406 CHAPTER 16
of substitutions that were fixed. As the time since divergence becomes greater, the
number of differences begins to saturate, and the number of differences between
groups of species that diverged further back in the past will be no greater than
those that diverged more recently. One solution to this problem is to use different
parts of the genome to estimate different parts of a phylogeny. Rates of molecular
evolution vary among DNA sites, among genes, and among lineages of organisms
[13]. In protein-coding sequences, for example, the third positions in codons evolve
most rapidly, and the second positions most slowly (FIGURE 16.5). The reason is
that mutations at second positions inevitably cause amino acid substitutions, many
of which are eliminated by purifying selection, whereas a large fraction of muta-
tions in third positions are synonymous and selectively neutral (see Chapter 7).
Likewise, some proteins evolve much faster than others, largely because of differ-
ences in purifying selection. We therefore use more rapidly evolving parts of the
genome to estimate phylogenies in the recent past, and more slowly evolving parts
for deeper evolutionary time.
RAPId dIvERSIfICATIoN If several species arise from a common ancestor
over a short time, it can be difficult to determine the phylogenetic relationships
among them. A group of such species is called a radiation (see Chapter 2). Phylo-
genetic relationships among species in a radiation are difficult to determine for
two related reasons. The simple and obvious reason is that during the short time
between two successive speciation events, few new mutations are fixed.
The second reason is that incomplete lineage sorting, or ILS, may occur during
rapid diversification. It is critical to distinguish between a phylogeny (the evolution-
ary relations among species) and a gene tree (the genealogical history of a group
of gene copies at the same locus—see Chapter 7). Many loci across the genome
are polymorphic when a speciation event happens. The gene trees for these loci
are sometimes consistent with the species tree. But sometimes they are not: the
copies of a gene at a locus sampled from the two most closely related species can
have a MRCA further back in time than the copies sampled from more distantly
related species. As a result, copies from the most closely related species are less
similar—they differ by more mutations—than are copies sampled from more dis-
tantly related species (FIGURE 16.6). When we have data from multiple loci, the
picture can be confusing: different genes suggest different phylogenies.
ILS can be surprisingly common among closely related species. FIGURE 16.7
shows the gene trees for six different genes that were sequenced in a study of four
closely related species of grasshoppers. Each gene displays ILS, with two or more
gene lineages persisting from one speciation event through another. Although a
single gene can disagree with other genes and with the species phylogeny, com-
bining the information from all the genes yields a good estimate of the species
Futuyma Kirkpatrick Evolution, 4e
Sinauer Associates
Troutt Visual Services
Evolution4e_16.05.ai Date 02-02-2017
Sequence divergence (%)
100 200 300 400 500 600 700
Time (Mya)
Second
First
Third
0.1
0
0.2
0.3
0.4
FIGURE 16.5 Different sites in the genome evolve at different rates. 0.5
Shown here are the proportions of base pairs that differ in the DNA
sequences of the mitochondrial gene COI between pairs of verte-
brate species, plotted against the time since their MRCA (estimated
from the fossil record). Sequence differences evolve most rapidly at
third positions and most slowly at second positions within codons.
Divergence at third positions increases rapidly at first, then levels off as
a result of multiple substitutions at the same sites. Thus these positions
provide no phylogenetic information for taxa that diverged more than
about 75 Mya. More slowly evolving sites in the sequence (such as sec-
ond positions) are useful for analyzing older relationships. (After [25].)
16_EVOL4E_CH16.indd 406 3/22/17 1:33 PM