Almost all plants grow with their roots colonised by fungi called
mycorrhizas, but the fungi are in most cases not there to help their hosts
extract nutrients from the soil, as was previously supposed, say ecologists in
Britain and Canada. Instead, they think that the main benefit derived by most
plants from the relationship is that the mycorrhizas mount a guard against
pathogenic fungi and tiny nematode worms.

Mycorrhizal fungi clearly benefit from associating with plants: botanists
have shown that sugars move from a plant’s roots to the surrounding fungus.
What plants get from the deal is less clear, but experiments indicate that the
fungi can help plants extract phosphorus from soil. But this explanation never
satisfied Alastair Fitter, an ecologist at the University of York.

Across most of Europe and North America, Fitter says, soils are rich in
phosphorus, which means that plants should have no problems getting enough of
the element, unless they have poorly developed root systems. Yet mycorrhizas
are common in these phosphorus-rich soils.

Fitter and his colleagues Kevin Newsham, now at the Institute of
Terrestrial Ecology at Abbots Ripton near Huntingdon, and Andrew Watkinson of
the University of East Anglia in Norwich, suspected that the main role of
mycorrhizal fungi is to defend their hosts against pathogenic fungi. They grew
seedlings of an East Anglian grass called Vulpia ciliata under sterile
conditions, and then planted them in sandy soil near Mildenhall in
Suffolk.

They inoculated some of the seedlings with Glomus, a mycorrhizal fungus,
some with the pathogenic fungus Fusarium, some with both, and others with
neither fungus. After 62 and 90 days, the plants inoculated with both fungi
had larger and healthier roots than those infected with Fusarium alone. And
plants inoculated with just Glomus were less likely than seedlings that were
inoculated with neither fungus to have succumbed to native pathogenic fungi
from the Mildenhall soil (Journal of Ecology, vol 83, p 991).

Similar results had already been obtained in greenhouse experiments, but
many ecologists argued that what goes on under glass may have little bearing
on what actually happens in the field. “This is the first ecological
demonstration,” says Fitter. “The case is now unequivocal.”

Meanwhile, Richard Little and Anwar Maun of the University of Western
Ontario in Canada have investigated the relationship between marram grass,
which grows on sand dunes, and its mycorrhizal fungi, which include species of
Glomus.

Marram survives better when it is continually buried by sand, and some
botanists had suggested that growing upwards into fresh sand may allow it to
escape from the nematode worms that eat its roots. But Maun was unconvinced:
“Nematodes move through sand quite easily.”

Little and Maun planted marram seedlings, again raised under sterile
conditions, in pots. The seedlings were were either inoculated with a cocktail
of mycorrhizal fungi, the root-munching nematodes Heterodera and Pratylenchus,
both fungi and nematodes, or neither. Some seedlings in each group were buried
under a 5-centimetre layer of fresh sand. But burial did little to protect the
plants from the stunting effects of nematode infestation, whereas the presence
of mycorrhizal fungi did, Little and Maun report in the latest issue of the
Journal of Ecology (vol 84, p 1).

How mycorrhizas protect their hosts from other fungi and nematodes is not
clear, but Fitter and his colleagues are planning further experiments to find
out. “It’s likely that there’s some chemical involved,” says
Watkinson.

P. mexicanum’s main host is a lizard called Sceloporus occidentalis that
lives in Northern California. Female sandflies feast on lizard blood, but
spend most of their time roosting in ground squirrels’ burrows.

Less than 2 per cent of female sandflies live to take a second meal. It is
vital that P. mexicanum completes its development inside the sandfly quickly,
so it is ready to infect a lizard if it gets the chance.

P. mexicanum can develop more quickly in the warm, and Roberto Fialho and
Jos Schall of the University of Vermont in Burlington say that flies infected
with P. mexicanum roost nearer to the mouths of the squirrels’ burrows, where
it is warmer.

They showed that the temperature preferred by sandflies increases by some
1.6°C after a blood meal. If the blood contained P. mexicanum, their
preferred temperature rose by nearly 4°C (Journal of Animal Ecology, vol
64, p 553). The parasites must produce a substance that resest the flies’
natural thermostat. “This would make for a very nice biochemistry project,”
says Schall.

Mark and Diane Littler of the Smithsonian Institution, Washington DC, and Phillip Taylor of the US National Science Foundation’s Biological Oceanography Program in Arlington, Virginia, studied the association between the chiton Choneplax lata and the alga Porolithon pachydermum, which live on a section of the 250 kilometre coral reef off Belize. In particular, they checked what happened when they removed Choneplax from its burrows in certain areas.

When the researchers returned up to two and a half years later, they found that the alga had grown better where the molluscs were present. The grazing seemed to have stimulated the growth of the alga, but also removed any minute new plants of other algae that tried to settle on it. This not only eliminated competition for light and space, but discouraged the attentions of parrotfish, which damage the underlying Porolithon when feeding on these algae (Ecology, vol 76, p 1666).

Choneplax definitely feed on Porolithon, as the researchers found that the alga makes up over half of the mollusc’s gut contents. Scanning electron microscope pictures also revealed a close match between the rasping teeth on the chiton’s radula, or tongue, and scars on the surface of the alga. However, these scars are only some 10 micrometres deep, while the really important tissues of the alga, concerned with its growth, photosynthesis and reproduction, are more than 20 micrometres below the surface, safe from damage.

The steady growth of the Belize Barrier Reef is a tribute to the success of this partnership, which provides the mollusc with a reliable source of food and a safe place to make its burrows. The same researchers have now also noted similar associations in reefs in the Pacific, off Fiji and Papua New Guinea, suggesting that this harmony between eater and eaten has a widespread ecological significance.

Neils Lindquist and Mark Hay of the University of North Carolina in Chapel Hill studied the Caribbean sea squirt Trididemnum solidum. Like most sedentary marine animals, T. solidum releases large numbers of larvae into the water, where they float around near the surface before coming to settle on a rock. The tissues of T. solidum are laced with noxious chemicals called didemnins. These are thought to protect the adults from predators, but many researchers had reasoned that the tadpole-like larvae could not contain enough of the toxin to cause plankton-feeding fish any problems – especially as they could only account for a few per cent of a fish’s total diet.

Lindquist and Hay suspected otherwise. They added the toxins to pellets of squid meat, which they fed to captive pinfish. The pellets contained only small amounts of didemnins, to mimic the dose that fish might receive if they ate the larvae in the wild.

When Lindquist and Hay gave 15 of the fish didemnin-treated pellets, all of them vomited within two hours. Pinfish rarely live in the same waters as T. solidum, so they would be unlikely to have eaten didemnins before. But three days later, all but four of the fish had learnt to avoid the treated pellets.

Because the fish cottoned on so quickly, the researchers could not use them to study the long-term effects of eating didemnins, so they switched to a less intelligent animal, the sea anemone Aiptasia pallida. The anemones came from the North Carolina coast, so again they had not encountered T. solidum.

The anemones were fed on commercial fish-food pellets, supplemented with a single pellet of squid meat each day. Half of the animals were fed squid pellets containing the amount of didemnins found in 15 T. solidum larvae. The other half received untreated pellets. Many of the anemones regurgitated the treated pellets, but none learnt to reject them. This proved costly: daily doses of didemnins reduced the anemones growth by 82 per cent, even though the tainted pellets made up just 2 per cent of their diet (Ecology, vol 76, p 1347).

Lindquist and Hay believe that wild fish learn to reject T. solidum larvae. Among animals too stupid to learn this association, the toxic effects of didemnins should soon select for the evolution of strains which instinctively avoid the larvae.

The researchers used a microscope to watch blood cells moving through the gills of anaesthetised trout and roach. The blood passes through structures called lamellae, two thin membranes held apart by cells which look like the pillars supporting a roof. Video recordings of this flow showed that red blood cells become deformed as they squeeze between the pillar cells.

Normally, the red blood cells of a trout or roach are oval measuring 13.5 by 8.4 micrometres. As the cells flow through the lamellae, however, they can stretch to more than 18 micrometres in length, and take the shape of a letter C or S. Some of the red blood cells get jammed between the pillar cells, blocking the progress of other blood cells. This means that red cells passing through the gill lamellae travel about 50 per cent farther than the shortest path (Journal of Experimental Biology, vol 198, p 1151).

This helps to explain why fish gills are so good at picking up oxygen from water. First, the deformed and jammed blood cells will create turbulence, preventing stagnant layers of blood plasma forming around each cell, a process notorious for slowing oxygen uptake. The deformation will also increase the cells’ surface area, relative to their volume. And when a red blood cell is contorted the haemoglobin molecules inside are churned up, increasing the likelihood that each will come near to the cell’s outer membrane and combine with oxygen. Finally, the slow progress of cells through the lamellae means that they get fully loaded with oxygen before moving off round the body.

Together with two Japanese scientists, they set about discovering what made the pagoda unpalatable to pests. When they tested chemicals from the type of cypress tree that was used to construct the building they found that some are so repellant that rodents will not gnaw anything coated with them (Journal of Chemical Ecology’, vol 21, p 263).

Starting with 42 kilograms of sawdust from Thujupsis dolabrata var. hondai, the researchers extracted 500 grams of oil. From this they separated various compounds, including terpenoids called thujopsene, carvacrol and betathujaplicine. They then dipped lengths of electric cable into solutions of the chemicals dissolved in ethanol, and checked whether caged mice gnawed these as much as they did cable dipped in ethanol alone. They found that the mice did not gnaw cable treated with the crude oil, and they were particularly reluctant to chew on wire coated with the three terpenoids, especially carvacrol.

Ahn and Lee say that the compounds could replace the poisons that some of the animals are becoming resistant to.

Poisons may also harm other animals or people. The anti-gnawing terpenoids should have none of these drawbacks, and the evidence from Nara suggests they may also control microorganisms and termites.

Unfortunately, the most effective compound, carvacrol, is only active for a few days, probably because it evaporates. So it will have to be mixed with a carrier that releases it slowly. The Japanese timber industry produces some 4000 tonnes of T. dolabrata sawdust each year.

If disturbed, giant tarantulas emit a hissing noise, which they make by rubbing together their first two pairs of legs and their pedipalps – smaller appendages on their heads specialised for feeding. Uetz and his colleagues suspected that the noise is made by feathery bristles carried on the spiders’ appendages, so they anaesthetised a spider with carbon dioxide and plucked off its bristles. When the spider came round, it rubbed its legs and pedipalps together as normal, but was silent.

A close examination under the electron microscope revealed why these bristles are noisy. Their tips carry a series of tiny hooks, while much of the shaft of each bristle is covered with thin filaments. When the spiders bring their opposing legs or pedipalps together, the terminal hooks of the bristles on one appendage become entangled with the middle filaments of the bristles on the other. The hissing noise is made when the entangled hooks and bristles are pulled apart (Journal of Zoology, vol 235, p 587).

Uetz and his colleagues learnt the hard way that the giant tarantula’s warning hiss is not an idle threat. If handled, the spiders shed numerous small hairs that cause itching. Other researchers have fed giant tarantulas to hand-reared coatimondis, which prey on large spiders in their native Venezuela. Although a naive coati attacks the first T. leblondi it is given, it rarely repeats this mistake. Uetz and his colleagues suggest that the hissing sound serves to jog a potential predator’s memory of its first unpleasant encounter with a giant tarantula.

This finding puts a new slant on a debate that has raged among zoologists for years. Bees seem to know how far they must fly to get to sources of food that they have visited before, and some scientists have argued that bees measure distance by evaluating the energy they have used in flying there. Others, however, have objected that this would not allow for varying wind speed and direction.

Chittka and Geiger trained bees to collect sugar solution from a feeder more than 250 metres from their hive, which was in the centre of a large featureless meadow. Then they placed a sequence of obvious landmarks – tents about 3.5 metres high – along the bees’ line of flight from the hive to the feeding station.

To start with there were four tents, spaced 75 metres apart, so the feeder was between the third and fourth landmarks. Even if sugar was also available between the second and third tents, most of the bees flew to the original feeder.

The researchers then changed the number or position of the landmarks. They set up the tents so that there were either five or six landmarks placed close together, or four landmarks which were spaced more widely than they had been previously. Although some of the bees still went to the original feeder, the number of landmarks had a definite effect. Many of the bees simply stopped at the first feeder they encountered after the third landmark, even though this meant that they were nowhere near the feeder on which they had been trained (Animal Behaviour, vol 49, p 159).

Chittka and Geiger say that bees clearly react to the number of landmarks they have passed, rather than to the distance they have flown, which means they must have the beginnings of an ability to count. Presumably, they say, bees use this ability under natural conditions, measuring the distance to patches of flowers by counting bushes or trees.

Jens Rydell of the University of Aberdeen and Raphael Arlettaz of the University of Lausanne in Switzerland found that the European free-tailed bat (Tadarida teniotis) searches for its prey using sound at a frequency of between 11 and 12 kHz, a frequency too low for most insects to hear. The disadvantage to the bats is that this frequency corresponds to a wavelength of about 3 centimetres, and since echolocation is insensitive to objects smaller than the wavelength of the sound being transmitted the bats are unlikely to detect small insects.

Rydell and Arlettaz decided to check the possibility that T. teniotis feeds mainly on relatively large insects that could not pick up its own lower-frequency pulses but could detect the echolocation pulses of most other bats (Proceedings of the Royal Society of London, Series B, vol 257, p 175). To do this, they examined bat droppings for insect remains. The researchers collected the droppings from under bat roosts in two locations: Sisteron in southeastern France and at an altitude of 850 metres in Kirghizia in central Asia.

In both sets of droppings Rydell and Arlettaz found that the majority of insect remains were of moths, or Lepidoptera – 68.3 per cent in the French collection and 86.8 per cent in the Kirghizian bats. Neuroptera such as lace-wings, which can also hear typical bats, accounted for 24 per cent of the European sample but were rare in the Asian sample, 12 per cent of which contained an unidentified insect. This is quite different from the diet of other bats, which feed mostly on flies. Moths make up less than 10 per cent of their diet.

These results clearly support the researchers’ idea that low-frequency echolocation allows T. teniotis to locate the relatively large insects, which are alerted to other bats by hearing their calls. It also explains why these bats are relatively rare. If they contributed too much to the selective pressure on the insects, they might trigger off the evolution of a wider hearing span in the moths.

T. teniotis also incurs an evolutionary cost. The low frequency of its echolocation call means that it cannot locate and eat small insects, such as flies, which are usually the more abundant than larger insects.

In the 1980s, laughing gulls (Larus atricilla) flew into New York’s
John F. Kennedy Airport from nearby marshland and started bathing and feeding
on runways and grass. The authorities were distinctly worried because a
DC-10 had crashed at the airport in 1975 after being struck by several
Canada geese (Branta canadensis).

The officials’ first idea was to kill the beetle larvae the gulls were
thought to be eating by spraying the entire airport with a pesticide. However,
P. A. Buckley of the University of Rhode Island and Molly McCarthy of Rutgers
University in New Jersey found a less dramatic solution (Journal of Applied
Ecology, vol 31, p 291).

Buckley and McCarthy marked out six experimental blocks of grassland
near the airport’s runways and taxiways. Each block had three plots of short
grass about 5 centimetres high, and three of long grass about 45 centimetres
high. The researchers counted the gulls and various insects on the blocks.

Buckley and McCarthy found that scarabs, the most common beetles, and
their larvae were not affected by the height of the grass. However, in each
block, the gulls selected the plots of short grass and avoided the long.
They also fed on the adult beetles only on the plots of short grass, even
when there were lots of beetles in the long grass. Also, the gulls found
areas of standing water particularly attractive. Because the gulls migrated
to the wildlife refuge only in April, staying until October, it was clear
that they were taking advantage of summer insects and water supplies.

So, to solve the airport’s laughing gull problem, Buckley and McCarthy
suggested making the airport unattractive to the birds. This involves simply
keeping the airport’s grass long (and planting grass in areas which are
now bare), draining off all standing water, and controlling the scarab
beetles. This last does not involve blanket spraying with insecticides,
but integrated pest management, possibly using milky spore disease (Bacillus
popilliae) or parasitic wasps.