YES, sharks do attack humans - but it is important to put such relatively rare events in perspective and to realise that sharks do not single out the human species as the object of a vicious predetermined vendetta. The interaction between sharks and humans is the same as that between sharks and any other floating or swimming object that attracts their attention. But human self-centredness rarely appreciates this. If a shark bites and eats a seal, that is fine (except for the seal). If a shark eats another shark that is even better. But if a shark bites a human, the shark is promptly branded as a savage killer deliberately attempting to slaughter peaceful bathers and surfers who are just enjoying themselves in the sea. It is also quite acceptable for us to eat shark meat, but not for the shark to eat human flesh.

Shark attacks are rare, and only some 30% of the people bitten die. Death is mostly due to loss of blood and shock, which can result in the victims drowning before rescue. Instances of a shark eating any substantial portion of a human prey are very few. It seems that sharks actually do not like human flesh. Most bites, as might be expected, are on the lower part of the legs, from the knee downwards, thighs are the next most vulnerable part of the body, followed by the arms, middle of the trunk, chest, back, shoulders. Least attacked of all is the head. In other words, most attacks are directed to the rear part of the swimmer. The hands and arms seem mostly to be bitten when the victim tries to fend off the shark. The most frequent wounds are deep lacerations, as if the taste of the preliminary bite is enough to tell the shark it has chosen the wrong sort of food. It is rare for much human flesh or appendages to be swallowed, and even rarer for a large part, or the whole, of a corpse to be ingested.

Often the shark will take an investigative bite, or circle slowly maybe bumping the victim. But mostly once the shark has bitten it will let go as it has discovered that it doesn't like the taste of human flesh or fibreglass as pictured right.

Most shark attacks occur in warm waters at temperatures of about 70ºF (21ºC). Some authorities believe this is to be because the sharks are commoner, and more active at these temperatures. Other authorities argue that it is because there are more people in the sea in warmer climates. Both, of course, make sense. Throughout the world, shark incidents are relatively numerous off east and south east Australia and the Pacific islands; the South African coast; the Caribbean and both seaboards of the USA; the Persian Gulf; and with decreasing frequency, south and east Asia, the Mediterranean and South America. The distribution of shark attacks must, though be interpreted with great care. For example, there are very many more reports from the eastern coast of South Africa than from the western coast. This is probably because the western coast is lapped by the cold Benguela current, so fewer people are in the sea. Off the west coast of the USA most attacks are close to large centres of population because (a) there are more swimmers, and (b) the attacks are more likely to be reported. Very few attacks have been recorded from South America, probably because attacks in isolated areas are not reported or because reports in non-European languages are less accessible. Overall, the reported locations of shark attack show little more than where there are concentrations of European-language speakers in the sea and not where there is an abundance of particularly inquisitive sharks.

The timing of shark attacks throughout the year varies from place to place and leads to what are again obvious conclusions. For instance, around New Zealand no attacks have been reported between April and November. That span of months represents New Zealand winter and few people go swimming in the cold seas. Similarly on the east coast of the USA about 80% of shark attacks occur during the five warmest months (May to September) of the year. On a smaller scale, there are more attacks at weekends than on weekdays. During the day, there is a late morning peak, a trough in attacks at lunchtime, and another peak between 2-4pm. Surveys of the human population show that these times are again when most bathers are in the water.

Nearly two thirds of attack happen in water less than 1.5m (5ft) deep, while only 1% take place in water more than 45m (150ft) deep. Also one third of attacks occur less than 15m (50) from the shore and just 3% from 45-60m (150-200ft) out. All that these statistics reflect is the abundance of people. More people are in shallow water near the shore than in deep offshore water.

If a person is determined to swim in tropical waters, what can be done to minimise the risk of shark attack? In general, the best 'protection' for a swimmer is to find out where the sharks are densest and avoid that area. For example, sharks often congregate around sewage outfalls or where offal is regularly discharged into the water. Avoid areas where people are fishing - the bait may have already attracted sharks. Swim in groups, the larger the number of people in a group, the less chance of any individual being attacked. Keep clear of areas where the depth changes rapidly - i.e. near a deep channel. Keep an eye open for any signs of unusual behaviour in the fishes of the area. It is advisable not to swim with seals or sea lions. Avoid wearing bright or reflective garments or decorations that could arouse the sharks curiosity. Do not become paranoid about being bitten by a shark: the risk is infinitesimally small. Remember: sharks do not really like the taste of human flesh. For the really paranoid the only safe way to avoid becoming a shark's snack is to stay on the beach. There are no recorded attacks on dry land.