Keywords
Applied wildlife management; Belief functions; Hybridization; Probabilistic reasoning; Risk analysis and assessment; Uncertainty
Abstract
Object Oriented Bayesian Networks (OOBNs) are a semi-quantitative modeling approach that can be utilized to represent complexities of management tradeoffs and spillovers within a conservation and ecological context. However, computation expense and gradual learning curve result in their underutilization in ecological and environmental disciplines. This is despite the reoccurring need for decision-makers to adapt wildlife management protocols while constraint by limited resources and scarce data. We provide guidelines to identifying and prioritizing uncertainties surrounding complex ecological processes. Empirical data and expert explicit understanding of uncertainties are utilized. We put forth two OOBNs, each accurately representing a snapshot of the moving parts in the complex wildebeest hybridization conservation case study in South Africa. We identifying and clustered key variables impacting the probability of hybridization in either spatial, biological, or market domains. Specifically, (i) blue wildebeest male to black wildebeest male ratio, and (ii) spatial connectivity. Ecologists facing similar constraints worldwide may utilize our stepwise procedural framework so that resources are maximized. This study promotes global collegially research by bridging the boundaries of applications across disciplines, so that their advantages may be extrapolated. The construction of the suggested prototypes is explained in detail so that they may be adapted modified to quantify similar ecology-related uncertainties worldwide.
Citation
: Benjamin-Fink N and Reilly BK. Breaking through Artificial Disciplinary Barriers; Guidelines for applying Bayesian Networks to the Ecology Discipline. SM J Biometrics Biostat. 2018; 3(2): 1031.