Currently the non-replacement curve substitutes a in the rogersII model with a <- (b * X^q).
X in this scenario is the initial density.
It is not entirely clear (to me) if this give the desired outcome for a so-called "decreasing prey" / "non-replacement" function. It is consistent with similar functions (namely rogersII). But is it correct.
There is some discussion in this PDF from Ben Bolker that this may not give the same solution as an ODE or iterative approach.
I am happy that FRAIR implements the approach used by others. But that is not really the point.
One solution is to replace flexpnr with an alternate formulation that doesn't use LambertW.
Currently the non-replacement curve substitutes
ain therogersIImodel witha <- (b * X^q).Xin this scenario is the initial density.It is not entirely clear (to me) if this give the desired outcome for a so-called "decreasing prey" / "non-replacement" function. It is consistent with similar functions (namely
rogersII). But is it correct.There is some discussion in this PDF from Ben Bolker that this may not give the same solution as an ODE or iterative approach.
I am happy that FRAIR implements the approach used by others. But that is not really the point.
One solution is to replace
flexpnrwith an alternate formulation that doesn't use LambertW.But is this the right thing to do?