxon of all other Bilateria (called Eubilateria), with the one-way gut being a central autapomorphy of Eubilateria. However, many subsequent analyses, morphological as well as molecular, make it more likely that Platyhelminthes and Gnathostomulida are part of the Spiralia (see Chapter 2). This implies two (p.221)

                   Intestinal systems

Fig. 12.2. Evolution of the intestine in Acoelomorpha. Nemertodermatida have epithelial (dark) and glandular (light) cells, Paratomella species have a bilayered intestinal system without gland cells and species of Euacoela have a central digestive syncytium. Modified after Ax (1996).

possible scenaria of intestine evolution. Either the sac-shaped intestine was conserved in Platyhelminthes and Gnathostomulida, then a one-way gut evolved four times independantly (in Deuterostomia, Nemathelminthes/Ecdysozoa, Syndermata + Limnognathia and in Euspiralia), or the one-way gut evolved in the bilaterian ancestor and a sac-shaped intestine evolved secondarily in Platyhelminthes and Gnathostomulida (Fig. 12.3.). Both scenaria appear to me to be difficult to weigh up against each other and a simple counting of evolutionary steps might be misleading. A one-way gut has advantages over a sac-shaped intestine and can develop in different ways (see below), therefore it might arise several times in parallel. On the other hand, the development of almost every organism goes through a gastrula, which is a stage with a sac-shaped intestine. If the development of the intestinal system arrests in this stage, adults may secondarily receive a sac-shaped intestine.

It should be mentioned here that there are few structures that have been indicated as something like an anus in flatworms and gnathostomulids.

                   Intestinal systems

Fig. 12.3. Alternative scenarios for the evolution of the one-way gut, see text for explanation.

In two species of Haplognathia (Gnathostomulida) Knauss (1979) found a special region, in which epidermal and intestinal cells are interdigitating and are not separated by ECM. This structure might function as a temporary anus, but any further observation substantiating this interpretation is lacking. Some flatworms possess a connection between intestine and the bursa, which is a part of the female reproductive system. This bursa has been interpreted as a derivative of an ancestral anus (Steinböck 1924, Remane 1951). The main argument against this hypothesis is that the bursa always develops from the anlage of the reproductive system and gains contact to the intestine later in development, therefore there is no indication that the bursa originally did belong to the intestinal system (see Reisinger 1961). Additionally, the intestinal system in Ctenophora opens with two pores close to the apical pole. Whether these can be regarded as something like an anus is, however, questionable.

During development, a one-way gut can develop in three different ways. In addition to the blastopore, a second opening of the archenteron can break through and develop either as the anus (‘protostomy’) or as the mouth (‘deuterostomy’), (p.222) making the blastopore an anus or mouth, respectively. The blastopore can also elongate, its margins can approach and finally fuse in the centre, leaving an opening at either end (Fig. 12.4.). This process is called ‘amphystomy’ and occurs in representatives of several taxa (Nematoda, Onychophora, Annelida). Because of this scattered distribution and because amphystomy elegantly fits into some problems such as dorsoventral inversion (Arendt & Nübler-Jung 1997) or nervous system evolution (Nielsen & Nørrevang 1985), it is regarded as ancestral for protostomes (Nielsen 2001) or bilaterians (Arendt & Nübler-Jung 1997). I regard it as difficult to reconstruct which of the three modes of archenteron development can be regarded as ancestral. Even when amphystomy appears to be ‘logically’ ancient, parsimony would favour its convergent evolution. Reconstruction of the ancestral mode also depends on the questionable phylogenetic position of the tentaculate taxa which show protostomy.

The one-way gut has some advantages over the sac-shaped intestine. While in a sac-shaped intestine, the passage of food can be guided only in a limited way and food likely passes the same regions of the intestine more than once, the one-way gut allows a directed passage of food, where each ‘station’ is passed only once. This allows a regionalization with a much more specialized digestion than is possible in a simple sac. Consequently, in one-way guts we find a diversification of intestinal structures. It has to be added that sac-like intestines do not always have to be shaped like a ‘simple’ sac, but can be branched to differing degrees, this is realized in triclad, polyclad, and digenean Platyhelminthes (see Fig. 10.2.).

One-way guts are composed of three parts. While the central part is of endodermal origin, ectoderm invaginates at the anterior and posterior ends to a varying degree to form the anterior and posterior regions of the intestinal system. These parts differ considerably in structure. The anterior ectodermal part is often specialized for food uptake. Here we often find a strongly developed musculature, which is named the ‘pharynx’. This can be understood as a further

                   Intestinal systems

Fig. 12.4. Three possible fates for blastopore development: protostomy, amphystomy, and deuterostomy.

(p.223) development of the subepidermal musculature underlying other parts of the ectoderm. Ectodermal cells are also capable of the secretion of extracellular material, and consequently hard structures can be formed in the anterior ectodermal part of the intestine. The central, endodermal part is specialized for digestion and nutrient uptake.

Many specializations of the intestinal system are unique features of certain taxa and I will only concentrate here on the discussion of three pharyngeal structures with potential for phylogenetic analyses: the structure of the pharynx, pharynges hard structures, and pharyngeal gill slits. A brief overview on the general structure of the intestine is given in Table 12.1.

Table 12.1 General structure of the intestinal system. The source of information is, if not otherwise indicated, taken from general textbooks (Westheide & Rieger 2004, 2007, Ruppert et al. 2004) and from the chapters in the series ‘Microscopic Anatomy of Invertebrates’. The terms ‘pharynx’ and ‘esophagus’ always indicate ectodermal structures, ‘stomach’ and ‘intestine’ are entodermal and ‘hindgut’ is ectodermal. When mouth or anus are located subterminal, the terms ‘ventroterminal’ or ‘dorsoterminal’ are used.

Structure of the intestinal system

Cnidaria

Sac-shaped intestine with 1 opening. In polyps opening leads into a central gastrovascular cavity which may be divided by 4 (Scyphozoa) 6, 8 or a multiple of these (Anthozoa) septa into gastral pockets. Anthozoan polyps with ectodermal mouth tube (‘pharynx’), including two ciliated grooves (siphonoglyphs). Medusae with stalked mouth opening (whole structure named manubrium) leading into a central ‘stomach’ and by radial canals to the periphery and into the tentacles. Radial canals connected by ring canal.

Ctenophora

Terminal mouth opening – ectodermal mouth tube (‘pharynx’), which mechanically and chemically macerates food – stomach, from which 2 pharyngeal, 2 tentacular, and 2 transverse canals originate, the latter ones leading to 8 meridional canals running under the comb rows. Aboral canal leads from stomach to 4 short canals around the statocyst, 2 of which open by anal pores. The majority of food remnants are expelled through the mouth opening, only very little through anal pores.

Xenoturbella

Ventral mouth opening with short pharynx (musculature similar to subepidermal musculature, Raikova et al. 2000a), sac-shaped intestine without further specializations.

Acoelomorpha

Ventral mouth opening, simple pharynx is present in several acoels, its absence in some acoels is probably derived (Tyler 2001, Todt & Tyler 2006). Epithelial intestine in Nemertodermatida (lumen often occluded), bilayered intestine in Paratomella, central digestive syncytium and peripheral cells in Euacoela (Ax 1996).

Gastrotricha

Terminal mouth opening – pharynx with triradiate lumen (Y in Chaetonotida, inverted Y in Macrodasyida), composed of myoepithelial cells, pharyngeal pores in Macrodasyida – straight intestine – ventroterminal anus.

Nematoda

Terminal mouth opening – buccal tube, often with teeth or stylets – long pharynx with triradiate lumen (Y-shaped), composed of myoepithelial cells – straight intestine – ventroterminal anus (cloaca in males).

Nematomorpha

Ventroterminal mouth opening – esophagus – straight intestine, blindly ending in Nectonema – ventral (males) or terminal (females) cloacal opening in Gordiida.

Priapulida

Terminal mouth opening – pharynx (epithelium + muscular layer), with cuticular pharyngeal teeth, muscular polythyridium in Tubiluchus (Rothe et al. 2006) – straight intestine – terminal anus.

Kinorhyncha

Terminal mouth opening – pharynx (epithelium and muscular layer), lumen round (Cyclorhagida) or triradiate (inverted Y, Homalorhagida) – straight intestine – ventroterminal anus.

Loricifera

Terminal mouth opening – pharynx with triradiate lumen (Y), composed of myoepithelial cells – straight intestine – terminal anus.

Platyhelminthes

Mouth opening either anterior or ventral in midbody – muscular pharynx, which is eversible in some taxa (e.g. Tricladida). Shape of pharynx is phylogenetically informative (see Ehlers 1985a) – intestine sac-shaped, but often divided into few to numerous branches. Intestine completely reduced in e.g. cestodes.

Gnathostomulida

Ventroterminal mouth opening – muscular pharynx including specialized musculature of cuticular jaw apparatus (Sterrer 1969, Müller & Sterrer 2004) – straight intestine, anus lacking or probably temporal (Knauss 1979).

Limnognathia

Ventroterminal mouth opening – muscular pharynx including specialized musculature of cuticular jaw apparatus – maerski straight intestine, tapering towards end, probably forming a temporal anal pore (Kristensen & Funch 2000).

Eurotifera

Mouth opening in center of wheel organ – muscular pharynx including specialized musculature of cuticular jaw apparatus – short esophagus – stomach – intestine – dorsoterminal cloacal opening.

Seisonida

Ventroterminal mouth opening – muscular pharynx including specialized musculature of cuticular jaw apparatus – long esophagus – stomach – short intestine only in Seison nebaliae (Ricci et al. 1993, Ahlrichs 1995). Protonephridial system and female reproductive system join the intestine in S. nebaliae. Cloaca (S. nebaliae) or anus (S. annulatus) dorsal.

Acanthocephala

Intestinal system completely absent.

Nemertini

Ventroterminal (Anopla) or almost terminal (Enopla) mouth opening in anterior end, joined by proboscis in Enopla – esophagus – stomach – intestine with serially arranged diverticula in most species – terminal anus.

Kamptozoa

U-shaped intestinal tract, mouth and anus inside tentacle crown. Mouth – esophagus – stomach – intestine – hindgut – anus.

Mollusca

Ventroterminal mouth opening – buccal cavity with radula (reduced in bivalves) – esophagus with salivary glands forming a mucous food string – stomach, in Eumollusca with a pair of digestive caeca (digestive glands), in Conchifera including a crystalline style (composed of enzymes), gastric shield, ciliated grooves, and other structures – intestine – hindgut – dorsoterminal anus. Digestion in stomach and digestive caeca, formation of fecal pellets in intestine.

Sipunculida

Terminal mouth opening – esophagus – long and coiled intestine – hindgut – middorsal anus.

Echiurida

Ventroterminal mouth opening at base of proboscis – esophagus (sometimes further divided into pharynx, esophagus, crop and gizzard based on diameter and surface structure) – long and coiled intestine – hindgut with pair of anal sacs – terminal anus. Bypass by a tube, the siphon, running parallel to the intestine.

Annelida

Variable structures due to different food sources. Ventroterminal mouth opening – pharynx – esophagus – midgut, often separated into stomach and ‘intestine proper’ – hindgut – anus in pygidium. Diversity of pharyngeal structures (ciliated folds, muscular parynges, eversible probosces, jaws in polychaetes, sucking pharynx with radial musculature and triradiate lumen in leeches) (Tzetlin & Purschke 2005). Plesiomorphic condition are probably ciliated folds (Purschke & Tzetlin 1996).

Tardigrada

Terminal mouth opening – buccal tube with cuticular stylets, muscular pharynx (myoepithelial cells, triradiate, Y-shaped lumen) – short esophagus – intestine – hindgut receives malpighian tubules and, in Eutardigrada, the female gonoduct – cloacal opening (female eutardigrades) or anus (other tardigrades) ventroterminal.

Onychophora

Ventroterminal mouth – oral cavity with cuticular ‘mandibles’ – pharynx (at least in juveniles with triradiate lumen, Schmidt-Rhaesa et al. 1998) – esophagus – straight intestine – hindgut – ventroterminal anus.

Euarthropoda

Ventroterminal mouth opening – buccal cavity – pharynx – esophagus, often with specialized regions as crop (Insecta), proventriculus (Insecta, Xiphosura, Malacostraca) or pumping pharynx (Arachnida) – midgut with one or more pairs of caeca – hindgut – anus.

Chaetognatha

Ventral mouth – pharynx – straight intestine – hindgut – ventral anus.

Phoronida

Intestinal tract U-shaped. Mouth within tentacle lophophor – short esophagus – midgut divided into prestomach, stomach and intestine – hindgut – anus located close to, but outside the lophophor.

Brachiopoda

Mouth at base of lophophor – muscular pharynx – short esophagus – stomach with branching diverticula (caeca, this is the place for digestion) – intestine – anus in inarticulate brachiopods, intestine blindly ending in articulates.

Bryozoa

Intestinal tract U-shaped. Mouth within lophophor – sucking pharynx with triradiate lumen – esophagus – tripartite stomach (cardia, caecum, pylorus, cardia can be a strong crushing gizzard) – intestine – hindgut – anus outside of lophophor.

Echinodermata

Diverse feeding structures. Crinoida: mouth central on oral disc – esophagus – intestine, sometimes with diverticula – hindgut – anus excentrically on oral disc. Asterioda: mouth – stomach (eversible cardia, pylorus with diverticula into arms) – aboral anus. Ophiuroida: mouth with 5 teeth – esophagus – blindly ending stomach. Echinoida: either with complex jaw apparatus (Aristotle's lantern) and aboral anus (regular echinoids), with jaw apparatus and anus in 90° angle to mouth (irregular echinoids except Spatangoida) or without jaws and anus in 90° angle (Spatangoida). Intestinal tract coiled, divided into pharynx – esophagus – stomach – intestine – hindgut. Often with bypass, the siphon, branching from esophagus and rejoining at border of stomach and intestine. Holothuroida: terminal mouth – muscular pharynx – short esophagus – stomach only in some species – coiled intestine – terminal cloaca.

Enteropneusta

Ventroterminal mouth opening – buccal cavity with anterior extension, the stomochord – pharynx with branchial pores, dorsal epibranchial ridge and ventral food channel – esophagus (in some species with esophageal pores) – straight intestine with diverticula (‘hepatic caeca’) – terminal anus. Food collection in pharynx, packing with mucous in esophagus, digestion in intestine.

Pterobranchia

Intestinal tract U-shaped. Mouth opening below oral shield – pharynx with anteriorly directed stomochord, pharynx in Cephalodiscida with 1 pair of pores – stomach, probably the main place for digestion – intestine – short hindgut – anus at base of tentacles (opposite side compared to mouth).

Tunicata

Intestinal tract U-shaped. Mouth apical – pharynx with gill slits, ventral endostyle and dorsal food groove – esphagus – stomach – intestine – hindgut – anus opening into peribranchial cavity.

Acrania

Ventroterminal mouth – buccal cavity – pharynx with gill slits, ventral endostyle and dorsal epipharyngeal groove – stomach with diverticle, the hepatic caecum – iliocolon – intestine – ventroterminal anus. Digestion primarily in stomach and iliocolon, absorption in the hepatic caecum.

Craniota

Ventroterminal mouth – buccal cavity, in Gnathostomata with teeth-bearing jaws – pharynx with gill slits used for nutrition only in larval Petromyzontida (ammocoetes-larva), respiratory gill slits in fish-like craniotes and larval amphibians – esophagus – stomach (mechanical, chemical and in some cases symbiotic digestion) – intestine (enzymatic digestion, absorption), divided in small intestine and colon – hindgut – anus (in some cases cloaca).

(p.224)

(p.225) The evolution of pharynges

Musculature can be present everywhere along the intestinal tract, but the anterior ectodermal part appears to be especially well suited to form dominant muscular structures. The reason may be that this region is more or less an invaginated piece of body wall and as such already equipped with complex subepithelial musculature. There-fore, the presence of a muscular anterior part appears to be a common phenomenon which can already be found in cnidarians, more precisely in anthozoan polyps. ‘Simple’ pharynges, composed of musculature hardly more dominant than the adjacent body wall musculature, are present in Xenoturbella and Acoela (Doe 1981, Raikova et al. 2000a, Hooge 2001, Todt & Tyler 2006) and this can be interpreted in accordance with a potential basal position of these two taxa within Bilateria (see Chapter 2). Among other bilaterians, two trends in pharynx evolution are recognizable. The pharynx can become a large muscular bulbus, which is eversible in some cases. This trend is followed in platyhelminths and annelids, and in some species the everted pharynx can be of considerable length (Fig. 12.5.). Basal flatworms, such as species belonging to the taxa Catenulida and Macrostomida, have a

                   Intestinal systems

Fig. 12.5. Eversible pharynges in A. Goniadides falcigera (Polychaeta) and B. Dugesia polychroa feeding on a tubificid annelid. The everted pharynx of Goniadides measures about 700 μm. A. after micrograph in Tzetlin & Purschke (2005), B. after Odening in Gruner (1984).

‘simple’ pharynx (Doe 1981), showing that more complicated pharynges evolved within Platyhelminthes. Also within polychaetes, it can be assumed that dominant pharynges evolved within the taxon and that weakly muscular pharynges are ancestral (Purschke & Tzetlin (p.226) 1996, Tzetlin & Purschke 2005). A similar kind of pharynx appears to be ancestral in sipunculids, which supports a closer relationship of sipunculids and annelids (Tzetlin & Purschke 2006). In Platyhelminthes, different types of pharynges can be recognized and used as phylogenetic markers (Ehlers 1985a, Ax 1996).

In many cases, the muscular wall of the pharynx creates a sucking force, which is essential

                   Intestinal systems

Fig. 12.6. Cross section through a sucking pharynx showing the triradiate lumen and radially arranged musculature in the macrodasyid gastrotrich Dactylopodola baltica. Photo by Birgen H. Rothe & A. Schmidt-Rhaesa.

for food uptake (especially liquid or small organisms) in a number of taxa. Although sucking can be performed with different pharyngeal constructions, the ‘smartest’ solution for a sucking pharynx is a radial orientation of muscle fibres in combination with a triradiate lumen (Fig. 12.6.). When muscles run from the border of the lumen (e.g. in myoepithelial cells from the cuticle bordering the lumen) to the pharyngeal periphery, a comparatively short contraction will create a triangular or almost round lumen rapidly (Fig. 12.7.). The rapid creation of a considerable lumen causes underpressure, by which liquid, small particles, and organisms can be sucked in. A sucking pharynx with a triradiate lumen can be found in a number of taxa: Gastrotricha, Nematoda, Kinorhyncha, Loricifera, in some polychaetes (Microphthalmus), leeches, tardigrades, juvenile onychophorans, some euarthropods (Pycnogonida, Amblypygi, Acari, Mystacocarida), and bryozoans (see Table 12.2.). Differences are found in the orientation of the lumen (Y or inverted-Y) and in the general composition of either myoepithelial cells or of a non-muscular epithelium and a separate muscular sheet (Table 12.2.). The distribution of triradiate sucking pharynges is (at least partially) so scattered that the functional constraints make it very likely that it evolved some times in parallel (this probably accounts for Microphthalmus, leeches, and bryozoans). The remaining taxa are,
                   Intestinal systems

Fig. 12.7. Function of triradiate sucking pharynges: the contraction of radial musculature quickly creates a large lumen.

(p.227)

Table 12.2 Occurrence of muscular sucking pharynges with a triradiate lumen.

Orientation of lumen

Myoepithelial (ME) versus epithelium + muscular layer

References

Gastrotricha Macrodasyida

Inverted Y

ME

Ruppert 1982, 1991b

Gastrotricha Chaetonotida

Y

ME

Ruppert 1982, 1991b

Nematoda

Y

ME

Wright 1991

Kinorhyncha Homalorhagida

Inverted Y

Epithelium + musculature

Kristensen & Higgins 1991

Loricifera

Y

ME

Kristensen 1991

Tardigrada

Y

ME

Dewel et al. 1993

Onychophora (juveniles)

Y

Epithelium + musculature

Schmidt-Rhaesa et al. 1998

Euarthropoda Pycnogonida

Y

Epithelium + musculature

Miyazaki 2002

Euarthropoda Acari (anactinotrichid mites)

Y

Epithelium + musculature

Alberti & Coons 1999, Coons & Alberti 1999

Euarthropoda Amblypygi

Y

?

Millot, J. 1968

Euarthropoda Derocheilocaris

Y

Epithelium + musculature1

Herrera-Alvarez et al. 1996

Polychaeta Microphthalmus

Inverted Y

?

Smith et al. 1986

Hirudinea Rhynchobdellida

Y

Epithelium + musculature (Sawyer 1986)

Sawyer 1986, Ax 1996

Hirudinea Arhynchobdellida

Inverted Y

Moser & Desser 1995, Ax 1996

Bryozoa

?

Epithelium + musculature

Bullivant & Bils 1968, Matricon 1973

(1) According to the investigation of Herrera–Alvarez et al. (1986), the triradiate lumen is surrounded only by circular musculature.

however, at least potentially, closely related (as Cycloneuralia, Nemathelminthes, or Ecdysozoa, see Chapter 2). Although the presence of triradiate sucking pharynges in further taxa and the functional constraints on its construction should raise some caution in using it as a phylogenetically informative character, I regard it as justified to assume the homology of sucking pharynges in Ecdysozoa, as long as such a relationship does not rest alone on the character ‘sucking pharynx’, but also on other characters or data. The differences may be easily explained. An evolution from a myoepithelium towards a separation of epithelium and musculature is a general trend that can also be found in the body wall (see Chapter 5) and in the coelom epithelium (see Chapter 8). A reversal in orientation of the lumen might not require a complicated scenaria, but rather a comparably simple change during pharynx development.

Pharyngeal hard structures

In contrast to the endodermal parts of the intestinal system, the ectodermal parts are capable of cuticle secretion. This can happen even when there is no body cuticle, as for example in gnathostomulids and eurotifers. Cuticular structures can form a variety of teeth, jaws, or stylets and can be important tools for prey capture, cell piercing, or algal scraping. Cuticular differentiations of the buccal cavity are found in nematodes and tardigrades, pharyngeal structures are present in priapulids, molluscs, polychaetes, and in Gnathifera. The pharyngeal teeth of macroscopic priapulids and the radula of molluscs are unique in structure and are likely to be autapomorphies of the respective taxa. In annelids, teeth or jaws occur within the taxa Phyllodocida, Eunicida (Fig. 12.8.), and in a few species of Ampharetidae (Tzetlin & Purschke 2005, Struck et al. 2006). The jaws, at least those in the eunicids Ophryotrocha labronica and Diopatra aciculata, are replaced during growth in a moult (Paxton 2004, 2005). Because the cuticular teeth and jaws in polychaetes are quite different in structure, it appears as if they all evolved within polychaetes.

In the taxon Gnathifera, however, the presence of jaws is the central character to unite gnathostomulids, Limnognathia, eurotifers (Fig. 12.9.), and seisonids (because acanthocephalans share other characters, they belong to Gnathifera (p.228)

                   Intestinal systems

Fig. 12.8. The polychaete Ophryotrocha sp. with jaws in the pharyngeal region of the intestinal system.

                   Intestinal systems

Fig. 12.9. Scanning electron micrograph of the cuticular jaw apparatus from the eurotifer Cephalodella hyalina. Photo by courtesy of Wilko Ahlrichs, Oldenburg.

despite the absence of jaws, see Chapter 2). Although the jaws are variable concerning their composition of hard elements (Sørensen 2000, 2002, 2003, Sørensen & Sterrer 2002), the
                   Intestinal systems

Fig. 12.10. Evolution of the branchial pharynx according to conflicting phylogenetic hypotheses, see text for details.

ultrastructure is similar in all cases (Ahlrichs 1995, Rieger & Tyler 1995, Herlyn & Ehlers 1997, Kristensen & Funch 2000).

Gill slits in deuterostomes

A pharynx with gill slits is found among deuterostome taxa, but this character has produced differing hypotheses on phylogeny. Among recent taxa, such a ‘branchial pharynx’ can be found in enteropneusts, in Cephalodiscida among pterobranchs, and in chordates (tunicates, acranians, and basal craniotes). On the one hand, Ax (2003) has hypothesized a sequential evolution of gill slits starting from one pair with the consequence that the taxa Pterobranchia and Hemichordata are paraphyletic (see Chapter 2; Fig. 12.10.). Regarding Pterobranchia and Hemichordata as monophyletic taxa, and (p.229) hemichordates related to echinoderms (e.g. Dohle 2004), requires that the absence of gill pores in recent echinoderms and in Rhabdopleura is a secondary reduction or that gill slits evolved convergently in enteropneusts, cephalodiscid pterobranchs, and chordates (Fig. 12.10.).

It is quite likely that gill slits are also present in fossil deuterostomes. This is most evident in the cornute Cothurnocystis (Jefferies 1968; Fig. 11.11.). Gill slits have also been hypothesized for other fossil deuterostomes (see Gee 1996 for a summary and, more recently, Shu et al. 2001, Lacalli 2002 for vetulocolids and Shu et al. 2002 for vetulocystids), but in neither of these cases is the evidence as convincing as in Cothurnocystis. Because the fossils in question have a stereom, which is a particular three-dimensional structure of the endoskeleton present among recent taxa only in echinoderms, there are two hypotheses. Concentrating on the stereom as the important character, the fossil organisms would be stem lineage echinoderms, which in consequence means that gill slits were present in their common ancestor and therefore also in the deuterostome ancestor (see, e.g., Philip 1979). Concentrating on the presence of gill slits, the fossils would be stem lineage chordates and the stereom would be a character of the deuterostome ancestor (the ‘calcichordate hypothesis’, see Jefferies 1968, 1975, 1981, Cripps 1991; Fig. 12.11.).

Feeding in larvae

Many marine species have microscopically small pelagic larvae that usually secure dispersal of the species. In freshwater, there are often other larval types and swimming larva are of course reduced in terrestrial animals. The mode of nutrition in marine larvae as well as the method by which food particles are gathered, are both important characters from which phylogenetic conclusions have been drawn. Not all larvae feed, they may receive nutrients from internal yolk and therefore survive a certain time period in the water without additional nutrition. This mode is called ‘lecithotrophy’ and is contrasted by ‘planctotrophy’, in which larvae feed on dissolved particles

                   Intestinal systems

Fig. 12.11. Different interpretations of the evolution of stereom and branchial pharynx due to different positions of fossil organisms such as Cothurnocystis. Hemichordates are not included into these trees.

or other planctonic organisms. Planctotroph larvae usually collect food with the aid of their cilia, although other mechanisms such as capturing food with a string of mucous can also occur (Fenchel & Ockelmann 2002). Some larvae are completely ciliated, while many larvae develop a heterogeneous ciliation, in which particular ciliary bands are either especially pronounced among a complete ciliation (e.g. by being longer) or are the only ciliated cells of the larva. Cilia are used for locomotion and nutrition. For a summary of larval features see Table 12.3.

What is a larva? A larva is considered a stage during development that differs from the adult morphology (Hickman 1999). It transforms to the adult by metamorphosis. Development including larvae is also termed ‘indirect development’, in contrast to ‘direct development’, where embryos and juveniles resemble the adults in most respects. Larvae show particular larval features, which are adapted to the special requirements of this particular developmental stage.

Larval patterns are central in several hypotheses on metazoan evolution and therefore it is important to know whether the possession of a larval stage is comparable among animals. For example, Nielsen and Nørrevang (1985) have (p.230) proposed a continuous evolution of larvae starting from a holopelagic blastea (in the metazoan ancestor) over a holopelagic gastraea with an archenteron (in the eumetazoan ancestor) to a holopelagic trochaea with one ciliary band (in the bilaterian ancestor, see also Nielsen 1985, 1995). In later publications, Nielsen (1998, 2001) assumes an early holopelagic evolution of metazoans, but is more cautious about a trochaea in the bilaterian ancestor. He assumes that both protostomes and deuterostomes have their own characteristic types of larvae. Such hypotheses imply that larvae (or at least one particular type of larva) evolved only once and became modified or lost during evolution. Is there evidence for this?

Most authors agree that two large groups of larvae can be distinguished, the ‘trochophora’ of Trochozoa (Mollusca, Kamptozoa, Annelida, Sipunculida, Echiurida) and the ‘dipleurula’ of echinoderms and enteropneusts (Figs. 12.12., 12.14.). As von Salvini-Plawen (1980b) and Rouse (1999a) explain, the name trochophore has often been used uncritically and has then been applied to a diversity of other taxa. A trochophore is regarded by von Salvini-Plawen (1980b) to have an apical plate, a pair of photoreceptors, protonephridia, an intestinal system with mouth and anus, and two ciliary bands called prototroch and metatroch. The prototroch is

                   Intestinal systems

Fig. 12.12. The two basic larval types among protostomes trochophore and dipleurula. The figured trochophore is from the polychaete Serpula vermicularis (after photo in Westheide and Rieger 2007), the remaining figures are drawn after Westheide & Rieger (2007) and Young (2002).

(p.231) the characteristic feature of a trochophore according to Rouse (1999a). The prototroch is characterized by a special cell lineage (Damen & Dictus 1994), starting in the third cleavage (Damen et al. 1996). Food particles are collected by the ciliary bands with a ‘downstream collecting system’. This means that a water current is created by the cilia that brings in and accumulates particles on the downstream side of the ciliary band, from where they are transported by shorter cilia to the mouth (Fig. 12.13.). The dipleurula is a hypothetical larvae, but all larval forms of echinoderms and the tornaria of enteropneusts can be derived from it (e.g. Dohle 2004). Even derived larval morphologies such as the doliolaria of crinoids and holothurians can be derived from the auricularia, which resembles the dipleurula (Semon 1888, Lacalli 1988, Nakano et al. 2003). The dipleurula has a circumoral ciliary band that functions in a different way compared to the trochophore. The cilia create a water current away from the mouth and retain particles on the upstream side of the cilia, therefore this is called an ‘upstream collecting system’ (Fig. 12.13.). The more problematic aspect now is whether and how marine ciliated
                   Intestinal systems

Fig. 12.13. Comparison of downstream and upstream collecting systems, figures after Nielsen (1987) and Lacalli (1996).

larvae from the remaining taxa relate to these two types of larvae.

Maslakova et al. (2004a) showed that in the completely ciliated larva of the palaeonemertean Carinoma tremaphoros several cells have a cell lineage corresponding to the prototroch cells of Trochozoa. This can be interpreted as an indication that larvae in Trochozoa and Nemertini are homologous, that these two taxa are closely related and that their common ancestor had a larva. Additionally, this observation supports the opinion that the ciliated early stages in Palaeonemertini and Enopla can be regarded as larvae (several authors call this a direct development), because the prototroch is a larval character. This evolutionary scenario (Fig. 12.14.) would imply that in the completely ciliated larvae of the ancestor of Nemertini + Trochozoa a certain subset of cells was created by a unique cell lineage. This stage was conserved in basal nemertean taxa, while another larva, the pilidium, evolved within Nemertini, in the taxon Heteronemertini. In the trochozoan ancestor, part of the cells lost their ciliation, but the particular subset of ciliated cells remained ciliated to form the prototroch.

The integration of further taxa of Protostomia makes the discussion of larval evolution more problematic. Most taxa are directly developing and have no larva (Gastrotricha, Nematoda, Kinorhyncha, Gnathostomulida, Eurotifera, Seisonida, Limnognathia) or have unique larvae not comparable to other forms (Nematomorpha, Loricifera, Priapulida). Only the interpretation of the presence of larvae in Platyhelminthes is problematic. While larval forms in parasitic flatworms are certainly derived, polyclads have ciliated larvae called Müller's or Götte's larva. All other flatworms are directly developing. Polyclads are not the basalmost taxon among Platyhelminthes (compare Ehlers 1985a, Littlewood & Olson 2001), making the possession of a larva in the platyhelminth ancestor not a parsimonious scenario. Additionally, the possession of larvae in polyclads appears to be explained by ecological constraints. While the microscopically small flatworms can leave the sediment to be dispersed with the water current (p.232)

                   Intestinal systems

Fig. 12.14. Occurrence of marine ciliated larvae on the phylogenetic tree. As assumed ancestral states, only the dipleurula, the trochophor and a kind of ‘pre-trochophora’ in the common ancestor of Trochozoa and Nemertini are convincing. Images of larvae after figures in Dilly (1973), Ruppert (1978), Westheide and Rieger (2007), Young (2002) and cover photo of Invertebrate Biology 121 (3).

(Armonies 1989, 1994), polyclads are too large to colonize new habitats in this way. Therefore, they had to develop, comparable to many other taxa with large species, planctonic larvae. Polyclad larvae are completely ciliated, with the cilia along the ‘arms’ being longer. Therefore, the cells forming these cilia were called prototrochal cells (Ruppert 1978a) or a ciliary band (Lacalli 1982), but it is not yet clear whether they originate in a cell lineage comparable to that of the prototroch in Nemertini and Trochozoa. A common character in polyclad larvae and (p.233) trochophores (as well as other larval types) is an apical organ, a tuft of longer cilia and sensory cells, which is directly above the larval brain in in polyclads (Lacalli 1983) and polychaetes (e.g. Lacalli 1981). Lacalli (1982) noted that the ‘ciliary band’ in Müller's larva is innervated by a peripheral nervous system originating from cells within the ciliary band and not from the central nervous system. This was also shown to be the case in a nemertean pilium larva (Lacalli & West 1985). It is hard to judge whether the similarites between polyclad larvae and nemertean and trochozoan larvae are homologies, but according to the position of polyclads within flatworms I would tend to regard them as convergences. Assuming that the platyhelminth ancestor did not have a larva, it becomes likely that ciliated marine larvae evolved within protostomes in the ancestor of Trochozoa + Nemertini and in polyclads, and that the protostome ancestor therefore had no larva. When, however, the polyclad larva and the trochophora are homologous, we must assume that a larva was present in the spiralian ancestor. As no (ciliated) larva is present in gastrotrichs, cycloneuralians, and arthropods, one protostome branch in this scenario has a larva while the other has not. Therefore, reconstruction of the bilaterian life cycle depends on the question of whether the dipleurula was present in the deuterostome ancestor, whether dipleurula, trochophora, and the larvae of the tentaculate taxa are homologous (see Fig. 12.14.).

The dipleurula is usually assumed to be the typical larva of Deuterostomia, but considering a possible close relationship between Echinodermata and Hemichordata (see Chapter 2), it might as well be an autapomorphy of the common ancestor of these two taxa, because chordates have no comparable larva (Fig. 12.14.). The larva of pterobranchs is an exception, it is a completely ciliated, lecithotrophic larva (Dilly 1973). The larvae of Bryozoa, Phoronida, and Brachiopoda cannot unambiguously be compared with either the trochophora or the dipleurula. Bryozoa have different larval forms; either as completely ciliated, lecithotrophic larvae in stenolaemates; or as planctotrophic cyphonautes

                   Intestinal systems

Fig. 12.15. Scanning electron micrograph of an actinotrocha, the larva of Phoronida.

or other forms in gymnolaemates (Woollacott 1999, Nielsen 2002c, Temkin & Zimmer 2002). The triangular cyphonautes has a ciliated lower ridge. Particles are captured by long laterofrontal cilia together with a varied water current created by lateral cilia and a flicking of the laterofrontal cilia towards the mouth (Strathmann 2006). Filtering by long and stiff cilia is also present in the phoronid actinotrocha (Riisgård 2002; Fig. 12.15.) and in planctotrophic brachiopod larvae (Glottidia pyramidata, Strathmann 2005). All three taxa have upstream collecting systems, which makes them comparable with echinoderms and hemichordates, but while in phoronids and brachiopods there are monociliary cells (as in echinoderms and hemichordates), cells are multiciliate in bryozoans (Nielsen 1987). Therefore, bryozoan larvae share characters of both protostomes and of deuterostomes. Larvae would support a relationship of at least phoronids and brachiopods to deuterostomes, and this would imply that a larva with an (p.234)

Table 12.3 Summary of features concerning ciliated larvae in marine taxa. LE = lecitotrophic, PL = planctotrophic, DCS = downstream collecting system, UCS = upstream collecting system.

Larval type

LE/PL

DCS/UCS

Ciliation

Reference

Porifera

‘Blastea-larva’

LE

Complete, monociliate

Nielsen 1987, Maldonado 2004

Cnidaria

Planula

PL and LE

Complete, monociliate, rarely multiciliate

Nielsen 1987, 2001

Platyhelminthes Polycladida

Müller's larva Götte's larva

Probably PL

?

Complete, bands of multiciliate cells along lobes

Lacalli 1982, Ballarin & Galleni 1987

Nemertini

Enopla and Palaeonemertini

Ciliated larva

LE

Complete, with ‘prae-prototroch’

Cantell 1989, Maslakova et al. 2004a

Heteronemertini

Pilidium

PL

?1

Ciliary band along lobes, partly double band, some cilia compound, multiciliate

Lacalli & West 1985, Nielsen 1987, Cantell 1989

Kamptozoa

Trochophora

PL

DCS

Proto- and metatroch with compound cilia and multiciliate cells

Nielsen 1971, 1990a, 2002c, see also Haszprunar et al. 1995

Mollusca

Solenogastres

Trochophora or larva derived

LE

2 Ciliary rings, fine structure unknown

McFadien-Carter 1979, Pearse 1979, Nielsen 1987, Haszprunar et al. 1995

Polyplacophora

from trochophora

LE

Prototroch with compound cilia, multiciliate cells

Gastropoda

LE and PL

DCS

Prototroch and metatroch with compound cilia, multiciliate cells

Bivalvia

LE in protobranchs, PL in others

DCS

Prototroch and metatroch with compound cilia, multiciliate cells

Scaphopoda

LE

?

Sipunculida

Trochophora in egg envelope, sometimes followed by pelagosphaera

Trochophora LE, pelagosphaera PL

Trochophora –, pelagosphaera does not use cilia for feeding

Trochophora ?, pelagosphaera with compound cilia

Nielsen 1987, 2001, Rice 1989, Jaeckle & Rice 2002

Echiurida

Trochophora, derived larva in Bonellia

PL, LE in Bonellia

DCS

Probably multiciliate (Nielsen 1987)

Davis 1989, Pilger 2002

Annelida

Trochophora

LE and PL

DCS

Multiciliate, rarely monociliate (Owenia)

Nielsen 1987, Heimler 1988

Bryozoa

Cyphonautes and other larval types

cyphonautes PL other larvae LE

UCS

Ciliary band with multiciliate cells (not compound)

Nielsen 1987, Nielsen 1990b, Woollacott 1999, Temkin & Zimmer 2002

Phoronida

Actinotrocha

PL

UCS

Monociliary cells (rarely biciliary), not compound

Strathmann 1973, Nielsen 1987, Emig 1990, Johnson & Zimmer 2002

Brachiopoda

Lingulacea, Discinacea

Tentaculate larva

PL

UCS

Monociliary cells, no compound cilia

Nielsen 1987, Chuang 1990, Pennington & Stricker 2002

Craniacea Articulata

‘Lobed’ larva

LE

Pterobranchia

‘Planula-like’

LE

UCS

Completely ciliated, no ciliary bands

Dilly 1973, Halanych 1993

Enteropneusta

Tornaria

PL

UCS

Monociliary cells, no compound cilia

Hadfield 1975, Nielsen 1987

Echinodermata

Doliolaria-type larvae

LE or PL

UCS

Monociliary cells, no compound cilia

Nielsen 1987

(1) Feeding has been assumed to be by DCS (Nielsen 1987), but only few cilia around the mouth contribute to feeding, the rest are involved in locomotion (Strathmann 1987, Haszprunar et al. 1995).

(p.235) upstream collecting system of monociliate cells was present in a common ancestor.

Concluding from this brief review of larvae, it is not certain that larvae in general can be compared with each other. Probably the single character that is present in almost all types of larvae is an apical concentration of longer, sensory cilia. This, however, may reflect functional needs for orientation rather than a homologous character. This is supported by a different use of transcription factors in the formation of apical organs in a sea urchin and a gastropod (Dunn et al. 2007). There is evidence that several larval features are variable and depend on external constraints. For example, within the sea star genera Asterina and Patiriella, there are species with larvae as well as directly developing species, and lecithotrophic as well as planctotropic larvae (Byrne & Cerra 1996, Hart et al. 2004, Byrne 2006). Byrne (2006) estimates that lecithotrophy evolved independantly at least six times within Asterinidae (the taxon to which Asterina and Patiriella belong). Ciliation patterns are important for feeding as well as locomotion, and the final form of a larva is often a compromise between these two functions (Emlet 1994, Strathmann & Grünbaum 2006). During evolution, the constraints can change, that is when changing from feeding to non-feeding. Pernet (2003) showed that in echinoderms such changes usually cause rapid morphological changes in echinoderms, while in sabellid polychaetes the ‘response’ is slower.

Taking the uncertainties about the homology of larvae and the dynamics of changes in structure into account, the decision about whether planctotrophy (e.g. Strathmann 1978, 1985, Havenhand 1995, Nielsen 1998) or lecithotrophy (e.g. Haszprunar et al. 1995) is the ancestral mode, is almost impossible to make. It is easier, however, when considering particular groups of taxa. Let us start with diploblastic animals. While sponges have lecithotrophic larvae (Nielsen 1987), both lecithotrophic and planctotrophic larvae exist in cnidarians. There are comparatively few observations of feeding, but it appears that the planula larvae of anthozoans are planctotrophic and feed with the aid of mucous and cilia (Siebert 1974, Tranter et al. 1982, summary in Strathmann 1987). Because anthozoans are the basal cnidarian taxon, it appears probable that within cnidarians, lecithotrophy was derived from planctotrophy, whereas in metazoan evolution planctotrophy might be derived from lecithotrophy (as in sponge larvae).

In protostomes (at least in Trochozoa + Nemertini) and in Echinodermata + Hemichor-data, larval feeding is performed by complex ciliary bands and particle capture mechanisms. This makes it likely that planctotrophy is ancient in these taxa. Indeed, there are no convincing examples, where planctotrophy is derived from lecithotrophy, but there are many examples of the reverse case (see Nielsen 1998). Downstream and upstream collecting systems appear to be restricted to particular taxa (Trochozoa and Deuterostomia, see, e.g., Nielsen 1987, 1994, 2001), but the potential position of the tentaculate taxa among protostomes would mean that upstream collecting systems evolved twice.

Feeding in adults

The ancestral mode of feeding in adult animals is filtration, in the broadest definition. This means that food is somehow extracted from a large volume of water. This is realized, with the exception of Trichoplax adhaerens, in all diploblastic taxa. However, filtration is performed in a variety of ways. While in sponges a water current produced by ciliary action brings in food particles, cnidarians and ctenophores ‘comb’ the water or use water currents to come into contact with food. Through their capture mechanisms, cnidocytes (Holstein & Tardent 1984, Mariscal 1984, Hessinger & Lenhoff 1988, Tardent 1995) or colloblasts (Benwitz 1978), are specialized to capture larger prey than sponges can (Table 12.4.).

Within Bilateria, filter feeding is also widely distributed. Some taxa such as molluscs, annelids, and echinoderms show a wide variety of modes of nutrition, among which filtration evolves secondarily (in bivalves, sedentary polychaetes, and in some ophiuroids and holothurians). In many cases, cilia are important (p.236)

Table 12.4 Feeding and food source in adults. The source of information is, if not otherwise indicated, taken from general textbooks (Westheide & Rieger 2004, 2007, Ruppert et al. 2004) and from the chapters in the series ‘Microscopic Anatomy of Invertebrates’.

Food source, capture of food and transport to mouth

Porifera

Food: fine, suspended particles, small planctonic organisms (< 50 m), captured by choanocytes from water incurrent. Exception: carnivorous sponges capture zooplanctonic organisms (Vacelet & Boury-Esnault 1995).

Trichoplax adhaerens

Food: flagellates (Cryptomonas) in laboratory cultures, in sea presumably protozoans and bacteria colonizing hard substrates. Food is captured in pockets between substrate and ventral epithelium.

Cnidaria

Food: zooplanctonic organisms. Prey captured by nematocyst explosion and moved to mouth by tentacles.

Ctenophora

Food: zooplanctonic organisms. Prey captured by contact with sticky colloblasts and moved to mouth by contraction of tentacles.

Xenoturbella

Food: not known with certainty. Probably bivalves (Nucula, see Bourlat et al. 2003), probably also dead animals (Westblad 1950). Prey ingested through ventral mouth.

Acoelomorpha

Food: no data are available for Nemertodermatida, very little is known for Acoela, but probably a range of very small (diatoms, protozoans, algal spores) to larger (small annelids and crustaceans) food can be used. According to observations on Convoluta convoluta (Jennings 1957), small food is engulfed by syncytial intestinal tissue protruded through the mouth while large food is trapped by lifting the anterior end and moving it over approaching prey.

Gastrotricha

Food: small organisms (bacteria, diatoms, small protozoans). Prey is sucked in by pharyngeal action.

Nematoda

Food (of free-living forms): bacteria, larger organisms, content of plant or fungal cells. Food is sucked in by pharyngeal action, in nematodes with stylets or teeth after or with simultaneous mechanical treatment.

Nematomorpha

No feeding as adults. Only in the (juvenile) parasitic phase, nutrients are absorbed, mainly over body surface (Hanelt et al. 2005).

Priapulida

Food incompletely known. Macroscopic forms probably predators of polychaetes and other prey, microscopic forms deposit-feeders or microcarnivoures. In macroscopic species, prey may be captured and swallowed with the aid of pharyngeal teeth.

Kinorhyncha

Food: diatoms, bacteria which are, at least in Homalorhagida, sucked in by pharyngeal action.

Loricifera

Food: bacteria, sucked in by pharyngeal action.

Platyhelminthes

Food: diversity of prey organisms, ranging in size almost up to the flatworm's body size. Pharyngeal action is central in prey capture and uptake. According to the flatworm taxon, small prey is swallowed as a whole, large prey is damaged by pharyngeal action and sucked out (Jennings 1968). Parasites feed on host's blood or tissue (e.g. Digenea) or on host's intestinal content, taken up over the body surface (e.g. tapeworms).

Gnathostomulida

Food: probably algal or bacterial films on sand grains, scraped off with jaw apparatus (Sterrer 1971).

Limnognathia maerski

Food: probably bacteria or diatoms, collected by preoral ciliary field and taken over by jaw apparatus (Kristensen & Funch 2000).

Eurotifera

Food: suspended particles (in suspension feeders), algae or zooplanctonic organisms (in predatory eurotifers). Suspension feeders (e.g. bdelloids, monogononts such as Brachionus, Keratella) create water current by action of wheel organ cilia, particles are sucked in by pharynx and ground by jaw (mastax). Selective feeders capture prey (algal cells, animal prey) by a combination of sucking and piercing (e.g. Notommata, Polyarthra, Synchaeta) or by grasping with forceps-like jaws (Asplanchna) (Wallace & Ricci 2002). Jaws are adopted to respective modes of feeding and can be distinguished into several types (e.g. ramate, virgate, malleate, forcipate; see e.g. Wallace & Snell 1991).

Seisonida

Food: according to stomach contents, Seison nebaliae feeds on bacteria and S. annulatus on hemolymph of the host (Ahlrichs 1995). Bacteria are probably picked up with jaw.

Acanthocephala

Food: nutrients from host's intestine, absorbed over the epidermis.

Nemertini

Food: prey organisms, particularly annelids and crustaceans, in some cases also dead animals. Prey is captured with proboscis, either by wrapping around prey or, in Enopla, with the aid of proboscidal stylet. Toxins are applied in both cases.

Kamptozoa

Food: suspended particles. Particle capture with tentacles (downstream collecting system), ciliary transport to mouth.

Mollusca

Caudofoveata: selective deposit feeders. Solenogastres: carnivores on cnidarians. Polyplacophora: algal scrapers. Neopilinida: deposit feeders. Gastropoda: variety of food sources, abundantly redula scraping on plant material, but also carnivory. Cephalopoda: carnivores, prey capture with tentacles, maceration with beak and radula. Bivalvia: deposit or suspension feeders with targeted (in deposit feeders) or untargeted (in suspension feeders) water incurrent, trapping of particles by mucous on gills, transport with labial papilla and cilia to mouth. Scaphopoda: microcarnivours or microomnivoures, prey or food captured and transported to mouth by captacula. (See also von Salvini-Plawen 1988).

Sipunculida

Food: fine particles, either suspended in water, within or on sediment. Suspension and deposit feeders capture particles with tentacles by mucous and transport them by cilia to the mouth, sediment feeders (e.g. Sipunculus nudus) ingest entire sediment.

Echiurida

Food: Fine particles on sediment. Deposit feeders capture particles in mucous on extended proboscis and transport it by cilia to the mouth. Exception is Urechis caupo who captures fine particles and plancton by a water current through a mucous net, which is ingested periodically.

Annelida

Food: diversity ranging from fine organic particles (suspended or on sediment) to plant material, animal prey, entire sediment and blood (gnathobdellid leeches). Food is either collected with anterior appendages (e.g. in sedentary polychaetes), grabbed with the pharynx (or jaws) or directly taken up into the mouth.

Tardigrada

Food: content of plant cells or animal prey. Plant cells (algae etc.) as well as animals (nematodes, eurotifers, sometimes even other tardigrades) are pierced with stylets and either sucked out or ingested completely by pharyngeal action.

Onychophora

Food: smaller euarthropods. Food capture by squirting of slime gland secretions over the prey, followed by maceration with ìmandiblesî. Prey is partly digested externally with saliva, which is then sucked in.

Euarthropoda

Food: diversity of food sources, many euarthropods are carnivorous, but also herbivory, filtration, blood sucking and other utilizations of food sources exist. Food uptake and manipulation with specialized mouth parts, arachnids have extraintestinal digestion.

Chaetognatha

Food: zooplanctonic organisms. Grasping of prey with hooks, teeth may help in maceration.

Phoronida

Food: fine, suspended particles. Particle capture with tentacles of lophophor (upstream collecting system), ciliary-mucous capture and transport to mouth.

Brachiopoda

Food: fine, suspended particles. Particle capture with tentacles of lophophor (upstream collecting system), ciliary transport to mouth (probably without mucous).

Bryozoa

Food: suspended particles, phytoplancton, exceptionally zooplancton. Particle capture with tentacles of lophophor (upstream collecting system), ciliary transport to mouth.

Echinodermata

Crinoida: suspension feeders, particle capture with mucous on tentacular ambulacral feet, ciliary transport to mouth. Asteroida: microvorous or macrovorous predators. Ophiuroida: microphagous or macrophagous carnivores, suspension feeders or mixture of mechanisms. When suspension feeding, particles are trapped by mucous. Echinoida: algal scrapers (regular echinoids) or deposit feeders using ambulacral feet for selective uptake (irregular echinoids). Holothuroida: suspension feeders (with branched, filtering buccal podia), deposit feeders (selective uptake of particles with buccal podia) or endobenthic sediment feeders (Apodida, Molpadiida).

Enteropneusta

Food: fine particles (detritus, suspended particles). Species are either suspension, deposit or sediment feeders. Suspension feeders create water current with proboscial and pharyngeal cilia, particles are trapped by mucous on proboscis and in pharynx. Deposit feeders (e.g. Saccoglossus species) trap particles on the sediment with mucous and transport this to the mouth. Sediment feeders (e.g. Balanoglossus species) feed on sediment rich in organic particles.

Pterobranchia

Food: fine, suspended particles. Particle capture with tentacles (upstream collecting system), ciliary transport to mouth.

Tunicata

Food: planctonic organisms. Cilia create a water current through the mouth (buccal siphon) and through the gill slits. Particles trapped in pharynx by mucous net created in the ventral endostyle, this net is collected by the dorsal groove, concentrated into a cord and transported posteriorly.

Acrania

Food: fine, suspended particles. Cilia create a water current into the mouth, water leaves through the branchial slits, atrium and atriopore. Food capture similar to tunicates.

Craniota

Food: diversity of food. Originally presumably fine suspended particles captured with branchial pharynx similar to Acrania (realized in larva of Petromyzontida). Ancestor of Gnathostomata likely switched to large animal prey, specializations on other food sources (e.g. plant material) are secondary.

(p.237) (p.238)
                   Intestinal systems

Fig. 12.16. Comparison of food capture in Bryozoa and Kamptozoa. To the left the general water current through the tentacles is shown. To the right are cross sections of single tentacles, with indicated direction of ciliary beat, water current, and particle capture. The frontal cilia direct towards the internal of the tentacular crown. Figures modified after Ryland (1970), Nielsen (1998) and Ruppert et al. (2004).

to create water currents and transport captured particles to the mouth, but the actual capture is often performed with mucous, to which dissolved food particles adhere. In some taxa, cilia play the dominant role in filtration and here we can distinguish, similarly to the larvae, downstream and upstream collecting systems (see Table 12.4.). For example, despite having quite similar tentacular crowns on first view, kamptozoans and bryozoans use completely different mechanisms for particle capture (Fig. 12.16.). Kamptozoans have two sets of cilia along their tentacles: longer lateral cilia creating a water current from outside into the tentacular crown (atrium), and shorter frontal cilia that transport the particles along the tentacle to the mouth (Nielsen & Rostgaard 1976, Emschermann in Westheide & Rieger 2007). Particles can be captured because turbulences occur in the water current while it passes the tentacles, and particles are caught in the ‘current shadow’ on the frontal (atrial) side of the tentacles (i.e. downstream). Bryozoans (Fig. 12.17.) also have lateral and frontal cilia on their tentacles, but they additionally have long and stiff laterofrontal cilia (Nielsen 2001). The lateral cilia create a water current from the atrium to the outside, that is, in the opposite direction as in kamptozoans. The stiff laterofrontal cilia mechanically filter particles from the water current (Riisgård & Manriquez 1997, Nielsen & Riisgård 1998, Riisgård et al. 2004), upstream of the filtering structure. These are ‘kicked’ by a flicking of the laterofrontal cilia and/or of the tentacle towards (p.239)
                   Intestinal systems

Fig. 12.17. Colony of Flustrellidra hispida (Bryozoa, Gymnolaemata), showing lophophor with tentacles in circular arrangement. Photo by courtesty of Georg Mayer, Berlin.

the mouth. The tentacles (lophophor) of brachiopods and phoronids function in a similar way. The presence of upstream collecting systems in adults is, as in larvae, an argument for a closer relationship of the tentaculate taxa to deuterostomes. In this case, the tentacle apparatus of phoronids, brachiopods, and bryozoans might be homologous to the pterobranch tentacles. If, however, bryozoans or all tentaculate taxa belong to the protostomes, then such a tentacular collecting system would be of convergent origin (see Halanych 1996b).

Conclusions

There are three basic steps in nutrition: the capture of food, its breakdown (when the captured particles are not already very small), and endocytosis—the uptake of particles into the body. It appears that only capture and uptake are ancestral patterns, and that an intestinal tract evolved particularly to extend the range of food to larger items. These are treated in a compartment separated from the environment mechanically and chemically. The intestinal tract is originally sac-shaped, with only one opening, but within bilaterians it becomes, probably several times convergently, an unidirectional one-way-gut with two openings. This allows a more specialized treatment of food and therefore more effective nutrition. With a diversification of bilaterian animals goes a diversification in the utilization of food sources and a diversification of feeding structures adopted to sucking, swallowing, scraping, piercing, and so on (see Table. 12.4.).

In marine, ciliated larvae, some patterns with some phylogenetic impact can be realized in the mode of nutrition (lecithotrophic versus planctotrophic) and the presence as well as the function of ciliary bands (downstream versus upstream collecting bands). While these patterns are very informative and helpful when considering particular taxa, broader hypotheses (e.g. for bilaterian or metazoan evolution) depend on several ambiguous considerations and are therefore not well substantiated.