Supplemental habitat is reservoir dependent: Identifying optimal planting decision using Bayesian Decision Networks is published in the Journal of Environmental Management! PDF

Nice bit of work by Dave Norris and coauthors, here are citation, highlights, and abstract.

Norris, D. M., M. E. Colvin, L. E. Miranda, and M. A. Lashley. 2022. Supplemental habitat is reservoir dependent: Identifying optimal planting decision using Bayesian Decision Networks. Journal of Environmental Management 304:114139.

Highlights

  • Flood control reservoirs experience drawdowns that create mudflats devoid of habitat.
  • Planting agricultural crops on mudflats is one strategy to increase mudflat habitat.
  • Plantings are inexpensive but success is risky due to water level uncertainty.
  • Bayesian decision networks were used to identify optimal decisions in several reservoirs.
  • The optimal planting decision varied across reservoirs of similar function and location.

Abstract Environmental management often requires making decisions despite system uncertainty. One such example is mudflat mediation in flood control reservoirs. Reservoir mudflats limit development of diverse fish assemblages due to the lack of structural habitat provided by plants. Seeding mudflats with agricultural plants may mimic floodplain wetlands once inundated and provide fish habitat and achieve habitat management objectives. However, planting success is uncertain because of unpredictable water level fluctuations that affect plant survival and growth. Decision support tools can account for uncertainty that influences decision outcomes and reduce the risk in reservoir mudflat planting decisions. We used Bayesian decision networks and sensitivity analyses to quantify uncertainty surrounding mudflat plantings as supplemental fish habitat in four northwest Mississippi reservoirs. When averaged across all uncertainty, planting was the optimal decision only in Enid Lake. Response profiles indicated planting decisions depended on elevation contours within Enid, Sardis, and Grenada reservoirs. No planting was optimal at all elevations for Arkabutla Lake. These results provide a quantified basis for establishing best management practices and identify key system states that influence decision outcomes. The process used in this study to evaluate planting decisions can be applied to any reservoir by modifying reservoir dependent inputs to evaluate planting decisions to provide supplemental fish habitat.