Reply to comment

In our discussion of Type A and Type B curves (ecosystem function versus species diversity), we came to the issue of order.  With a Type A curve, the order in which species are lost from the system does not matter, the curve orientation will not change.  However, with Type B, order can be important.  For some cases, where species function is shared (redundantly) among a few species in the ecosystem, the order in which they are lost is unimportant; all that matters is when you come to the last one.  For other cases, where there are a few species that perform fairly different functions from the rest, the Type B curve can be very steep at low diversity, leveling out at greater diversity (that is the different-functioning species are not lost until later); or the Type B curve can be very leveled at low diversity and steep at high diversity (signaling the loss of the different-functioning species earlier on). 

Following this train of thought, we could introduce a third curve type (again order is important) whose shape is parabolic with maximum functioning for the middle range of diversity, and minimum functioning at the extremes (high and low) of diversity.  The problem with this curve is that it depends heavily on what is being measured or defined as ecosystem functioning.  For example, one could find a curve like this if the function was biomass, and if animals were lost first, leaving just plants (maximizing biomass).  As the animals die, ecosystem functioning increases, but as the plants die, functioning decreases.  Of course, if our function was predation for the same situation, then we would not find this curve type.

Returning to Ruri’s point about the relevance of knowing where the ecosystem is on the curve for conservation purposes, order might also an important parameter in conservation decisions.  For the type B curve, if we seem to be in a level situation in which losses of diversity is not heavily changing ecosystem functioning, then one might want to know whether there is a point in the future when that will change, or whether that point has actually already been passed.