Mixed-method study on medicinal herb collection in relation to wildlife conservation:the case of giant pandas in China

Medicinal herb collection has historical and cultural roots in many rural communities in developing countries.Areas where herb collection occurs may overlap with biodiversity hotspots and crucial habitat of endangered and threatened species.However,impacts of such practices on wildlife are unknown and possibly underestimat­ed,perhaps due to the elusive nature of such activities.We examined this phenomenon in Wolong Nature Re­serve,China,a protected area in the South-Central China biodiversity hotspot that also supports a community of Tibetan,Qiang and Han people who use herb collection as a supplementary source of livelihood.We adopted a participatory approach in which we engaged local people in outlining spatial and temporal dynamics of medici­nal herb collection practices.We found that the overall spatial extent of herb collection increased in the past two decades.We then overlaid herb collection maps with localities of giant panda(Ailuropoda melanoleuca)feces collected over two time points in the reserve.Using a Bayesian parameter estimation,we found evidence for de­clined giant panda occurrence in the areas most recently impacted by emerging medicinal herb collection.Our methodology demonstrates the potential power of integrating participatory approaches with quantitative meth­ods for processes like herb collection that may be difficult to examine empirically.We discuss future directions for improving explanatory power and addressing uncertainty in this type of mixed-method,interdisciplinary re­search.This work has implications for future attempts to understand whether and how prevalent but subtle hu­man activities may affect wildlife conservation.

Particle flters for probability hypothesis density flter with the presence of unknown measurement noise covariance

In Bayesian multi-target fltering,knowledge of measurement noise variance is very important.Signifcant mismatches in noise parameters will result in biased estimates.In this paper,a new particle flter for a probability hypothesis density（PHD）flter handling unknown measurement noise variances is proposed.The approach is based on marginalizing the unknown parameters out of the posterior distribution by using variational Bayesian（VB）methods.Moreover,the sequential Monte Carlo method is used to approximate the posterior intensity considering non-linear and non-Gaussian conditions.Unlike other particle flters for this challenging class of PHD flters,the proposed method can adaptively learn the unknown and time-varying noise variances while fltering.Simulation results show that the proposed method improves estimation accuracy in terms of both the number of targets and their states.