In addition, they are looking at the genetic make-up of different species, which may point towards some unlikely relatives.

First-year PhD students at the National University of Singapore's department of biological sciences have been everywhere - from fish markets to forests - to do this.

Their discoveries could eventually lead to commercial riches because a good understanding of, say, a particular breed of a rare crab could help a farmer breed them for the seafood industry.

Miss Lai Chiu Yun, 25, and her colleagues had to battle the stench of a mountain of rotting fish in Indonesian fish ports to sift for rare treasures.

They have been lucky a few times, pulling out a newly-discovered goby and a rare elbow crab recently, finds that are more than worth the trouble.

Besides, she joked, 'how many scientists have tans like ours?'

Once they locate a rare find, it is preserved in formalin or alcohol and brought back to Singapore for further study.

In the lab, genetic fingerprinting has become part of the researcher's arsenal.

Sequencing bits of DNA from different species yields clues to how closely related the species are.

Miss Lai has used it on the flower crab, a widely-dispersed species that she believes may be made up of two or more different species.

The technique also throws up species that do not seem to have close relatives.

The mammary crab, named for the distinctive markings on its shell, is believed to be the only species in its family.

Learning more about the crabs' family trees may seem a purely academic exercise, but it has important implications for the seafood trade.

For example, Australian scientists recently found that the mud crab - long thought to be a single species - is actually four, each with very different habits and habitats.

'This helps us learn more about how to find and breed them, and which grow the biggest and fastest,' she said.

Miss Zeehan Jaafar, 25, is more interested in gobies, of which more than 2,000 species exist.

One that she dubs the 'garbage can' consists of 20 species that have been lumped together, but are not necessarily related.

She blames this on the scant knowledge about the fish.

'These fish are hardy, and you can find them all over the world.

'But they are also small and hard to catch, and we don't even know how they spawn.'

By patiently studying their appearance, structure and DNA, she is coming up with a clearer picture of which species are related, and how closely.

Because so little is known about them now, all but one or two species are trapped in the wild, not farmed. By learning more about gobies, they could be farmed commercially for the aquarium trade.

As for 27-year-old Alfredo Amiel Leonardia, it is the miniature world of moss, not fish, that fascinates him. He is examining the degree of variation in two species of mosses found in Singapore and Malaysia.

This simple and ancient group of land plants is a microcosm of the forest, he said.

Greater variation in the genetic code of a single species may mean it is less susceptible to change, and less likely to be wiped out.

'These mosses can be used as an index of how the forest is affected by destruction and degradation, and how it affects even the smallest organism,' he said.

His preliminary work in the forests of Singapore and Johor has uncovered some good news - there is still some variation of mosses over small areas.