A meeting in London on the 8th and 9th July (ESPA Complexity Workshop) brought together mathematicians, environmentalists, development experts and computer scientists. They discussed how studies aimed at reducing poverty through improved management of ecosystems could (and should) incorporate ideas from complexity science, like tipping points. My talk " Engaging with complexity science in ecosystem services: what is the value-added of new empirical approaches?" focused on how we have been trying to do this in China. The bleak message is that agricultural intensification over the past 2-3 decades in eastern China has driven some ecosystems to the point of collapse, particularly lakes. In long records it's possible to see the gradual loss of resilience in the environment to withstand the impact of climate, land use changes and urbanisation. There's even evidence that crop yields are declining. It seems the very basis of sustainable agriculture and fisheries is being rapidly eroded. You can see the slides here.