Back in May, I said that there was nothing extaordinary about the maximum rate of branching that occurred early in the evolution of perching birds and that I would show my calculations in my next post. I want to go to the trouble of backing up my statistics on this so that I can use the same approach on different groups of plants and animals. In the end, I think we will be able to show which speciation events were out of the ordinary and which were part of an ongoing process.
The most common branching pattern in any group of organisms is no branching at all. Whether branching into families, genera or species, the most common outcome is to have groups with only one kind of plant or animal. The next most common outcome is to have groups with two kinds of plants or animals. After that, just about any number of branches can happen. This is called a multinomial distribution.
To compare two branching patterns over evolutionary time, we can make a model that reliably creates a pattern similar to one branching pattern, in this case the pattern of perching bird families dividing into genera, and see how it compares with an earlier pattern, that of the first bird orders dividing into families.
The number of genera in perching bird families
The number of families in all bird orders
By trial and error, I fit the perching bird data to a distribution with a 23 % chance of having one branch, a 13% chance of having 2 branches, a 50% chance of having 3-19 branches and a 14 % chance of having 20- 110 branches. Then I produced 99 random versions of the first birds dividing into families based on the perching bird model, calculating an average and a maximum number of branches for each. The rank of the real branching pattern among these 99 simulated ones is an estimate of the chance that it was a precursor of the later bird evolution. The question being asked is whether a group of plants or animals has an inborn ability to divide into groups or does this potential change with conditions over evolutionary time.
The early birds’ average of 6.5 branches was only duplicated once in 99 tries by the perching bird model. The real distribution has an average that ranks 2nd out of 100. This translates into a 2% chance of coming from the same distribution as the perching birds. On the other hand , the maximum number of 82 branches for the early birds is quite typical for the modelled distribution as it comes 45th out of 100, having a 45% chance of coming from the same distribution.
As I concluded before , it is the lack of branching in other orders that stands out here, not the prolific branching of the perching birds. The average number of genera arising in the perching birds was unprecedented, even though the maximum number of branching events was not. We still cannot tell if this was because these little birds with the perching feet had hit upon a better means of survival or a more reliable mechanism for splitting into groups. I will try to use this approach on other prolific groups and determine if they are superior organisms or just lucky to be in the right place at the right time.