Phylogeny and
CHAPTER 3 Diversification of
Mammals
Relationships and Classification Relationships and Classification
of Living Mammalian Orders of Living Mammalian O rders
Simpson’s Classification of
Mammals
Monotremes and Prototheria
Marsupials SIMPSON’S CLASSIFICATION OF MAMMALS
Insectivores
Euarchontoglires The comprehensive classification of mammals put forward by Simpson
Xenarthra (1945) is a landmark in the history of mammalogy. Critically synthesiz-
Cetartiodactyla ing and building upon generations of taxonomic work by previous
Afrotheria authors—some pre-evolutionary or speculative, some grounded in evi-
Laurasiatheria dence and careful reasoning—Simpson’s scheme (Table 3.1) exerts a
Boreoeutheria strong influence on our thinking about mammalian diversity to this day.
The Root of Placentalia Although much of mammalian classification has changed as a result of
subsequent research, many of Simpson’s groups and the names he as-
Timing of the Mammalian signed them remain integral to the vocabulary of vertebrate systematics.
Radiations Here we focus on the identity and groupings of living mammalian o rders
Divergence of Monotremes, Marsu- (information on intraordinal classification is provided in Chapters 10–
pials, and Eutherians 21). Understanding the historical development of mammal ordinal clas-
The Origin of Crown-Group sification from Simpson to the present day (Table 10.1) illustrates the
Marsupials ineluctable connection between taxonomy and phylogeny, the trans-
Placental Mammal Diversification forming influence of phylogenetic methodology, the illuminating power
of fossil discoveries, and the breakthroughs of molecular systematics.
George Gaylord Simpson (1902–1984) was an American paleontolo-
gist and among the most influential evolutionary biologists of the
20th century. Along with Ernst Mayr, Theodosius Dobzhansky,
G. Ledyard Stebbins, and others, he was an architect of the “New Syn-
thesis” in evolutionary biology—the integration of population genet ics
theory with natural history data to form a comprehensive, explanatory
model of the evolutionary process. His classification of mammals was
only one of many significant works. In addition to classic evolutionary
writings, he provided a textbook of systematic zoology (Simpson 1961)
that formalized many of the principles underlying his classification of
mammals. Along with Mayr and o thers (e.g., Mayr et al. 1953), Simpson
was a champion of what came to be known as “evolutionary taxonomy,”
46
, Chapter 3 Phylogeny and Diversification of Mammals 47
Table 3.1 Simpson’s (1945) classification of Such a review, constrained to just a few pages, is possible
living mammalian o
rders largely because Simpson’s (1945) monograph so thoroughly
summarized the character evidence and arguments of pre-
Class Mammalia vious taxonomists (Szalay 1999).
Subclass Prototheria
Order Monotremata (platypus, echidna)
Subclass Theria MONOTREMES AND PROTOTHERIA
Infraclass Metatheria
Order Marsupialia (marsupials)
Infraclass Eutheria
Species of platypus and echidna are highly distinct from
Cohort Unguiculata other mammals. Their egg-laying reproductive mode and
Order Insectivora (hedgehogs, shrews, moles, tenrecs, other unusual characteristics (Chapter 10) are a mixture of
golden moles, solenodons, elphant shrews, tree shrews) primitive and specialized conditions (Hand 2006), though
Order Dermoptera (flying lemurs)
Order Primates (prosimians, monkeys, apes)
derived features of their nasal and frontal bones have been
Order Chiroptera (bats) considered indicative of monophyly (Augee 1983). More
Order Edentata (sloths, anteaters, armadillos) problematic has been the “Subclass Prototheria” and the
Order Pholidota (pangolins) relationships of monotremes to other living mammals. Al-
Cohort Glires
Order Rodentia (rodents) though Simpson (1945) restricted Prototheria to mono-
Order Lagomorpha (rabbits, hares, pikas) tremes, the larger taxon has been thought by o thers to
Cohort Mutica include fossil groups such as triconodonts, docodonts, and
Order Cetacea (whales, dolphins, porpoises) multituberculates (Rose 2006). Evidence for this view was
Cohort Ferungulata
Superorder Ferae discredited by Kemp (1983), and subsequent paleontologists
Order Carnivora (dogs, cats, pinnipeds) have struggled to place monotremes reliably among extinct
Superorder Protoungulata mammal clades. Luo and coworkers (2001) suggested that
Order Tubulidentata (aardvarks)
monotremes form an ancient clade (Australosphenida) with
Superorder Paenungulata
Order Proboscidea (elephants) two Cretaceous genera from the Southern Hemisphere,
Order Hyracoidea (hyraxes) deeply divergent from therian mammals and implying par-
Order Sirenia (manatees, dugongs) allel origins of tribosphenic molars (see Chapter 4). Thus
Superorder Mesaxonia
Order Perissodactyla (horses, tapirs, rhinoceroses)
it is not clear at present which nonmonotreme taxa, if any,
Superorder Paraxonia should be considered prototherians.
Order Artiodactyla (cattle, deer, camels, hippos, pigs, This situation was complicated by Janke and colleagues
pecarries, tragulids, giraffes, pronghorns) (1996 and subsequent papers), who recovered a phylogene
tic tree from mitochondrial genomes, which suggested that
monotremes and marsupials are sister groups apart from
eutherians. This relationship was first posited by Gregory
a systematic philosophy holding that classification should (1947), whose “palimpsest theory” united monotremes and
reflect phylogeny and shared adaptations (Ridley 1986). marsupials in a clade he called “Marsupionta.” If correct,
This is distinct from the dictum of phylogenet ic system- the theory would imply that the many derived traits shared
atists (Chapter 2) that classification should reflect phylog- by therian mammals (marsupials and placentals) are con-
eny alone. Although modern systematics embraces the lat- vergent rather than homologous. The ensuing controversy
ter position, Simpson’s classification of mammals is very prompted Kirsch and Mayer (1998), whose DNA hybrid-
much a work of evolutionary taxonomy. Specifically, the ization study also favored Marsupionta, to review the ana-
classification admits the possibility that some groups are tomical evidence for mammalian subclasses and conclude
paraphyletic (i.e., constituted by some but not all descen- that the case for therian monophyly was in fact equivocal.
dants of a single common ancestor, often characterized by They argued, for example, that the reproductive systems
shared primitive characters). Admittedly paraphyletic of marsupials and placentals are so distinct that “live birth”
orders, such as “Insectivora,” w ere certain to be reclassi- in the two groups cannot be considered a homologous
fied by phylogenet ic systematists as knowledge of evolu- character. This uncomfortable situation persisted even
tionary relationships improved. O thers that seemed as contradictory (but inconclusive) molecular evidence
monophyletic, such as Artiodactyla, have been seriously mounted, u ntil van Rheede and colleagues (2006), Kullberg
challenged by molecular or other data. Most of Simp- and coworkers (2008), and Hutley (2009) presented large
son’s orders, as we w ill see, have survived decades of sub- DNA- sequence data sets from nuclear protein- coding
sequent research and are now recognized as monophy- genes, which strongly rejected Marsupionta in favor of
letic groups (Table 10.1). The same cannot be said of his Theria. W hether the initial mitochondrial and DNA hy-
supraordinal groups. bridization results w ere idiosyncratic or based on insuffi-
In the sections that follow, we explore the basis for cur- cient data for such ancient divergences, the molecular Mar-
rent departures from Simpson’s classification, as well as supionta episode seems to have resolved itself in support
some provocative challenges to it that ultimately failed. of Simpson’s (1945) arrangement. Even so, it encouraged a
CHAPTER 3 Diversification of
Mammals
Relationships and Classification Relationships and Classification
of Living Mammalian Orders of Living Mammalian O rders
Simpson’s Classification of
Mammals
Monotremes and Prototheria
Marsupials SIMPSON’S CLASSIFICATION OF MAMMALS
Insectivores
Euarchontoglires The comprehensive classification of mammals put forward by Simpson
Xenarthra (1945) is a landmark in the history of mammalogy. Critically synthesiz-
Cetartiodactyla ing and building upon generations of taxonomic work by previous
Afrotheria authors—some pre-evolutionary or speculative, some grounded in evi-
Laurasiatheria dence and careful reasoning—Simpson’s scheme (Table 3.1) exerts a
Boreoeutheria strong influence on our thinking about mammalian diversity to this day.
The Root of Placentalia Although much of mammalian classification has changed as a result of
subsequent research, many of Simpson’s groups and the names he as-
Timing of the Mammalian signed them remain integral to the vocabulary of vertebrate systematics.
Radiations Here we focus on the identity and groupings of living mammalian o rders
Divergence of Monotremes, Marsu- (information on intraordinal classification is provided in Chapters 10–
pials, and Eutherians 21). Understanding the historical development of mammal ordinal clas-
The Origin of Crown-Group sification from Simpson to the present day (Table 10.1) illustrates the
Marsupials ineluctable connection between taxonomy and phylogeny, the trans-
Placental Mammal Diversification forming influence of phylogenetic methodology, the illuminating power
of fossil discoveries, and the breakthroughs of molecular systematics.
George Gaylord Simpson (1902–1984) was an American paleontolo-
gist and among the most influential evolutionary biologists of the
20th century. Along with Ernst Mayr, Theodosius Dobzhansky,
G. Ledyard Stebbins, and others, he was an architect of the “New Syn-
thesis” in evolutionary biology—the integration of population genet ics
theory with natural history data to form a comprehensive, explanatory
model of the evolutionary process. His classification of mammals was
only one of many significant works. In addition to classic evolutionary
writings, he provided a textbook of systematic zoology (Simpson 1961)
that formalized many of the principles underlying his classification of
mammals. Along with Mayr and o thers (e.g., Mayr et al. 1953), Simpson
was a champion of what came to be known as “evolutionary taxonomy,”
46
, Chapter 3 Phylogeny and Diversification of Mammals 47
Table 3.1 Simpson’s (1945) classification of Such a review, constrained to just a few pages, is possible
living mammalian o
rders largely because Simpson’s (1945) monograph so thoroughly
summarized the character evidence and arguments of pre-
Class Mammalia vious taxonomists (Szalay 1999).
Subclass Prototheria
Order Monotremata (platypus, echidna)
Subclass Theria MONOTREMES AND PROTOTHERIA
Infraclass Metatheria
Order Marsupialia (marsupials)
Infraclass Eutheria
Species of platypus and echidna are highly distinct from
Cohort Unguiculata other mammals. Their egg-laying reproductive mode and
Order Insectivora (hedgehogs, shrews, moles, tenrecs, other unusual characteristics (Chapter 10) are a mixture of
golden moles, solenodons, elphant shrews, tree shrews) primitive and specialized conditions (Hand 2006), though
Order Dermoptera (flying lemurs)
Order Primates (prosimians, monkeys, apes)
derived features of their nasal and frontal bones have been
Order Chiroptera (bats) considered indicative of monophyly (Augee 1983). More
Order Edentata (sloths, anteaters, armadillos) problematic has been the “Subclass Prototheria” and the
Order Pholidota (pangolins) relationships of monotremes to other living mammals. Al-
Cohort Glires
Order Rodentia (rodents) though Simpson (1945) restricted Prototheria to mono-
Order Lagomorpha (rabbits, hares, pikas) tremes, the larger taxon has been thought by o thers to
Cohort Mutica include fossil groups such as triconodonts, docodonts, and
Order Cetacea (whales, dolphins, porpoises) multituberculates (Rose 2006). Evidence for this view was
Cohort Ferungulata
Superorder Ferae discredited by Kemp (1983), and subsequent paleontologists
Order Carnivora (dogs, cats, pinnipeds) have struggled to place monotremes reliably among extinct
Superorder Protoungulata mammal clades. Luo and coworkers (2001) suggested that
Order Tubulidentata (aardvarks)
monotremes form an ancient clade (Australosphenida) with
Superorder Paenungulata
Order Proboscidea (elephants) two Cretaceous genera from the Southern Hemisphere,
Order Hyracoidea (hyraxes) deeply divergent from therian mammals and implying par-
Order Sirenia (manatees, dugongs) allel origins of tribosphenic molars (see Chapter 4). Thus
Superorder Mesaxonia
Order Perissodactyla (horses, tapirs, rhinoceroses)
it is not clear at present which nonmonotreme taxa, if any,
Superorder Paraxonia should be considered prototherians.
Order Artiodactyla (cattle, deer, camels, hippos, pigs, This situation was complicated by Janke and colleagues
pecarries, tragulids, giraffes, pronghorns) (1996 and subsequent papers), who recovered a phylogene
tic tree from mitochondrial genomes, which suggested that
monotremes and marsupials are sister groups apart from
eutherians. This relationship was first posited by Gregory
a systematic philosophy holding that classification should (1947), whose “palimpsest theory” united monotremes and
reflect phylogeny and shared adaptations (Ridley 1986). marsupials in a clade he called “Marsupionta.” If correct,
This is distinct from the dictum of phylogenet ic system- the theory would imply that the many derived traits shared
atists (Chapter 2) that classification should reflect phylog- by therian mammals (marsupials and placentals) are con-
eny alone. Although modern systematics embraces the lat- vergent rather than homologous. The ensuing controversy
ter position, Simpson’s classification of mammals is very prompted Kirsch and Mayer (1998), whose DNA hybrid-
much a work of evolutionary taxonomy. Specifically, the ization study also favored Marsupionta, to review the ana-
classification admits the possibility that some groups are tomical evidence for mammalian subclasses and conclude
paraphyletic (i.e., constituted by some but not all descen- that the case for therian monophyly was in fact equivocal.
dants of a single common ancestor, often characterized by They argued, for example, that the reproductive systems
shared primitive characters). Admittedly paraphyletic of marsupials and placentals are so distinct that “live birth”
orders, such as “Insectivora,” w ere certain to be reclassi- in the two groups cannot be considered a homologous
fied by phylogenet ic systematists as knowledge of evolu- character. This uncomfortable situation persisted even
tionary relationships improved. O thers that seemed as contradictory (but inconclusive) molecular evidence
monophyletic, such as Artiodactyla, have been seriously mounted, u ntil van Rheede and colleagues (2006), Kullberg
challenged by molecular or other data. Most of Simp- and coworkers (2008), and Hutley (2009) presented large
son’s orders, as we w ill see, have survived decades of sub- DNA- sequence data sets from nuclear protein- coding
sequent research and are now recognized as monophy- genes, which strongly rejected Marsupionta in favor of
letic groups (Table 10.1). The same cannot be said of his Theria. W hether the initial mitochondrial and DNA hy-
supraordinal groups. bridization results w ere idiosyncratic or based on insuffi-
In the sections that follow, we explore the basis for cur- cient data for such ancient divergences, the molecular Mar-
rent departures from Simpson’s classification, as well as supionta episode seems to have resolved itself in support
some provocative challenges to it that ultimately failed. of Simpson’s (1945) arrangement. Even so, it encouraged a
