EvoWiki is now a project of the
We are moving all content to RationalWiki.
See the for details! .
As was noted in Taxonomy and Phylogeny, Charles Darwin himself had noted that one expects a treelike pattern of relationships from descent with modification. However, there are numerous cases where sets of organisms seem to have mixtures of shared and different features, a circumstance which seems contrary to that expected treelike pattern. Examples:
- Wings for powered flight: birds, bats, pterosaurs, insects
- Anteating: South American anteater, aardvark, pangolin, echidna (spiny anteater)
- Antishness: ants, termites
- Camera-like eyes: vertebrates, squid/octopuses
- Tail flukes for propulsion in water: fish, cetaceans (whales and dolphins)
In these cases, adaptation for similar function has induced similar overall structure -- but in all of these cases, there are enough differences in detail to enable us to recognize convergence, or homoplasy (also see parallelism). Bird wings have a characteristic architecture, one that differs from the characteristic architecture of bat wings, and likewise for pterosaur and insect wings. This indicates that wings evolved four times, once for each of these groups of animals. And we find that each of these groups of animals is closer to some nonflying animals than to any of the other ones:
- Birds - theropod dinosaurs
- Bats - other placental mammals
- Pterosaurs - early archosaurs
- Insects - crustaceans
The various anteaters have similar adaptations for consuming their meals, notably, long snouts and tongues. However, the rest of the animal looks different enough to suggest convergence. Echidnas are monotremes as opposed to placentals and SA anteaters are edentates/xenarthrans, closest to armadillos and sloths. And recent molecular evidence shows that aardvarks are in a group called "afrotherians" that includes tenrecs, elephants, and sea cows, and pangolins are probably closest to carnivores.
Their diet, ants and termites, are another example of convergent evolution. They have similar lifestyles, living in subterranean or subterranean-like fashion and with their only flyers being reproducers dispersing. However, ants are descended from wasps and termites are probably descended from cockroach-like insects.
Camera-like eyes also fit this pattern; they are different in detail between characteristic architectures, such as whether or not the nerves are in the path of the light to the photoreceptor cells, and their owners are most closely related to animals with much poorer-quality eyes.
Tail flukes are an effective means of propelling a creature through water. Yet fish tails have vertical flukes, while cetaceans have horizontal ones. Note that fish such as eels wriggle from side to side to swim; side-to-side is the fish way of moving. And note that land mammals tend to have a dorso-ventrally (up and down) flexible spine, for quadrupedal running. If cetaceans derived from land animals, as evolution claims, then it is no surprise that they would move their tails up and down, and therefore that their tail flukes would be horizontal.
All these examples suggest that convergent evolution can be recognized in cases where enough distinctive features can be found -- if they can be found. If they cannot be found, then one is likely to reach ambiguous or even false conclusions about relationships. This has been the case for certain birds, for example; convergence was only recognized when additional features were found -- gene sequences.
Biological homology and biological analogy
The difference between homology (detailed similarity of organization that is not explainable by similarity in function) and analogy (gross similarity explainable by similarity in function) was in fact recognized before Darwin. It was formalized by Richard Owen. With Darwin, it was recognized that homology is explained by descent from common ancestors, and that analogy is due to similar selective pressures acting on different starting organisms, i.e. convergent evolution. If biology was "intelligently designed" in a manner similar to human design, we would expect that similar functions would be served by fundamentally the same structures -- like GPS units from the same company have been placed in cars, boats, airplanes, etc. But instead, biology commonly shows a pattern of different functions served by the same basic structures (homology) and the same functions served by different structures (analogy). It is as if the "designer" is continually re-designing the same system from scratch with no knowledge of an already-existent solution to the problem on a continent only a few thousand kilometers away.
A few complications of the homology/analogy distinction should be mentioned. One occurs when organisms that are relatively closely related have similar adaptations. Did the adaptation occur before or after these groups diverged? Since the organisms would have very similar "starting points" even if the adaptations were independently acquired, it may be more difficult to clearly distinguish homology and analogy. In many cases this may be due to our lack of sufficiently detailed information (i.e., developmental and genetic) about the structures in question, or our intentional atomization of the structures into cladistic characters (two species might share the same individual character by common descent, common selective pressures, or chance; the homology vs. analogy question can only be intelligibly assessed by looking at a large number of characters). One term for this situation is parallelism.
A similar confusion may occur where nucleotide sequences are being compared and the level of sequence similarity is close to the similarity expected by random chance. In such cases, factors that bias sequence composition (such as a mutation bias or a temperature favoring DNA stability via a particular G+C content) might make the apparent sequence similarity appear more significant. In the opinion of this author, similarities attributable to chance or to a non-selective deterministic factor are neither homology nor analogy and should be described with another term. --oplo