Evolution, 4th Edition

404 CHAPTER 16

Our discussion of these few species and a few characters may convey the basic

idea of how we can infer relationships. But in practice, the number of possible phy-

logenetic trees grows rapidly with the number of species. With 10 species, more

than 34 million trees are possible, and with 52 species, the number of possible trees

is larger than the number of protons in the universe! So even with the world’s fast-

est computers, it is impossible to consider all possible trees. Sophisticated statistical

methods have been developed to enable phylogenetic inference, even with hun-

dreds of species and huge numbers of characters (variable sites in DNA sequences).

But this is a mere technical challenge. The process of evolution itself poses difficul-

ties that can make it hard to infer phylogenetic relationships among species.

Why estimating phylogenies can be hard

Most characteristics are not as complex as vertebrae, and are more likely to evolve

more than once: we know that many characteristics do evolve repeatedly. (Think

of black coloration in various snakes, birds, bears, and black widow spiders—see

Figure 6.29.) That is certainly true of individual mutations as well. In practice, no

biologist would base a phylogeny only on body color, or only on a single base pair

difference among species. Repeated independent mutations and several other evo-

lutionary phenomena can make it difficult to determine relationships and phylo-

genetic history. We review these difficulties before describing some phylogenetic

methods that take them into account.

HoMoPlASy Homoplasy refers to the indepen-

dent evolution of similar traits. It results from con-

vergent evolution, parallel evolution, and evolution-

ary reversal (return to an earlier, ancestral character

state) Homoplasy creates problems for estimating

phylogenies because the similarity is not caused by

shared ancestry, or homology (see Chapter 2).

Insect wings illustrate how homoplasy compli-

cates building the phylogeny for a group of species

(FIGURE 16.3). The bristletails (Microcoryphia)

and silverfishes (Zygentoma) are wingless orders of

insects that branched off from the lineage that later

(about 300 million years ago [Mya]) evolved wings.

All the other orders of insects are descended from a

winged ancestor. In more than half of these orders,

some lineages lost their wings secondarily, and

some orders, such as fleas (Siphonaptera) and lice

(Phthiraptera), have no winged species at all. These

groups have independently undergone rever-

sal to the wingless state. Although lack of wings

is a derived state in both fleas and lice, those two

groups are not a clade. We know this because we

have information about many other morphological

characteristics (as well as DNA sequences) in many

other kinds of insects. This information shows that

lice are related to true bugs (order Hemiptera) and

that fleas are related to flies (order Diptera). Like-

wise, wingless crickets and grasshoppers (order

Orthoptera) are closely related to their winged

counterparts. By including many characters of

many species in an analysis, we avoid being misled

by homoplasy.

Futuyma Kirkpatrick Evolution, 4e

Sinauer Associates

Troutt Visual Services

Evolution4e_16.03.ai Date 01-20-2017

Loss

Loss

Loss

Insects

Origin of

wings

Fleas

Flies

Lice

Bugs

Crickets

(wingless)

Crickets

Dragonies

Silvershes

Bristletails

Crustaceans

FIGURE 16.3 Winglessness in many insects has resulted from evolu-

tionary reversals. The basal orders of insects, bristletails (Microcoryphia)

and silverfishes (Zygentoma), are descended from the wingless ances-

tor of all insects, and have never had wings. The other orders of insects

are descended from an ancestral insect that had wings. In many orders,

however, some species have reverted to the wingless condition, such

as some species of crickets and grasshoppers (order Orthoptera). Lice

(order Phthiraptera) are entirely wingless but are related to winged suck-

ing insects such as true bugs and cicadas (order Hemiptera). The entirely

wingless fleas (order Siphonaptera) are related to true flies (order Diptera).

16_EVOL4E_CH16.indd 404 3/22/17 1:33 PM