Evolutionary Anthropology 23:27–29 (2014)
ISSUES
Species: Beasts of Burden ALFRED L. ROSENBERGER
Ernst Mayr (1904–2005) was the twentieth century’s most influential writer to wrestle with the species problem.1–4 The following draws heavily on his work, albeit without presumptuously claiming to mirror his thinking or present any original ideas. As a personal meditation, I am thinking mostly of platyrrhines. Following Mayr, I adhere to what is commonly called the Biological Species Concept (BSC) as a way of thinking about a species in the real-world biosphere as a taxon. I also hold to the idea that the Linnaean category called species has the same function as other categories: a linguistic tool for organizing and retrieving information about biodiversity while embodying evolutionary hypotheses. In other words, alpha taxonomy, the area of systematics that involves identifying, naming, and classifying species, is not purely an exercise in either biology or inventory because it involves communication as well. The burdensome work of the species category stems partly from tension created by the several purposes associated with the concept: the objective observation and examination of a fundamental biological phenomenon, the collection and interpretation of data in a selective context of relevance, and the intention to deploy scientific decisions as a form of communication within a dynamic but highly structured language system.
The biological entity we call species is a bearer of names steeped in language traditions. The binominal labels we give to species are subject to whim, imagination, cultural influences and other idiosyncrasies. This is why name use, name giving, and name change are dynamic conditions, even when the science behind
One of the author’s current projects examines how and why many studies of platyrrhine phylogenetic interrelationships produce both corresponding and conflicting hyptheses. In the Dominican Republic, Rosenberger’s cave-diving scuba team is collecting fossil primates from bone-rich freshwater caverns. With Tim Smith and Valerie DeLeon, he is also working on the influence of cranial development on primate orbital morphology. E-mail: [email protected]
Key words: platyrrhines; taxonomy; systematics; Biological Species Concept; New World monkeys
C 2014 Wiley Periodicals, Inc. V
DOI: 10.1002/evan.21392 Published online in Wiley Online Library (wileyonlinelibrary.com).
nomenclatural decisions is intended to be objective, rigorous, and enduring. As a taxonomic category, the operation of identifying and ranking of a set of animals as a species is also inevitably subject to case-bycase utilitarian arbitrariness because every species has its own character, phylogenetic setting, and history, much like every classification. In practice, that slot in the Linnaean hierarchy is delimited through a context-specific analysis that often involves comparing the biodiversity of closely related species-level groups belonging to the same superordinated genus. This makes across-the-board standardization of operating procedures – the biology part of the species problem – difficult. For example, we all agree that the monotypic species Callimico goeldii has no intrinsic properties underlying its characterization as a species apart from the same set we use at the genus level, whereas Alouatta seniculus is delimited by biologically meaningful characteristics designed to
distinguish it uniquely from perhaps a half-dozen other howler monkeys. These same traits are not meant to simultaneously distinguish A. seniculus from, say, Ateles, Lagothrix of Callimico. A totally different yardstick is properly used in the alpha taxonomy of Cebus, an animal built from a different blueprint with different historical and ecological constraints and different genetic potential. This means that the information we use to sort species need not always be the same as long as it reflects, in theory at least, reproductive isolation per se or its consequences. Because this is a relational concept, the difficulty of imposing consistent, nonarbitrary criteria to delimit species in the real world is further complicated by the likelihood that populations that are relevant for comparison probably are also evolving at different rates of species formation and senescence, while their phenotypes are simultaneously experiencing mosaic evolution. As a morphologist, I favor anatomical criteria during the discovery process and in assessing the power of diagnoses and differentia. I also welcome nonmorphological evidence, when available, to assess the broader picture. That picture includes genes, chromosomes, karyotypes, vocalizations, fixed-action patterns, mating strategies, social organization, diet, locomotion, activity cycle, hybridization, and sympatry – attributes that are currently and historically important in situating a species in nature, which is the point of evolution. The presence and incidence of directly observable phenotypic and genotypic characteristics are byproducts of reproduction and/or reproductive isolation, making them perfectly valid indicators of species status under the BSC, as are the inferred
28 Rosenberger
behaviors and adaptations arising from them. Sympatry has a different status; it is a nongenetic auxiliary criterion used in combination with breeding information. As a practical matter, we ideally assume that species are discrete entities, but also relational constructs, which makes operational rules easier to express than to implement. As a starting point, my approach is to develop a differential diagnosis by attempting to establish group-specific morphological patterns within a comparative taxonomic framework. These patterns are presumed a priori to be homologous structures. Why patterns and why homologies? Because species are, in effect, distributed networks of reproductively compatible individuals bearing unique combinations of genes that are likely to be manifest or mirrored at the phenotypic level as a design. In comparing samples of two potentially different but related “maybe species,” if the degree of pattern discontinuity within and/or across systems surpasses their evident continuity, a species difference may be indicated as a hypothesis. But the final determination of status also must square with taxonomic practice, meaning that reasonable taxonomic compromises may be in order to maintain nomenclatural stability for purposes of classification and dialogue. In a morphological study of three allopatric populations of Leontopithecus, for example, Coimbra-Filho and I found that craniodental metrics, body proportions, and pelage consistently separated the northern form, chrysopygus, from the two southern populations, rosalia and chrysomelas, which were less distinct from chrysomelas and from one another.5 Incisor morphology suggested that this separation reflected adaptive dietary differences, and this inference shaped our taxonomic thinking. These observations paralleled vocalization evidence discovered later.6 The taxonomic rank of Leontopithecus populations was then in question: species or subspecies?7 We elected to rank all three groups as species in order to maintain a balanced, stabile nomenclature for the genus, even though, from a purely biological perspective, we felt most comfortable with the hypothesis that, minimally, there were two.
ISSUES
Advances in the implementation of three-dimensional scanning technologies, geometric morphometrics, and other analytical tools promise to revolutionize these sorts of taxonomic morphological comparisons by making it possible to document mathematically complex anatomical patterns, such as those on the occlusal surface of molar teeth, and to make quantitative comparisons of the true geometries of shape among populations. During decades without such highpowered approaches, beginning with G.G. Simpson, simple statistics of tooth size for fossil populations served as an important benchmark for delimiting species boundaries.8 In studies of
Why patterns and why homologies? Because species are, in effect, distributed networks of reproductively compatible individuals bearing unique combinations of genes that are likely to be manifest or mirrored at the phenotypic level as a design.
fossil platyrrhines, however, a tendency to rely exclusively on simple metrics without considering modern ecology – a species’ place in nature – has, in several cases, created a false sense of taxonomic rigor. The basic problem is that sympatric extant platyrrhine species, across the gamut of size classes, overlap considerably in tooth size and body mass. The late Oligocene Branisella boliviana and Szalatavus attricuspis are a case in point. Specimens come from the same area (not locality) and overlap in size, but the type specimens’ upper molars differ in occlusal morphology. Thus, pooling all the known specimens of molars (actually very few) and showing that they are grossly similar in size is insufficient reason to merge these species into one.9 A prior critique that sought to sink Szatalavus also conflated more
than it clarified by erecting a false equivalence among samples. Rather than demonstrating continuity in the occlusal anatomy of the two types, it claimed that a new two-toothed fossil, which the authors simply assumed to be Branisella, undermined Szalatavus because its M1 “closely resembled that of ‘Szalatavus,’ while M2 of the same specimen has a well developed distolingual cingulum and a moderate hypocone, being different from M2 of “Szalatavus.”10 Reasoning differently even on this basis alone, one might recognize that the new fossil was Szalatavus instead of Branisella or that all three samples differ from one another in asymmetrical ways, adding credibility to the hypothesis that more than one taxon exists at La Salla, Bolivia. A similar situation occurs with Aotus dindensis from the middle Miocene of Colombia. Designating it as synonymous with Mohanamico hershkovitzi because the molar teeth of these taxa overlap in size does not account for documented differences in the morphology of the incisors, premolars, and mandible. It also does not account for the instructive parallel of the anatomically similar Aotus and Saimiri: same body size; grossly similar lower molars; often sympatric; very distinct genera.11 Both these examples are laudable for their efforts to be objective through quantification and conservative in trying not to overestimate taxonomic abundance at fossil sites where faunas have yet to be adequately sampled. But they are also weakened by taking a narrow approach to morphology and presuming that fossil sites can be well interpreted without regard to the evidence of species packing and biodiversity among contemporary related animals. Obviously, rare and fragmentary fossils pose inevitable taxonomic difficulties. But the species-level systematics of living platyrrhines is being roiled today by work, some methodologically quite sophisticated, which seems like a throwback to preModern Synthesis typological thinking.12,13 A re-classification frenzy is taking place, ostensibly based on adoption of the Phylogenetic Species Concept (PSC), a reworking of Simpson’s Evolutionary Species Concept, which focuses on lineages.14 This reclassification seizes on populations previously listed as subspecies distinguished by key characters, initially meant for identification purposes,
Species: Beasts of Burden 29
ISSUES
then raises them to species status without contributing much more biology.15 Taxonomically, one wonders where, exactly, this will lead us. Flattening the Linnaean hierarchy below the genus into a single rank that does not distinguish the subgenus, species, and subspecies makes the interconnected genus- and species-level categories outliers in the hierarchy, the only ones not employing nested sets. This makes it harder to communicate interrelationships, similarities, and disparities. It also gives the impression that all these species entities exist in an equivalent steady state, while we fully expect that the populations in question are and have been evolving at different rates toward extinction, speciation, or phyletic evolution. Biologically, what is to be gained? One leading advocate of the PSC criticizes the BSC as being unscientific because reproductive isolation is said to be untestable in cases of allopatry.16 There are several reasons why this position is subjective and hardly defensible. For starters, it is not justifiable to accept the scientific reality of species under conditions of sympatry but simultaneously deny the broader concept of species because certain geographic or temporal situations make testing difficult. Furthermore, the criticism of nontestability sets forth the exclusive requirement to perform direct tests of reproductive isolation, thus rejecting any other form of indirect evidence of genetic separation that can reflect infertility. This is not too different from disqualifying as unscientific the hypothesis that evolution by natural selection is responsible for the high-powered human brain because no one has been eyewitness to the underlying neuron-specific gene changes that occurred across the generations. It seems retrograde to redefine a biological reality that has been tested countless times in order to tidy up method (just as the phylogeny concept has been unjustifiably redefined as cladistic affinities so that the only relevant evidence of descent is synapomorphy). The preferred evidence in applying the PSC is said to consist of any phenotypic trait that defines a population as a discrete genetic linage or, preferably, a cladogram that can be used to test the existence of lineages
by revealing a topology. Here again, the logic is soft. Lineages and species are not equivalent phenomena. The different streams of phylogeny identified in these studies may be demes, species, or anything in between, some on their way to becoming fullfledged reproductive isolates, some careening toward extinction before achieving that status. There is no way to discern whether a simple or randomly selected trait difference in phenotype or a branch on a tree is a consequence of meaningful genetic separation in the sense of species, which are fundamental units of evolution as the highest level of biological organization determined by the genome. There is no logical connection between an assertion that the cladistic affinities of populations are known and the operation of ranking them, biologically or taxonomically. Moreover, the reliability of molecular cladograms, on which much species inflation relies for its justification, is coming under increasing scrutiny because of concern that they often represent gene trees rather than species trees. Like evolution by natural selection, the BSC is built on many testable interrelated hypotheses. The same can be said of the PSC. But making part of the species concept as easy as possible to refute by robbing the entire phenomenon of its evolutionary richness contributes little to our understanding of nature. Primatologists can take heart that other fields of science, which, by their very nature, are more solidly ingrained in mathematics and prediction, and powered by the most advanced instruments and computers ever built, share some of the same quandaries we do. Remember when Pluto was a planet? Well, not any more. What changed? Perhaps the view of prominent Berkeley astronomer Geoff Marcy offers us solace in his response (http://www. space.com/3142-planets-defined.html): “Your questions imply that a definition of the word ‘planet’ is useful scientifically. That is a view not shared by many professional planetary scientists . . . . The astrophysics of planetary bodies is so rich and complex that defining ‘planet’ has never been an issue under discussion among professionals. The taxonomy of asteroids, comets, moons, planets and brown dwarfs is far too limited to capture the
diversity of their origins and internal constitutions.” (Postscript: Marcy is credited with identifying some 250 planets outside our solar system.)
ACKNOWLEDGMENTS Thanks to John Fleagle for inviting me to offer thoughts on the species problem and for his insightful critique of the draft manuscript, which was significantly improved by his editorial input. Nevertheless, I remain responsible for the stubborn errors and sprinkling of fuzzy logic.
REFERENCES 1 Mayr E. 1963. Animal species and evolution. Cambridge: Harvard University Press. 2 Mayr E. 1982. The growth of biological thought. Cambridge: Harvard University Press. 3 Mayr E. 1991. Principles of systematic zoology. New York: McGraw-Hill. 4 Mayr E. 1977. Evolution and the diversity of life: selected essays. Cambridge: Harvard University Press. 5 Rosenberger AL, Coimbra-Filho AF. 1984. Morphology, taxonomic status and affinities of the lion tamarins, genus Leontopithecus (Callitrichinae, Cebidae). Folia Primatol 42:149–179. 6 Snowden CT, Hodun A, Rosenberger AL, et al. 1986. Long-call structure and its relation to taxonomy in lion tamarins. Am J Primatol 11:253–261. 7 Hershkovitz P. 1977. Living New World monkeys (Platyrrhini): with an Introduction to Primates. Chicago: University of Chicago Press. 8 Simpson GG, Roe A. 1939. Quantitative zoology. New York: McGraw Hill. 9 Takai M, Anaya F, Shigehara N, et al. 2000. New fossil materials of the earliest New World monkey, Branisella boliviana, and the problem of platyrrhine origins. Am J Phys Anthropol 111:263–281. 10 Takai M, Anaya F. 1996. New specimens of the oldest fossil platyrrhine, Branisella boliviana, from Salla, Bolivia. Am J Phys Anthropol 99:301–317. 11 Rosenberger AL, Setoguchi T, Shigehara N. 1990. The fossil record of callitrichine primates. J Hum Evol 19:209–236. 12 Zachos FE, Apollonio M, B€ armann EV, et al. 2012. Species inflation and taxonomic artefacts; a critical comment on recent trends in mammalian classification. Mamm Biol 78:1–6. 13 Rosenberger AL. 2012. New World monkey nightmares: science, art, use, and abuse (?) in Platyrrhine taxonomic nomenclature. Am J Primatol 74:692–695. 14 Simpson GG. 1951. The species concept. Evolution 5:285–298. 15 Groves C. 2000. Primate taxonomy. Washington DC: Smithsonian Institution Press. 16 Groves C. 2012. Species concepts in primates. Am J Primatol 74:687–691.
C 2014 Wiley Periodicals, Inc. V
ISSUES
Species: Beasts of Burden ALFRED L. ROSENBERGER
Ernst Mayr (1904–2005) was the twentieth century’s most influential writer to wrestle with the species problem.1–4 The following draws heavily on his work, albeit without presumptuously claiming to mirror his thinking or present any original ideas. As a personal meditation, I am thinking mostly of platyrrhines. Following Mayr, I adhere to what is commonly called the Biological Species Concept (BSC) as a way of thinking about a species in the real-world biosphere as a taxon. I also hold to the idea that the Linnaean category called species has the same function as other categories: a linguistic tool for organizing and retrieving information about biodiversity while embodying evolutionary hypotheses. In other words, alpha taxonomy, the area of systematics that involves identifying, naming, and classifying species, is not purely an exercise in either biology or inventory because it involves communication as well. The burdensome work of the species category stems partly from tension created by the several purposes associated with the concept: the objective observation and examination of a fundamental biological phenomenon, the collection and interpretation of data in a selective context of relevance, and the intention to deploy scientific decisions as a form of communication within a dynamic but highly structured language system.
The biological entity we call species is a bearer of names steeped in language traditions. The binominal labels we give to species are subject to whim, imagination, cultural influences and other idiosyncrasies. This is why name use, name giving, and name change are dynamic conditions, even when the science behind
One of the author’s current projects examines how and why many studies of platyrrhine phylogenetic interrelationships produce both corresponding and conflicting hyptheses. In the Dominican Republic, Rosenberger’s cave-diving scuba team is collecting fossil primates from bone-rich freshwater caverns. With Tim Smith and Valerie DeLeon, he is also working on the influence of cranial development on primate orbital morphology. E-mail: [email protected]
Key words: platyrrhines; taxonomy; systematics; Biological Species Concept; New World monkeys
C 2014 Wiley Periodicals, Inc. V
DOI: 10.1002/evan.21392 Published online in Wiley Online Library (wileyonlinelibrary.com).
nomenclatural decisions is intended to be objective, rigorous, and enduring. As a taxonomic category, the operation of identifying and ranking of a set of animals as a species is also inevitably subject to case-bycase utilitarian arbitrariness because every species has its own character, phylogenetic setting, and history, much like every classification. In practice, that slot in the Linnaean hierarchy is delimited through a context-specific analysis that often involves comparing the biodiversity of closely related species-level groups belonging to the same superordinated genus. This makes across-the-board standardization of operating procedures – the biology part of the species problem – difficult. For example, we all agree that the monotypic species Callimico goeldii has no intrinsic properties underlying its characterization as a species apart from the same set we use at the genus level, whereas Alouatta seniculus is delimited by biologically meaningful characteristics designed to
distinguish it uniquely from perhaps a half-dozen other howler monkeys. These same traits are not meant to simultaneously distinguish A. seniculus from, say, Ateles, Lagothrix of Callimico. A totally different yardstick is properly used in the alpha taxonomy of Cebus, an animal built from a different blueprint with different historical and ecological constraints and different genetic potential. This means that the information we use to sort species need not always be the same as long as it reflects, in theory at least, reproductive isolation per se or its consequences. Because this is a relational concept, the difficulty of imposing consistent, nonarbitrary criteria to delimit species in the real world is further complicated by the likelihood that populations that are relevant for comparison probably are also evolving at different rates of species formation and senescence, while their phenotypes are simultaneously experiencing mosaic evolution. As a morphologist, I favor anatomical criteria during the discovery process and in assessing the power of diagnoses and differentia. I also welcome nonmorphological evidence, when available, to assess the broader picture. That picture includes genes, chromosomes, karyotypes, vocalizations, fixed-action patterns, mating strategies, social organization, diet, locomotion, activity cycle, hybridization, and sympatry – attributes that are currently and historically important in situating a species in nature, which is the point of evolution. The presence and incidence of directly observable phenotypic and genotypic characteristics are byproducts of reproduction and/or reproductive isolation, making them perfectly valid indicators of species status under the BSC, as are the inferred
28 Rosenberger
behaviors and adaptations arising from them. Sympatry has a different status; it is a nongenetic auxiliary criterion used in combination with breeding information. As a practical matter, we ideally assume that species are discrete entities, but also relational constructs, which makes operational rules easier to express than to implement. As a starting point, my approach is to develop a differential diagnosis by attempting to establish group-specific morphological patterns within a comparative taxonomic framework. These patterns are presumed a priori to be homologous structures. Why patterns and why homologies? Because species are, in effect, distributed networks of reproductively compatible individuals bearing unique combinations of genes that are likely to be manifest or mirrored at the phenotypic level as a design. In comparing samples of two potentially different but related “maybe species,” if the degree of pattern discontinuity within and/or across systems surpasses their evident continuity, a species difference may be indicated as a hypothesis. But the final determination of status also must square with taxonomic practice, meaning that reasonable taxonomic compromises may be in order to maintain nomenclatural stability for purposes of classification and dialogue. In a morphological study of three allopatric populations of Leontopithecus, for example, Coimbra-Filho and I found that craniodental metrics, body proportions, and pelage consistently separated the northern form, chrysopygus, from the two southern populations, rosalia and chrysomelas, which were less distinct from chrysomelas and from one another.5 Incisor morphology suggested that this separation reflected adaptive dietary differences, and this inference shaped our taxonomic thinking. These observations paralleled vocalization evidence discovered later.6 The taxonomic rank of Leontopithecus populations was then in question: species or subspecies?7 We elected to rank all three groups as species in order to maintain a balanced, stabile nomenclature for the genus, even though, from a purely biological perspective, we felt most comfortable with the hypothesis that, minimally, there were two.
ISSUES
Advances in the implementation of three-dimensional scanning technologies, geometric morphometrics, and other analytical tools promise to revolutionize these sorts of taxonomic morphological comparisons by making it possible to document mathematically complex anatomical patterns, such as those on the occlusal surface of molar teeth, and to make quantitative comparisons of the true geometries of shape among populations. During decades without such highpowered approaches, beginning with G.G. Simpson, simple statistics of tooth size for fossil populations served as an important benchmark for delimiting species boundaries.8 In studies of
Why patterns and why homologies? Because species are, in effect, distributed networks of reproductively compatible individuals bearing unique combinations of genes that are likely to be manifest or mirrored at the phenotypic level as a design.
fossil platyrrhines, however, a tendency to rely exclusively on simple metrics without considering modern ecology – a species’ place in nature – has, in several cases, created a false sense of taxonomic rigor. The basic problem is that sympatric extant platyrrhine species, across the gamut of size classes, overlap considerably in tooth size and body mass. The late Oligocene Branisella boliviana and Szalatavus attricuspis are a case in point. Specimens come from the same area (not locality) and overlap in size, but the type specimens’ upper molars differ in occlusal morphology. Thus, pooling all the known specimens of molars (actually very few) and showing that they are grossly similar in size is insufficient reason to merge these species into one.9 A prior critique that sought to sink Szatalavus also conflated more
than it clarified by erecting a false equivalence among samples. Rather than demonstrating continuity in the occlusal anatomy of the two types, it claimed that a new two-toothed fossil, which the authors simply assumed to be Branisella, undermined Szalatavus because its M1 “closely resembled that of ‘Szalatavus,’ while M2 of the same specimen has a well developed distolingual cingulum and a moderate hypocone, being different from M2 of “Szalatavus.”10 Reasoning differently even on this basis alone, one might recognize that the new fossil was Szalatavus instead of Branisella or that all three samples differ from one another in asymmetrical ways, adding credibility to the hypothesis that more than one taxon exists at La Salla, Bolivia. A similar situation occurs with Aotus dindensis from the middle Miocene of Colombia. Designating it as synonymous with Mohanamico hershkovitzi because the molar teeth of these taxa overlap in size does not account for documented differences in the morphology of the incisors, premolars, and mandible. It also does not account for the instructive parallel of the anatomically similar Aotus and Saimiri: same body size; grossly similar lower molars; often sympatric; very distinct genera.11 Both these examples are laudable for their efforts to be objective through quantification and conservative in trying not to overestimate taxonomic abundance at fossil sites where faunas have yet to be adequately sampled. But they are also weakened by taking a narrow approach to morphology and presuming that fossil sites can be well interpreted without regard to the evidence of species packing and biodiversity among contemporary related animals. Obviously, rare and fragmentary fossils pose inevitable taxonomic difficulties. But the species-level systematics of living platyrrhines is being roiled today by work, some methodologically quite sophisticated, which seems like a throwback to preModern Synthesis typological thinking.12,13 A re-classification frenzy is taking place, ostensibly based on adoption of the Phylogenetic Species Concept (PSC), a reworking of Simpson’s Evolutionary Species Concept, which focuses on lineages.14 This reclassification seizes on populations previously listed as subspecies distinguished by key characters, initially meant for identification purposes,
Species: Beasts of Burden 29
ISSUES
then raises them to species status without contributing much more biology.15 Taxonomically, one wonders where, exactly, this will lead us. Flattening the Linnaean hierarchy below the genus into a single rank that does not distinguish the subgenus, species, and subspecies makes the interconnected genus- and species-level categories outliers in the hierarchy, the only ones not employing nested sets. This makes it harder to communicate interrelationships, similarities, and disparities. It also gives the impression that all these species entities exist in an equivalent steady state, while we fully expect that the populations in question are and have been evolving at different rates toward extinction, speciation, or phyletic evolution. Biologically, what is to be gained? One leading advocate of the PSC criticizes the BSC as being unscientific because reproductive isolation is said to be untestable in cases of allopatry.16 There are several reasons why this position is subjective and hardly defensible. For starters, it is not justifiable to accept the scientific reality of species under conditions of sympatry but simultaneously deny the broader concept of species because certain geographic or temporal situations make testing difficult. Furthermore, the criticism of nontestability sets forth the exclusive requirement to perform direct tests of reproductive isolation, thus rejecting any other form of indirect evidence of genetic separation that can reflect infertility. This is not too different from disqualifying as unscientific the hypothesis that evolution by natural selection is responsible for the high-powered human brain because no one has been eyewitness to the underlying neuron-specific gene changes that occurred across the generations. It seems retrograde to redefine a biological reality that has been tested countless times in order to tidy up method (just as the phylogeny concept has been unjustifiably redefined as cladistic affinities so that the only relevant evidence of descent is synapomorphy). The preferred evidence in applying the PSC is said to consist of any phenotypic trait that defines a population as a discrete genetic linage or, preferably, a cladogram that can be used to test the existence of lineages
by revealing a topology. Here again, the logic is soft. Lineages and species are not equivalent phenomena. The different streams of phylogeny identified in these studies may be demes, species, or anything in between, some on their way to becoming fullfledged reproductive isolates, some careening toward extinction before achieving that status. There is no way to discern whether a simple or randomly selected trait difference in phenotype or a branch on a tree is a consequence of meaningful genetic separation in the sense of species, which are fundamental units of evolution as the highest level of biological organization determined by the genome. There is no logical connection between an assertion that the cladistic affinities of populations are known and the operation of ranking them, biologically or taxonomically. Moreover, the reliability of molecular cladograms, on which much species inflation relies for its justification, is coming under increasing scrutiny because of concern that they often represent gene trees rather than species trees. Like evolution by natural selection, the BSC is built on many testable interrelated hypotheses. The same can be said of the PSC. But making part of the species concept as easy as possible to refute by robbing the entire phenomenon of its evolutionary richness contributes little to our understanding of nature. Primatologists can take heart that other fields of science, which, by their very nature, are more solidly ingrained in mathematics and prediction, and powered by the most advanced instruments and computers ever built, share some of the same quandaries we do. Remember when Pluto was a planet? Well, not any more. What changed? Perhaps the view of prominent Berkeley astronomer Geoff Marcy offers us solace in his response (http://www. space.com/3142-planets-defined.html): “Your questions imply that a definition of the word ‘planet’ is useful scientifically. That is a view not shared by many professional planetary scientists . . . . The astrophysics of planetary bodies is so rich and complex that defining ‘planet’ has never been an issue under discussion among professionals. The taxonomy of asteroids, comets, moons, planets and brown dwarfs is far too limited to capture the
diversity of their origins and internal constitutions.” (Postscript: Marcy is credited with identifying some 250 planets outside our solar system.)
ACKNOWLEDGMENTS Thanks to John Fleagle for inviting me to offer thoughts on the species problem and for his insightful critique of the draft manuscript, which was significantly improved by his editorial input. Nevertheless, I remain responsible for the stubborn errors and sprinkling of fuzzy logic.
REFERENCES 1 Mayr E. 1963. Animal species and evolution. Cambridge: Harvard University Press. 2 Mayr E. 1982. The growth of biological thought. Cambridge: Harvard University Press. 3 Mayr E. 1991. Principles of systematic zoology. New York: McGraw-Hill. 4 Mayr E. 1977. Evolution and the diversity of life: selected essays. Cambridge: Harvard University Press. 5 Rosenberger AL, Coimbra-Filho AF. 1984. Morphology, taxonomic status and affinities of the lion tamarins, genus Leontopithecus (Callitrichinae, Cebidae). Folia Primatol 42:149–179. 6 Snowden CT, Hodun A, Rosenberger AL, et al. 1986. Long-call structure and its relation to taxonomy in lion tamarins. Am J Primatol 11:253–261. 7 Hershkovitz P. 1977. Living New World monkeys (Platyrrhini): with an Introduction to Primates. Chicago: University of Chicago Press. 8 Simpson GG, Roe A. 1939. Quantitative zoology. New York: McGraw Hill. 9 Takai M, Anaya F, Shigehara N, et al. 2000. New fossil materials of the earliest New World monkey, Branisella boliviana, and the problem of platyrrhine origins. Am J Phys Anthropol 111:263–281. 10 Takai M, Anaya F. 1996. New specimens of the oldest fossil platyrrhine, Branisella boliviana, from Salla, Bolivia. Am J Phys Anthropol 99:301–317. 11 Rosenberger AL, Setoguchi T, Shigehara N. 1990. The fossil record of callitrichine primates. J Hum Evol 19:209–236. 12 Zachos FE, Apollonio M, B€ armann EV, et al. 2012. Species inflation and taxonomic artefacts; a critical comment on recent trends in mammalian classification. Mamm Biol 78:1–6. 13 Rosenberger AL. 2012. New World monkey nightmares: science, art, use, and abuse (?) in Platyrrhine taxonomic nomenclature. Am J Primatol 74:692–695. 14 Simpson GG. 1951. The species concept. Evolution 5:285–298. 15 Groves C. 2000. Primate taxonomy. Washington DC: Smithsonian Institution Press. 16 Groves C. 2012. Species concepts in primates. Am J Primatol 74:687–691.
C 2014 Wiley Periodicals, Inc. V
