Most studies considering optimal control of invading species have taken the perspective of a single decision-maker. However, invasions are rarely confined to a single management unit, but rather spread across property boundaries. In this study we examined the effect of land divisions among multiple managers on the spread of invasive species, and compared the predictions of a simple analytical model to case study data.

Our model assumes a) an invasive plant spreads at a constant rate across a linear strip of land, b) constant marginal damages of the invasion, and c) increasing marginal control costs. We compared the control choices of managers and the spread rate of the species under three management scenarios: 1) a single manager; 2) multiple managers in the absence of cooperation; and 3) multiple managers with cooperation between managers.

We compared model results to interview data of ranchers (n=40) in California’s Sierra Nevada foothills. Our survey area is managed by many managers who individually decide how much to control yellow starthistle (Centauria solstitialis), an invasive plant with adverse economic and ecological consequences. Because ranchers manage much of the current and potential range of yellow starthistle in California, their control decisions affect its rate and extent of spread.

Results/Conclusions

Our model shows that dividing an equal-sized area among greater numbers of managers leads to faster invasions for two reasons. First, each manager only considers damages on his own land when making control decisions, reducing his likelihood for control. Second, if one manager opts not to control the invasion, his neighbors’ control costs increase due to the constant source for reinvasion, thus reducing his neighbors' likelihood to control. In this way, division of land promotes invasions relative to the single manager case. However, results also show that cooperation among managers helps to mitigate these effects.

How do ranchers respond to yellow starthistle spread across the landscape? Three quarters of interviewed ranchers explained that spread across property boundaries directly, negatively affects them. One quarter of interviewees explained that they control less than they would if they didn’t have a neighboring source because reinvasion increases their control costs. Two thirds of interviewees stated the need for a coordinated control effort in order for individuals’ control efforts to be successful, and some interviewees had developed cooperative weed control solutions with their neighbors. These interview results provide evidence for our modeling outcomes and support the importance of social-geographic factors for influencing invasion rates.