The Tumut Fragmentation Study is a relatively rare beast in the world of ecological research – it's a large-scale, long-term ecological research programme that incorporates observational studies that have run for over a decade, experimental manipulations and integrated demographic and genetic research. The research programme has provided a platform for a range of simulation modelling studies. The Tumut Fragmentation Study has also served as a wonderful test bed for a broad range of ecological theory and a platform for modelling. And it all began with a glance out of a plane window.

In the early 1990s, I was flying to Melbourne from Canberra when I saw a fascinating landscape near the timber town of Tumut in southern New South Wales, south-eastern Australia. I saw vast areas of land in which the native forest had been cleared for pine plantation, and set in this plantation were numerous pockets of native forest that were now ‘islands in a sea of pines’. The exciting thing was that these ‘islands’ were not the unwanted parts of the landscape that you often find in pine plantations where, for example, native remnants are only on rocky hills. Nor were these remnants only representing riparian areas. They were, in fact, patches of forest of the same type as the nearby intact native forest.

And from this observation immediately sprang a series of questions: How does the biodiversity of these native remnants compare with the biodiversity of the intact native forest?