Fig 1.
Influences on farmers’ decisions and their impacts.
A schematic illustrating the influences on farmers' decisions on what varieties of maize to grow, and how this impacts the population dynamics of the European corn borer and the profitability of farming at a landscape scale. The width of the green arrows indicates the approximate appropriation of agricultural resources.
Fig 2.
The functions used to model the dispersal of the European corn borer.
The dispersal functions for adult moths pre- and post- mating in spring and summer.
Fig 3.
Average numbers of overwintering lavae from Minnesota over time (solid black line) during a period where the proportion of Bt maize broadly increased (dashed red line). Our simulation model (solid blue line) captures the behavior observed in the field with a cycle in the population of similar wavelength to that observed in the data. The introduction of Bt maize results in this cycle being damped but still persisting (the cycle is under-damped in this case—see S2 Text).
Fig 4.
Results from the landscape-network, neighbor-network, and varying-response-network simulations.
The top pane of each pair shows the proportion of Bt maize and bottom panes show the average number of overwintering larvae per plant across the two areas of the landscape, one in Wisconsin and the other in Minnesota. The simulation was started with 1% of the maize as Bt distributed randomly in the landscape.
Fig 5.
The spatial distribution of crops and larvae in a single year of the simulation.
(a) The land use in year 73 of simulated landscape under the neighbor-network. The left half of the landscape represents Minnesota (abscissa from 0 to 350 km) and the right Wisconsin (abscissa from 350 to 700 km); (b) shows the corresponding average number of overwintering larvae per plant. Enlarged sections show the spatial distributions in more detail.
Table 1.
The average losses and the average proportion of the crop that is Bt between year 20 and 100 under each simulation according to communication network type and value of the parameter β, which changes the responsiveness of the farmer to loss.
The standard deviations are given in parentheses.
Fig 6.
The loss in profit incurred by growing conventional maize compared with growing Bt maize plotted against the percentage of maize that is Bt.
The arrow indicates the direction of time. Subplots (a) and (b) are based on data from states in the Corn Belt and subplots (c) to (e) are based on simulations.