So working out whether these insects follow the broad path, paved with pollinating riches - or instead strike along the straight and narrow - has big implications for the survival of the Gomortega tree. And it could help find the best way to manage this important hotchpotch of a landscape.

Connecting flights

Now a paper in the latest Current Biology has demonstrated that, when it comes to arranging 'connecting flights' between the few surviving stands of Gomortega, the broad path of flower-rich farmland can be too much of a distraction. It seems that narrow, rocky, resource-poor road may be better, if pollinating nirvana is to be achieved.

That runs counter to many previous, simpler ways of looking at how broken-up conservation areas can be stitched back together, so as to improve the flow of pollen between them. And because hoverflies are such generalist pollinators, the implications of the study is likely to spill far beyond Chile's Valdivian Forests.

Dr Dan Bebber, in charge of climate change research at Earthwatch, and co-author of the paper said: 'This study complements our forest research in the UK, where scientists from Earthwatch and the University of Oxford are investigating how moths and small mammals are using landscape corridors such as hedgerows and trees to move among habitat types. Landscape connectivity through corridors can enhance species migration, and could prevent the extinction of vulnerable species.'

Tracking the gene flow

From this they produced a map of the Gomortega's 'gene flow' - which was effectively a map of the preferred routes taken by the hoverflies. Two computer models were then used to tease out how the varied landscape influenced the choices made by the hoverflies.

What the team found was something of surprise. The land between the remaining fragments natural forests contained some areas rich in flowers and pollen - including farmland. It was thought that the hoverflies would be attracted from the forest into these - and so be drawn to other fragments of forest on the other side. But instead, the results showed that it was the resource-poor pine plantations that best helped to funnel the insects between the Gomortega trees.

The Circe Principle

The team called this type of behavior the 'Circe principle'. It reminded them of a tale from the Greek legend of Odysseus; Odysseus' crew were waylaid for months, on their journey home, by the lavish feasts laid on by the enchantress, Circe. As an approach to describe gene flow, it makes a break with previous simple ecological models.

They tend to look at the problem in black-or-white - natural landscapes are seen as 'permeable' to gene flow, whereas most man-made ones are seen as 'barriers'. But as a general principle, it does need more research. Dr. Bebber, proposes a novel way of testing this out, involving tagging of pollinating insects. 'The Circe Principle does need further validation,' he said. 'Direct measurements of pollinator movement, for example by tiny GPS tags, would be one way to achieve this.'

Not all habitats pristine

Looking at conservation in this way - modeling the complex interlayering of man-made and natural landscapes - has often been shied away, from according to Dr Bebber. 'Scientists and conservationists have tended to focus attention on 'pristine' ecosystems. Understanding natural ecosystems is hard enough: including human disturbance makes things even more difficult. However, we need to do this because most ecosystems are affected by people to some degree.'

But doing so could be positive for getting better results in saving endangered species, whether trees, butterflies or tigers. 'It is often assumed that managed habitats are just simply bad for conservation, but this is not necessarily the case. There are many different ways of managing a landscape, which provide different levels of productivity and conservation benefit. Scientists need to find out what those trade-offs are.'

Top Image: Hoverfly on blue daisy © CS