Above: a Pov-Ray model of a sponge. Sponges belong to the phylum Porifera, which literally many 'many pores' since the
surface of a sponge is covered in minute pores that suck in water and nutrients, which the sponge filters before expelling
the water from a large opening or osculum. Sponges are usually brightly coloured - red, orange, purple, green and yellow
are common sponge colours, though deep sea forms are often dark brown and drab or glassy. The sponge above is
definitely an individual - it has a single large opening, or osculum, which carries a jet of water out of the sponge's body,
but many sponges form colonies of many individuals fused together with many oscula, in which case it is not very sensible
to talk about an individual sponge (although the tissue surrounding each osculum may define an individual). Sponge
bodies may be ball, vase, basket, cup or club-shaped like the one above, or they may be flat and encrusting or branching
and tree-like. Indeed, the shape of a sponge is an adaptation to its environment - in rough waters, such as along a rocky
coastline, sponges will be flat and encrusting, clinging tightly to any nook and cranny they can find in the rock surface,
whilst in calm waters they tend to be more upright and present a larger surface to the water to enhance the rate at which
they can sieve food from the water. Sponges may be tiny forms, which are easily overlooked, or they may be large
enough for a man to stand inside! Sponges are very unusual animals, representing an off-shoot of the animal kingdom
that evolved along its own lines, separate from the vast majority of animal types. It is instructive, therefore, to see how the
sponge body is put together and how it works!
The simplest type of sponge is the asconoid type. This sponge has many minute pores opening directly into a single
central cavity, or atrium (spongocoel), and one or more large oscula (singular: osculum). The sponge sucks water into the
atrium through the many tiny incurrent or inhalent pores and pumps it out through the excurrent or exhalent osculum as a
forceful jet of water that may travel 10 feet or more from a large sponge. As this water flows through the sponge body,
food particles are filtered from it. These particles include mostly bacteria and other microscopic organisms and organic
debris. The simplified anatomy of a small asconoid sponge is shwown below (most sponges will contain far more cells and
many more pores than this simple model!):
The asconoid sponge body consists of two layers of cells with a layer of gelatinous material (called mesogloea)
sandwiched in-between. The outermost layer consists of flattened paving-slab like cells, called pinacocytes. These cells fit
together tightly to form the outer surface or pinacoderm (a type of covering tissue or epidermis). They typically have wavy
contours to strengthen the connections between them (by increasing the surface area along which the cells can be
fastened together). Inside is a layer of so-called collar cells or choanocytes, each with one flagellum pointing toward the
atrial cavity and surrounded by a ring of short appendages called cilia. The flagellum undulates or beats, expelling water
out into the atrium and sucking it in through the pores that penetrate both cell layers. Each of these incurrent pores is a
channel travelling through a cylindrical cell called a porocyte. Collectively, many thousands or millions of choanocytes can
produce a powerful jet of water that leaves the atrium from the large excurrent osculum (or oscula if more than one is
present). The choanocytes work very much like a group of organisms called choanoflagellates. Choanoflagellates may
live as single cells or in multicellular colonies and are rather like protosponges - sponges lacking the pinacocytes and
mesogloea and other packaging cells that make up the sponge body. The structure of a solitary choanoflagellate is
Above: the structure and function of a choanoflagellate. These microscopic creatures beat their flagellum, driving a
current of water away from them and sucking in water from behind which passes between the collar of microvilli to fill
the void left by the water pumped away. (Note: some refer to the appendages making the collar as cilia, others as
microvilli, but although superficially similar these are very different organelles). The flagellum has wings or flanges
which help it push against the water more effectively (it must be remembered that on this microscopic scale water
behaves as quite a thick sticky liquid). As water passes between the collar, potential food items, like bacteria, are
sieved out and ingested by the cell. This is essentially similar to how a choanocyte in a sponge works (except that
choanocytes lack the flanges on their flagella). Thus, a sponge's body contains thousands of microscopic pumps that
also filter food from the water. It is thought that choanoflagellates are related to sponges, and that perhaps sponges
evolved from colonies of these cells.
Another general cell type is typical of sponges - the amoebocyte. Amoebocytes are amoeboid cells that are free to
crawl around (in and on) the body of the sponge in much the same way as do amoebae. (It would be interesting to
examine the amoeboid locomotion of these cells in detail to see if it more closely resembles that of amoebae or of
animal cells). Amoebocytes have a variety of functions, including maintaining the mesogloea (mesenchyme) and
removing foreign organisms and debris from the sponge body. One of the main functions of the wandering
amoebocytes is secretion of the sponge skeleton.
The mesenchyme is a transparent gelatinous matrix (mesogloea) containing free amoebocytes. The mesenchyme
may be a collenchyma (meaning it has few cells and a lot of material between the cells), or a parenchyma (with a high
cell density and little material between the cells as they are tightly packed together). The amoebocytes are free to
wander about the sponge and fall into two main classes, lobopodous amoebocytes and collencytes. Lobopodous
amoebocytes include pigmented chromocytes, thesocytes that store food reserves and scleroblasts that secrete the
skeleton. Scleroblasts are further divided into calcoblasts, silicoblasts and spongioblasts, depending on the nature of
the skeletal material secreted (calcium carbonate, silicon or spongin protein). Lobopodous amoebocytes have
lobopods - locomotory appendages or psedudopods that are blunt, rounded and finger-like - the so-called lobopod
type of pseudopod. Collencytes have slender branching pseudopods (called filopods) and may form a syncytial
network (a syncitium is a group of neighbouring cells whose cell membranes are fused together to form a continuous
mass of cytoplasm. However, distinguishing between a syncitium and a collection of distinctly 'separate' cells with
junctions linking them together is not easy!).
Archaeocytes are lobopodous amoebocytes. Archaeocytes are possibly undifferentiated cells and produce the sex
cells (as may choanocytes in some sponges?) i.e. the spermatocytes and oocytes (sperm and egg cells) and are
involved in regeneration since they can give rise to all other sponge cell types. Indeed, one of the striking
characteristics of sponges is their ability to regenerate from a few cells - mince a sponge and each piece can grow
back into a new sponge!
More complex sponge designs
Syconoid sponges have a more complicated body plan than asconoid sponges. Syconoids are vase-like with a single,
terminal ostium. They have many finger-like out-pushings which form radial canals continuous with the central atrium
(spongocoel). These external projections pack the surface and may be free and surrounded by sea-water or they may
be covered by an outer epidermal covering which contains dermal pores, which open into the channels between the
projections, called incurrent canals, which pass through pores called prosopyles into the radial canals which are lined
by choanocytes. The radial canals then open into the spongocoel through internal ostia and the water flows through
the osculum to the outside.
Leuconoid sponges have the most complex structure. There is usually no central spongocoel, instead the sponge
cavity is highly branched and divided into clusters of small round or oval chambers lined by choanoflagellates.
Mesenchyme fills the spaces around these chambers. Leuconoid sponges have an indefinite form permeated by a
maze of water channels. Nevertheless, a given unit of water only flows through one choanocyte chamber as the
chambers operate in parallel rather than in series (there is little point trying to filter food from water that has already
been filtered!). Most sponges are of the leuconoid type, as this permits the sponge to develop the most efficient water
current and to attain a larger size (as it filters food from water more efficiently). The leuconoid architecture is shown in
the diagrams below (click the images to enlarge). This figure was redrawn from Libbie Henrietta Hyman's classic text:
The Invertebrates (Vol 1: Protozoa through Ctenophora). Note how the choanocytes are restricted to numerous small
chambers formed by the two-fold evagination of the spongocoel.
|Left: a model of an infaunal
sponge. These sponges vary
considerably in form between
species and many undoubtedly
remain unknown to science.
Siphons emerge above the
sediment which buries the bulb-like
body and root-like appendages
extend into the substrate. Right:
the sponge as seen in its natural
habitat - with only the porous tips
of the siphons visible above the
The Sponge Skeleton
The sponge skeleton deserves special attention. The mesenchyme secretes and contains the skeleton. The skeleton
consists of spicules and/or spongin fibres. The spicules may be principally calcium carbonate in calcareous sponges
and are made of silica in siliceous sponges; spongin is a protein. The spicules (sclerites) are tiny (mostly microscopic)
crystalline bodies and each is a spine or a number of spines radiating from a point. Each spicule consists of an
organic axis surrounded by calcium carbonate or hydrated silica. Megascleres are the larger spicules that from the
main supporting framework. Microscleres are smaller flesh spicules strewn throughout the mesenchyme. However,
such a size distinction does not hold for calcareous sponges and some other groups.
Spicules are classified according to the number of spines or axes as follows (this topic gets very technical and has
been much simplified!):
1. Monaxon spicules have a single axis, straight or curved. There are many types according to their shape and how
they form. They may be lance-like, C-shaped, bow-shaped, thorny or knobbly, rod-shaped, twisted spirals or spiny.
Some have a pointed end projecting to the exterior of the sponge, making its surface rough and spiny, presumably for
2. Tetraxons (tetractines, quadriradiates) have four rays radiating from a common point. These rays are not in the
same planes and may resemble jacks, though some of the rays may be reduced, absent or modified into discs.
3. Triaxons (hexactinal spicules) have three axes crossing at right-angles to give six rays, some of which may be lost
or reduced or branched or curved and may have spines or knobs, etc. These spicules occur only in the class
4. Polyaxons have several equal rays that radiate from a central point and may be star-shaped or resemble spiny
5. Spheres form from concentric growth around the centre.
6. The desma is a megasclere formed from a minute monaxon, triradiate or tetraxon spicule, forming a central
structure upon which layers of silica are deposited. These deposits develop branches and tubercles. Desmas are
usually united into a network to form a net-like skeleton (lithistid).
Spongin is a protein that forms a branching network. In some sponges, spongin often binds the siliceous spicules
together. In the keratosan sponges the skeleton consists entirely of spongin (and embedded foreign particles).
Spicules are secreted by scleroblasts, which are a sub-class of amoebocyte. Silicoblasts are a type of scleroblast that
secrete siliceous spicules in the siliceous sponges. Spongin is secreted by spongioblasts. The various spicule types
(of which there are many sub-types I have not mentioned here) have different skeletal roles within the sponge body
and the types present also depends upon species. I think that you get the point that the sponge skeleton is actually
rather complex! The spicules make the tissue of many sponges very hard, prickly or stony and quite difficult to cut
with a knife. Traditional bath sponges (not the artificial type!) have spongin skeletons and no mineral spicules, which
gives them a springy and spongy texture.
Glass Sponges (Hexactinellids)
Glass sponges have a skeleton consisting of a lattice of glass-like siliceous fibres. These strange sponges are deep-
sea sponges, occurring at 100 m to 5000 m depth. Most are 10-30 cm tall, but some are over 1 metre. They are pale
in colour and may have projecting spicules (which may be up to 30 cm long) which may give them a 'glass wool' like
covering. In at least one form it would appear that the projecting glass fibres (the spicules) act as fibre-optic wires,
transmitting light from a bioluminescent crustacean resident that lives inside the spongocoel, which must create a
spectacular light display. It is quite possible that many glass sponges are bioluminescent. It has also been suggested
that the glass spicules may be part of a light-sensing system, whether for releasing or receiving light, they are very
good at transmitting light. These mysterious sponges are not very well studies and many glass-sponge wonders
surely await discovery. One well known example is the Venus's flower basket - the beautiful glassy skeletons of the
The internal structure of glass sponge tissues is very different from that of other sponges. They are covered by a
dermal membrane, comprising a syncytium with underlying membrane of spongin (a type of collagen) and have a very
open body plan with large open spaces, rather like a honeycomb. The 'archaeocytes' are generally connected to one
another by slender cytoplasmic bridges, forming a meshlike or trabecular syncytium that constitutes the bulk of the
sponge body. This syncytium is continuous with choanablasts, forming a choanosyncytium. Each choanoblast gives
off one or more stalk-like stolons that connect to collar bearing units, resembling choanocytes but lacking nuclei
(there cytoplasm is continuous with that of the choanoblasts). Stolons may also connect these collar units to one
another. This arrangement is reminiscent of certain colonial choanoflagellates, in which the cells are connected by
stalk-like or root-like stolons. The only cells that are routinely seen separated from the syncytium are the scleroblasts,
which secrete many of the spicules. All the other cells appear to be connected to one another by cytoplasmic bridges,
though they may have rounded bodies, making them resemble distinct cells. It has been argued that glass sponges
are so different to all other sponges that they should be in a separate phylum. However, their strange internal
architecture stems from the tendency of their cells not to separate fully at cell division, but to remain in contact with
one another by cytoplasmic bridges (which may be occluded by pore plates that allow certain materials to pass from
'cell' to 'cell' directly as the cytoplasm is continuous across the pores) with a single continuous cytoplasmic membrane
surrounding the whole structure. Considering that some choanoflagellates are solitary, whilst others form colonies of
connected individuals, I don't personally think that this difference is great enough to warrant placing glass sponges in
a separate phylum.
Demosponges either have no spicules at all or they have the siliceous type. All are of the leuconoid type. There are
at least three types, including the horny sponges (keratosans) which possess spongin skeletons but have no spicules
(unless these are made of spongin), though they may incorporate rock grains and other materials into their bodies.
The spongin forms a lattice or tree-like branching network throughout the keratosan sponge body. Keratosans are
usually black in colour and have a smooth or warty leathery texture. They live attached to hard materials, glued in
place by spongin secretion.
These have calcareous spicules and include all the asconoid types, some syconoids and some leuconoids. The
calcium carbonate spicules are usually separate but are sometimes fused into a network or are enclosed in
calcareous cement. Projecting spicules often give these sponges a bristly texture.
Development of tissues in sponges
Sponges possess tissues (aggregates of cells) but few or no obvious organs. Contractile muscle cells called
myocytes, along with the epidermal cells which may be contractile, permit limited movement and muscle cells may form
a sphincter around the osculum which may close or open the osculum. This sphincter and the choanocyte chambers
are the closest thing sponges have to 'organs'. In general they are said to be at the tissue-grade level of organisation
and lack the complex organ systems that make up most animals. They do form definite structures, however, for
example, long spicules may project to form an oscular fringe surrounding the osculum and a sieve membrane or sieve
plate may cover the osculum. The alien-ness of sponges compared to other animals will become clear after studying
aspects of their physiology below. Some biologists argue that sponges do not possess tissues and are at the 'cellular
stage' of organisation. Though the central mesenchyme contains wandering cells, rather than cell aggregates as
such, this is very similar to the connective tissue of vertebrates. Also the outer covering layer of cells form an
epidermis (lacking a definite basement membrane), and the inner choanocyte layer form a choanoderm (also lacking
a definite basement membrane). A dermis is a type of tissue. It may well be that these cell layers are not connected
by electrical gap junctions, as in many animals, but some sponges possess syncytia - fusions of cells into a large
multicellular mass, e.g. the meshlike tissue of glass sponges (hexactinellids) is syncytial. Indeed, many biologists use
the word 'tissue' to describe sponge material. Personally, I consider sponges to be made up of tissues, albeit tissues
in which the cells are mostly only loosely integrated.
Physiology – Movement
Adult sponges are immotile and exhibit limited movement. The only movements exhibited by some sponges are
changes in the porocytes which may open or close the incurrent pores. Most sponges, however, are capable of local
or general contraction due to forces produced by the pinacocytes, desmacytes or the myocytes. Such contractions
are especially noticeable in some ascons. Most calcareous sponges exhibit slight or no contractility. Many
Desmospongiae can contract their whole surface as a result of contractions of fibre cells in the cortex and along the
main channels. Sphincters control the aperture sizes of the oscula. Dermal pores can open and close as a result of
sphincter action or changes in the encircling cytoplasm, except in glass sponges in which the pores appear non-
contractile but remain open. Indeed, most glass sponges seem incapable of movement.
Physiology – Sensitivity
Sponges have no nervous system and no specialised sensory cells. Disturbances result in general body contraction.
Oscula close on exposure to air, in response to injury, lack of oxygen, chemical irritation, extremes of temperature
and sometimes also to touch. The oscula also close when the sponge is in a small volume of water that is inadequate
for waste removal. The dermal pores are less responsive, but close on injury to the sponge. Apparently neither of the
behaviours of oscula, dermal pores and choanocytes is correlated (?). Reactions are slow, taking place over one to
several minutes. Obnoxious stimuli are transmitted either not at all or at up to 3-4 mm maximum.
The osculum is the most conductive area, especially in the direction away from the osculum, and the oscular rim is the
most sensitive part. In the freshwater sponge Ephydatia each osculum is borne on the summit of a chimney-like tube.
If the oscular rim is stimulated, a signal is transmitted down the chimney causing the chimney to contract or collapse.
In glass sponges, the cells are connected to one another as a syncytium to form cobweb-like meshes and electrical
signals can travel quite rapidly from one part of the sponge to another, presumably via the cytoplasmic bridges
connecting the 'cells' together.
Physiology – Canal System
In the well-studied sponge Leucandra, the speed of the oscula excurrent has been measured at 8.5 cm per second,
and travels as a jet for up to 25-50 cm. A Leucandra sponge 10 cm high and 1 cm in diameter has some 2.25 million
flagellated chambers and filters 22.5 litres of water per day. On this basis a large sponge would filter about 100 litres
per day. The flow rate inside the flagellated chambers is very slow at about 0.01 mm per second due to a slight
pressure gradient regulated by the sizes of the oscula apertures. This slow flow rate gives sufficient time for the
filtration process, which removes food particles from the water.
The flagella do not beat in coordination. Each undergoes a spiral undulation that travels from the base to the tip and
ensures that the water current travels in a single direction. The collars point towards the apopyle (the opening of the
flagellated chamber to the excurrent canal in non-asconoid sponges with complex architecture) and the resulting
current exits through the larger aperture. Smaller flagellated chambers are more efficient since they permit less
In leuconoids the water current is very slow in the flagellated chambers, allowing time for filtering and exchange of
materials to occur. The current then speeds up as it leaves the sponge.
Physiology – Nutrition
Sponges are filter feeders and sieve out microorganisms and organic debris from the water flowing through the
flagellated chambers. Dissolved nutrients are possibly also taken up directly by the sponge (?). Food particles in the
water current adsorb to or are filtered by the choanocyte collars and ingested into the cytoplasm. The microvilli of the
collars of all sponges seem to be connected to one another by strands of mucus or glycoprotein, forming a very fine
mesh capable of filtering bacteria from the water. In the large choanocytes the particles maybe wholly digested, but
are usually either partially digested or passed straight onto the amoebocytes of the mesenchyme for intracellular
digestion. The smaller choanocytes of other sponges possibly have no digestive role but pass food directly onto the
amoebocytes (?). It is also possible that amoebocytes receive some particles absorbed directly through the walls of
the incurrent passages (?). As they are motile (in all but glass sponges) the amoebocytes may transport food around
the sponge body and pass it on to other cells (?). The amoebocytes eject waste that leaves the sponge via the
excurrents. Other cell types also appear capable of capturing and ingesting food particles directly, should they come
into contact with them (such as the cells lining the canals). The collar units and choanoblasts of glass sponges are
continuous with the trabecular (mesh-like or web-like) syncytium that forms the bulk of the sponge and nutrients may
pass directly from the collar units to the rest of the sponge tissue via the connecting cytoplasmic bridges. Sponges
can also absorb dissolved nutrients directly from the sea water, though in many sponges bacteria constitute the main
food item, and some sponges can live entirely off bacteria.
Physiology – Respiration
Sponges are aerobic. Sycon consumes 0.16 – 0.04 cubic cm of oxygen per gram of fresh-weight per hour, larger
specimens consuming less oxygen per unit mass than smaller ones, as is the case for animals generally. Oxygen
consumption is reduced by 80% when the oscula are closed and this is compensated for by supernormal oxygen
consumption when the oscula reopen.
Physiology – Excretion
Ammonia is excreted and waste-laden amoebocytes are possibly discharged (?). Gland cells may put out long
strands to the sponge surface and may secrete slime. Amoebocytes also secrete slime. Some sponges exude mucus
or slime that has an unpleasant odour and may cause irritation to the skin on contact.
Many animals live on or in sponges, exploiting their sessile habit. Coelenterates, bryozoans and barnacles may grow
on the surface of sponges, whilst annelids and crustaceans often inhabit the water canals. In one large specimen of
loggerhead sponge (Spheciospongia) were found 16 352 Synalpheus shrimps. Shrimps of the family Stenopidae,
especially Spongicola, live in pairs in the spongocoel of glass sponges like Euplectella and Hyalonema. The shrimps
enter the sponge when young and grow such that they can no longer escape through the oscular sieve plate and
must remain trapped alive within the sponge.
Sponges of the family Suberitidae grow on the snail shells inhabited by hermit crabs. The sponge grows to enclose
the shell, the shell is dissolved and the crab occupies a spiral cavity in the sponge, which is lined by smooth fibrous
tissue. The larvae of this sponge possibly only grow when they land upon a shell containing a hermit crab (?).
Freshwater sponges (and marine?) may be parasitised by mites. The female mites lay eggs in the sponge’s tissue.
Crabs, such as Dromia and others, break off pieces of living sponge (and other objects) and hold them over their
backs with their last pair of legs, for camouflage. Some crabs, e.g. from the family Majidae, stick sponges, algae and
hydroids onto their back and legs with an adhesive secretion. These implanted organisms may continue to grow.
Sponges are seldom eaten since they are prickly and have a bad taste/odour. Some crustacean inhabitants feed on
sponge tissue as parasites. Some nudibranchs (sea slugs and their kin) also eat sponges. Sponges may smother and
kill other sessile animals, as they grow, including oysters. Boring sponges may kill barnacles and other shelled
animals by slowly boring through their shells.
Large horny sponges live for 50+ years. Smaller sponges live for several months or years. Freshwater sponges are
Different sponge morphologies suite different waters. Sponges in calm or silty water have elevated oscula. Sponges
that live in rough waters are low and encrusting. Flabellate (fan-shaped) and lamellate (vertical and sheet-like)
sponges thrive in water with a constant current, and grow with their inhalent surface facing the current and their
oscula point downstream. Deep-water sponges living on muddy bottoms have anchoring root tufts or long projections
of spicules. Most marine sponges live in shallow water, from the tidal zone down to 50-m depth, and do not live in
brackish water. The glass sponges are deep-water sponges. Some desmosponges live in the deep ocean oozes.
Spongin skeletons prevail in tropical / subtropical waters, whilst mineral skeletons (of calcium or silica) are more
abundant in cold waters.
Asexual Reproduction: Some sponges constrict off the ends of branches, which then round up into a ball and
regenerate. This may occur regularly or under adverse conditions. All freshwater, and some marine, sponges form
gemmules. A gemmule is a group of amoebocytes (archaeocytes?) with other amoebocytes forming a covering layer
(of columnar cells) that secretes thick and hard inner and outer membranes. The amoebocytes receive glyco- or
lipoprotein food reserves from special nurse cells (trophocytes). Scleroblasts may deposit spicules between the
membranes. The result is a round ball of archaeocytes, trophocytes and columnar cells, with a micropyle (pore) outlet.
The freshwater sponges produce gemmules in autumn and then die. The gemmules are able to resist winter drying
and freezing and hatch in spring. Upon hatching, cells stream out through the micropyle and form a young sponge in
about one week.
In marine gemmule producing sponges, the gemmule consists of an amoebocyte aggregation (archaeocytes?) with a
flagellated columnar dermal covering and the gemmule becomes a free-living flagellated larva that swims as the
flagella beat. The larva attaches near its posterior non-flagellated pole, loses its flagella, and develops into a young
Sexual Reproduction: All sponges can reproduce sexually. Ova and spermatozoa are produced, possibly from
archaeocytes, other amoebocytes or from choanocytes (?). The egg mother cell (ovocyte) is an enlarged
amoebocyte that grows either by engulfing other amoebocytes or from nutrients supplied to it by trophocytes. And
then undergoes maturation divisions (the cell divides into daughter cells that mature into egg cells (ova)).
Each sperm mother cell is an enlarged amoebocyte, or possibly a transformed choanocyte since whole flagellated
chambers have been observed to transform into spermatozoa. The sperm mother cell becomes covered by one or
more flattened covering cells, which derive from divisions of the sperm mother cell or from other amoebocytes. The
whole structure is called a spermatocyst. The enclosed spermatogonium undergoes 2-3 divisions to produce
The sperm are ejected in the out-going water current, often forming milky, smoke-like jets that may extend for several
metres above the oscular vent. In most species it appears that the eggs are also ejected, in which case fertilisation
occurs in the external water, but in some species, the sperm enter the recipient sponge in the water current and
fertilisation is internal. In the Calcarea (calcareous sponges), the sperm enters a choanocyte nurse cell. The nurse
cell then fuses with the egg, releasing the contained sperm. In the species Reniera the sperm enters an amoebocyte,
which transfers the sperm to the egg.
The fertilised egg undergoes unequal holoblastic cleavage to produce a blastula (a hollow ball of cells), inside the
recipient sponge. The blastula is flagellated and exits the sponge via the osculum. This larva swims for several hours
before attaching and developing into a sponge.
Most sponges are hermaphroditic (they produce both eggs and sperm), but some are dioecious (have separate
sexes). Hermaphrodite individuals often produce eggs and sperm at different times, whilst some produce sperm
apically and eggs basally. Shallow water forms reproduce seasonally.
Individuality in sponges
Many sponges are colonial - many tubes, each bearing an osculum vent may be fused together at their bases to form
a single mass. It may be that each unit, comprising a single osculum and surrounding tissue is one individual and that
one sponge produced new individuals that failed to separate completely. On the other hand, it is known that some
sponges if placed next to one another will fuse together to form a single individual, whilst others may fuse temporarily
before rejecting one another, in which case a space (zone of non-coalescence) appears between them. It turns out
that usually only sponges of the same species and the same strain will fuse together. A sponge will fuse with pieces of
itself - if a piece of tissue is removed and then placed near to the source sponge, or inside a hole cut into the sponge
(an autograft), then the fragment will fuse with its parent body. If a graft (say a cube of tissue) from another sponge is
placed inside a hole cut into the recipient sponge, then only if the graft was from another sponge of the same strain will
it fuse with the recipient. If the graft is from a different strain, even one of the same species, then rejection will occur -
the graft will not fuse and may become black and necrotic and shrink away as tissue from the recipient grows into the
wound to replace the rejected graft. Other times the graft will grow and enlarge at the expense of the recipient sponge.
Biochemicals, called sponge factors, are known to be secreted by sponge tissues when a graft is introduced, if the
strains are incompatible, then these factors will reduce the adhesiveness of the cells to one another and the graft will
be unable to stick and fuse to the recipient. Whether we define an individual sponge as a physically separate sponge
(even if one grew from a fragment of the other) or whether we define the compatible strain as the individual is a matter
Sponges have very high regenerative powers. Any piece can regenerate into a whole sponge, but the process is
slow, requiring months or years for the new sponge to reach full size. If a sponge is broken into cells and cell-clumps,
then the amoebocytes aggregate to form a reunition mass. Some reunition masses contain collar cells without collars
and various types of amoebocyte. Some of the amoebocytes form an epidermis and a whole sponge is reformed.
Reunition masses composed entirely of choanocytes cannot reform a sponge. Cells from different species, mixed
together, may temporarily form a reunition mass before separating.
In adverse conditions, many marine sponges and freshwater sponges, collapse and disintegrate to leave a reduction
body remnant comprised of a covering epidermis and an internal amoebocyte mass with partially de-differentiated
choanocytes. This will grow into a sponge when and if favourable conditions return.
A peculiar type of sponge that is little known and little studies is the infaunal sponge that lives mostly buried in soft
sediment at the bottom of the sea or below reef slopes on the continental margin. For example, the model sponge
shown below (click thumbnails to enlarge) is similar to species of Oceanapia consists of a dark brown spheroidal
central bulb, some 6 cm by 4 cm, buried beneath 5-10 cm of sediment. From this bulb extend 4-9 tubes or siphons,
6-28 cm in length and about 1 cm in diameter and whose white tips extend 3-8 cm above the sediment surface. In
Oceanapia peltata, found off the Colombian coast, these siphons may bear a number of stacked horizontal discs
partially enclosing them (not shown). Also from the bulb extend a number of tubular 'roots' which penetrate more or
less vertically into the sediment to a distance of several cm. The description given is intended to assist your
imagination, but infaunal sponges come in a diverse variety of shapes and forms. Experiments with Oceanapia have
shown that the tips of the siphons are porous and draw water in to a series of channels that permeate the non-porous
bulb. Food particles are filtered from the water and then the water is expelled into the sediment by the root-like
excurrent tubes (which bear pores of an uncertain nature as it is hard to collect the specimens without damaging the
roots). In life, these sponges are firm and somewhat elastic, but they become brittle when dry. The secret nature of
these sponges and the technical difficulties involved in collecting intact specimens has hampered the study of these
sponges. Remember, that despite their appearance, texture and limited movements, sponges are animals!
Above: three tubular demosponges, one of which is ejecting sperm in
the excurrent jetting from its osculum (a 'smoking sponge'). These
sponges may be 1.5 m or more in height and can eject sperm for
several metres. Sponges that live rooted in soft sediment benefit
from being tall as this prevents them being easily buried by shifting
sediments. Click image to enlarge.
The shapes of sponges
Sponges appear in a huge variety of forms: spherical, conical, club-shaped, vase-like, tubular, goblet or cup-like,
encrusting sheets, upright sheets, plate-like, fans and treelike forms, etc. Sometimes individuals of the same species,
but growing in different locations, may show very different shapes, although each species tends to adopt a particular
range of shapes. There are several physical reasons why sponges are the shapes they are:
1. Treelike forms branch in such a way as to maximise the filtration of water. Just as an oak tree has to maximise the
light it intercepts and the carbon dioxide it removes from the atmosphere, so a sponge has to maximise the efficiency
with it extracts food from water by filtration. If two branches of an oak tree are too close together, then their leaves may
overlap and the shaded leaves may receive insufficient light for photosynthesis and they are a waste to the plant.
Additionally, the leaves of one branch absorb carbon dioxide from the atmosphere and create a zone of air around the
branch depleted in carbon dioxide which is only slowly replaced by diffusion in slow air, but is replaced more quickly in
windy conditions. With two branches too close together the branches will compete for carbon dioxide and most likely
there will not be enough to go around. This would be the case if the crown was a solid ball with no branches - all parts
of the crown would be competing for insufficient resources and this would be wasteful. To overcome this problem, tress
branch according to one of several optimum branching patterns (the exact solution used depends upon species, which
is one reason why different species have different branching patterns) which spreads the branches apart so that they
do not compete too much with one another and the tree can maximise its exploitation of light and carbon dioxide
without wasting resources (like wood).
It is possible to use a computer to solve the mathematical problem of supply and predict the optimum branching
patterns. This has been done for sponges (using the diffusion equation to model the diffusion of food particles into the
volume of water around the sponge that the sponge can exploit by sucking this water through its body). The
assumption is that sponge tissue will not grow into regions of water where it finds insufficient food (such as too close to
other sponge tissue which have already removed the food from the water). These models correctly predict many
sponge-like shapes, including the branching tree-like forms. One also has to consider whether or not the sponge lives
in the open where there are strong currents to bring in fresh food rapidly, or whether it lives in a sheltered region of
more stagnant water.
Another factor which effects sponge shape is the diameter of supply. This diameter is the furthest that a sponge can
eject water from itself before some of that water recirculates and gets taken up again by the sponge, assuming
stagnant conditions. Ideally, the sponge would take up fresh water without mixing in any of the water it has already
filtered - filtered water has had the food and much of the oxygen removed from it by the sponge and is now of little use
and hence the further the sponge can eject it, the better, as then it is less likely to return or is thoroughly mixed and
diluted with fresh water. In strong currents, this is not a problem, the currents carry away the waste water and bring in
fresh water. Such a sponge may be an encrusting film that coats the surface of a rock and puts up very short
chimneys, each with an osculum on top. In more stagnant water, however, the sponge may have taller chimneys
through which to eject its waste, which may help the ejected water reach the currents higher above the surface (a
boundary layer of stationary or slow-moving fluid always covers the surface of objects, and the depth of this boundary
layer increases in more stagnant conditions and reduces in strong currents; above the boundary layer is the turbulent
layer where currents mix the water). Factories use the same principle - taller chimneys carry the waste higher into the
winds where they carried further away.
A third factor determining sponge shape is mechanical. A tall fan-shaped sponge may be swept away in strong
currents, whereas a slender conical sponge or a low-lying encrusting sponge stuck to the rocks over a large surface,
are more likely to survive. On the other hand, in milder currents, the flat encrusting form may be trapped inside the
boundary layer, away from the currents higher up that bring in fresh water, oxygen and food.
Different species are best adapted to different conditions, however, the fact that sponges can alter their shape to
some degree proves that sponges are sensitive creatures - able to respond to their environment and grow to a shape
that best suites the environment. For example, the sponge Hymeniacidon perleve grows as a thin film that encrusts the
surface and puts up a number of very short chimneys, but sometimes it can grow tall chimneys, which may branch or
fuse into a single large cluster of chimneys (for greater strength allowing them to grow taller). Presumably, these
different forms are adaptations to the local environment. Other sponges may grow on elongated stalks, enabling them
to reach the currents in the turbulent zone and perhaps to avoid sucking up silt from the sediment, which may clog
Left: a model of a goblet sponge -
several species adopt a cup-like form,
complete with stalk and attachment
disc. This is one example of an open
sponge architecture, in which the
outflowing water enters a common
spongocoel chamber before being
expelled through the osculum. Other
sponges have a solid architecture, in
which canals permeate the whole body
and coalesce to open directly into one
or more oscula. Click the image to
Above: the principle types of sponge internal architecture. Asconoid sponges are the simplest and smallest and probably the most
ancient in design. The whole inner surface of the asconoid is lined by choanocytes (the choanoderm) as indicated in red. The syconoid
type represents an array of asconoid types arranged around the cylinder of the sponge (the incurrent pores are not shown, but the red
layer is porous in all these sponge types). In type 2 syconoid sponges an additional layer of dermal tissue, perforated by large pores,
covers the structure. The leuconoid type is the most advanced and consists of an array of syconoid units arranged around a cylinder.
Only the layers shown in red, the choanocyte chambers, bear choanocytes, which line the inside only of each chamber. Syconoid and
leuconoid sponges necessarily have a more complex system of pores and channels to convey water through the sponge. The
arrangement is such that water only flows through one choanocyte chamber on its passage through the sponge, otherwise energy would
be wasted in filtering water that has already been filtered. The more complex syconoid and leuconoid architectures are more efficient and
so are able to achieve greater size - an asconoid becomes increasingly less efficient at larger size. Asconoid sponges are small, and
rarely exceed 10 cm in height, but the more complex and efficient leuconoid sponges may reach two metres in height. The architectures
shown above are all open architectures, in which water exits through a large central spongocoel cavity. Many sponges have solid
architectures, in which a series of channels coalesce to expel water through one or more oscula and the large central spongocoel is
essentially absent. Leuconoid sponges are the most common and their architecture may be complicated, such that the sponge's tissue
becomes permeated by a complex maze of water channels.
One sponge or three? Click image to enlarge.
Above: a glass sponge.
Typical examples of spicule (sclerite) types are illustrated below with 3D computer-generated models:
Six-rayed triaxon (6 points and 3 axes)
Tetraxon (4 axes/rays)
Triaxon (3 rays/axes)
Diaxon (2 rays/axes)
Monaxon (one ray/axis)
An open cubical cage. Several such
subunits may be joined into a larger
The sponge spicule skeleton is a truly
remarkable structure that ought to be the
envy of engineers. It is interesting to look
at sections of sponges and see how the
various spicules fit together like mechano
into diverse large-scale meshworks
designed to take the weight of the animal,
give its tissues hardness and protect it
against predators, and all with minimum
material costs and light-weight
construction. Megascleres are larger
and often mesh together to form the bulk
skeleton; microscleres are smaller and
generally free in the tissues, e.g.
microcalthrops - a small calthrops.
Triaenes are tetraxons with one long ray,
called a rhsbdome and 3 smaller clads
that form the cap-like cladome.
Pentaradiate (5 points and 3 axes)
Tetraxon (calthrops - equal rays)