Evolution, 4th Edition

530 CHAPTER 20

rates are measured by comparing trait means over just a few generations (see

Chapter 6). It has been calculated that at these rates, evolution of body mass

from a 20-g mouse to a 2-million-g elephant would take fewer than 100,000

generations [22, 31]. But these high rates are not sustained for very long, and the

average rates of change calculated for fossil lineages are much lower. These rates

are measured by comparing means in the same lineage sampled at two points

in geological time that may be separated by many thousands or even millions

of years. Extinct populations were likely evolving as fast as living populations,

but their rates are averaged over longer periods, giving the appearance of slower

evolution. Josef Uyeda and colleagues compiled data on body mass of terrestrial

vertebrates, and found that the net change in body size within a lineage is the

same for both very short and much longer time intervals, up to 1 million years

(FIGURE 20.12B) [108].

But even these low rates can produce big changes, because the time spans are so

long. The body mass of the largest species in the horse family (Equidae) increased

by a factor of ten during the last 25 My (FIGURE 20.13). Data from the fossil record

show that the body mass of terrestrial mammals has been able to increase 100-fold

in about 1.6 million generations, and 5000-fold in 10 million generations [22].

The long-term rates of morphological evolution, as measured either in fossil lin-

eages or among living species, are usually so low that they almost always could be

Futuyma Kirkpatrick Evolution, 4e

Sinauer Associates

Troutt Visual Services

Evolution4e_20.12.ai Date 01-30-2017

(B)

Ratio of body masses

100 101 102 103 104 105 106 107 108 109

Interval (years)

0.1

0.01

1

10

100

Field studies

Fossil time

series

Phylogenetic

divergence

Evolutionary rate (haldanes)

Time interval (generations)

1

1 10 102 103 104 105 106 107

10 –1

10 –2

10 –3

10 –4

10 –5

10 –6

10 –7

(A)

FIGURE 20.12 Rates of phenotypic evolution vary greatly, espe-

cially depending on the time interval over which they are mea-

sured. (A) This graph shows two important patterns. The first is

that the rates measured on living species (red dots) are hundreds

to millions of times faster than those seen in the fossil record (blue

dots). Rates are measured in haldanes (on a log scale). A haldane

is a rate of evolution equal to the change in the mean of a trait, in

units of standard deviations of change per generation. (See the

Appendix for a discussion of standard deviations.) The second

pattern is that the evolutionary rate is inversely correlated with

the time interval over which the rate is measured (shown on the

x-axis, again on a log scale). The inset images show some of the

species that appear in these data: for living species, house mouse

and guppy; for extinct species, Hyracotherium (an ancestral

horse) and Globorotalia (a foraminiferan). (B) A similar analysis of

changes in body mass in lineages of terrestrial vertebrates. Each

point is the ratio of mean body mass of a later (descendant) to

earlier (ancestral) sample. Field studies represent changes within

individual populations; fossil time series are ancestor-descen-

dant pairs of fossils; phylogenetic divergence is the difference

between two living taxa, divided by time since their common

ancestor. Remarkably, changes in body mass accumulate and be-

come steadily greater only over time intervals greater than about

1 million years. Before then, size may fluctuate rapidly, but with

little net change. The authors of the study suggest that cumulative

change may occur only after a lineage adapts to a substatntially

different ecological niche. (A after [31]; B after [108].)

20_EVOL4E_CH20.indd 530 3/22/17 1:44 PM