Very nice explanation Dan. Three points:
1) Another limitation is that performing cruel experiments with animals is that nowadays it is  impossible unless you use flies, slugs, of water fleas.
2)  It is possible to use very fundamental first principles of biophysics and physiology to obtain part of the fundamental niche from calculations based on temperature and water stresses. Very complicated. Worse than the frogs experiment, but feasible.
3) It is not straightforward to move from lab-variables to climatic variables. A certain amount of data juggling and assumptions are required.

Otherwise, this is a great explanation of the basics of the fundamental niche (Nf), which is TRULY a fundamental concept. I am very glad that you started with it. A question for you and the other readers. How many variables are there in the Nf? 

Jorge Soberon
 


Comments

Dan Warren link
11/29/2011 14:14

Thanks! I agree on all three counts. I'm going to start working on a project soon that involves more process-based modeling, so I'm looking forward to learning more about that. It's something I know relatively little about.

As for the question of how many axes are in the fundamental niche, it's something I've worried about quite a bit. The funny thing about the concept is that almost any environmental variable will be limiting for a given species at some (often unrealistic) extreme, but it may have little effect on the species' fitness or the suitability of habitat within the range of values seen in the real world, or in the area accessible to the species.

It's particularly interesting to think about in terms of ENM/SDM when you consider that a species' distribution may be limited by different factors at different points in space and time. Something that appears unimportant using a correlational approach in a particular point in space and time may be the limiting factor for the species elsewhere or elsewhen.

Reply
Dan Warren link
11/29/2011 15:28

Actually the answer is apparently two. (J. Lozier, pers. comm.)

Reply
João Paulo de Cortes
11/29/2011 19:03


Another point I´d like to raise, which is something that just occured to me is,
When we consider adding phisiology based variables into the ENM process don´t we take a risk in mixing different niche concepts (eltonian and grinnellian)?

Thanks for the high level discution

João

Reply
João Paulo de Cortes
11/29/2011 19:23

Let me correct myself. I was actually thinking in bionomic vs scenopoetic variables(both inside a grinnellian approach), so a better question would be:
How to handle with occasional differences among phisiological tolerances inside (normally wide distributed) species?

Reply
João Paulo de Cortes
11/29/2011 19:29

Let me correct myself. I was actually thinking in bionomic vs scenopoetic variables(both inside a grinnellian approach), so a better question would be:
How to handle with occasional differences among phisiological tolerances inside (normally wide distributed) species?

Reply
Sam Veloz
11/29/2011 21:19

I think Dan is right on when thinking about space and time. A lot of work with invasive species shows different limiting factors in native vs invaded ranges. I also have just finished a fair bit of work looking at how niches of Eastern North American plants have changed over the last 21 thousand years, and the punchline is that yes, it does appear that certain variables can be more or less important for limiting a species distribution/abundance through time.

Reply
William Godsoe link
11/29/2011 22:11

I think Jorge's question about how many variables are in the fundamental niche is promising, but needs to be narrowed down. My worry is that no matter how much we know about the environmental requirements of a species, it is always possible that there is one more environmental variable that we have not measured. If we carry this logic to its extreme, we may not know the actual number of dimensions that affect the fundamental niche of a species, but we should always know that we haven't measured them all.

Is it possible that we can narrow down this question further? I'm often interested in how many variables we will need to predict the extant or future distribution of a species? or alternatively how many variables we characterize before our models become too complex to be useful?

Reply



Leave a Reply