Management strategies for a large-scale mountain pine beetle outbreak: Modelling impacts on American martens

We assessed the effectiveness of alternative forest management strategies for maintaining American martens (Martes americana) in a sub-boreal landscape subject to an extensive mountain pine beetle (Dendroctonus ponderosae) outbreak, associated salvage logging, and climate change in north-central British Columbia, Canada. We structured the analysis in a Bayesian network (BN) meta-model that incorporated the results of spatially explicit modelling of landscape conditions (natural and logging disturbance, habitat quality, number of potential territories, and connectedness of territories) with analytical population modelling. The BN meta-model was then used to examine habitat and population size responses (adult females only are presented) to management scenarios, in the context of uncertainty of model parameters, management objectives, and climate change. Status quo management is dominantly clear-cutting with 3–20% of each harvest unit retained as mature patches, with reforestation by planting in the remainder. Management options we examined were: (1) the status quo, (2) varying the total annual timber harvest on the landscape (100%, 80% or 50% of current long-term sustained yield estimates), (3) the protection of understory trees during logging, and (4) 30–70% retention of overstory (partial cutting in distinct patches <1 ha in size) in each harvest unit, for 33% or 50% of the annual timber harvest. We found that marten habitat and population size declined substantively with the beetle outbreak and associated salvage cutting. The choice of management strategy then had a long-term effect on the potential for marten recovery after the beetle outbreak. Partial cutting scenarios had the greatest average long-term marten population levels, followed in order by reduced harvest rates, understory protection, and the status quo. Management scenarios with the best chance of meeting conservation goals without over-protecting habitat (and thus unnecessarily constraining timber management) varied with the population objective chosen. The choice of management strategy will depend on the weighting of marten outcomes against the economic desirability of timber harvest strategies, willingness to gamble on climate change, and the time-frame of interest.