Introduction to Biological Anthropology: Notes 9
Classification and Phylogeny: Evolutionary family trees
 Copyright Bruce Owen 2009
− Species have an odd property:
− species can be lumped by similarity into groups
− those groups can be lumped into larger groups by similarity
− and so on, in ever larger, more inclusive groups
− in most cases, there is only one way to lump a given lot of species so that they form sets and
subsets that are similar
− if you gave several careful people the task of lumping species into groups, and those
groups into larger groups, they will all come up with pretty much the same scheme
− this pattern of nested groupings is the basis of the Linnaean taxonomic system: the system
that we use to scientifically name organisms
− one or more species are lumped together into each genus, then one or more genera lumped
together into each family, one or more families lumped together within each order, etc.
− this nested grouping is not an arbitrary creation by biologists
− instead, like the existence of species themselves, it seems to be a fact of nature, "out there"
for us to observe
− To illustrate the same thing in a different way:
− Animals can be divided into those that have backbones (vertebrates) and those that do not
(invertebrates).
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 2
− Animals with backbones can be divided into those that nurse their young (mammals), and
those that do not (reptiles, bony fish, amphibians, etc.).
− but the same division does not apply to invertebrates, because there simply are no
invertebrates that nurse their young
− every single animal that nurses its young has a backbone
− Animals that nurse their young (mammals) can be divided into those that have placentas
(placental mammals), and those that do not (marsupials and monotremes)
− placenta: a structure in certain mammals that connects the bloodstream of the mother to
that of the fetus in the womb
− again, this division only applies to mammals, because there are no placental animals that
do not also nurse their young
− The pacental mammals can be divided into numerous groups, and so on
− These groupings are not arbitrary, because each subtype is found only in one group
− for example, every single placental animal is a mammal
− among all the reptiles, fishes, amphibians, etc., there is not a single animal with a
placenta
− every single living thing can be fit neatly into this hierarchical system of subgroups nested
within subgroups
− there are no in-between cases
− and there is only one way that all the species can be arranged like this
− of course, there is debate about certain cases and details
− but the biologists that argue about these cases don't think they disprove the general
pattern, only that a colleague has made an error in his or her observation of it
− This may seem obvious, but the fact that species can be classified in this way is actually a very
specific, unusual feature of living things
− most objects in the world can not be classified in an unambiguous hierarchy like this
− for example, imagine classifying the screws and bolts at a hardware store
− you might first divide them into tapered (screws) vs. straight (bolts)
− and then subdivide each into round head and hex head
− and then subdivide each of those into brass and steel
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 3
− unlike with living things, these categories are cross-cutting
− that is, there are round-headed fasteners in both the tapered and the straight groups
− and brass fasteners with round heads, and with straight heads
− worse yet, you could make an equally good typology dividing first by brass vs. steel
− then by head shape
− then by straight vs. tapered
− So fasteners, and most other non-living things, do not have the property that you can
classify them in a single, unambiguous set of nested groups
− no matter how you sort them, some fasteners will always have characteristics similar to
others in different groups
− there is no single classification scheme that is simpler or better; instead, there are many
equally good hierarchies that you could create
− Another way to look at this: you could organize the fasteners into a table
Brass
Steel
Straight
Tapered
Straight
Tapered
Round
Hex
Round
Hex
Round
Hex
Round
Hex
− all the possible combinations are made
− that is, most or all of the combinations of traits actually do occur
− But this is not true of the classification of living things
− let's organize the animals into a table, as we did with the fasteners
Have vertebral columns
Lack vertebral columns
Nurse young
Don't nurse
Don't nurse
Nurse young
Placenta
No plac.
No
plac.
No
plac.
Placenta
Placenta
Placenta No plac.
− many of the imaginable combinations of traits simply do not occur
− each distinctive trait (nursing young, having a placenta) is found only in a single
category
− placentas are found ONLY in animals that nurse their young
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 4
− nursing the young is found ONLY in animals with a vertebral column
− in other words, the hierarchical classification of living organisms reflects the fact that some
combinations of traits occur, but others which seem equally reasonable (like a hypothetical
animal without a backbone that nurses its young) simply do not occur
− the pattern of which traits are found together and which are never found together creates
this nested-category arrangement
− this is not something arbitrary, invented by biologists
− it is “out there” to be observed in nature
− there is only one way to arrange species in a hierarchy like this
− if several observant, intelligent people independently tried to develop a set of nested
categories for living organisms, they would all come up with roughly the same system
− any disagreements would be small, and most could be resolved by more careful
observation
− incidentally, this is an interesting argument against a "designer" of life
− you would think a designer could mix and match any useful characteristics she wanted
− yet the vast majority of conceivable combinations of features do not exist
− not a single flying insect has feathers
− not a single reptile produces milk for its young
− and so on
− so, why are the features of living organisms parceled out in this oddly limited way?
− This pattern makes sense if species are created by a series of speciation events
− that is, cladogenesis: the evolution of a new species from part of an existing population
− speciation (cladogenesis) is the splitting of a population of one species into two populations
of different species
− the splitting apart of lineages over time, like a branching family tree
− the species after each split are modifications of the last ancestral species that they had in
common
− if a trait, for example nursing the young, develops in one of these groups (clades), only the
animals that descend from that group will nurse their young
− any animals that descended from groups that had already split off before nursing arose will
not nurse
− they are reproductively isolated, and cannot benefit from this development in a different
population
− the only exception would be if nursing arose independently in the two different lineages
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 5
− but the odds are astronomically small that the exact same feature would arise on two
separate occasions, although similar features certainly could evolve
− in that case, close inspection ought to show some differences, even if the feature is
superficially similar in both lineages
− for example, even though both mammals and spiders have what looks like hair, mammal
hair is structurally and chemically different from the hair-like stuff on spiders; despite
their similar appearance, they are clearly separate evolutionary "inventions"
− Darwin was well aware of this patterned way that features are distributed in living things,
and that a sequence of splitting descent could explain it
− this is one of the main reasons why Darwin (and others) believed in the process of
evolution, even before he came up with a theory to explain how the changes could occur
− We can reconstruct the pattern of splitting lineages that resulted in existing organisms
− such a "family tree" is called a phylogeny
− often drawn as an upside-down version of the classification hierarchy
− that is, a branching structure representing change over time:
−
−
−
−
−
the letters represent modern species (gorilla, chimp, human, etc.)
the vertical axis is time, with the earliest time at the bottom and the present at the top
the lines represent "family lines" of descent
species exist all along these lines, not just at the ends or the splits.
even if we have only the evidence of modern species, we could fill in the implied ancestral
species at different times in the past:
− the lineage, or sequence of ancestral populations, from Q through S through B led to the
modern species B
− at different times, there were different numbers of species present
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 6
− around 6 mya (6 million years ago), there were only two: Q and R
− around 3 mya, there were 3 species, S, C, and R
− today, there are 5 species, A, B, C, D, and E
− in this case, species R split into species D and E.
− such a split is called a "speciation event", or cladogenesis, because it creates a new
species
− in fact, it is unlikely that R disappeared and two new species appeared
− this would require two populations to evolve rapidly, and one to go extinct, all at the
same time
− more likely, R was probably very much like either D or E. If it was like E, the actual
pattern would look like this:
− and the species D arose when some population of species E evolved enough differences
from the others to become a new species
− maybe they were isolated on an island with a markedly different environment (allopatric
speciation)
− or any of various possible speciation scenarios
− but unless we have other information, we can't know whether D or E is like the ancestral
species, so we just give it a different name (R)
− we have the same ambiguity here with species Q and S.
− S might have been essentially the same as A or B
− Q might have been essentially the same as A, B, or C
− since we don't know the answer, we use a neutral placeholder to stand for a species that
was probably similar to one or the other of its descendants
− evolutionary changes may occur in species even if a speciation event (split) does not occur
− this would be microevolution (evolution within a single species)
− like the progressive deepening of finch beaks
− recall that evolution that causes a considerable change, but does not lead to speciation (a
split in the phylogeny), is called anagenesis
− if it progressed far enough, we might want to call the later form a different species from
the earlier form, even though there was no split in the lineage
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 7
− this kind of categorization results in naming chronospecies (or paleospecies)
− as we discussed earlier, chronospecies are arbitrary creations by scientists
− unlike biological species, which are real, natural categories of different kinds of
organisms living together in the same world, like species D and E above
− chronospecies simply divide up a gradual, continuous change for our own convenience
in discussing it
− the creatures from different times in a single lineage might seem to be so different
(based on their fossils, for example) that we don't want to treat them as being the same
− so we just draw an arbitrary line in the gradual development over generations, and say
that at some point they constituted a new species
− why would anyone want to reconstruct phylogenies?
− 1. Oldest but least interesting reason: phylogeny is the basis of the Linnaean taxonomic
system, how we name organisms
− The system of scientific names for organisms was invented around 1730 by Carolus
Linnaeus
− the Linnaean system of naming living organisms lumps species together into genera, then
genera together into families, and so on
− The nested, named groups of species, genera, families, etc. originally reflected how similar
the species were
− but now we see that this, in turn, really reflects the successive speciation events, or splits
in the lines of descent that led to the currently living species
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 8
− So we have to get the phylogeny right in order to construct a consistent naming system
with no ambiguous cases
− A note about scientific names of organisms
− called "binomial nomenclature", because species are usually called by the most specific
two levels of this hierarchy, the genus and species.
− for example, there are two species of chimpanzees
− troglodytes (common chimp)
− and paniscus (bonobo, also called pygmy chimp)
− both are members of the genus that includes all chimps: Pan
− the normal way of referring to a species is to use both the species name and the genus
name
− so the common chimpanzee is Pan troglodytes
− and the bonobo is Pan paniscus
− there is a specific, accepted way to write these names correctly:
− both the genus and the species are always italic (or underlined, if you can't do italics)
− the genus name is capitalized, but the species name is not
− example: Pan paniscus
− If it is clear what genus you are referring to, you can abbreviate it with just the first
letter: P. paniscus
− 2. More interesting reason for reconstructing phylogenies: phylogenies lay out the historical
process by which each species evolved
− the order of development of different species and traits
− which implies the kind of organism in which each trait developed
− example: which evolved first, bipedalism or big brains?
− that is, did big brains evolve in animals that swung from the trees, or that walked erect?
− getting the order of branching and changes right helps in understanding why each trait was
advantageous at the time it appeared
− if big brains evolved among animals that swung in the trees, we might suggest that big
brains initially had some advantage related to swinging in trees
− while if big brains evolved among animals that walked erect, we might have to suggest
some other reason why big brains were advantageous
− these are necessary steps to understanding why organisms (like humans) are as they are
− 3. Understanding the phylogenetic relationships of species helps us understand the
similarities and differences of different organisms by putting them into a logical scheme
− in the coming classes, we will look at the species of living primates
− they make a lot more sense if you consider how they are related
− The study of how organisms are related using these evolutionary phylogenies is called
cladistics
− because each group of related species (species that share a single common ancestor) is called
a clade
− How do we construct phylogenies?
− you need to understand the logic in order to use phylogenies, but I won't ask you to construct
them yourself
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 9
− In general: assume that the more similar two species are to each other, the more recently they
have split from their last common ancestor
− last common ancestor of two species: the last species (or really, population) that existed
just before the split that separated the two species (the last ancestor that they had in
common)
− species that differ more will be on lineages that split apart longer ago
− thus allowing more time for evolutionary changes to accumulate in both lineages, making
them more different from each other
− another way of saying this: the most likely phylogeny is the simplest one
− the arrangement that accounts for all the differences among the existing species using the
minimum number of evolutionary changes
− think of these changes as "inventions" of new characteristics, or evolutionary
"innovations"
− such as the first appearance of fur, or the loss of tails
− that is, the most likely phylogenetic tree is the one that:
− minimizes the number of independent duplications of the same innovation
− ideally, each trait is only "invented" once
− minimizes the number of innovations that backtrack or undo a previous one
− ideally, there will be no backtracking at all
− in general, we prefer to believe hypotheses that are simplest, that involve the fewest
coincidences
− but occasionally that may not be the way things actually happened, so we have to remain
open to new evidence…
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 10
− usually if you consider numerous features, the simplest arrangement will be clear
− Some concepts for discussing and understanding phylogenies
− derived trait: a trait that has changed since the time of a given ancestral population
− especially: a trait that all the species in a clade have in common, but that is not found
outside the clade
− that is, a trait that defines the clade, making it different from others
− that arose in the population from which all members of the clade descended
− example of a derived trait:
− humans and chimps are closely related
− at some point, a single population split, with one part becoming modern chimps, and the
other part becoming modern humans
− since only humans have huge brains, the development of the large brain must have
happened after that split
− the last common ancestor of humans and chimps had a more chimp-sized brain
− relative to that ancestor, large brains are a derived trait: a new development since the
time of that ancestor
− ancestral trait: a trait that has been retained from the time of a given ancestral population
− example of an ancestral trait:
− chimp brains are probably about the same size as the brains of the last common ancestor
of chimps and humans
− thus chimp-sized brains in chimps are an ancestral trait: a trait retained from an
ancestral population
− some people use the term "primitive" trait instead of "ancestral" trait
− "primitive" is a loaded word for many people, so I won't use it here
− We can see on the phylogeny where and (relatively) when these changes must have occurred
− for example, the large brain must have evolved somewhere along the lineage leading to
humans, after it split off from chimps
− We can also see what the ancestral species must have been like
− for example, we can see that the last common ancestor of humans and chimps must have
had a small brain
− so all the evolution up to the chimp-human split was in primates with small brains
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 11
− How can we know this?
− try out the two possibilities: what if the last common ancestor had one trait, and what if
it had the other
− whichever possibility lets you explain the modern species with the fewest evolutionary
changes or reversals is the most likely
− example: are big brains in humans derived, or ancestral, relative to chimps?
− First, consider the possibility that it is ancestral.
− that would mean that the last ancestor we shared with chimps had a big brain:
− for chimps to have small brains, their lineage must have developed smaller brains
since we split from that common ancestor
− now go further back to the common ancestor of humans, chimps, and baboons
− for baboons to have small brains, there are two options:
− either the common ancestor had a large brain, and the baboon lineage lost it
− or the common ancestor had a small brain, and the large brain developed later on
the way to chimps and humans
− either way, two changes would be required to produce the three species alive today
− If a big brain was ancestral, it would have had to have either disappeared twice, or
appeared once and then disappeared once
− Now consider the possibility that big brains in humans are derived, that is, that they
were not present in a common ancestor:
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 12
− the large brain would have developed only in our lineage, after the split from the
lineage leading to chimps
− chimps and baboons would have small brains simply because their ancestors did
− so if a small brain was ancestral, it would take only one change to produce the brain
sizes of the three modern species
− We simply pick the least unlikely scenario: the one that requires the fewest “inventions”
to have occurred
− big brains being ancestral requires two brain size changes in this phylogeny
− big brains being derived requires only one brain size change
− so we think that the scenario that the big brain is derived in humans, is most likely
− and thus that the last common ancestor of humans and chimps had a small brain
− Whether a given trait is ancestral or derived depends on which species you are discussing
− for example, humans and chimps do not have tails:
− so the last common ancestor of humans and chimps likely did not have a tail, either
− so relative to humans and chimps, being tailless is an ancestral trait
− but baboons do have tails
− and the last common ancestor of humans and baboons had a tail (just assume this for the
moment)
− the lineage that led to chimps and humans lost the tail some time after splitting from the
one that led to baboons
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 13
− so relative to humans and baboons, being tailless is a derived trait
− and having a tail is the ancestral trait
− notice that how frequent a trait is among different species has no connection to whether the
trait is ancestral or derived
− students often think that ancestral traits should be common, and derived traits should be
rare, but this is not true
− for example, consider baboons, chimps, and humans
− among these, being tailless is a derived trait
− and it is also the more common trait in these three species
− having a large brain is a derived trait
− but this time, the derived trait is rare
− derived traits can be found in many species or just a few; ancestral traits can be found in
many species or just a few
− it depends only on how many splits occurred in the lineage after the derived trait appears
− We have been using one shape of phylogenetic branching, but there are many, many other
possibilities
− and there are often multiple ways of drawing the very same branching structure
− These three phylogenies are just different ways of drawing exactly the same thing
− notice how A and B are the most closely related in every case
− and the group of A, B, and C have a common ancestor more recent than the other two…
etc.
− the order of letters across the top does not matter, only the order of their branching
relationships do
− This phylogeny is different from the previous ones, though…
− Finally, let’s head off a common misunderstanding illustrated in the next phylogeny:
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 14
− Many students assume that E is somehow more closely related to Z than, say B is
− this is incorrect
− both E and B have had exactly the same amount of time to diverge from Z
− remember, time is the vertical axis
− the number of kinks in the path does not matter; that depends on how you happen to
draw the chart.
− Swapping A and B gives an equivalent phylogeny, but drawn that way, B is a straight
shot from Z
− A and B are closely related because they share a common ancestor only a little while
ago (a little way below them)
− A and D are more distantly related because their common ancestor is further back in
time
− all the species shown on the top are exactly equally distant from Z
− All this classifying is based on comparing similarities and differences between species
− but similarities between species can exist for two different reasons
− homologous traits are similar in two species because they have been inherited from a
common ancestor
− example: the five fingers of a human and the five fingers of a chimp are homologous
traits
− since our last common ancestor had five fingers
− both humans and chimps have simply maintained this trait that was found in our
common ancestor
− another form of the word: homologous traits show homology
− traits may be homologous even if they superficially appear very different
− example: forelimbs of humans, whales, and birds
− these limbs have different shapes and functions
− but they have the same specific arrangement of bones
− they are clearly modifications of the same basic design that was present in their last
common ancestor
− so we consider the bone structure of human, whale, and bird forelimbs to be
homologous traits
− analogous traits are similar in two species because they happen to have independently
evolved into a similar state
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 15
− example: wings of birds and wings of insects are analogous traits
− they are similar in many ways
− but they evolved independently
− the common ancestor of birds and insects did not have wings
− wings evolved separately in the two lineages after they split apart
− so we consider bird wings and insect wings to be analogous traits
− analogous traits are often the result of convergent evolution
− convergent evolution is the process in which similar structures evolve independently
from different origins because natural selection favors similar features in both species
− analogous traits can also result from parallel evolution
− example: a bird population splits into two species
− since they have recently split, they are still fairly similar
− if the environment changes, natural selection may act on both of the species in almost
the same way, since they are so similar
− for example, a drying climate might select for larger beaks in both species
− so both species may then change in roughly the same way
− each starting from a similar starting point
− even though each species is evolving completely independently
− in this example, both species might end up with large beaks, but these would be
analogous, not homologous, traits
− because they evolved independently, rather than being inherited from the last common
ancestor
− why do we care about this difference?
− because homologous traits tell us about patterns of descent, but analogous traits do not
− if we used the analogous traits of bird wings and insect wings to lump birds and insects
into one clade, the rest of the branching structure would never work out properly
− so in order to construct and understand phylogenies, we have to be able to determine
whether similarities between species are due to homology or analogy
− How can we tell whether a shared trait is homologous or analogous?
− usually this can be worked out by careful observation, chemical testing, and so on
− example homologous traits: human hand and whale flipper
− careful examination of the bones of both will show many odd but specific similarities
− it would be a huge coincidence if hands and flippers shared all these features by chance
− but it is easy to understand if they both derived from a five-fingered forelimb of a
common ancestor
− example analogous traits: hair on mammals and "hair" on insects
− microscopic analysis shows that they actually look quite different
− chemical analysis shows that they are made of different substances
− so we conclude that they are not modifications of the same ancestral structure
− sometimes the fossil record can help to show the evolutionary route that produced the
traits came from a single source or not
− this is usually not too hard to resolve… but sometimes it is
− How can we tell whether a shared trait is derived or ancestral?
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 16
− one method: by checking the fossil record
− fossils may show which traits appeared earlier
− these are more ancestral
− and which traits appeared more recently
− these are more derived
− another method: by considering an outgroup
− an outgroup is a species that is more distantly related than any of the species we are
actually concerned with, used to determine whether a trait is ancestral or derived
− example of using an outgroup
− baboons have tails; chimps and humans do not
− is having a tail ancestral? or is being tailless ancestral?
− we can't tell if we only look at these species.
− If we consider that it did have a tail, then:
− if having a tail was ancestral, it would take only one change to produce the three
modern species.
− but consider if the last common ancestor did not have a tail:
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 17
− if being tailless was ancestral, then that would also require only one change to
produce the three modern species
− so we can't tell which scenario is simpler
− to solve the problem, we add an outgroup to the phylogeny: dogs
− dogs are so different that we assume they split from the lineage that led to the three
primates long before the primates split up
− now we can try both possibilities and see that one is simpler:
Intro to Biological Anthro S 2009 / Owen: Classification and Phylogeny p. 18
− if the last common ancestor had no tail, it would take either two "inventions" of tails,
or one "invention" and one loss, to produce the four living species
− if the common ancestor of baboons and humans had a tail, it would take only one
change to produce the four living species
− so it is simpler to suggest that having a tail is ancestral among the three primates
− baboons have maintained the ancestral trait
− thus the alternative, taillessness, must be a derived trait
− that appeared after the split between baboons and the lineage that led to chimps and
humans
− We have been discussing obvious traits like tails and brain size
− but the exact same reasoning applies to traits in DNA
− like the presence or absence of the DNA that codes for a specific variant of a protein
− This approach, combining the logic we used here with traits determined from DNA testing,
is resolving many of the remain questions about the phylogeny of life on Earth
− We may look at this later in the course
− Point: These principles of phylogeny will help us to understand the classification and evolution
of the living and fossil primates that we look at through the rest of the course