Introduction

So what is taxonomy?

Taxonomy is the way we order and classify the natural world. It is more of a specialised field than many people realise. Concepts like “how different does it have to be to be a distinct species/genus/family?” are not relevant; what we want to know is information about evolution and phylogeny.

The basic unit of taxonomy is the species. Species have a real existence: they are evolutionary lineages (de Queiroz, 2007). What we seek is evidence for these lineages, and the way to recognise them is what has been called the Phylogenetic Species Concept (Groves, 2001; Cotterill et al., 2014) which, following de Queiroz's (2007) analysis, would be better dubbed the Diagnosability Criterion. If one population (or group of populations) is consistently (diagnosably) different from others in heritable characters, this is absolute evidence that it forms a separate evolutionary lineage. The proposal that a given population, sample, cluster, or whatever, constitutes a species is therefore a testable hypothesis.

The so-called Higher Categories do not, on the other hand, have a real existence. Since Hennig (1950/1966) there has arisen a general consensus that a genus or a family must be monophyletic, but beyond that there is subjectivity. There has been a proposal to link these categories to time-depths (see, for example, Groves, 2001); thus, a genus should have been separated around the Miocene–Pliocene boundary, and a family should have been separated around the Oligocene–Miocene boundary. There has been surprisingly little discussion about this, which is distressing, because we are supposed to be doing science, and science is about testability.

The concept of Cheirogaleidae

Early zoologists took a while to recognise that dwarf and mouse lemurs are related to each other, rather than, for example, mouse lemurs being Madagascan galagos. More modern arrangements were beginning to be expressed by the mid-1800s. In Gray (1870), the genus Cheirogaleus was included in the Lemuridae, and two years later Gray (1872) recognised a tribe Cheirogaleina with more or less its present composition.

INTRODUCTION

Fieldworkers are often unaware of the value of making accurate records of their study animals' morphology. External morphology includes measurements, observations of glandular activity and detailed description of the pelage; internal morphology may include skull and postcranial measurements, recording of suture closure, epiphyseal fusion, dental eruption and wear, and observations on the gut. All these observations will yield information on taxonomy, age–sex class (Table 9.1), reproductive status, individual variation, growth, development, sexual dimorphism and so on. It may also be interesting to look at individuals' behaviour in the light of their external differences: does facial coloration correlate with behaviour in mandrills (Mandrillus sphinx), does flange development correlate with behaviour in orangutans (Pongo spp.), and so on?

Measurements may be made either on dead specimens, or on living animals when they are captured (Chapter 7) and, preferably, anaesthetized (Chapter 8). Valuable descriptive information, other than measurements, can be gathered from simple observations, or from photographs (Chapter 17). On occasion, measurements can even be made in this way. For example, Markham and Groves (1990) cite a personal communication from Herman Rijksen of how he weighed four wild orangutans (worth repeating here because it is such a wonderful example of lateral thinking) ‘… by “measuring” the bending arc of particular branches when supporting the full weight of the animal and hoisting up buckets of sand to match the same arc’.

Introduction

Variation in cranial measurements between species has been commonly used to shed light on controversies in the classification, evolution, phylogeny, and functional adaptation of primates. Variation within species has also been used frequently to reveal differences between populations or sexes in morphology, social activities, behavior, ancestral heritage, size, and sexual selection (e.g., Leutenegger and Kelly, 1977; Oxnard, 1983a; Cheverud et al., 1985; Albrecht and Miller, 1993). Studies involving both inter- and intraspecific variation simultaneously are, however, rare (Pan, 1998; Pan and Oxnard, 2000, 2001a). This approach has proven useful in revealing patterns of variation in different functional units or anatomic regions in the same organ (e.g., skull) and in the analysis of the relationship between species or species groups in terms of shape and functional adaptation of various structures. Such studies, although clearly phenetic, have the potential to shed useful light on controversies relating to phylogeny and classification, especially of closely related primate taxa. Previous studies examining inter- and intraspecific variation were carried out on macaques (Pan, 1998; Pan and Oxnard, 2000, 2001a). In this study, the same approach is applied to the Asian colobines because they appear to have evolved in the same places and roughly at the same time as the macaques, and their evolution was probably influenced by many of the same environmental changes during the Pliocene and Pleistocene (Pan and Jablonski, 1987; Jablonski, 1993).

Asian colobines have been a highly successful radiation, as judged by their diversity and wide distribution.

The aye-aye and the question of strepsirhine monophyly

The two suborders of primates, Haplorhini and Strepsirhini, are distinguished clearly by four well-studied and anatomically complex characters:

1. The possession by Strepsirhini of a rhinarium. A rhinarium is widespread among mammals generally, and its possession by Strepsirhini among primates is undoubtedly plesiomorphic.

2. The presence in Strepsirhini of a tapetum lucidum in the eye. Although a tapetum is likewise widespread among mammals, its structure differs from one group to another, so its polarity among primates is unclear.

3. The existence of a fovea in the retina in the Haplorhini. A retinal fovea appears to be unique to Haplorhini, except that an approach to it is seen in Lemur catta and Hapalemur griseus (Pariente, 1970). The most parsimonious interpretation is that it is a haplorhine apomorphy, and that there has been some convergent development of one in the two lemurid taxa cited.

4. A complex of characters of the fetal membranes, such that the placenta is hemochorial in Haplorhini but epitheliochorial in Strepsirhini. Martin (1990), noting that some trophoblastic invasion occurs in two species of galago, argued that the primitive state is endotheliochorial, and that consequently both epitheliochorial and hemochorial states are derived; but the cladistic analysis of Luckett (1993) concluded that the condition in the Strepsirhini is primitive.

It would appear, therefore, that of these four characters the Strepsirhini are questionably derived in only one, the presence of a tapetum. Otherwise, they are defined only by plesiomorphic retentions. What, then, is the evidence that the Strepsirhini really are monophyletic?

INTRODUCTION

Fieldworkers are often unaware of the value of making accurate records of their study animals' morphology. External morphology includes measurements, observations of glandular activity and detailed description of the pelage; internal morphology may include skull and postcranial measurements, recording of suture closure, epiphyseal fusion, dental eruption and wear, and observations on the gut. All these observations will yield information on taxonomy, age–sex class (Table 9.1), reproductive status, individual variation, growth, development, sexual dimorphism and so on. It may also be interesting to look at individuals' behaviour in the light of their external differences: does facial coloration correlate with behaviour in mandrills (Mandrillus sphinx), does flange development correlate with behaviour in orang-utans (Pongo spp.), and so on?

Measurements may be made either on dead specimens, or on living animals when they are captured (Chapter 7) and, preferably, anaesthetised (Chapter 8). Valuable descriptive information, other than measurements, can be gathered from simple observations, or from photographs (Chapter 16). On occasion, measurements can even be made in this way. For example, Markham & Groves (1990) cite a personal communication from Herman Rijksen of how he weighed four wild orang-utans (worth repeating here because it is such a wonderful example of lateral thinking) ‘by “measuring” the bending arc of particular branches when supporting the full weight of the animal and hoisting up buckets of sand to match the same arc’.

Introduction

Gorillas, perhaps because of their size, always seem to be the subject of spectacularly divergent interpretations. Views of their behavior have ranged from the rapacious, vicious giant ape of the nineteenth century, through Robert Ardrey's (1961) view of them as lethargic, depressed, evolutionary dead ends, to the current view of them as gentle giants, albeit with infanticidal tendencies. Views of gorilla systematics have been no less diverse over the past century and a half (see review by Groves, this volume). Beginning with the initial description of Gorilla gorilla by Savage and Wyman in 1847 through the 1920s, ten separate species of Gorilla were described by systematists from all over the world, often from a single skull. The modern systematics of Gorilla stems from the work of Harold Jefferson Coolidge in 1929. Coolidge reviewed all of the previously described species and provided measurements and graphs of 213 specimens from seven major geographic regions. He placed all gorillas in a single species, Gorilla gorilla, in accordance with others such as Rothschild (1906), Elliot (1913), and Schwarz (1928), but went even further in identifying only two subspecies – Gorilla gorilla gorilla for the gorillas of western and central Africa, and Gorilla gorilla beringei for gorillas from the Virunga mountains and adjacent regions.

More recently, Gorilla systematics has derived almost exclusively from Groves's (1970, 1986; also Groves and Stott, 1979) study of 747 skulls and over 100 skeletons.

Prologue

In the fifth century BC the Carthaginian admiral Hanno was commissioned to sail down the west coast of Africa and found Carthaginian colonies. After dropping off colonists at intervals, he sailed on, and eventually came to a fiery mountain, near which was a bay in which was an island. In the island was a lake, and within this another island, full of savage hairy people whom the interpreters called “gorillas”. The Carthaginians tried to catch them; the men escaped and threw stones from the cliffs; but they caught three women and, finding them untameable, killed them, skinned them and took their skins back to Carthage.

Hanno's voyage has been endlessly discussed: did he get to Cameroon? – to Sierra Leone? – just to southern Morocco? Were they really gorillas? – or chimpanzees? – or baboons? – or even Neandertalers? What language was this word “gorillas”, and who were these interpreters?

Heuvelmans (1981) was the first to point out what should have been obvious: the account of the voyage which has come down to us is in Greek, not Punic (the language of Carthage), and it was written some centuries after the voyage was said to have taken place. In the interval, who knows how much it has been embellished, abbreviated, and perhaps modified to accord with various accounts of Greek, Phoenician and Egyptian ocean voyages? There seems little hope of ever establishing what those so mercilessly slaughtered “gorillas” actually were, or where they lived.

Introduction

It is time for a fresh look at the phylogeny of the Old World monkeys. There have been recent attempts to analyse the group by molecular or chromosomal means, and it is time that “traditional” methods of analysis caught up, holding-out the prospect of new macroscopic diagnostic tools. The purpose of this chapter is to examine Cercopithecoid phylogeny by cladistic analysis, using primarily skull characters, but including other sources of evidence where these are available. The results should be regarded as preliminary, in that characters were taken from the literature and tested on specimens, rather than new characters being sought. However, a few new insights have emerged.

Traditionally, the Cercopithecoidea have been contained within a single family, Cercopithecidae, with two subfamilies, Colobinae and Cercopithecinae. The work of Benefit (1993) has established a second family, Victoriapithecidae, to include plesiomorphic early Miocene fossils. The Victoriapithecidae will not be further considered in the present review, which is concerned with reconstructing the phylogeny of extant cercopithecoids and their immediate fossil relatives.

Hill (1966, and elsewhere) supported upgrading the two extant subfamilies to family level. The differences are not extreme, but highly characteristic; adopting full family status for them would give more taxonomic flexibility within each to recognise fine degrees of relationship among genera (Groves, 1989). It is clear that the two form a clade with respect to the Victoriapithecidae.