Optimal fish passage barrier removal – Revisited

Infrastructure, such as dams, weirs and culverts, disrupt the longitudinal connectivity of rivers, causing adverse impacts on fish and other aquatic species. Improving fish passage at artificial barriers, accordingly, can be an especially effective and economical river restoration option. In this article, we propose a novel, mixed integer programing model for optimizing barrier mitigation decisions given a limited budget. Rather than simply treating barriers as being impassable or not, we consider the more general case in which barriers may be partially passable. Although this assumption normally introduces nonlinearity into the problem, we manage to formulate a linear model via the use of probability chains, a newly proposed technique from the operations research literature. Our model is noteworthy in that it can be readily implemented and solved using off-the-shelf optimization modelling software. Using a case study from the US State of Maine, we demonstrate that the model is highly efficient in comparison with existing solution methods and, moreover, highly scalable in that large problems with many thousands of barriers can still be solved optimally. Our analysis confirms that barrier mitigation can provide substantial ecological gains for migratory fish at low levels of investment.