Current Biology

Magazine

R782 Current Biology 25, R775–R792, September 21, 2015 ©2015 Elsevier Ltd All rights r

an individual animal may contain reproduction-competent germ cells in its gonads from two genetically different individuals: its own germ cells as well as those of its sibling. Hence, their offspring may not be the genetic descendant of the physiological parents.

Does chimerism affect parental care? Chimerism has been noted to affect parental care especially male parental care. It has been observed that fathers provide signifi cantly higher care to chimeric infants than non-chimeric infants.

Baby marmosets benefi t from male parental and alloparental care, what do genetics say? Very recently, a novel variant form of the so-called love hormone, oxytocin, has been reported in marmosets and some other close relatives. Interestingly, this variant of oxytocin showed a clear correlation with litter size. Therefore, it is believed that this variant form of oxytocin might be responsible for male parental and alloparental care of marmoset babies.

How did the tiny marmosets evolve from a larger bodied ancestor? In contrast to other primates, marmosets and their tamarin relatives have undergone a secondary decrease in body size from a larger platyrrhine ancestor. The marmoset genome suggested this may be due to positive selection in fi ve genes related to the growth hormone/insulin-like growth factor axis, which have possible roles in body-size control. Furthermore, eight genes associated with adaptations for challenges of small body size were found. These genes are related to subunits of the respiratory complex I, which plays an important role in energy transduction and hence affects metabolic rate and body temperature.

Are marmosets endangered? Due to its wide distribution and adaptability, the common marmosets — as well as some other marmoset species — are not in danger of immediate extinction. However, many other marmoset species have a threatened conservation status. Wied’s black-tufted-ear marmoset (Callithrix kuhlii) is near threatened, while golden-white

bare-ear marmoset (Mico leucippe), black-crowned dwarf marmoset (Mico humilis), Rondon’s marmoset (Mico rondoni), buffy tuft-eared marmoset (Callithrix aurita) and black-headed marmoset (Callithrix nigriceps) are at vulnerable status in the International Union for Conservation of Nature (IUCN) red list of threatened species (http://www.iucnredlist.org/search). The buffy-headed marmoset (Callithrix fl aviceps) is listed as an endangered species by IUCN. Habitat destruction due to human encroachment as well as the pet trade are among the major threats for extinction of these marmoset species.

Where I can fi nd out more?Aeckerle, N., Drummer, C., Debowski, K.,

Viebahn, C., and Behr, R. (2015). Primordial germ cell development in the marmoset monkey as revealed by pluripotency factor expression: suggestion of a novel model of embryonic germ cell translocation. Mol. Hum. Reprod. 21, 66–80.

Eng, C.M., Ward, S.R., Vinyard, C.J., and Taylor, A.B. (2009). The morphology of the masticatory apparatus facilitates muscle force production at wide jaw gapes in tree-gouging common marmosets (Callithrix jacchus). J. Exp. Biol. 212, 4040–4055.

Fereydouni, B., Drummer, C., Aeckerle, N., Schlatt, S., and Behr, R. (2014). The neonatal marmoset monkey ovary is very primitive exhibiting many oogonia. Reprod. Camb. Engl. 148, 237–247.

Ford, S.M., Porter, L.M., and Davis, L.C. eds. (2009). The Smallest Anthropoids (Boston, MA: Springer US). Available at: http://link.springer.com/10.1007/978-1-4419-0293-1.

Mustoe, A.C., Cavanaugh, J., Harnisch, A.M., Thompson, B.E., and French, J.A. (2015). Do marmosets care to share? Oxytocin treatment reduces prosocial behavior toward strangers. Horm. Behav. 71, 83–90.

Ross, C.N., French, J.A., and Ortí, G. (2007). Germ-line chimerism and paternal care in marmosets (Callithrix kuhlii). Proc. Natl. Acad. Sci. USA 104, 6278–6282.

Rylands, A.B. ed. (1993). Marmosets and Tamarins: Systematics, Behaviour, and Ecology (Oxford, Oxford University Press).

The Marmoset Genome Sequencing and Analysis Consortium (2014). The common marmoset genome provides insight into primate biology and evolution. Nat. Genet. 46, 850–857.

Vargas-Pinilla, P., Paixão-Côrtes, V.R., Paré, P., Tovo-Rodrigues, L., Vieira, C.M. de A.G., Xavier, A., Comas, D., Pissinatti, A., Sinigaglia, M., Rigo, M.M., et al. (2015). Evolutionary pattern in the OXT-OXTR system in primates: Coevolution and positive selection footprints. Proc. Natl. Acad. Sci. USA 112, 88–93.

Vinyard, C.J., Wall, C.E., Williams, S.H., and Hylander, W.L. (2003). Comparative functional analysis of skull morphology of tree-gouging primates. Am. J. Phys. Anthropol. 120, 153–170.

Stem Cell Biology Unit, German Primate Center, Leibniz Institute for Primate Research, Kellnerweg 4, D-37077 Göttingen, Germany. *E-mail: fwahab@dpz.eu

Neural nets

Andreas Hejnol and Fabian Rentzsch

“The nerve-net of the lower animals contains the germ out of which has

grown the central nervous systems of the higher forms.”

— G.H. Parker: The Elementary Nervous System, 1919

Although modern evolutionary biology has abandoned the use of ‘lower’ or ‘higher’ for animals, the quote of G.H. Parker captures quite well the current understanding of the nerve net as the evolutionarily oldest organization of the nervous system, the major organ system responsible for processing information and coordinating animal behaviour. The degree of complexity of a nervous system — in particular its organization into substructures such as brains and nerve cords — shows fascinating variations between animals. Even within an individual, the nervous system can show parallel existing types of organizations that are only partially connected, illustrated by the well-known central and peripheral nervous system. In general, the architecture of the nervous system is adapted to the specifi c needs and lifestyle of the individual species. How these diverse and complex nervous systems evolved is an ongoing debate among zoologists and evolutionary biologists.

The simplest organization of a nervous system is commonly referred to as the nerve net. The term nerve net is often used for a mesh-like nervous system in which signals can be transmitted in any direction. Here we use this term only in a morphological and not in a functional sense. A basic nerve net consists of an irregular arrangement of neurites from monopolar, bipolar or multipolar neurons (Figure 1A). The neurites form a planar sheet that is connected to underlying myocytes or other contractile cells. On the opposite side, the nerve net is connected to epithelial sensory cells — for example, mechanoreceptors — and can also involve neurosecretory cells (Figure 1B). A nerve net thus can be suffi cient

Primer

eserved

Current Biology

Magazine

B

Longitudinalmusculature

A

Epidermis

Ringmusculature

Sensory cell

Nerve net

Figure 1. Schematic drawings of nerve nets.(A) Ectodermal nerve plexus from the cnidarian Rhizostoma (from Hanström 1928, taken from Bozler 1927; with permission from Springer Science and Business Media). Multipolar neurons in blue, bipo-lar neurons in grey. (B) Location of the nerve net in an idealized Bilaterian. The neurons of the nerve net are in red. Connections to the muscular sheets and sensory cell are shown. Panel B reproduced with permission from Pavans de Ceccatty, M. (1974). Perspect. Biol. Med. 17, 379–391.

to form the integrative part between reception of the environment of an organism and its ability to react to changes in these environmental cues.

This architecture of a mesh-like arrangement seems to be quite versatile, as more or less elaborated nerve nets can be found in most animal taxa, albeit at different scales, constituting either most of an animal’s nervous system (as in cnidarian polyps) or innervating a particular organ or part of it (as in the vertebrate intestine). At the level of a whole animal, however, nerve nets in a strict sense — without any ganglion-like accumulations of nerve cells or bundles of neurites — are rarely found. Even in the cnidarian polyp Hydra, which has a classic textbook example for a nerve net, some species have a distinct bundle of neurites that encircles its single body opening.

With the aim of discussing the evolution of neural nets, we focus here mainly on animals in which nerve nets form a major part of the nervous system and that have positions in the animal tree of life that are informative for considerations of how nervous systems have evolved (Figures 2 and 3).

The phylogenetic distribution of nerve netsFor a better understanding of the origin of this ancestral organization of nervous systems, it is important to examine which extant lineages of animals have retained a nerve net and how it is integrated into the body. The gelatinous comb jellies (ctenophores) and jellyfi sh (cnidarians) are listed in common zoology textbooks as ‘nerve net’ animals. As mentioned above, however, there are no members in these groups that exclusively possess a nerve net without any condensations at all. Instead, we fi nd in both groups highly elaborated neural organs: ctenophores have a complex neural apical organ with gravitation and light sensory systems; some cnidarian medusae have sophisticated sensory organs along their umbrella, called rhopalia, that in the case of box jellyfi sh include eyes with lenses.

While the uniformity of the neurons in some nerve nets encouraged morphologists to describe it as ‘a nightmare telephone system in

Curren

which there is no proper exchange but only a tangle of wires’, we know from molecular studies that the morphologically rather uniform neurons in animal nerve nets in fact belong to different cell types. Recent studies indicate that differentially expressed molecular markers — transcription factors as well as neuropeptides and neurotransmitters — assign specifi c neurons to different identities and functions.

It seems reasonable to assume that the fi rst neurons of animals were organized in a simple nerve net. The emergence of the nerve net is thus connected with the evolution of the neuron. Whether the nerve cell of animals has a single origin, or possibly originated more than once, is currently a robustly debated topic among evolutionary neurobiologists, one stimulated by the (controversial) possibility that comb jellies may be the sister taxon to all remaining metazoan animals. Sponges (Porifera) and placozoans lack neurons, so if comb jellies are a sister taxon to all other metazoans, then either those two taxa have lost neurons during evolution or neurons (and nerve nets) evolved twice independently (see Figure 3). The analyses of two recently sequenced ctenophore genomes has not resolved this dispute, which is set to continue, hopefully being informed by further work on the developmental genetics and comparative physiology of the nerve cells of comb jellies and those of other animals.

t Biology 25, R775–R792, September 21, 201

More detailed studies of the nervous system of another interesting group of animals, the Xenacoelomorpha, have revealed a basic organization of the nervous system in form of a net (Figures 2 and 3). These rather small, mostly marine worms are completely ciliated and their nervous system is mainly composed of a basiepidermal nerve net. This nerve plexus forms a two-dimensional mesh, which is connected to intraepidermal sensory cells and the underlying muscle sheet, composed of longitudinal and ring muscles. In fact, the nervous system of Xenoturbella, which is likely the sister group to all remaining xenacoelomorphs, has negligible condensations and provides a textbook example of a pure nerve net. Interestingly, within the Xenacoelomorpha, the taxon Acoela is the only group that has a subepidermal brain, with internalized neurite bundles arranged in an orthogonal-like architecture. Xenacoelomorphs thus showcase how nerve nets can be evolutionarily transformed into neurite bundles and brains, independently from the nervous system centralization seen in remaining Bilateria.

The other bilaterian groups — protostomes and deuterostomes, together Nephrozoa — also possess a more or less elaborated nerve net, albeit in most groups only in addition to condensed neurite bundles and brains (Figure 2). One can only speculate as to the precise function of these nerve nets, as detailed studies of the

5 ©2015 Elsevier Ltd All rights reserved R783

Current Biology

Magazine

tyr-tubhoechst

acet-tubHT5

HT5tyr-tub

tyr-tubHT5

A C

D E F

B

elav::mOrange

Figure 2. Neural nets.Confocal microscopic images of nerve nets from different animals labelled with different stain-ings. (A) Basiepidermal nerve net of the epidermis of the ctenophore Mnemiopsis leidyi labelled with the nuclear stain Hoechst and an antibody against tyrosinated tubulin. (B) (endodermal) Nerve net of the cnidarian Nematostella vectensis labelled with elav::mOrange. (C) Basiepi-dermal nerve net of the xenacoelomorph Meara stichopi labelled with antibodies against sero-tonin (HT5) and tyrosinated tubulin. (D) Basiepidermal nerve net of the hemichordate Sacco-glossus kowalevskii labelled with an antibody against serotonin. (E) Basiepidermal nerve net of the chaetognath Sagitta setosa labelled with an antibody against acetylated tubulin. (F) Dorsal basiepidermal nerve net of the nemertean Lineus ruber labelled with antibodies against tyrosi-nated tubulin and serotonin. Panel A, courtesy of Kevin Pang; panel D courtesy of Ariel Pani; E reproduced from Richter, S. et al. (2010), adapted from Harzsch, S. and Müller, C.H.G. (2007). Front. Zool. 4, 14.

neurophysiology of these non-model species are lacking.

The nerve net as a starting point for condensations and specializationsAssuming the nerve net is the earliest neural tissue in which interwoven neurons connect with epithelial sensory cells and internal muscle cells, we might be able to postulate a pathway leading to derived nerve condensations, such as neurite bundles, medullary cords and brains.

The integration of sensory information received from different parts of the body can be facilitated by an accumulation of heavily interconnected nerve cells in a nerve cord or a brain-like structure. Similarly, longer neurites assembled into bundles can facilitate fast, coordinated responses of the animal, such as a turn of the swimming direction to escape a predator. Centralized nervous systems can be found distributed throughout many animal taxa, with a variable

R784 Current Biology 25, R775–R792, Septe

extent of centralization and the nerve condensations placed either within or below the epidermis (basiepidermal or subepidermal, respectively). In some animals with prominent subepidermal longitudinal nerve cords — for example, vertebrates, the fruit fl y Drosophila melanogaster and the annelid Platynereis dumerilii — the molecular and functional organization of the nerve cord show very striking similarities, which has been argued to refl ect an ancient origin of the nerve cord. In contrast, comparative morphology and recent advances in solving animal relationships with molecular tools suggest that internalizations from a basiepidermal to a subepidermal condensation happened multiple times independently, for example inside the ribbon worms (Nemertea) and segmented worms (Annelida).

Resolving these confl icting observations will require detailed analysis of the expression and function of genes that pattern the nerve

mber 21, 2015 ©2015 Elsevier Ltd All rights rese

cords in those animals that possess prominent nerve nets instead of single nerve cords. It will also be important to acquire a better understanding of the functional properties of different nerve nets. What types of behaviour do they allow, and what advantages might this confer for life in a particular environment? Only by such a multiplicity of studies on a broad range of species will it be possible to understand how nerve nets can be transformed during evolution into more complex architectures, and whether there might be a common mechanism that can explain how similar-looking central nerve cords evolved independently several times.

MotilityOne common characteristic of animals is motility, which of course is achieved in different ways by different species. Some of the different types of animal motility can be controlled by a nerve net. Nerve nets trigger, for example, the rhythmic beating of the umbrella of many pelagic cnidarian medusae. It has been shown that the motor nerve net of scyphomedusae uses bidirectional chemical synapses to allow diffuse conduction of signals that originate from peripherally located pacemaker neurons. This pulsed information results in the synchronized contraction of the underlying muscle sheet and propulsion of the animal by expulsion of water. But the same mode of locomotion can also be achieved without the help of a nerve net: in some hydromedusae, the contraction of the subumbrella muscle sheet is regulated by gap junctions, protein channels that electrically couple the muscle cells, without involvement of nerve cells at all.

Such electrical coupling is also involved in the bending of the body of hydra; however, here neurons of the nerve net communicate via gap junctions to trigger contraction of the epithelial muscle cells in the ectoderm. In combination with ring and longitudinal muscle sheets, a nerve net can even coordinate the rhythmic contractions and extensions of peristaltic movements. Such principles are used during the digestive movements of the human intestine or of even an entire animal, as in enteropneust hemichordates.

rved

Current Biology

Magazine

Platyhelminthes

Tardigrada

Phoronida

Nemertea

Cycliophora

Micrognathozoa

Loricifera

Nematoda

Mollusca

Cephalochordata

Cnidaria

Ctenophora

Xenacoelomorpha

Echinodermata

Nematomorpha

Kinorhyncha

Gnathostomulida

Craniata

Brachiopoda

Porifera

Priapulida

Gastrotricha

Hemichordata

Annelida

Placozoa

Urochordata

Entoprocta

Bryozoa

Chaetognatha

Rotifera

Onychophora

Ambulacraria

Gnathifera

Ecdysozoa

Scalidophora

ArthropodaCurrent Biology

Panarthropoda

Protostomia

ChordataBilateria

Nephrozoa

Animals

Deuterostomia

Nerve

net

-

+

+

+

-

+

+

-

+

---

-+

?

+

+

+

++

+

+

+

?

??

?

?

?

?

N.A.

N.A.

Brain

s

Longitudin

al

nerve

s

+

++

-

+

+

+

-

-

-

+

+

+

+

+

+

(+)

N.A.

N.A.

+

+

(+)

(+)

+

(+)

+

+

+

+

+

+

+

+

++

+

+

+

-

-

+

+

+

+

+

+

N.A.

N.A.

+

+

+

+

+

+

+

+

+

+

(+)

+

+

-

-

-

Spiralia

Figure 3. Animal phylogeny and distribution of nerve nets.Current animal phylogeny with the occurrence of nerve nets, brains and longitudinal nerves mapped on the higher taxonomic ranks: presence/absence coding after Schmidt-Rhaesa, A. (2007).

Again, this mode of locomotion is not restricted to animals with prominent nerve nets: earthworms (Annelida) and the mud-dwelling priapulids also move by peristalsis, but their nervous system is clearly more centralized than that of hemichordates. Thus, nerve nets can control different modes of locomotion, but there appears to be no mode of locomotion that can only be controlled by nerve nets.

Furthermore, some animals, such as xenacoelomorphs and ctenophores, have a prominent nerve net but do not use it for locomotion. The benthic xenacoelomorphs use the cilia that cover their whole body to swim and glide: here the nerve net likely is operational as a sensory system. Pelagic ctenophores (comb jellies) propel themselves through the water column with the row of composite cilia arranged in ‘combs’ — the major action of the nerve net is in triggering catching and swallowing the prey, although also here it is likely involved in the reception of environmental clues.

We see that nerve nets are good for many things and are quite versatile systems that can be integrated in varied ways into animal bodies.

DevelopmentA key feature of nerve nets is that the cell bodies of the nerve cells are distributed throughout most of the tissue. There are two ways in which this can be achieved. A homogenous distribution of nerve cells can refl ect a spatially broad neurogenic potential of a tissue: the neurons are located close to where they or their progenitors have been generated. An alternative, but not mutually exclusive, mechanism involves a spatially restricted area in which neurons or neural progenitors are specifi ed and subsequent migration that distributes them throughout the tissue.

Evidence for the fi rst of these possibilities —a nerve net-like nervous system arising from tissue with broad neurogenic potential — can be found in the anthozoan cnidarian Nematostella vectensis and in the hemichordate Saccoglossus kowalewskii (Figure 4). In Saccoglossus, the transcription factor soxB1, which labels neural ectoderm in several bilaterians, is expressed broadly

Current

in the ectoderm of the gastrula. Successively, expression of the early and late neural differentiation markers elav and synaptotagmin 1, respectively, can be detected in cells throughout the soxB1-expressing territory. Nematostella has fi ve soxB genes and expression of three of them together covers the entire blastoderm/ectoderm at blastula and gastrula stages. The use of a transgenic reporter line allowed the identifi cation of neural progenitor cells, which can be detected as broadly distributed cells from the blastula stage onwards in a pattern that prefi gures the subsequent expression of early and late neural differentiation markers. Based on the

Biology 25, R775–R792, September 21, 2015

expression pattern of the potential early neural differentiation gene soxC, the sea star Patiria miniata also appears to have a broad ectodermal neurogenic potential. In embryos of this species, however, differentiated neurons are only found in the apical domain and along the ciliary bands, which suggests that migration of neural precursors may help to prevent the formation of a non-centralized nervous system.

In Hydra and other hydrozoans, most neurons are generated by a pluripotent stem cell population, the interstitial cells (i-cells). During embryogenesis i-cells, but also differentiating neural cells, migrate from the endoderm to the ectoderm;

©2015 Elsevier Ltd All rights reserved R785

Current Biology

Magazine

Nematostella Hydra

Embryonic progenitor cells

Cell migration

- Ectoderm- Endoderm

- Endoderm

? Yes

- Ectoderm

Current Biology

?

Saccoglossus

Juvenile/adult

Embryo

Figure 4. Development of nerve nets.Schematic drawings illustrate the structure of the juvenile/adult nervous system and the distribu-tion of neural progenitor cells during embryogenesis for the cnidarians Nematostella vectensis and Hydra sp., and for the hemichordate Saccoglossus kowalewskii. The tentacle nerve nets in Nematostella and Hydra have been omitted. Note that neither adults nor embryos are drawn to scale. Nerve cells and progenitors in black are located in the ectoderm, those in orange in the endoderm. For Saccoglossus the anterior end is oriented to the left and top left, respectively. The location of neural progenitors in Saccoglossus is inferred from gene expression studies and immunohistochemistry. The bracket and the lighter color in Saccoglossus indicate ciliary band, which prevents clear in situ hybridization signals. In adult Hydra, interstitial stem cells that gener-ate nerve cells are found in the ectoderm.

and in the adult polyp, neural precursors migrate into the i-cell-free tissue at both extremities of the body column. Thus, while broad neurogenic potential likely facilitates the formation of a nerve net-like nervous system, there is no strict correlation between the mode of neurogenesis and nervous system centralization.

A broad neurogenic potential also represents a distinct starting point for the establishment of neural connectivity in nerve nets. The formation of functional neural circuits typically requires an intrinsic ability of neurons to extend dynamic cellular processes, the attraction or repulsion of these processes by target-derived cues and the extracellular matrix, the formation of synaptic contacts with other neurons or effector cells and the activity-dependent consolidation of these synapses. For the formation of a nerve net, random outgrowth of neuronal processes in a permissive

R786 Current Biology 25, R775–R792, Septe

environment might be suffi cient — there might be no requirement for target-derived neurite guidance. Classical axon guidance molecules like netrins or semaphorins are present in the genomes of animals with nerve nets, but there is currently no information as to whether they have comparable functions in these organisms.

Thus, the formation of nerve nets presents specifi c challenges at several levels and it appears that different organisms employ different developmental mechanisms to overcome these challenges and eventually end up with a nerve net-likenervous system.

ConclusionNerve nets are a multi-purpose type of nervous system organization that can be found in most animal taxa and at different scales; they control a multitude of behaviours

mber 21, 2015 ©2015 Elsevier Ltd All rights res

and can be generated by different developmental programs. The biology of nerve nets remains a fascinating and poorly understood topic and it is clear that comparative studies of neural development and physiology of non-model systems embedded in an ecological context are paramount to fi nally understand nervous system evolution.

FURTHER READING

Bozler, E. (1927). Untersuchungen über das Nervensystem der Coelenteraten. I. Teil: Kontinuität oder Kontakt zwischen den Nervenzellen? Zeitschr. Zellforsch. Mikros. Anat. 5, 244–262.

Bullock, T.H., and Horridge, G.A. (1965). Structure and Function in the Nervous Systems of Invertebrates, Volume I., (San Francisco: W.H. Freeman and Company).

Denes, A.S., Jékely, G., Steinmetz, P.R.H., Raible, F., Snyman, H., Prud’homme, B., Ferrier, D.E.K., Balavoine, G., and Arendt, D. (2007). Molecular architecture of annelid nerve cord supports common origin of nervous system centralization in Bilateria. Cell 129, 277–288.

Dunn, C.W., Giribet, G., Edgecombe, G.D., and Hejnol, A. (2014). Animal phylogeny and Its evolutionary implications. Annu. Rev. Ecol. Evol. Syst. 45, 371–395.

Dunn, C.W., Leys, S.P., and Haddock, S.H. (2015). The hidden biology of sponges and ctenophores. Trends Ecol. Evol. 30, 282–291.

Hanström, B. (1928). Vergleichende Anatomie des Nervensystems der wirbellosen Tiere (Berlin: Springer).

Holland, L.Z., Carvalho, J.E., Escriva, H., Laudet, V., Schubert, M., Shimeld, S.M., and Yu, J.K. (2013). Evolution of bilaterian central nervous systems: a single origin? Evodevo 4, 27.

Katsuki, T., and Greenspan, R.J. (2013). Jellyfi sh nervous systems. Curr. Biol. 23, R592–R594.

Lowe, C.J. (2008). Molecular genetic insights into deuterostome evolution from the direct-developing hemichordate Saccoglossus kowalevskii. Phil. Trans. R. Soc. Lond. B 363, 1569–1578.

Northcutt, R.G. (2012). Evolution of centralized nervous systems: two schools of evolutionary thought. Proc. Natl. Acad. Sci. USA 109 (Suppl 1), 10626–10633.

Richter, S., Loesel, R., Purschke, G., Schmidt-Rhaesa, A., Scholtz, G., Stach, T., Vogt, L., Wanninger, A., Brenneis, G., Döring, C., et al. (2010). Invertebrate neurophylogeny: suggested terms and defi nitions for a neuroanatomical glossary. Front. Zool. 7, 29.

Satterlie, R.A. (2015). Cnidarian nerve nets and neuromuscular effi ciency. Integr. Comp. Biol., epub ahead of print, http://dx.doi.org/10.1093/icb/icv067.

Schmidt-Rhaesa, A. (2007). The Evolution of Organ Systems (Oxford: Oxford University Press).

Schmidt-Rhaesa, A., Harzsch, S., and Purschke, G. eds. (2015). Structure and Evolution of Invertebrate Nervous Systems (Oxford: Oxford University Press).

Sars International Centre for Marine Molecular Biology, University of Bergen, Thormøhlensgate 55, 5008 Bergen, Norway. E-mail: andreas.hejnol@uib.no, fabian.rentzsch@uib.no

erved