Like most consensus scientific theories, common descent is supported by patterns of evidence rather than by a few disconnected observations. As Stephen Jay Gould said, “We know that the sun is hub of our little corner of the universe, and that ties of genealogy connect all living things on our planet, because these theories assemble and explain so much otherwise disparate and unrelated information—not because Galileo trained his telescope on the moons of Jupiter or because Darwin took a ride on a Galápagos tortoise.”
In my opinion, one of the best illustrations of this is in the olfactory receptors (OR). These molecules come in two major classes, and as the mechanics of smell differ greatly between the water and the open air, the structure of these proteins differs greatly between fish and tetrapods. The order of Cetacea includes whales and dolphins, as well as porpoises. These are all mammals, and all mammals are themselves tetrapods (it’s worth noting that sort of classification pattern – a nested hierarchy – is itself a broad prediction of common descent).
Most of these animals lack nostrils to smell with. The nostrils are homologous to the blowhole (a well-documented transition). So, lacking nostrils, what would they need olfactory receptors for at all? Well, there is no real mechanism in evolution for the selective deletion of unused genes. What happens in almost all cases is that these genes are simply degraded over time by neutrally-accumulating mutations because natural selection is not actively “protecting” them. If cetaceans evolved from other mammals, whose genomes all feature a suite of ~1000 ORs, we would predict that at least a large number of these be present in cetaceans. In fact, they are; the whole lot of them, in fact. But wait, there’s more! Remember, cetaceans are marine animals. Comparative anatomists since long before Darwin pointed out that they were mammals at their core, not fish as one might naively guess based upon their lifestyle. From an evolutionary point of view, this must mean that cetaceans are secondarily adapted to an aquatic lifestyle, but have also inherited traits from their terrestrial mammalian ancestors. So, time for another prediction: The OR genes of the cetacean genomes will be most closely related to those of terrestrial vertebrates, rather than the fish with whom they share the underwater world. Again, this prediction is borne out. Now this all says a lot, but there are further tests and we can learn a lot more from studying these genes. For example, one could examine whether the cetacean ORs most closely resemble those of the tetrapods with which they have the most in common with in other ways (morphology, other genes, fossil data, etc). Indeed, they do. Even within cetacea, it is possible to build a phylogenetic tree for the order based upon the haves and have-nots of mutations in the OR genes. That is, one could construct a tree based upon the passively-accumulating loss-of-function mutations in the gene. There is no reason why multiple species should share the same set of errors except if they had inherited them from a common ancestor.
At each of these levels, getting to a progressively finer scale, common descent can be tested. I picked the cetacean ORs because they are especially interesting and you can make a lot of points about evolution with that one example, but this kind of examination could be done for most any gene or gene family. For example, in humans and other anthropoid primates, the functional OR repertoire is similarly diminished, probably due to an increased reliance on sight over smell (we belong to a very small set of animals with such wide-ranging color vision). While ~40% of the human ORs are still functional, the remaining 60% are pseudogenes and enable the same sort of comparison between ours and those of the other primates (and more distant mammals) as between the various cetacean species and terrestrial vertebrates.