Taxonomy and Evolution of Gibbons

28 Primatology and AnthropologyTaxonomy and Evolution of GibbonsTHOMAS GEISSMANNThree different data sets of approximately equal size were collected in order toassess their relevance for a reconstruction of gibbon phylogeny with cladisticmethods. Set 1 uses characteristics of fur coloration, set 2 consists mainly ofmorphological and anatomical data, and set 3 consists of vocal data.Because the fossil history of gibbonsis virtually unknown, 1,2 gibbonevolution can only be reconstructedfrom a comparative analysis of evolutionarilyinformative characteristicsof modern gibbons and, to some degree,of related primate taxa that canbe used as outgroups.Several studies have tried to reconstructgibbon phylogeny, using furcoloration, morphological, vocal, ormolecular data. 3–10 Each study produceda different result. As a consequence,the relationships among thevarious hylobatids are under debate,and even the evolutionary relationshipsamong the major groups of gibbonsremain unresolved. 11For the present study, three differentdata sets of approximately equalsize were collected in order to assesstheir relevance for a reconstruction ofgibbon phylogeny with cladistic methods.Set 1 uses characteristics of furcoloration, set 2 consists mainly ofmorphological and anatomical data,and set 3 consists of vocal data.Thomas Geissmann is at the Institut fürZoologie, Tierärztliche Hochschule Hannover,Bünteweg 17, D-30559 Hannover,Germany (E-mail: thomas.geissmann@gibbons.de).Key words: gibbons; taxonomy; evolution; Hylobatidae;phylogeny; vocalizations; fur colorationEvolutionary Anthropology, Suppl 1:28–31 (2002)DOI 10.1002/evan.10047Published online in Wiley InterScience(www.interscience.wiley.com).© 2002 Wiley-Liss, Inc.MATERIALS AND METHODSMost recent studies agree that thereare four distinct groups of gibbons,which are here referred to as genera.Table 1 presents the classificationused for the present study. 12,13The three large data sets that wereanalyzed are listed in Table 2. Theywere analyzed separately and comparedwith one another using partition-homogeneitytests. 14Phylogenetic analyses were conductedwith unweighted charactersusing the PAUP program, version4.0b3a. 14 Two types of maximumparsimonytrees were generated: abranch-and-bound search of the samedata set in PAUP yielded the shortesttree, i.e., a single most parsimoniouscladogram, and the bootstrap optionof PAUP was used to examine the robustnessof internal nodes.A hypothetical “ancestor” was usedas an outgroup. This “ancestor” wasassembled using primitive characterstates wherever they could be reconstructedor plausibly assumed.The following standard measureswere calculated in order to assess the“quality” of trees: consistency index(CI), retention index (RI), and therescaled consistency index (RC), 15,16as well as the number of bootstrapvalues larger than 50. The indices basicallyrange from 0–1, with higherRC values indicating that charactersin the data set are more congruentwith each other and the tree.Mean pairwise distances betweentaxa (i.e., the sum of their characterstate differences, divided by the numberof characters 14 ) were calculatedwith each data set. Three types of distanceswere compared: 1) distancesamong subspecies of the same species(n 4), 2) distances among species ofthe same genus (n 38), and 3) distancesamong members belonging todifferent genera (n 63). In order toincrease the sample of subspecies distances,H. muelleri was separated intoits subspecies (H. m. muelleri, H. m. funereus,and H. m. abbotti) for this partof the analysis.RESULTSBecause the three data sets are statisticallydifferent (P 0.01) whencompared with the partition-homogenetitytest, they are not combined inthis study.The trees based on fur coloration(Fig. 1a) strongly contradict all previouslypublished gibbon phylogenies.3–10 For instance, members ofthe lar group appear in the mostbasal positions of the two shortest trees(not shown). The lar group is not identifiedas a monophyletic group: Hylobatesklossii is shown as the sister taxonof the siamang (Symphalangus syndactylus),obviously because both gibbonsare completely black. In contrast, mostpreviously published phylogenies identifyH. klossii as a member of the genusHylobates, whereas the siamang isplaced in the genus Symphalangus.The trees based on “noncommunicatory”data (Fig. 1b) resemble thetrees of earlier studies far moreclosely. The data set has a relativelylow resolution because of the manymissing values in the data matrix.Many of the anatomical data werecollected during very early studies,when several gibbon taxa were simplynot known. As a result of theempty cells in the matrix, there isnot one single most parsimonioustree, but an ensemble of 117 shortesttrees.

Primatology and Anthropology 29Diploid NumberTABLE 1. Main Divisions of Genus Hylobates aOther DivisionGenusof Chromosomes Names Species English-Language NameHylobates 44 lar group H. agilis b Agile gibbonH. klossii Kloss’s gibbonH. lar White-handed gibbonH. moloch Silvery gibbonH. muelleri c Müller’s gibbonH. pileatus Pileated gibbonBunopithecus 38 B. hoolock HoolockNomascus 52 concolor group, N. concolor Western black crested gibboncrested gibbonsN. sp. cf. nasutus Eastern black crested gibbonN. gabriellae Yellow-cheeked crested gibbonN. leucogenys d White-cheeked crested gibbonSymphalangus 50 S. syndactylus Siamanga Raising the four main groups of gibbons to genus rank follows the consensus reached at the workshop “Primate Taxonomy forthe New Millennium” (February 25–29, 2000, Orlando, Florida), to be presented in a future publication, and reference 20.b Including H. agilis albibarbis.c Including H. muelleri abbotti and H. muelleri funereus.d Including N. leucogenys siki.The trees based on vocal characteristics(Fig. 1c) are similar to those ofdata set 2. Bunopithecus hoolock appearsin an unusually basal position inthe tree, however, because of its manyapparently primitive vocal characteristics.Hylobates klossii is identified asthe sister taxon of H. moloch.The various standard measures calculatedin order to assess the “quality”of trees are shown in Table 2 (bootstrapvalues, CI, RI, and RC). Data set1 scores distinctly worse in all variablesthan the other two data sets. Thedifference between data sets 2 and 3 isless distinct in these variables. Nevertheless,data set 3 has higher values inall of these variables (especially inbootstrap values) than set 2, indicatingthat particularly “good” trees aregenerated with vocal data.The mean pairwise distances betweengibbon taxa are shown in Figure2. Fur coloration data are able todifferentiate not only between speciesand genera, but also between subspecies.Both the range and standard deviationof the distances are extremelylarge, however, and show a broadoverlap among different systematiclevels. This means that differences infur coloration provide little informationon the genetic distance betweenany two taxa. Two subspecies may bemore different in fur coloration thanmembers of two different genera.“Noncommunicatory” data, on theother hand, do not differentiate wellbetween subspecies, and the standarddeviations for subspecies and speciesoverlap. These data differentiate betweensome, but not all, species. Thepairwise differences between membersof different genera are distinctlygreater than those between species,and standard deviations for speciesand genera do not overlap.The vocal data produce similar resultsto the “noncommunicatory”data. This data set does not differentiatebetween subspecies at all, butdifferentiates well between speciesand genera. Standard deviations ofsubspecies, species, or genera do notoverlap.The three data sets also differ in thespecies groups they support (Table 2).TABLE 2. Various Parameters of Trees in Figure 1, and Topologies of Gibbon Taxa Supported by These Trees aData Set Set 1 Set 2 Set 3Number of Characters 37 34 34Type of Tree S B S B S BTree length 158 191 73 78 88 90Number of bootstrap values above 50 5 5 10Number of shortest trees 2 117 1Consistency index (CI) 0.49 0.41 0.66 0.62 0.67 0.66Retention index (RI) 0.63 0.47 0.80 0.77 0.81 0.80Rescaled consistency index (RC) 0.31 0.19 0.53 0.47 0.54 0.52Monophyly of agilis and albibarbis Monophyly of concolor group Monophyly of lar group (44-chromosome gibbons)H. hoolock as sister group of lar group H. klossii sister of all other members of largroup a et 1, fur coloration data; Set 2, “noncommunicatory” data; Set 3, vocal data. S, shortest tree; B, Bootstrap 50% majority-ruleconsensus tree; , supported; , not supported.

30 Primatology and AnthropologyFigure 1. Maximum parsimony trees (bootstrap 50% majority-rule consensus) of three differentdata sets. a: Fur coloration data. b: “Noncommunicatory” data. c: Vocal data.The sets do agree, however, in supportingmonophyly of the concolorgroup at least to some degree, and innot supporting H. klossii as the sistertaxon to the other members of the largroup.differ among individuals of the sametaxon (polymorphisms).“Noncommunicatory” data wouldseem to be much better suited for reconstructingthe gibbon phylogeny,but of the data sets under study, thevocal data produce the most reliablephylogeny. The trees generated usingvocal data suggest that:Gibbons of the genus Hylobates(lar group) and the genusNomascus (crested gibbons orconcolor group) are each monophyleticgroups. There is weaksupport for a sister-group relationshipbetween the concolorand lar groups.Hylobates klossii is neither thesister taxon of the siamang(Symphalangus syndactylus), assuggested by some early studies,17,18 nor the sister taxon orthe most basal group of the largroup, in contrast to many previousstudies. 4–6,9,19 This speciesis a fully integrated member ofthe lar group, 7 and apparentlythe sister taxon of H. moloch.The same conclusion was independentlyreached by a studyusing a much smaller set of vocalcharacteristics (including degreeof sex-specificity of the vocalrepertoire, occurrence of solosongs, and preference for a specifictime of day for song production).10Bunopithecus hoolock may bemore basal than previously believed.Most earlier studies recognizedthis species as the sisterDISCUSSIONA cladistic analysis suggests that thetempo of evolutionary change differsamong the data sets under study, similarto DNA sequences derived fromdifferent parts of the genome. Fur colorationcharacters appear to changeconsiderably faster than either “noncommunicatory”or vocal characters.The three data sets produce differentresults, and each set appears to besuited to the analysis of different levelsof resolution within the hylobatidradiation. Fur coloration characteristicsappear to provide little informationfor a gibbon phylogeny, but maybe valuable tools for subspecies identification,in contrast to most charactersof the other two data sets. In addition,many fur coloration charactersFigure 2. Pairwise character differences between (left to right) different subspecies (Ssp.) ofthe same species, species (Sp.) of the same genus, and genera (G.) of the same genus.a: Fur coloration data. b: “Noncommunicatory” data. c: Vocal data. Each boxplot showsmean value (horizontal line through box), standard deviation (box), and range (“whiskers”).

Primatology and Anthropology 31taxon to the lar group, 4,5,9whereas a more basal positionhas only rarely been suggested.6,10,21The radiation of the four genera andof the species within the genera is notreliably resolved with these preliminaryresults. In future studies, the reliabilityof the trees may be improvedby integrating data from several datasets, by weighting the characters, andby supplementing the missing variablesin data set 2.REFERENCES1. Fleagle JG. 1999. Primate adaptation and evolution.San Diego and London: Academic Press.2. Geissmann T. 1993. Evolution of communicationin gibbons (Hylobatidae). Ph.D. thesis, Universityof Zürich, Zürich, Switzerland.3. Groves CP. 1972. Systematics and phylogenyof gibbons. In: Rumbaugh DM, editor. Gibbonand siamang, volume 1. Basel and New York:Karger. p 1–89.4. Chivers DJ. 1977. The lesser apes. In: PrinceRainier III of Monaco, Bourne GH, editors. Primateconservation. New York: Academic Press. p539–598.5. Haimoff EH, Chivers DJ, Gittins SP, WhittenAJ. 1982. A phylogeny of gibbons (Hylobates spp.)based on morphological and behavioural characters.Folia Primatol (Basel) 39:213–237.6. Creel N, Preuschoft H. 1984. Systematics ofthe lesser apes: a quantitative taxonomic analysisof craniometric and other variables. In:Preuschoft H, Chivers DJ, Brockelman WY, CreelN, editors. The lesser apes. Evolutionary and behaviouralbiology. Edinburgh: Edinburgh UniversityPress. p 562–613.7. Garza JC, Woodruff DS. 1992. A phylogeneticstudy of the gibbons (Hylobates) using DNA obtainednon-invasively from hair. Mol PhylogenetEvol 1:202–210.8. Hayashi S, Hayasaka K, Takenaka O, Horai S.1995. Molecular phylogeny of gibbons inferredfrom mitochondrial DNA sequences: preliminaryreport. J Mol Evol 41:359–365.9. Purvis A. 1995. A composite estimate of primatephylogeny. Philos Trans R Soc Lond [Biol]348:405–421.10. Geissmann T. 2002. Duet-splitting and theevolution of gibbon songs. Biol Rev 77:57–76.11. Hall LM, Jones DS, Wood BA. 1998. Evolutionof the gibbon subgenera inferred from cytochromeb DNA sequence data. Mol PhylogenetEvol 10:281–286.12. Geissmann T. 1995. Gibbon systematics andspecies identification. Int Zoo News 42:467–501.13. Geissmann T. 1997. New sounds from thecrested gibbons (Hylobates concolor group): firstresults of a systematic revision. In: Zissler D,editor. Verhandlungen der Deutschen ZoologischenGesellschaft: Kurzpublikationen—short communications, 90. Jahresversammlung1997 in Mainz. Stuttgart: Gustav Fischer. p 170.14. Swofford DL. 2000. PAUP*. Phylogeneticanalysis using parsimony (*and other methods).Version 4. Sunderland, MA: Sinauer Associates.15. Kitching IJ, Forey PL, Humphries CJ, WilliamsDM. 1998. Cladistics. Second edition. Thetheory and practice of parsimony analysis. Oxfordand New York: Oxford University Press.16. Maddison WP, Maddison DR 1992. Mac-Clade: analysis of phylogeny and character evolution.Version 3.0. Sunderland, MA: Sinauer Associates.17. Elliot DG. 1913. A review of the primates,volume 3: Anthropoidea, Miopithecus to Pan.New York: American Museum of Natural History.18. Miller GS Jr. 1903. Seventy new Malayanmammals. Smithson Misc Coll 45:1–73.19. Schultz AH. 1933. Observations on thegrowth, classification and evolutionary specializationof gibbons and siamangs. Hum Biol5:212–255, 385–428.20. Roos C, Geissmann I. 2001. Molecular phylogenyof the major hylobatid divisions. Mol PhylogenetEvol 19:486–494.21. Zehr SM. 1999. A nuclear and mitochondrialphylogeny of the lesser apes (Primates, genusHylobates). PhD thesis, Harvard University, Cambridge,MA.