When Pseudomonas syringae pv. actinidiae (Psa-V) devastated New Zealand's kiwifruit industry in 2010, the spread pattern was puzzling. The bacterium can travel short distances via wind and rain, but the nationwide outbreak moved too fast and too far for weather alone to explain. Something was carrying it between orchards separated by hundreds of kilometers.
Rao et al. demonstrate that the missing vector is human movement. Using a high-resolution mobility network — actual travel data between New Zealand regions — they build a metapopulation model where pathogen transmission follows the routes that people drive. The model reproduces the observed spatiotemporal spread of Psa-V without invoking any mechanism beyond human-mediated dispersal.
The key finding is a structural one about the network itself. Most disease transmission is local, as every plant pathology textbook would predict. But the local-spread model fails to produce a nationwide outbreak. What bridges the gap are sporadic long-range connections — the occasional orchard worker or equipment truck moving between distant regions. Remove these rare long-distance links and the epidemic fragments into isolated local clusters. Keep them and the disease goes national.
This is the small-world property applied to agriculture. A few long-range edges in an otherwise local network can transform a contained outbreak into a catastrophe. The mathematics is the same as in human disease epidemiology (SARS, COVID), but the implementation is different: the susceptible populations are orchards, not people, and the mobility data describes workers and vehicles, not commuters.
The sensitivity analysis reveals something actionable. Outbreak severity depends sharply on when and where the initial importation occurs. Early-season introductions in high-connectivity regions produce dramatically worse outcomes than late-season introductions in peripheral ones. This asymmetry means that surveillance resources are not fungible — watching the right places at the right times matters more than watching everywhere all the time.
The seasonal labor connection is the most concerning finding. High-traffic seasons — when workers move between orchards for pollination and harvest — correlate with elevated epidemic risk. The people who make kiwifruit production possible are also the mechanism by which kiwifruit pathogens spread. The vector and the value chain are the same network.