TABLE OF CONTENTS Introduction 2 Taxonomic Concepts: Nominalist or Realist? Cladistic Methodology Parsimony. Computers and Cladistics Applications to Historical Biogeography A Numerical Cladistic Analysis for the Genus Homo Origin and Biogeography of the Hominids Outgroup. Description of Characters. Hominid Groups Utilized in Analysis Results of Analysis Biogeographic Speculation on the Origin of the Hominids A Critique of Cladism Types of Characters Conclusions References Cited
Biologists classify biological entities for two main reasons (Nelson and Platnick 1981). The first is to put some order into their postulations about nature, while still providing a long-term and logically transferable system of information storage. The second is to be able to make some predictions about unknown and partially known taxa, especially with regard to the relationships between taxon. Whatever the methods utilized to generate these categorizations, the resulting classifications follow the system of binomial nomenclature developed by Linnaeus in the 17th century (Linnaeus 1735), which considers the relationships between organisms to be hierarchical. Classifications today still follow the general hierarchical scheme of Kingdom, Phylum, Class, Order, Family, Genus, and Species, with this descending order indicative of the intensity of the relationships between taxa.
There are two primary analytic methods utilized in the classification of organisms into groups: phenetic delineation and phylogenetic systematics. In the phenetic methodology, organisms are classified according to their phenetic characteristics. Phenetic characters are those that are phenotypically derived, and typically involve physical criteria such as morphology and physiology (Heywood and McNeill 1964). The phylogenetic methodology forms classifications of organisms based upon common ancestry, deduced from characters that have evolutionary significance (Wiley 1981).
This paper reviews one technique of classification in the phylogenetic methodology called phylogenetic systematics, or cladistics. This technique was pioneered by the German entomologist Hennig (1966), and the Venezuelan zoologist Croizat (1981), in the early 1960's (Wiley 1981). All phylogenetic classificatory schemes generally recognize that given any three taxa, two of them will be more closely related to each other than either group is to the third (Nelson and Platnick 1981). Members of a classificatory entity in a cladistic analysis share a more recent common ancestor with each other than with the members of any other group (Ridley 1993:632). Cladism follows a species concept "in which a species is a lineage of populations between two phylogenetic branch points (or speciation events)" (Ridley 1993:632, emphasis in original).
Systematics during the last three decades, the era of phylogenetic systematics, has experienced a transformation unprecedented in earlier decades of the twentieth century (Novacek 1993). The nature of this change has been the subject of enormous debate, on the scientific side as well as the philosophical and historical ones (Hull 1988). The central theme of phylogenetic systematics, with its emphasis on shared derived traits, is now widely applied, wherever practical, to the whole range of data from fossils and behavior to molecules and the genetic code itself (Novacek 1993). Albiet there are cases such as immunological distance or morphometric data, where the cladistic approach has constraints, the logical framework for modern cladistic analysis has broad appeal in the systematics community (Novacek 1993).
This is not to say that the debate is over. Major issues persist as to the complete employment of a cladistic program in systematics. But, whereas arguments in proceeding years appeared to have focussed on the logical validity of cladistics verses phenetics or the poorly defined "evolutionary taxonomy" (Eldridge & Cracraft 1980), current debate is largely influenced by some aspect of accounting for an ever increasingly complex and diverse database.
The problematic debate is now are focussed on the relative merits of fossil, ontogenetic, molecular, and other types of data and how to coordinate them all into a pragmatic yet scientifically justifiable program. These considerations address more methodologically orientated concerns, with less (or almost none) emphasis on the philosophical validation of the cladistic approach. Methodological questions always appear with an intensive data-gathering process, and even as such considerations pose difficulties, they are not seen as insurmountable ones, and there is a general feeling of getting on with the task (Novacek 1993).
This paper will present a review of some of the more recent literature concerning both the theoretical issues and the practical applications of a phylogenetic systematic program. In addition, I will attempt to complete a rudimentary cladistic analysis on a group near and dear to this author, the hominids. Through this direct application, I will explore potential associations of the cladistic taxonomic technique to biogeographic questions of interest and relevance.
Taxonomic Concepts: Nominalist or Realist?
Classification schemes are basically mechanical or methodological techniques for the assessment of order in nature. Given this constraint, an epistemological issue of fundamental importance is whether the classifications generated are nominalistic or realistic. A nominal classification would mean that the relationships deduced are artificial divisions of a natural continuum and can be ascribed a name only through the classificatory methods operational parameters. A classification scheme based on realism would mean the taxa generated would have direct meaning and significance in nature, no matter what the classificatory method utilized to discover them.
While this may seem a slightly tautological difference of little practical import, some considerations remain. If taxa are real natural entities simply awaiting discovery, then a final correct taxonomy is obtainable, if only we can discover the correct taxonomic technique. There is only one true classification possible in a realist's world. From a nominalist's perspective, it does not matter if the one true taxonomy is elucidated, because there isn't one. Researchers can generate several different taxonomies based on different criteria to meet the needs of a particular research problem. As long as the methodology is logically consistent and scientifically verifiable, one could generate taxonomies to suite the a priori specifications of each investigation.
Phylogenetic systemicists consider their taxonomies to be ones based on realism. The development of a cladistic program is based on modification of the cladistic technique to enable the discovery of true taxa. The phylogenetic species concept is based on the biological species concept considering reproductive isolation to be the primary criteria to the delineation between separate species. From this tenet that atemporal species do obviously exist (reproductive isolation is a testable hypothesis with extant species), the cladistic species concept considers that if atemporal species exist, then temporal ones must exist as well, and a cladistic analyses can discover them.
Cladistic methodology considers that classifications are generated only with the use of genealogical information, and that the resulting dendrograms, called cladograms, indicate the branching of lineages. A cladogram is a hierarchical dendrogram, with the divergent points designating clades (evolutionary branches in an organisms phylogeny). Clades are determined by identifying the branching sequence of differentiation via the analysis of shared derived characters (Brown and Gibson 1983).
A cladistic analysis proceeds under four basic assumptions, as follows, from Brown and Gibson:
1) speciation is allopatric in the majority of cases;
2) the features analyzed are homologous, i.e., descended from the same ancestors and the same ancestral genes;
3) parallel evolution of individual characters and suits of characters is rare; and
4) organisms with derived characteristics generally do not give rise to taxa with more primitive ones (1983:261).
Within the constraints following from these assumptions, a cladistic analysis proceeds in a logically consistent manner through a sequence of chronological steps. A cladistic analysis begins by defining the taxa to be classified. The number of species is ascertained, and a list of characters representative of each is drawn up. Each character, called a character state, is defined and assigned a transformation series, typically a linear set of numbers representing the observed evolutionary history of a feature. The series range from the most primitive condition, called a plesiomorphic character state, to the most highly derived condition, called an apomorphic character state.
A pleisiomorphic character is one in that a species' character is similar to that character in an ancestral species, and this state is considered the primitive or ancestral state in a cladistic analysis. An apomorphic character is one that the character has been derived from, and yet differs from, the ancestral condition, and this state is the derived state in a cladistic analysis.
The plesiomorphic state is determined via correlations with a closely related taxa called the outgroup. In a cladistic analysis, the outgroup is a group of organisms closely related to the target study organisms, that are observed to determine whether a characteristic is primitive or derived. Typically, if the character is shared with the outgroup it is considered primitive, if it is not shared it is considered derived. These outgroups share the same ancestry with the principle taxa to be analyzed, and hence have the same ancestral traits. Next, the character states are scored for each taxon and this scored data is input into a rectangular matrix format. This is typically a matrix of dissimilarity, and is subsequently converted into a lower half matrix of distance.
Several distance measures are available for the computation of this lower half matrix, including Pearson's, gamma, and percentage metrics, but convention utilizes normalized euclidean distance, which computes root mean-squared distances. On this distance matrix, a clustering operation is completed. A cluster analysis is a multivariate procedure for detecting natural groupings in data sets, and there are numerous procedures available to accomplish the clustering.
All clustering procedures seek to group taxa based on the dissimilarity between them, and the variation in clustering technique only involves modification of the linkage, or amalgamation method. Several are again available, including single, complete, average, median, centroid, and Ward's linkages. Each method involves producing a hierarchical tree with strictly increasing amalgamation distances.
The number of possible trees generated can become quite large, and a cladistic approach considers that the most parsimonious tree is the output desired. While there is no evidence nature is always parsimonious in producing an evolutionary lineage, practicality and logic indicate that the least number of steps required to reach the outcome of an event, the more likely that event is to occur (Nelson and Plantick 1981). It is clear that the use of parsimony has little relation with its dependency on an assumption about a particular mode of evolution. Rather, it is "...simply a logical way to proceed with the analysis of complex data for which laws of behavior are unknown" (Novacek 1993:5). Parsimony then, has a logic independent of a priori deductions about the mode or mechanics of evolution and specifically speciation events, and therein lies its appeal.
The resulting parsimonious cladogram serves as a hypothesis of the phylogenetic relationship between the taxa analyzed. Ideally, a cladogram would be a dichotomous statement of synapomorphy (a shared derived character state between taxa), so that at each branching point a test could be conducted to validate a divergence (likely also a speciation) event. This need not always be the case, however, as a cladogram only serves to hypothesize the relationship between the taxa that is indicated mathematically by the data (Brown and Gibson 1983). A branching point may have three or more emerging lines because the data do not offer sufficient discrimination between alternate choices of phylogeny, as well as the possibility that the groups phylogenetic divergence was not a dichotomous event (Brown and Gibson 1983).
The last step in a cladistic analysis is to fit the resulting cladogram into a phylogenetic classification. This is done by a straightforward transformation directly into the Linnaean system of subordination and phylogenetic sequences (Brown and Gibson 1983). Adjacent branches on a cladogram are considered sister taxa (such as several species of a particular genus), and when dichotomies occur they are indicators of taxonomic levels of increasing hierarchical import (Brown and Gibson 1983).
Computers and Cladistics.
A complaint in the literature about the cladistic program is its reliance on powerful computer software such as PAUP (Swofford 1990). It has been critically suggested (Cronquist 1987) that the use of computer procedures cuts the researcher off from the careful scrutiny of parsimonious alternatives and the deduction of homologous relationships. Cladistic analysis has been labelled "cookbookish" (Novacek 1993:5), and unsatisfactorily inductive. These critiques are an important philosophical notation to consider in the age of use and abuse of microprocessors, but they overlook the fact that the development of the character data matrix is far from a rote listing, and requires careful study and consideration of homologous relationships (Novacek 1993).
Furthermore, the opinion that the cladogram construction is sterile and overly inductive assumes that the initial taxonomy to be generated is basically the final result. While this is an easy trap for the lazy cladist to fall into, more experienced scientists would easily avoid such sloppy application of any technique to a hypothesis test. The most parsimonious tree generated should be but a starting point for investigations (Novacek 1993). As Novacek considers,
Experiments involving the additions or removal of taxa and characters, the substitutions of different outgroups, the use of topology constraints on independent data, or the inspection of tree distributions provide tests for inferences derived from the most parsimonious result (Novacek 1993:6).
The increased utilization of this type of methodology has allowed for the application of a rigorous technique and more opportunity for hypothesis testing using a cladistic program.
Applications to Historical Biogeography
Historical biogeography has traditionally focussed on the central problem of determining "the evolutionary histories of each taxon and to map these histories on the changing surface of the earth" (Brown and Gibson 1983:248). This goal includes the reconstruction of geophysical and climatological events in the history of the earth, as well as "recapturing the histories of evolutionary lineages and relating these to geography" (Brown and Gibson 1983:248). This latter goal of historical biogeography has been articulated through the use of phylogenetic systematic techniques, and considers several basic assumptions: 1) that evolution has proceeded to change the form of organisms; 2) that these altered characters follow observable patterns of inheritance; and 3) that the majority of characters can be considered evidentiary to the determination of evolutionary sequences (Brown and Gibson 1983).
Until the advancement of the phylogenetic technique, taxonomies were developed via purely phenetic classification. Phenetic classifications do not fulfill the needs of a historical biogeographer, in that they need to know which two taxa are most closely related, not which two look most alike. Additionally, a phenetic classification makes no attempt to predict the past, while a phylogenetic classification offers testable predictions about the historical relationships among groups of organisms and biogeographic regions (Brown and Gibson 1983).
The principle approach considers that at the event horizon of a speciation, some form of disjunction occurs to a group of organisms. This disjunction typically has an explanation that is discernable through biogeographical analysis. In essence, phylogenetic classifications must be developed before any biogeographic analysis (of the patterns in the distribution of taxa among various regions of the world) may proceed (Brooks and Wiley 1986). It is not possible to frame reasonable biogeographic hypothesis in the absence of a reliable representation of phylogenetic relationships (Brown and Gibson 1983).
Additionally, it is important to develop a classification technique that is independent of biogeographic information. This allows the avoidance of circular reasoning when biographic analyses are generated with the phylogenetic classifications. Finally, given this assertion regarding the necessity of independent generation, it is essential that there be constant feedback between systematics and biogeography. When biogeography can describe the distributional patterns of a group of organisms, these patterns can be utilized to as confirmational evidence of the validity of cladistically generated phylogenetic reconstructions (Brown and Gibson 1983).
A Numerical Cladistic Analysis for the Genus Homo
Origin and Biogeography of the Hominids.
The Family Hominidae first appeared at least 3 to 4 million years ago in the form known as Australopithecus afarensis (Chamberlain and Wood 1987). This individual was about 3½ feet tall and had a brain capacity of no more than 500 cc. (modern humans hover around 1200 cc.). The hominids unique adaptive strategy includes a mechanically efficient erect posture and bipedal locomotory pattern, well adapted prehensile hands, a generalized dentition pattern for an omnivorous diet, and probably of the most significance, highly developed cerebral functions capable of abstract thought, complex manipulation of the environment, and most probably the symbolic representation and transmission of information via language.
The higher cerebral functions enabled the hominids to manufacture tools (principly from stone, wood, and bone), and manipulate and manufacture fire. The principle adaptory mechanism of the hominid line has been the development of a complex cultural system that, among other things, allowed the hominids to successfully manipulate the environment.
Another, possibly less cerebral adaption, was the refinement of a mechanically efficient system of erect posture and bipedal locomotion (Chamberlain and Wood 1987). Hominids have the ability to outwalk, in terms of energy efficiency, virtually any other life form. As well as increasing the success of their omnivorous subsistence pattern, this ability contributed to the hominids rapid and extremely widespread geographical and adaptive radiation. Hominids have successfully adapted and acclimated to virtually every terrestrial environment, and with some of the more recent technological leaps, the potential exists for adaptive radiation to occur into the marine and extraterrestrial environs.
At present only one species of hominid remains, Homo sapiens sapiens, but this Family has gone through periods of extensive interspecific (intergenus) and interspecific competition (Kimbel 1991). The genus Homo first appeared in Southern Africa by (possibly) the middle to late Pleistocene (1.9 to 1.2 million years ago) in the form of Homo habilis, and most certainly by the terminal Pleistocene (1.5 million to 100,000 years ago) in the form of Homo erectus (Foley 1991).
As an example often offers the best illustration of the functionality of a technique, I attempted a cladistic analysis utilizing nine different hominid groups typically allocated to the genus Homo. From these groups, eleven essentially continuous characters are utilized for the analysis, and the most parsimonious trees were generated by the SYSTAT statistical packages clustering module. These cladistically generated trees are compared to "conventional" trees generated via phenetic methods of classification.
The selection of an outgroup for this analysis is one of the most problematic and arbitrary decisions involved in the procedure. It is complicated by the fact that there is no agreement as to which hominid line represents the most directly related ancestral group to the hominids. Several varieties of the genus Australopithecus have been ascribed this station by various authors (Olsen 1978; White, Johanson, & Kimbel 1981; Trinkaus 1983; Stringer 1985; Brauer & Leakey 1986), including the species A. africanus, A. robustus, A. boisei, and A. afarensis, but due to the lack of statistically viable samples and the problems innate in utilizing fossils for the determination of species, no agreement exists as to who is the direct line descendant.
In light of these disagreements, I have borrowed the tact of renowned hominid cladist C.B. Stringer, who constructed a "..hypothetical outgroup character set from the morphologies present in the whole "australopithecine" group, and refer to other hominoids (living "great" and "small" apes) where there are problems in resolving polarities" (Stringer 1987:139). In short, Stringer utilized a conglomerate of australopithecine characters, further modified with the addition of characteristics from extra-hominid sources such as the modern anthropoid apes.
Description of Characters.
The choice of characters again follows convention in hominid paleontology and systematics, in that almost all the characters are cranial morphometrics (Andrews and Martin 1987). Hominid evolution has exhibited an extreme case of what has been termed mosaic evolution (Mayr 1988). Throughout the entire hominoid line, the morphometrics of the crania has undergone a more complex and more dynamic series of evolutionary changes than has any aspect of the postcrania (Brauer 1984). Even at an inter-familial level, the postcranial skeletal anatomy of the hominoids exhibits relatively little variation in morphology (outside the pelvic girdle, but that's another story). Given this, the crania is seen to offer more opportunities for the delineation of taxa via the analysis of evolutionarily significant changes in morphometrics (Mayr 1988). The characters are characterized as follows, and their coded designation in the analysis can be seen in Figure 1 (after Stringer 1987):
1) Cranial Robusticity (CROB) - Cranial robusticity describes the overall musculature of the cranial bones, and is used as an indicator of dietary habits and overall physical morphology. The gracile crania of H. habilis is considered homologous to that form in H. sapiens, and they are scored 0 for that character while the more robust forms exhibited in H. erectus are scored with a 3.
2) Postcranial Robusticity (PROB) - Postcranial robusticity describes the overall musculature development of the postcranial bones, and is used as an indicator of overall physical morphology. This character ranges from the gracile form of the Skul-Qafzeh and modern groups, which are scored as 0, to the heavily built robust forms of the Asian H. erectus and Archaic H. sapiens, scored as a 2.
3) Supraorbital Torus Morphology (BROW) - This feature is commonly described as a brow ridge, and the size of this character decreases through time in the hominid line. This character ranges from the minimal torus development of living humans, scored as a 0, to the thick supraorbital torus found in the Asian H. erectus and Early Archaic H. sapiens, scored as a 3.
4) Cranial Flattening and Elongation (CRFL) - Cranial flattening and elongation describes the shape of the cranial vault. This character ranges from the short, high and rounded cranial shape found in the Skul-Qafzeh & Modern groups, scored 0, to the low, flat cranial form of the Asian H. erectus, scored 2.
5) Occipital Angulation (OCAN) - The occipital angulation considers the overall shape of the posterior of the crania. This character ranges from the rounded profile of H. habilis A and Modern groups, scored as 0, to the extreme angulation of the Asian H. erectus and Early Archaic H. sapiens, scored as a 3.
6) Midfacial Projection (MFPJ) - This index considers the overall amount of projection of the face, which is reduced over time in the hominid line. This character ranges from the flattened midface of H. habilis A, scored as 0, to the projecting midface of the Neanderthals, scored as a 3.
7) NLH/EKB Index [Nasal Height to Biorbital Breadth] (NIND) - The nasal index describes the overall shape of the nasal aperture which is thought to have reduced over time as well as being indicative of environmental variations. This character ranges from the low values in the Asian H. erectus, scored 0, to the high values for this index in the Neanderthals, scored as 2.
8) Reduced Facial Prognathism (REPN) - This index expresses the state of projection of the mouth and nose area and the amount of reduction is thought to increase over time in the hominid line. This character ranges from the prognathic condition in H. habilis A, scored as a 1, to the orthognathic condition found in Modern groups and scored 4.
9) Reduced Occipital Squama Height to Nuchal Length (REOH) - This index illustrates the overall shape of the distal proximal aspect of the crania, and is considered to be indicative of cerebral organization as well as posture and locomotion. This character ranges from the longer height of Neanderthals, scored 0, to the low height founds in Asian H. erectus, scored 2.
10) Endocranial Volume (ENDV) - This character, a more sophisticated measure of what was once called brain capacity, is considered to have significant import in many aspects of evolutionary development in the hominid line. This character ranges from the low values of H. erectus A & B, scored 1, to the high values of the Neanderthals, scored 4.
11) Dental Reduction (DNRE) - This index measures the size and morphology of the dentition, which has been given tremendous import in evolutionary circles as being indicative of dietary and subsistence patterns. This character ranges from the large dentition of H. habilis A, scored 1, to the small dentition of Modern Humans, scored 4.
Hominid Groups Utilized in Analysis.
The nine hominid groups utilized in this study again are after Stringer's 1987 cladistic analysis of the genus Homo, and include:
Results of Analysis.
All the cladograms grouped the Outgroup and the H. habilis forms either with each other in the same clad or at least the most closely related clades, and these groupings typically considered the H. habilis B form to have diverged first. The H. erectus forms appeared to be most closely related to either of the H. habilis groups (in the case of the Early African H. erectus), or to the Early Archaic H. sapiens (in the case of the Asian H. erectus).
The Neanderthals typically grouped with the Late Archaic and Modern forms, while appearing to have been the first form to diverge from the main group. The Modern and the Skhul-Quafzeh forms were always grouped together, with the next nearest group being the African Late Archaic H. sapiens.
Biogeographic Speculation on the Origin of the Hominids.
From this one can follow with a bit of speculative extension by plotting the geographical loci of these taxa along a hypothetical route of migrational evolution. If the cladogram is complex enough and correctly operated, this could be correlated with known geophysical events to generate a rather clear illustration of the phylogenetic-biogeographic evolution of an entire taxa. While the cladistic analysis attempted here is too limited for reliable results to be generated from it, some effort will be made none the less.
Determining the origin of the hominids is a complex and controversial endeavor, and a cladistic analysis is only a small part of the work required to draw the big picture. The hominids are traditionally considered to have developed in equatorial Africa during the late Pleistocene, and the archaeological recovery of H. habilis confirms this hypothesis. Paleobiogeographical studies indicate that H. erectus is the form recognized to have developed from the earlier H. habilis, and our cladistic analysis confirms their close relation. Most hominid paleontologists consider the Asian H. erectus to have diverged first, and this group appears to be most closely related to the European Early Archaic H. sapiens. The other form of H. erectus appears to be most closely related to the African Archaic H. sapiens.
Although the connection is not universally clear via this cladistic analysis, the neanderthals appear to be most closely related to European Early Archaic H. sapiens, which would add confirmational value to the rather narrow biogeographical range typically ascribed to the neanderthal group. Furthermore, the neanderthals appear to have no close relatives after this divergence. From the East African H. erectus to the African Archaic H. sapiens to the Middle Paleolithic Skhul-Qafzeh (Mount Carmel in Israel) H. sapiens seems to display a relatively clear path of divergence, both at the evolutionary level and the biogeographical level. The Middle Paleolithic H. sapiens are the most modern forms, and given their close temporal association with completely modern humans, it does not take much of an intuitive leap to deduce that they are the most closely related phylogenetically to the Modern group, as the cladistic analysis confirms.
A Critique of Cladism
Several problematic issues have arisen with the use of a phylogenetic classification, and they all have something to do with the types of characters a cladistic program chooses to utilize in its analytic technique (Cronquist 1987).
Types of Characters.
There are three types of character states possible, with respect to phylogenetic origin. The first is analogous characters, those that arise from convergent evolution where the shared character state is not derived from a common ancestor, but as the result of similar responses to an environmental condition. The second type of character state is a shared ancestral homology, where a character is found in a groups' ancestors' but not all of its descendants, with some taxa losing the character and some not. The third character type is a shared derived homology, where a character is unique to a particular group of species and their ancestors. Only shared derived characters are utilized in a cladistic analysis.
This leads to the first problematic issue concerning phylogenetic classifications, that is, they must be generated with the exclusive use of monophyletic groups (Cronquist 1987). While the determination of exclusively monophyletic groups is difficult, it can be seen to be necessary from our consideration that the types of characters must be shared derived homologies. A monophyletic group includes all the descendants of a common ancestor. Obviously this must be deduced from shared derived homologies, and no other types of characters.
This makes paraphyletic groups (some, but not all the descendants of a common ancestor, members being excludes via radical differences in overall character states that make them appear dissimilar to other members of the group), and polyphyletic groups (formed when two lineages convergently evolve similar character states although they have no recent common ancestor, typically analogous groups) not able to be utilized in the formation of unambiguous hierarchical arrangements, such as are required in a phylogenetic analysis. This however, is not a limitation on the particular groups that cladistics can be applied to, only that a certain amount of information must be gleaned about those groups. It is an epistomological constraint, and maybe a pragmatic one as well, but not a taxonomically limiting one.
Phylogenetic classifications can generate an unusually high number of levels, far more than can be accommodated by the Linnaean system (Simberloff et alia 1981). This problem, while upsetting conventional Linnaean taxonomists, can be solved rather easily by the consideration of groups as sister taxa, with several sister taxa existing at the same phylogenetic level. Even as the optimal is a dichotomous branching, this does not hamper the validity of a phylogenetic assessment.
Hybridization, quite common in certain groups of flowering plants, can cause some problematic issues in a phylogenetic analysis (Cronquist 1987). If the new hybrid and its parent are similar enough to belong to the same genus, no problem for cladistics. However, if the parents of the new hybrid were from very different taxonomic levels, the hybrid would have such a divergent origin as make it an anomaly within the Linnaean system (Cronquist 1987).
This is a real limit to a phylogenetic classification, which works impeccably on normal phylogenies but not on hybridized ones. A radical response has been to dump the Linnaean system altogether, and classify the groups as the system indicates. If the phylogeny formed is overlapping rather than hierarchical, then let the classification illustrate it as so.
The last problematic issue that I will discuss (although not the only ones left to be addressed) involves the rather rare phenomena of gene transfer. That is, when some genetic material is transferred between species and adopted by a new species. This is not uncommon in bacterial species, but is rare in higher organisms (an example could be the presence of a hemoglobin-like protein in some clover species). This obviously could derail a cladistic analysis, since part of a species ancestry would come from one lineage, and part from another. If this occurred with many of a species genes, cladistic analysis would be impossible, but even within the bacterial line, gene transfer is rare and occurs in a low frequency when at all.
The role of taxonomy is clearly a vital one to almost all aspects of the biological sciences, and some method of creating order from apparent chaos is at least endemic to the scientific mentality, if not the human one. Phylogenetic systematics offers the most comprehensive and scientifically rigorous program available at present. This is not to indicate that the technique is perfected or even nearly so, but when compared to other methods now available, it is the best there is. Phylogenetic systematics is still relatively new as a major paradigm forming technique, and it will clearly need many more years of adjustment and refinement if it is to be accepted by all.
TAXA HABA HABB EAFE ASE EAS NEA AAS SQ MOD OUTGROUP CHARACTER CROB .000 .000 .000 1.000 3.000 2.000 1.000 1.000 .000 .000 PROB1.000 1.000 1.000 1.000 2.000 2.000 1.000 1.000 .0001.000 BROW1.000 1.000 1.000 2.000 3.000 3.000 2.000 2.000 1.0001.000 CRFL1.000 1.000 1.000 2.000 2.000 1.000 1.000 1.000 .0001.000 OCAN .000 .000 1.000 2.000 3.000 3.000 1.000 1.000 .000 .000 MFPJ1.000 .000 1.000 1.000 1.000 2.000 3.000 1.000 1.0001.000 REPN1.000 1.000 1.000 1.000 .000 1.000 2.000 .000 .0001.000 REOH .000 1.000 2.000 2.000 2.000 3.000 3.000 3.000 3.000 .000 ENDV1.000 1.000 1.000 1.000 3.000 1.000 .000 .000 .0001.000 DNRE .000 1.000 1.000 2.000 2.000 3.000 4.000 3.000 4.000 .000 REDN .000 1.000 2.000 3.000 3.000 3.000 3.000 3.000 3.000 .000
Figure 1. Character Series Assignments for Hominid Groups
Figure 2. The nine hominid groups utilized in this study, from top to bottom in traditional monophyletic order of descent, after Stringer 1987.
Figure 3. Character Series Utilized (after Stringer 1987):
Andrews, Peter and Lawrence Martin
1987 Cladistic Relationships of Extant and Fossil Hominoids. Journal of Human Evolution, 16:101-118.
1984 A Craniological Approach to the Origin of Anatomically Modern Homo sapiens in Africa and Implications for the Appearance of Modern Europeans. Chapter 14, pp. 327-410 in: The Origins of Modern Humans, Edited by F. Smith and F. Spencer. Alan Liss Co. New York, NY.
Brauer, G. and R.E. Leakey
1986 The ES-11693 Cranium form Eile Springs, West Turkana, Kenya. Journal of Human Evolution, 15:289-312.
Brooks, Daniel R. and E.O. Wiley
1986 Evolution as Entropy: Towards a Unified Theory of Biology. The University of Chicago Press, Chicago, IL.
Brown, James H. and Arthur C. Gibson
1983 Biogeography. The C.V. Mosby Company, St Louis, MO.
Chamberlain, A.T. and B.A. Wood
1987 Early Hominid Phylogeny. Journal of Human Evolution, 16:119-133.
1981 Biogeography: Past, Present and Future. Chapter 12, pp. 501-523 in: Vicariance Biogeography: A Critique: Symposium of the Systematics Discussion Group of the American Museum of Natural History, May 2-4, 1979. Edited by Gareth Nelson and Donn E. Rosen. Columbia University Press, New York, NY.
Cronquist, Arthur 1987 A Botanical Critique of Cladism. The Botanical Review, 53(1):1-52.
Cox, C. Barry and Peter D. Moore
1993 Biogeography: An Ecological and Evolutionary Approach, Fifth Edition. Blackwell Scientific Publications, Oxford, U.K.
Eldridge, N. and J. Cracraft
1980 Phylogenetic Patterns and the Evolutionary Process. Columbia University Press, New York, NY.
1991 How Many Species Should There Be? Journal of Human Evolution, 20:413-427.
1966 Phylogenetic Systematics. University of Illinois Press, Urbana, IL.
Helwood, V.H. and J. McNeill
1964 Phenetic and Phylogenetic Classification. Publication No. 6, The Systematics Association, London, U.K.
1988 Science as a Process. University of Chicago Press, Chicago, IL.
Kimbel, William H.
1991 Species, Species Concepts and Hominid Evolution. Journal of Human Evolution, 20:355-371.
Mayr, Ernst 1988 Toward a New Philosophy of Biology: Observations of an Evolutionist. The Belknap Press of Harvard University Press, Cambridge, MA.
Nelson, Gareth and Norman Platnick
1981 Systematics and Biogeography: Cladistics and Vicariance. Columbia University Press, New York, NY.
Nelson, Gareth and Donn E. Rosen (editors)
1981 Vicariance Biogeography: A Critique: Symposium of the Systematics Discussion Group of the American Museum of Natural History, May 2-4, 1979. Columbia University Press, New York, NY.
Novacek, Michael J.
1993 Reflections on Higher Mammalian Phylogenetics. Journal of Mammalian Evolution, 1(1):3-30.
1978 Hominid Phylogenetics and the Existence of Homo in Member I of the Swartkrans Formation, South Africa. Journal of Human Evolution, 7:159-178.
1981 Methods of Paleobiogeography. Chapter 11, pp. 446-501 in: Vicariance Biogeography: A Critique: Symposium of the Systematics Discussion Group of the American Museum of Natural History, May 2-4, 1979. Edited by Gareth Nelson and Donn E. Rosen. Columbia University Press, New York, NY.
1993 Evolution. Blackwell Scientific Publications. Boston, MA.
Simberloff, D., K.L. Heck, E.D. McCoy, and E.F. Conner 1981 There Have Been No Statistical Tests of Cladistic Biogeographical Hypotheses. Chapter 2, pp. 40-93 in: Vicariance Biogeography: A Critique: Symposium of the Systematics Discussion Group of the American Museum of Natural History, May 2-4, 1979. Edited by Gareth Nelson and Donn E. Rosen. Columbia University Press, New York, NY.
1987 A Numerical Cladistic Analysis for the Genus Homo. Journal of Human Evolution, 16:135-146.
1985 Middle Pleistocene Variability and the Origins of Late Pleistocene Humans. Chapter 16, pp.289-295 in: Ancestors: The Hard Evidence. Alan Liss, New York, NY.
1990 PAUP: Phylogenetic Analysis Using Parsimony, Users Manual. Version 3.5, Illinois Natural History Survey, Champaign, IL.
1983 The Shadidar Neanderthal. Academic Press, New York, NY.
White, T.D., D.C. Johanson, and W.H. Kimbel
1981 Australopithecus africanus: Its Phyletic Position Reconsidered. South African Journal of Science, 77:445-470.
1981 Phylogenetics: The Theory and Practice of Phylogenetic Systematics. John Wiley and Sons, New York, NY.