THE tricky business of identifying insect species, far from being the hobby
of a few entomologists, is vital to agriculture and human health. In Africa in
1990, for example, it was only the rapid identification of the fly that spreads
the New World screwworm (Cochliomyia hominivorax) that enabled the
authorities to wipe out the parasite before it got a foothold.
Now scientists from the University of Sheffield and the Natural History
Museum in London have developed a computerised system that allows nonexperts and
researchers without access to specialist centres to find out exactly which bugs
they are dealing with. Applications could include not just pest monitoring, but
biodiversity inventories and environmental impact assessments say the
researchers.
The Digital Automated Identification System (DAISY) is capable of identifying
insect specimens and recognising their close relatives using only their wing
patterns. 鈥淧eople haven鈥檛 looked at wing patterns as a way of identifying
insects before,鈥� says Paul Weeks of the museum鈥檚 entomology department. 鈥淭o the
human eye it鈥檚 very difficult to see the differences. But the computer can.鈥�
Advertisement
The system was able to distinguish between five closely related parasitic
wasps of the Ichneumonidae family, which live on other insects in Costa Rica.
The system looks at the pattern of veins and the pigmentation on the insects鈥�
wings.
DAISY was 95 per cent accurate, and the researchers say this compares
favourably with other systems. But the system has an extra advantage鈥攖he
digital images can be captured on computer and transferred electronically to the
Internet.
鈥淎t the moment, there鈥檚 an enormous barrier to people in, say, the Amazon
trying to find out what species they鈥檙e dealing with,鈥� says Weeks. 鈥淏ut if we
develop the system for the Internet, they might be able to access the
颈苍蹿辞谤尘补迟颈辞苍.鈥�
He says DAISY could also allow nonspecialists to routinely and rapidly
identify insects. At the moment, many species can only be identified by a
handful of experts with access to extensive collections and libraries.
Colin Leake, a lecturer in entomology at the London School of Hygiene and
Tropical Medicine, welcomes the development. 鈥淭his is a very good effort,鈥� he
says. 鈥淭his seems to be the way forward. There are only a handful of experts in
the world who can identify some species. As these experts retire, unfortunately
they鈥檙e not being replaced.鈥� However, he warns that DAISY might have less
success in distinguishing between species of some insects like mosquitoes, which
are extremely similar.
