Kitakaze and all,
The Rowcliffe model is theoretical. The relationships of the model were developed using scientific principles and it was tested/vetted in an enclosed animal park with known abundances (but a limited number of animal species). Thus, I am cognizant that we can’t draw firm conclusions from this model. I liked this model because we don’t have to recognize the individual animal and still estimate animal density (if the cameras are placed with enough distance to each other and followed the specified methodology).
Nevertheless, models are used all the time to understand problems and to figure out the bounds of possible solutions and how the key variables of the problem affect the outcome. In this particular case, I was trying to figure out how many camera-trap days are required to capture Bigfoot, if you assume that BF is a real animal. I thought that this is the first question somebody should answer before attempting a camera-trap survey for BF.
We all agree that there is no data on BF, but we can always bound the problem and analyze. We don’t know the weight, but we can use a ceiling and a floor based on ranges reported and Fahrenbach’s work. We don’t know the animal day range, but we can use a ceiling and floor based on reports or analogies to known animals. The density relationship to weight are based on empirical data from many mammalian species, but I acknowledge that they don’t include BF data. However, again I looked at a range of equations to avoid anchoring and estimated a range of densities.
Using ranges for the 2 key parameters of the model, we end up with camera-trap days required that are reasonable and not impossible (< 10,000 camera-trap days). A very interesting observation from this model is that the higher the animal day range, the higher the likelihood of camera capture and the less cameras you need.
Personally, after doing this work, I was surprised by the results because I was expecting millions of camera trap days required to capture BF. If you believe these results and assume that project Forest Vigil and OP followed the methodologies Rowcliffe outlines, then you would expect capture. However, I don’t know what methodologies were applied for FV and OP. In the Canadian Rockies Carnivore project and Bobcat study in South Cascades, while they did not use Rowcliffe model (that was not their objective) they used very robust methodologies and I believe exceeded the 10,000 camera-trap days. (I did not calculate to check).
I consider the BF Density post, to be a first attempt at bounding the problem. BF proponents should attempt to estimate density since this helps understand the problem scope. The allometric scaling equations are all I had. I acknowledge that the upper-bound estimate of camera-trap days required could be wrong. Would I bet my salary that 10,000 camera-traps are sufficient to capture a BF? No, I am not that confident on this model and don’t have any experience setting camera-traps to really understand the field issues/problems that divide theory from practice. I wish others with more expertize in wildlife biology that do this kind of work for a living will jump in and develop a more robust theoretical model.
On the other hand, a key conclusion from that model that is supported by their test results, is that animal density could be estimated from camera-trap rate. So the ratio of BF density to Bear or cougar density, should provide an estimate for camera-trap rate expected for BF (given the obtained trap rate for bear and/or cougar). This assumes that BF behaves just like any other animal and does not have special abilities or powers. Also, it assumes that none of the more speculative reasons for lack of detection apply.
BTW, I am not a wildlife biologist. I am just an engineer who builds statistical models of real world problems and find the camera-trap modeling problem fascinating.
My conclusion on all this discussion is that we have a mystery (lack of BF capture on camera-trap studies) that is hard to explain if BF is a real animal.