Carefully read all the evidence here
http://www.talkorigins.org/faqs/comdesc/
and report back to us after you have pondered the cited sources and digested what they say.
And besides. It's pretty much rock solid that one species can evolve into another, or at least that multiple current species sprang from single earlier species. When you look at genetic data across living things, the signs of a tree-like structure are absolutely fucking unmistakable.
Sure, it's not a trivial problem to actually extract the entire tree of life (see http://en.wikipedia.org/wiki/Phylogenetics - the difficulty is that we only get to touch DNA from the leaves of the tree, and we have to infer the rest), but the tools biologists use are more than powerful enough to distinguish a set of leaves that likely came from an underlying but hidden tree structure from "leaves" that came about in some other way. I don't know numbers off the top of my head, but IIRC from conversations I've had with bioinformatics folks the confidence levels are extremely high, the only debate is over which particular tree should be inferred from a given set of leaves.
You have already accepted that A1 and A2 will adapt to their surroundings. What prevents those adaptations from eventually making A1 and A2 no longer able to mate?