–Say you are involved in modeling the lifecycle of a listed species. You and your colleagues rightly start out ambitious by aiming to develop and then integrate sub-models for species reproduction, period-to-period, region-based species survival, movements between regions, and juvenile/adult mortality due to exogenous factors, such as human-made disaster.
It doesn’t take long to confirm what you and your colleagues suspected anyway that not only do pertinent data not exist, but modeling uncertainties and errors work against integrating current sub-models into a comprehensive lifecycle model (LCM).
Thereafter with time and funding, you and your colleagues develop much reduced versions, called LCM1, LCM2 and now LCM3, each bringing to light further refinements and significant methodological and data issues. You embark on developing LCM4 in the hopes that the research team—again funding permitting—are moving closer to identifying management interventions for the species.
The many technical reports (now approaching 50 in number) produced during the decade of research track the refinements, improvements, insights and difficulties in modeling species reproduction, movement and survival rates. The peer-reviewed literature on lifecycle models has been advanced in the view of many outside experts by this research.
–Unfortunately for a variety of reasons, none of the reports identify modeling and data uncertainties in a way that they can be contrasted to the uncertainties and errors made in the existing comprehensive model for managing said species.
What “comprehensive model,” you ask? Didn’t I say there was no comprehensive lifecycle model? I forget to tell you that, during all the years the modeling research, real-time deliberations of interagency staff and scientists continued with really-existing decisions, period-by-period, over the management of said species.
From time to time the consequences of the management actions find their way into a technical report, but even here modeling uncertainties hold center-stage: “Though it is tempting to interpret declines in estimated [mortality] as evidence of management success, models of population dynamics are required to disentangle. . .”
–You’d think that the burden of proof has been on the modelers to demonstrate that reliance on life-cycle models would lead to better results compared to the next best alternative of current interagency deliberations of scientists and support staff. . .
But, not to worry: The judge who mandated the research in the first place asserted way back when: “All experts agree that application of a lifecycle model is the accepted method for evaluating the effects of an action upon a populations growth rate.”
This means all we need do is assume management isn’t improving faster than the modeling. And what could make more sense in reality than doing what is so needed in theory?