On December 31 last year, researchers trapping butterflies at Deep Water Bay on Hong Kong Island found a chestnut tiger with writing on its wings. In fine marker pen were numbers, letters and characters, including a code: "YSK541".
It was a thrilling find. The team from the Tai Po Environmental Association Fung Yuen Butterfly Reserve had caught hundreds of butterflies and only this one bore such writing. Checking information from elsewhere, they found it had been marked in Osaka, Japan, on October 10.
In the 83 days since, it had flown an astonishing 2,500 kilometres - the longest known journey of any butterfly species, after North America's monarch.
Monarchs are migrants that have been studied extensively. Some breed as far north as Canada, from where adults born in summer fly to Mexico, joining huge congregations. Early in spring, they leave for the southern United States to breed and die. Butterflies are born over the summer, with some moving north until members of the fourth or fifth generation reach Canada.
So do any of our local butterflies migrate in similar ways? No one knows, though there are several migratory species of tigers and crows in the danaid family.
The best information on Asian butterfly migrants is from Taiwan, where tens of thousands of purple crows travel to the south of the island for winter and north again in spring.
Some chestnut tigers marked in Taiwan have reached Japan in summer and others marked in Japan have made the reverse journey in autumn.
It is believed that relatively few chestnut tigers undertake such marathon flights, doing so only when monsoon winds are favourable. But there are annual gatherings of danaids in Hong Kong, which indicate substantial seasonal movements.
One of these is the common tiger, a close ally of the monarch. Like a monarch, its caterpillars feed on plants packing chemicals that are toxic to many creatures. It, too, has brilliant orange wings marked with black lines and white spots, warning predators to keep off. Common tigers begin gathering in October; in recent days, I've seen tens of them at Pui O and Fan Lau on Lantau.
At both sites, flowers have nectar for tigers to feast on. Favoured plants include Crotalaria, with nectar featuring a chemical the males use to make a pheromone for attracting females. Like several of Hong Kong's butterfly gathering sites, both sites are coastal: butterflies flying south or southwest in autumn might reach the sea, and then halt for days or weeks.
Later in the year, there may be mass congregations of crows - dark butterflies with white spots and blue flashes on their wings. In November, they feed on winter flowering trees. By December, when the largest assemblies form, most are focused on surviving the winter by clinging to trees sheltered from chill winds.
A favoured place near Lai Chi Kok became known as Butterfly Valley, home to "swarms" of these butterflies. The valley's woodland has been largely destroyed, but 40,000 butterflies were still counted there in the 2000-2001 winter.
Since then numbers have shrunk: Green Power surveyors found just 576 wintering butterflies at the same place last December. Global warming and developments in nearby China could be to blame.
Many butterflies may die during the winter, particularly if temperatures plunge to seven degrees Celsius or below. Yet the gatherings disperse by late January. Do most butterflies head further south or start spring migrations into south and southeast China? Again, no one knows.
Martin Williams is a Hong Kong-based writer specialising in conservation and the environment, with a PhD in physical chemistry from Cambridge University