¿ìè¶ÌÊÓƵ

A CORAL REEF is the biggest and most spectacular structure made by living
things. Yet although it looks permanent and indestructible, it is neither.
On a geological timescale, reefs are at the mercy of global climatic changes:
continual changes in sea level, in the patterns of circulation on the sea’s
surface and in temperature. On shorter timescales, reef-building corals
have been devastated repeatedly by regional events, both environmental –
cyclones for instance – and biological, such as attack by crown-of-thorns
starfish. In the shortest timescale, human activity can be just as damaging:
small increases in the amounts of nutrients in the water are potentially
lethal to most reef-building corals.

The organisms primarily responsible for building modern reefs are hermatypic
corals (from the Greek, herma=reef) and the microscopic photosynthetic algae
that live symbiotically in their tissues. These algae, dinoflagellates commonly
called zooxanthellae, are essential to the existence of coral reefs. But
there were symbioses long before there were reefs, and reefs long before
there were hermatypic corals. Reefs of all ages and types have probably
needed such partnerships: the primary builder and a photosynthetic symbiont
that enhances the process of calcification – either zooxanthellae or, perhaps,
cyanobacteria (blue-green algae).

The reef-building corals of today belong to a group called the Scleractinia,
the stony corals; the earliest representatives of this group were reef builders.
Before them, reefs were the preserve of tabulate and rugose corals and huge
sponge-like stromatolites. We do not know if these extinct groups were symbiotic,
but it seems likely that at least those that formed true reefs had algal
partners. The reason? Simply so that reefs could grow fast enough to overcome
the climatic and biological forces of change and destruction.

Possession of algae benefits modern corals in three ways. First, it
allows the hermatypic corals – and only them – with their very large numbers
of small polyps, to create the skeletal architectures that will withstand
the immense forces of ocean waves. Other types of corals, in contrast, generally
have larger polyps designed primarily for capturing prey; in their case
colonialism is relatively less important than survival of the individual.
Secondly, the algae export much of the carbon they fix by photosynthesis
to their host, helping the coral to grow and respire. Finally, the hermatypic
corals can take up inorganic nutrients, such as nitrogen (as ammonium),
from the environment, whereas corals without symbionts cannot. This is an
enormous advantage in the ‘nutrient deserts’ of tropical oceans. Clearly,
the biochemistry at the interface of host and alga is the key to understanding
the whole spectacular phenomenon of coral reef formation.

For the algae, the benefits of living inside a coral are less obvious.
One advantage is that coral tissue is rich in nutrients compared with the
environment. Many nutrients are present in sea water only at very low concentrations;
they may be orders of magnitude higher in coral tissue. Phosphate and ammonium,
for example, are usually present in reef waters at extremely low concentrations;
estimates put the concentration of ammonium in the host’s cytoplasm at between
five and 50 times as high. We recently found that the concentration of phosphate
in the cytoplasm at of host animals is more than a thousand times as high
as that of the water. The algae, then, are living in nutrient-rich niches
within a nutrient-poor environment. The concentrations of nutrients within
the host are also relatively stable; outside, free-living algae have to
contend with large fluctuations in supply.

Symbioses involving algae are not exclusive to hermatypic corals. An
enormous range of other shallow-water coelenterates, including soft corals
and representatives of most other important groups, have algal partners.
So, too, do sponges and many molluscs that live on reefs, most notably the
giant clams. The symbionts, algae and bacteria, are equally varied. The
nature of the relationship depends on the host: there is a fundamental difference
in the organisation of symbiosis between the corals and their algae and
the giant clams and theirs. In all coelenterate hosts, the algae live within
the endodermal cells, each enclosed within a ‘perialgal vesicle’, a structure
analogous to a normal coelenterate digestive vacuole. In clams, the algae
live outside the cells, in the haemal sinuses of the mantle, and lack these
extra membrane layers.

Freshwater coelenterates also harbour symbiotic algae. The symbionts
in these cases are Chlorella, a green alga, and the most familiar host is
the tiny Hydra. These green hydra symbioses are organised in a similar way
to those of hermatypic corals and soft corals. Largely due to the efforts
of David Smith’s group at the University of Oxford, the biochemical interactions
between Hydra and Chlorella are much better understood than are those of
any marine symbioses; and we may learn much from analogy.

Zooxanthellae grow much faster in culture, outside their hosts, than
they do in symbiosis. Inside the host, the number of algae is tightly regulated,
the system operating as a kind of continuous-flow culture system, or natural
chemostat. This implies that a single factor may limit the growth of zooxanthellae
in the host. The most likely contenders are nitrogen or phosphorus because
these nutrients are thought to limit the growth of most free-living populations
of algae.

Deficiency in either nitrogen or phosphorus produces a well-defined
set of physiological responses in free-living algae. Over the past two years,
working with David Yellowlees from James Cook University of North Queensland,
we have compared the physiology of zooxanthellae freshly isolated from corals
with those grown under defined conditions in the laboratory. The results
imply that algae may be starved of phosphorus in their hosts, but they are
almost certainly not limited by nitrogen.

Not everyone agrees. Many coral biologists remain convinced that algae
in symbiosis are limited by the availability of nitrogen. This is because
many still believe that zooxanthellae take up nutrients by the ‘depletion-diffusion’
mechanism, with the algae depleting the host of nutrients, allowing nutrients
to diffuse passively into the host from the environment. According to this
view, the coral association could take up ammonium from sea water only if
the algae were starved of nitrogen, which is not true for the algae in corals.
Even in the physiological pH range, below a pH of 8, some ammonium is in
the uncharged form and so diffuses freely across membranes. So any ammonium
present in the host’s cytoplasm is available to the algae by passive processes,
which means that ammonium is not likely to be a limiting nutrient.

Phosphorus is a much better candidate: as phosphate, it is fully ionised
over the whole physiological pH range, and so cannot simply diffuse across
membranes. The process must be active. Because the membrane around each
alga is a host membrane, the carriers may be controlled by the host, and
so enable the host to regulate the supply of phosphorus. Given that symbiotic
algae show symptoms of phosphorus starvation, the implication is that the
host may directly regulate the number of algae by controlling the supply
of phosphorus. But this may be an oversimplification.

Since the early part of this century, there has been good evidence that
symbiotic coelenterates – unlike their nonsymbiotic relatives – can take
up phosphate and ammonium from the sea water around them. That evidence
suggests that uptake requires light, implying that zooxanthellae are directly
involved. Until recently, biologists thought that nutrients were taken up
by the whole coral by the ‘depletion-diffusion’ mechanism. But we have found
that this does not hold for phosphate, and is probably not the case for
ammonium. Sea water contains less than 2 parts per million of phosphate,
whereas animal cytoplasm typically contains hundreds of parts per million.
We reasoned that unless their whole biochemistry was very unusual, the cytoplasm
of corals must contain more phosphate than the surrounding water. Recently,
we showed that, as predicted, the concentrations of phosphate in coral tissue
are typical of those for other animals. So, contrary to the ‘depletion-diffusion’
hypothesis, the coral is not simply a passive partner in the process, but
must actively take up phosphate. For ammonium, the host’s involvement is
more likely to be at the level of assimilation than uptake, because, in
this case, passive uptake can and will occur.

The fact that light regulates the uptake of nutrients clearly implicates
zooxanthellae in the process. Some researchers think that the relationship
is direct; we think it is indirect, that the algae simply provide the energy
for taking up nutrients such as phosphate, and perhaps enable the animal
to assimilate ammonia by providing the necessary carbon backbone (a-ketoacids
or their precursors).

Many reef-building corals grow at a wide range of depths, so the quantity
and quality of light available to them also varies. Where light is intense,
corals may acquire almost all the energy and carbon they need from their
photosynthetic partners. Where the light is dimmer, the algae contribute
less to the host’s carbon budget, and the corals must capture some of their
food from the water around them.

The mechanism by which carbon fixed by the algae is transferred to the
coral remains unclear. Free-living unicellular algae leak carbon compounds
into the water by passive processes, though rarely to the same extent as
zooxanthellae, which can export 90 per cent of the carbon they fix to their
hosts. Theories to account for this come in many shapes and degrees of credibility.
One class of theory is based on misdirected carbohydrate storage, or misdirected
synthesis of cell wall material. The idea here is that the alga ‘thinks’
that it is exporting carbon to one of its own storage vacuoles, or that
the exported compounds are destined for use in the alga’s cell wall. The
carbohydrate storage theory is plausible, but requires that an algal storage
vesicle becomes fused with the outside of the cell membrane at some stage,
and there is no evidence for this.

More plausible hypotheses for carbon release are based on a reversal
of the normal processes of uptake prompted by the pH or ionic environment
of the vacuole that surrounds the alga. Life inside a vacuole in an animal
cell is very different from life in the sea; sea water is mildly alkaline,
but the vacuole is almost certainly acidic. This means that some of the
nutrients the algae use may be in a different ionic state in symbiosis than
in the outside environment. The acidity of the vacuole will also affect
transport processes; transport proteins usually operate at close to equilibrium
(they can run in either direction), and transport processes often involve
protons. Thus, the acidic pH of the vacuole relative to sea water may induce
‘backflow’ via a transport protein that normally functions in the ‘uptake
mode’ when the algae are free living.

The elusive host factor

A third view of carbon transport revolves around what are referred to
as ‘host factors’. The idea of host factors stems from the observation that
the addition of homogenised host tissue to isolated zooxanthellae in the
laboratory in some cases apparently stimulates the algae to release carbon
compounds. According to theory, the hypothetical ‘host factor’ regulates
the release of carbon compounds within the host. Despite years of effort
by many groups of researchers, however, ‘host factor’ remains as elusive
as ever.

The complex and unsubstantiated host-factor models of transport between
alga and coral contrast sharply with the simplicity of communication between
the green hydra and its algal partner. In green hydra, there is good evidence
that the most important product of photosynthesis to pass from alga to host
is maltose, and that maltose release is simply a function of pH. When the
green hydra’s algal partner, Chlorella, is isolated from the hydra it releases
maltose at low pH; if pH rises above about 6, the algae release little or
no maltose. As these algae, like zooxanthellae in the coral system, are
enclosed in a vacuole, this means that the host can effectively manipulate
the release of carbon compounds by controlling the pH of the vacuole. Presumably
certain enzymes associated with the outside (the animal side) of the membrane
around the alga acidify the vacuole, stimulating the release of maltose.

The structural similarity between the symbiosis of green hydra and Chlorella
and corals and their zooxanthellae suggests that they may function in a
similar fashion. The release of fixed carbon by zooxanthellae is not directly
influenced by pH, but this is not surprising in view of the physiological
differences between free-living freshwater and marine algae. Active transport
systems that are coupled directly to proton movements in freshwater organisms
are often coupled to the movements of other ions in marine algae. For example,
in free-living Chlorella, sugar transport is coupled to the movement of
protons, but sodium is generally the ion involved in marine algae. It is
our view that the release of fixed carbon involves a simple mechanism, probably
analogous to the green hydra system but involving other ions, and that the
apparent effects of ‘host factors’ will turn out to be artefacts.

Coral reefs, the self-made habitats of a simple group of animals near
the bottom of the evolutionary ladder, are the largest structures made by
living things. Grand and imposing though they are, this whole phenomenon
depends for its existence on a subtle relationship between the coral and
an even more lowly form of life, the unicellular alga. In one sense, we
can imagine the coral reef to be the animal kingdom’s greatest attempt to
displace plants from their rightful place on Earth, with corals, not plants,
being the overwhelmingly dominant macroorganisms.

Shallow tropical seas provide the world’s only suitable environment
for this. Carbon, the raw building material, is in limitless supply. Continental
margins and island archipelagos provide millions of hectares of suitable
substrate. What is missing is energy: the only limitless source of energy
to match this limitless resource is sunlight. And only symbiosis offers
the opportunity to harness the energy from light.

There are two other key aspects to the coral reef phenomenon. The first
is simple. Because the rate at which corals lay down calcium carbonate depends
on temperature, reefs must be warm enough to outgrow the forces of erosion:
they can exist only in tropical, or near-tropical, conditions. The second
is anything but simple. Like all other major tropical ecosystems, coral
reefs are subject to continual change. And, like other systems, they cope
with change by diversifying. The range of species of invertebrates on a
coral reef probably has no equal on Earth. Environmental gradients down
the slope of a reef are pronounced and sudden; the range of habitats is
enormous, and hermatypic corals have evolved to exploit them.

Symbiosis, the linking of two types of organism to achieve what neither
can achieve alone, is the outcome of long evolutionary process in a highly
competitive ecosystem. Small wonder that the range of organisms involved
is so great. We are only just beginning to understand the extent of this
most intimate association of species. We still have only the crudest understanding
of the mechanisms they employ. By applying to the coral system the full
arsenal of biochemical techniques, we are slowly beginning to understand
the biochemical basis of the coral reef.

David Miller is a biochemist at James Cook University, Townsville, Queensland.

Charlie Veron is a scientist at the Australian Institute of Marine Science.
He has written several books on Australian and Indo-Pacific corals.