Response to Comments on : Robben Island penguin survival rates from Bayesian analysis by William Robinson and Doug Butterworth (Altwegg: MCM/2010/SWG-PEL/08)

Doug, Will and I are on the same page about the need for quantitative models to address this problem, and I commend them for taking this forward. However, at the last SWG meeting on 13 January I felt that Doug made claims about his and Will’s modeling approach that I felt were not warranted. The first was that they are fully accounting for all sources of uncertainty. An important source of uncertainty in this kind of work is structural uncertainty; the uncertainty whether ones modeling approach captures the system adequately. My first comment at the meeting therefore was to point out that by using one approach only, this type of uncertainty is not easily quantified. A different approach might well give rather different results. We should at least consider one alternative approach to be reasonably confident, and with others I am working towards one.