Multi-objective modeling as a decision- support tool for free-roaming horse management

Decisions related to controversial problems in natural resource management receive the greatest support when they account for multiple objectives of stakeholders in a structured and transparent fashion. In the United States, management of free-roaming horses (Equus caballus; horses) is a controversial multiple-objective problem because disparate stakeholder groups have varying objectives and opinions about how to manage fast-growing horse populations in ways that sustain both natural ecosystems and healthy horses. Despite much decision-support research on management alternatives that prevent excessive population size or cost, horse management decisions still receive resistance from a variety of stakeholder groups, potentially because decisions fail to explicitly or transparently account for multiple objectives of diverse stakeholders. Here, we used a predictive model for horse populations to evaluate the degree to which alternative management strategies involving removals and fertility control treatment with the immunocontraceptive vaccine PZP-22 maximize 4 objectives in horse management: maximize ecosystem health, maximize horse health, minimize effects on horse behavior, and minimize management cost. We simulated scenarios varying in management action, frequency, magnitude, and starting population size over a 10-year interval and evaluated scenario performance with a weighted multiple-objective utility reward function. Management involving high-magnitude removals along with PZP-22 treatment generally outperformed other alternatives by achieving higher reward relative to alternatives in 2 scenario analyses. Simulation of 1,372 scenarios at 5 starting population sizes generally found that management with biannual removals and 2 doses of PZP-22 treatment for half of eligible females during years 1 and 5 generated the most rewarding outcomes. However, a removal scenario with more frequent PZP-22 application generated the greatest reward when starting population size was already within target population size range. Our paper demonstrates how values and objectives of diverse stakeholders can be used to support management decisions in ways that might lead to greater acceptance of decisions by a broad array of stakeholder groups.