Reproductive success is measured by number of offspring, not the selective forces or lack thereof. But perhaps I am stretching the definition to the point that it no longer has meaning.
Remember, Mendelian genetics gives you the odds. With a heterozygous dominant trait and both parents are homozygous for the different alleles, the offspring will be 25% AA, 50% Aa, and 25% aa. That doesn't mean that, if you have 4 kids, you are guaranteed to have 1 AA, 2 Aa and 1 aa. Instead, chance may give you 2 AA and 2 Aa. In that case, the gene frequency of A has increased, hasn't it? But there has been no "differential reproductive success".
This can work thru an entire population such that there isn't really any "reproductive success", but the frequency of the gene will change. Purely by chance.
Natural selection, OTOH, does work by differential reproductive success because the environment is selecting particular designs.
Geographic isolation also allows different mutation to accrue in each population which is the basis for divergence. Don't get me wrong, sympatric speciation is a very important and well recognized mode of speciation, but it isn't the only one.
But mutations don't "accrue" unless there is selection. Remember Hardy-Weinberg. Gene frequencies stay the same from generation to generation unless there is some other process acting. The equations for genetic drift shows that it is not a factor when the effective population size >50.
BTW, sympatric speciation often also involves isolation. In this case it is behavioral or ecological, not geographical.
I am not a fan of wiki, but they have a pretty good description
here:
"In
philosophy,
systems theory and the
sciences,
emergence refers to the way complex systems and patterns, such as those that form a hurricane, arise out of a
multiplicity of relatively simple interactions."
OK
I could be totally off my rocker, but I think evolution does have some features of emergence. Simple rules (ie the mechanisms of evolution) result in very complex interactions and adaptations. Interactions within a both stable and dynamic ecosystems result in a complexity that overshadows the simplicity of those interactions.
Ecosystems are different from evolution. Yes, the adaptations are more "complex" that the simple change of base sequences. However, there is a series of links there, even if we haven't delineated all of them.
One good thing that evo-devo has done is move us farther away from a "one function/one gene " view of evolution. What I think evo-devo shows us is that adaptations can arise through changes in developmental pathways. This is important (at least in multi-celled eukaryotes) because developmental pathways are not directly affected by selective pressures. Development, in a way, is a black box who's only selectable function is the final product which gives a lot of freedom to developmental pathways.
No gene is
directly affected by selection. This is the problem Dawkins ran into with his selfish gene theory. Instead, the
individual is the unit of selection. Which means that selection works on the "final product". So evo-devo isn't any different from the rest of evolution that way.
And evolution has long known that selection works thru developmental pathways. The classic example are the development of teeth in birds and baleen whales. In both teeth develop in utero only to be resorbed in utero. An additional developmental step. Neoteny is also recognized as an evolutionary mechanism.
The contribution of evo-devo seems to be that the timing of expression of genes can have phenotypic effects and that some genes can coordinately control the expression of a whole series of genes.
Historically, evolution started out much more along the lines of evo-devo. This seemed to change with the advent of the Modern Synthesis but the tides are changing once again.
This is the strawman. Everyone likes to make the Modern Synthesis more restrictive than it is so they can look good "refuting" it. Gould tried that with Punctuated Equilibrium. Now it's the turn of evo-devo's.
This is my favorite example of Hox genes. Change just ONE base in the Ubx gene, and you change the number of legs:
1a. http://www.nature.com/cgi-taf/DynaPage.taf?file=/nature/journal/vaop/ncurrent/full/nature716_fs.html Hox protein mutation and macroevolution of the insect body plan. Ronshaugen M, McGinnis N, McGinnis W. Nature 2002 Feb 21;415(6874):914-7 Mutate one serine to alanine and change limb # from multiple limbs of crustaceans to 6 limbs of insects. "To test this, we generated mutant versions of Artemia Ubx in which C-terminal Ser/Thr residues were mutated to Ala. In the first such mutant (Art Ubx S/T to A 15), the first five Ser and Thr residues in the C-terminus are changed to Ala. This mutant Ubx has little limb-repression function, similar to wild-type Artemia Ubx (Fig. 3). However, the mutation of one additional Ser in a CKII consensus site (Art Ubx S/T to A 15 and 7) results in a Ubx that strongly represses embryonic limbs (Fig. 3)."
