Does ecology tell us that some species are worth more than others?

I just read a great paper by Michael Soulé et. al. discussing the management implications some ideas in ecology that have outpaced environmental policy.

The authors, a mix of ecologists and conservationists,  argue that some species, which they call "strongly interacting species," deserve higher priority in conservation because of their unique roles in ecosystems.  These species have gone by many names in the ecological literature, including "keystone species," and "ecosystem engineers."  Essentially, these species have a strong effect on the species around them and the functioning of their ecosystems.  For instance, sea otters eat sea urchins, preventing the urchins from razing kelp forests.   Prairie dogs turn over and improve soil, increase plant productivity, creates new habitat for other species.

One problem the authors point out is that our existing policy tools are not designed to handle this level of complexity.  The Endangered Species Act only requires that species be protected until their populations are large enough to prevent extinction.  But much larger populations are neccessary for these species to perform their ecological function effectively.

I think this raises some really interesting economic questions.  For instance, if a species is shown to have a greater ecological function, is it worth more from a social or economic point of view?  It seems to me that such species would have a greater value in delivering ecosystem services, as well as maintaining the "intrinsic" or "existence" value of an ecosystem to society.

Also, how do we measure and define "strongly interacting" species?  The authors put forward some qualitative questions one could ask.  (e.g., "Does the absence or decrease in abundance of the species change an important ecological process in the system?")  However, I think it's a pretty useful exercise to try to figure out what a standardized, quantitative measure would be.   One could theoretically measure the relative impact of one individual of a species on, say, a number that represents food web complexity, or system stability.  This might get us closer to finding effective ways to quantify "regulating" ecosystem services, which are the most difficult to value.

Is there any work out there that's attempted to compare the relative importance of "strongly interacting" species across different ecosystems?  I'd love to hear others' thoughts on this topic.