Int J Primatol (2008) 29:1313–1339 DOI 10.1007/s10764-008-9306-1

Evolutionary Acceleration and Divergence in Procolobus kirkii Katarzyna Nowak & Andrea Cardini & Sarah Elton

Received: 25 January 2008 / Accepted: 24 June 2008 / Published online: 10 October 2008 # Springer Science + Business Media, LLC 2008

Abstract We investigated the role of geographical insularity in divergence and speciation of Procolobus kirkii by examining cranial morphology. The sample (n= 369) included museum specimens of Procolobus spp. and recently deceased individuals of P. kirkii from the main island of Zanzibar and 2 smaller islands in the archipelago. Geometric morphometrics evinced pronounced divergence of Procolobus kirkii from mainland Procolobus, including members of P. badius ssp., P. pennantii ssp., P. rufomitratus, P. gordonorum and also representatives of the assemblage of red colobus populations from Central Equatorial Africa. Procolobus kirkii has a small cranium, consistent with the island rule for large mammals, reduced sexual dimorphism consistent with Rensch’s rule, and a distinct cranial form. Analyses of phenotypic variance of Procolobus kirkii gave no evidence for population bottlenecks in the history of the species, but there is a clear indication that the species has experienced accelerated morphological evolution of size, probably as a result of insularity. Their highly distinctive morphology lends weight to the argument that they are a unique insular endemic species in need of active conservation. Keywords endemism . geometric morphometrics . island rule . Procolobus . taxonomy

K. Nowak (*) Wildlife Conservation Research Unit (WildCRU), University of Oxford, Tubney House, Abingdon Road, Abingdon OX13 5QL, UK e-mail: [email protected] A. Cardini Museo di Paleobiologia e dell’Orto Botanico, Universitá di Modena e Reggio Emilia, 41100 Modena, Italy A. Cardini : S. Elton Hull York Medical School, The University of Hull, Hull HU6 7RX, UK

1314

K. Nowak et al.

Introduction The study of peripheral populations can help to explain ecogeographic patterns and determine the consequences of isolation and habitat fragmentation. In particular, islands afford special opportunities to study speciation, dynamism of morphological traits, and the mechanisms that influence changes in body size (Goltsman et al. 2005; Lomolino et al. 2006). Studies on islands can help to establish rates and patterns of adaptation and inform conservation decisions. In particular, the morphological study of young and isolated populations can provide clues to the evolutionary changes that occur after population crashes and periods of environmental change. Accurate quantitative comparisons based on modern morphometric methods are effective in understanding the factors that contribute to transformations in form, in identifying endemisms, and in reconstructing patterns of evolutionary change and population divergence (Cardini and O’Higgins 2004; Cardini et al. 2007b). Such studies have obvious relevance in reconstructing the morphological evolution and divergence of primate species, with important implications for the conservation of primate populations inhabiting fragmented habitats and subjected to anthropogenic pressure. The study of internal morphology, including crania, is fundamental to the recognition of biological variation and hence uniqueness. One primate isolate whose morphological evolution has remained largely unstudied is Zanzibar red colobus, Procolobus kirkii Gray 1868. It is 1 of 2 red colobus populations to inhabit islands (the other is Procolobus pennantii pennantii on Bioko Island, Equatorial Guinea). Though researchers conducted a few comparative morphological studies on red colobus (Colyn 1991; Schultz 1957; Verheyen 1962), their evolutionary history, including differentiation of peripheral populations, and taxonomy are poorly understood. Despite being a monophyletic group (Grubb et al. 2003; Ting 2008), red colobus are highly variable and patchily distributed across Africa in moist lowland tropical forests from the Zanzibar Archipelago in the Indian Ocean (Procolobus kirkii) to Senegal on the Atlantic coast (P. badius temminckii) (Oates and Davies 1994). A combination of climate change, vicariant speciation, competition with cercopithecines, and human disturbance are probably responsible for the current distribution of Procolobus (Burgess et al. 1998; Delson 1994), presumed to have previously had a wide and continuous distribution (Tappen 1960). Red colobus taxonomy has been in flux since the 1960s, and taxonomic revisions have been published by Groves (2001, 2007) and Grubb et al. (2003). Grubb et al. (2003, p. 1339) acknowledged that Procolobus represent “one of the thorniest taxonomic problems among the African primates” and stated that “if several species are to be recognized, it is difficult to determine where the lines between them are to be drawn” (p. 1341). They subdivided the genus into 2 subgenera, the monotypic Procolobus, which includes P. verus, and Piliocolobus. Piliocolobus comprises several species or subspecific assemblages: western red colobus Procolobus badius (3 subspecies distributed from Senegal to Ghana), P. pennantii (4 subspecies from Nigeria to Congo, including Bioko Island), the Central Equatorial assemblage (including P. sp. ellioti, P. sp. foai, P. sp. oustaleti, P. sp. tephrosceles, and P. sp. tholloni), P. rufomitratus (Tana River, Kenya), P. gordonorum (Udzungwa Mountains, Tanzania), and P. kirkii (Zanzibar Island, Tanzania: Grubb et al. 2003). For brevity, we follow the subdivisions of Procolobus (Piliocolobus) suggested by Grubb et al. (2003), and refer to the

Evolution in Zanzibar Red Colobus

1315

populations simply as species without necessarily implying that they need to be considered good biological species. Endemic to the Zanzibar Archipelago, and legally protected since 1919, Procolobus kirkii was categorized as Endangered by the IUCN in the late 1980s (Hilton-Taylor 2000) and listed on Appendix I of CITES in 2000. Their natural habitat has undergone significant reduction, and most populations now inhabit 3 discontinuous forest blocks (Nowak 2007). Procolobus kirkii is distinguished from other Colobinae on the basis of coat color and pattern, acoustics of male calls (Kingdon 1997; Napier 1985; Struhsaker 1975, 1981; Struhsaker and Leland 1980) and the medial frontal suture of the cranium, which remains open, a condition occasionally occurring in humans but suggested to be generally absent in mainland colobus (Schultz 1957; Verheyen 1962). Data on pelage and vocalizations, geographical distribution, and results from molecular studies indicate that Procolobus kirkii is phylogenetically closest to P. gordonorum (Kingdon 1981; Struhsaker and Leland 1980; Ting 2008; Ting et al. 2006), though on the basis of the studies by Ting (2008) it is unlikely that the 2 taxa were in contact in the very recent past as Rodgers (1981) suggested. Brandon-Jones (Rodgers et al. 1982) noted ≥1 similarity, a persistently unfused frontal suture, in Procolobus kirkii and another small-bodied East African red colobus, P. rufomitratus. However, there is no reason a priori to link the 2 species on this basis: the similarity may represent a small population effect such as drift (Brandon-Jones in Rodgers et al. 1982). In pelage and vocalization, Procolobus rufomitratus appears to be much more similar to the western P. tephrosceles than to the geographically more proximate P. kirkii and P. gordonorum (Struhsaker 1981; Struhsaker and Leland 1980) and these observations are consistent with results from molecular studies (Ting 2008; Ting et al. 2006). In addition to Procolobus kirkii, Zanzibar Island supports endemic and nearendemic populations of Ader’s duiker (Cephalophus adersi), Zanj elephant shrew (Rhynchocyon petersi adersi), Fischer’s turaco (Tauraco fischeri), and the presumably extinct Zanzibar leopard, Panthera pardus adersi (Kingdon 1990, 1997; Rodgers et al. 1982). Though researchers have tentatively suggested several hypotheses regarding the colonization of Zanzibar by Procolobus kirkii (Mturi 1991), there is little concrete evidence for any of them. It is likely that the Zanzibar Archipelago, and insular representatives of Procolobus kirkii, have been separated from the African mainland since at least the beginning of the major Holocene interglacial (Rodgers et al. 1982). However, recent molecular data indicate that Procolobus kirkii diverged from its closest relative, P. gordonorum, around 0.6± 0.2 Ma. Thus, the biogeographic patterns evident on Zanzibar may be linked to a complex history of vicariance due to connection and separation from the mainland. Little information exists on the past distribution of colobus on Zanzibar, but monkeys probably ranged more widely when forest was more widespread (Silkiluwasha 1981). Today, they are patchily distributed in mostly secondary vegetation from the northern Kiwengwa-Pongwe Forest Reserve to the southern peninsula in Muyuni on Unguja (Siex 2005, pers. comm.) and on 2 islands, Uzi and Vundwe, directly to the south of Unguja (Silkiluwasha 1981; Nowak 2007). The highest concentration of red colobus is in the Jozani-Chwaka Bay National Park (Siex and Struhsaker 1999), Zanzibar’s largest protected area. Procolobus kirkii also occurs as translocated populations in 2 forest reserves on Unguja and was introduced

1316

K. Nowak et al.

to the Ngezi Forest Reserve on Pemba (Silkiluwasha 1981; Struhsaker and Siex 1998). Persistence in these areas suggests that the colonizing ability of Zanzibar colobus may previously have been underestimated (Camperio Ciani et al. 2001) and may reflect their ability to exploit a greater variety of habitats than other red colobus species (Siex and Struhsaker 1999), a possible outcome of their island life. As the most geographically isolated Procolobus spp., studying the morphology of P. kirkii in comparison to that of their congeners may provide important clues about the patterns and processes of their evolutionary divergence. The study of evolutionary divergence in animal form on islands has a long history, dating back at least to Darwin’s time, and is crucial for understanding speciation models (Carson and Templeton 1984; Coyne and Orr 2004; Mayr 1963; Templeton 1980). Though primatologists have conducted several studies on island macaques (Fooden and Albrecht 1993), there is a surprising paucity of research on the morphological evolution of other insular primates, though the recent controversial discovery of Homo floresiensis (Brown et al. 2004) has renewed interest in the topic (Bromham and Cardillo 2007). Examining the morphology of Procolobus kirkii will therefore help to appreciate potentially unique aspects of their biology and facilitate further exploration of trends in morphological evolution of primates and other mammals on islands. One trend is the so-called island rule, the tendency toward gigantism of smaller species and dwarfism of larger species of mammals that inhabit islands (Foster 1964; Lomolino 1985; Van Valen 1973), which Lomolino et al. (2006) and Meiri et al. (2008) have recently challenged. Another is accelerated morphological evolution in insular mammals (Millien 2006, 2007), a phenomenon about which Perez-Claros and Aledo (2007) have raised doubts. Our principal aim was to estimate the morphological divergence of Procolobus kirkii with respect to its closest living relatives. We use the term divergence per Cardini and Elton (2008b), who followed a more relaxed convention than the rigorous one that implies evolutionary distance between a species and its ancestor. Thus, more simply, divergence indicates how much a species contributes to total disparity, i.e. morphological variation in a clade, relative to the grand mean of all species. To this end, we measured crania of most representatives of Procolobus via 3-dimensional coordinates of anatomical landmarks, and applied geometric morphometrics methods to investigate whether populations of P. kirkii 1) have significantly modified their size and shape over the last century; 2) show significant differences vs. congeneric populations on the mainland; 3) show a significant difference in variance, with phenotypic variance used as a proxy for genetic variance, which may indicate population bottlenecks that could have occurred either because of geographic isolation or in relation to recent, strong anthropogenic pressures; and 4) are characterized by an unusually divergent form, linked to insularity (Millien 2006).

Methods Sample Our sample comprised 369 adult specimens (230 female and 139 male) from collections of the National Museum of Natural History (Washington, DC), Museum of Comparative Zoology of Harvard University (Cambridge, MA), Museum für

Evolution in Zanzibar Red Colobus

1317

Naturkunde of Humboldt University (Berlin, Germany), Staatliches Museum für Naturkunde Karlsruhe (Karlsruhe, Germany), Senckenberg Natural History Museum (Frankfurt am Main, Germany), Royal Museum for Central Africa (Tervuren, Belgium), British Museum of Natural History and Hunterian Museum of the Royal College of Surgeons (London), Powell-Cotton Museum (Birchington, UK), and from Nowak’s field collections in Zanzibar. We determined the maturity of each specimen on the basis of third molar and canine eruption. Following the classification of Grubb et al. (2003), we identified taxa on the basis of geographic distribution (Colyn 1991) and published taxonomy of particular specimens (Colyn 1991). The assignments were generally congruent with information from museum catalogues. We excluded specimens that we could not taxonomically assign with confidence, and were cautious in specific identification and defining groups a priori. We included most groups of Procolobus in our sample (Table I). There was no available specimen of Procolobus pennantii pennantii, P. sp. langi, P. sp. lulindicus, or P. sp. parmentieri in the museums we visited. Only females of Procolobus gordonorum and P. p. bouvieri (n=1) are represented in our sample and a single male specimen represents P. p. epieni. Specimens of both sexes were available for Procolobus rufomitratus, but the male sample is a single individual. Thus, we excluded Procolobus pennantii bouvieri, P. p. epieni (for males), and P. rufomitratus from most analyses. We included P. (Procolobus) verus (sister clade of Piliocolobus) and 3 species of the closely related Colobus (Table I) as outgroups. Data Collection Cardini collected 3-dimensional coordinates of anatomical landmarks on crania and mandibles via a 3D digitizer (MicroScribe 3DX, Immersion Corporation). He Table I Taxa used in analysis and sample sizes Genus (Subgenus)

Procolobus (Piliocolobus)

Species

N Females

Males

verus

— badius temminckii waldroni — bouvieri epieni preussi — ellioti foai oustaleti tephrosceles tholloni —

33 36 11 15 4 1 — 31 5 16 3 9 7 4 20

10 23 4 6 — — 1 10 1 18 6 5 16 2 7

angolensis guereza polykomos

— — —

4 20 11

4 19 7

kirkii badius

pennantii gordonorum

rufomitratus sp.

Procolobus (Procolobus) Colobus

Subspecies

1318

K. Nowak et al.

digitized landmarks only on the left side to avoid redundant information in symmetric structures. The set (configuration) of 64 landmarks used for the analysis is in Fig. 1, and landmarks are in Table II. The landmarks correspond to a subset of cranial landmarks in previous studies on skull variation in Old World monkeys (Cardini and Elton 2007, 2008a, b, c; Cardini et al. 2007a). There were 1–4 missing landmarks in 5% of the specimens. We estimated values for them via intrasex species means per Cardini and Elton (2008a). Simulations by Cardini and Elton (2008a) on guenons and Cardini and Thorington (2006) on marmots showed that estimating such a small proportion of missing landmarks in a few specimens via means does not introduce any appreciable error in either size or shape. As in Cardini and Elton (2008a), we found negligible measurement error. Geometric Morphometrics We performed geometric morphometric analyses via Morpheus (Slice 1999), TPSSmall 1.20 (Rohlf 2006a), NTSYS-pc 2.2L (Rohlf 2006b), and Morphologika (O’Higgins and Jones 2006). Geometric morphometrics (Adams et al. 2004; Rohlf and Marcus 1993; Zelditch et al. 2004), now standard practice in much morphological research, compares forms via the information captured by Cartesian coordinates of sets (configurations) of topographically corresponding anatomical landmarks (Marcus et al. 2000). Differences in coordinates due to rotation and translation of specimens during data collection are removed (Procrustes superimposition, Rohlf and Slice 1990), and size and shape components of form are separated and analyzed via multivariate statistics. Size is measured as centroid size, which is the square root of the sum of squared distances between all landmarks and their centroid. The magnitude of shape differences between 2 configurations is measured via their Procrustes shape distance, the square root of the sum of squared differences Fig. 1 Landmark configuration (modified from Cardini et al. 2007a, study of vervet ecomorphology).

Evolution in Zanzibar Red Colobus

1319

Table II Definition and numbering of landmarks (L) L

Definition

1 2

Prosthion: antero-inferior point on projection of premaxilla between central incisors Prosthion2: antero-inferiormost point on premaxilla, equivalent to prosthion but between central and lateral incisors Posteriormost point of lateral incisor alveolus Anteriormost point of canine alveolus Mesial P3: most mesial point on P3 alveolus, projected onto alveolar margin Contact points between adjacent premolars/molars, projected labially onto alveolar margin Posterior midpoint onto alveolar margin of M3 Contact points between adjacent premolars/molars, projected lingually onto alveolar margin Posteriormost point of incisive foramen Meeting point of maxilla and palatine along midline Greater palatine foramen Point of maximum curvature on the posterior edge of the palatine Tip of posterior nasal spine Meeting point between the basisphenoid and basioccipital along midline Meeting point between the basisphenoid, basioccipital and petrous part of temporal bone Most medial point on the petrous part of temporal bone Most medial point of the foramen lacerum Meeting point of petrous part of temporal bone, alisphenoid and base of zygomatic process of temporal bone Anterior and posterior tip of the external auditory meatus Stylomastoid foramen Distal and medial extremities of jugular foramen Carotid foramen Basion: anterior-most point of foramen magnum Anterior and posterior extremities of occipital condyle along margin of foramen magnum Hypoglossal canal Center of condylar fossa Opisthion: posterior-most point of foramen magnum Inion: most posterior point of the cranium Most lateral meeting point of mastoid part of temporal bone and supraoccipital Nasospinale: inferior-most midline point of piriform aperture Point corresponding to largest width of piriform aperture Meeting point of nasal and premaxilla on margin of piriform aperture Rhinion: most anterior midline point on nasals Nasion: midline point on fronto-nasal suture Glabella: most forward projecting midline point of frontals at the level of the supraorbital ridges Supraorbital notch Frontomalare orbitale: where frontozygomatic suture crosses inner orbital rim Zygo-max superior: antero-superior point of zygomaticomaxillary suture taken at orbit rim Center of nasolacrimal foramen (fossa for lacrimal duct) Center of optic foramen Upper-most posterior point of maxilla (visibile through pterygomaxillary fissure) Frontomalare temporale: where frontozygomatic suture crosses lateral edge of zygoma Maximum curvature of anterior upper margin of zygomatic arch Zygo-max inferior: antero-inferior point of zygomaticomaxillary suture Zygo-temp superior: superior point of zygomaticotemporal suture on lateral face of zygomatic arch Zygo-temp inferior: infero-lateral point of zygomaticotemporal suture on lateral face of zygomatic arch Posterior-most point on curvature of anterior margin of zygomatic process of temporal bone Articular tubercle Distal-most point on post-glenoid process Posterior-most point of zygomatic process of temporal bone

3 4 5 6–9 10 11–14 15 16 17 18 19 20 21 22 23 24 25–26 27 28, 30 29 31 32, 35 33 34 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59

1320

K. Nowak et al.

Table II (continued) L

Definition

60 61

Foramen ovale (posterior inferior margin of pterygoid plate) Meeting point of zygomatic arch and alisphenoid on superior margin of pterygomaxillary fissure Meeting point of zygomatic arch, alisphenoid and frontal bone Bregma: junction of coronal and sagittal sutures Lambda: junction of sagittal and lamboid sutures

62 63 64

The terms anterior and posterior are used with reference to Fig. 1.

between corresponding landmarks of 2 superimposed landmark configurations. Zelditch et al. (2004). provided an extensive introduction to applications of geometric morphometrics in biology. Detailed mathematical descriptions of geometric morphometric methods are available in Bookstein (1991) and Dryden and Mardia (1998). Guidelines on how to implement linear statistical models in geometric morphometrics are provided in Rohlf (1998) and Klingenberg and Monteiro (2005). Statistics Owing to the large degree of sexual dimorphism in colobines, we performed all analyses using split-sex samples. Thus we avoided using correction factors for sexual dimorphism, which make results more difficult to interpret, and were able to recheck patterns we observed by comparing results for males and females. To investigate whether significant modifications occurred within skull morphology of Procolobus kirkii over the last century, we compared a sample of specimens that died of natural causes, probably in the last decade, and that Nowak collected in the field at 3 locations in Zanzibar between 2003 and 2005, with museum specimens that mostly dated from the first half of the 20th century. This is a very short time span for appreciable evolutionary differences to have occurred, but it is also a period during which direct (deforestation) and indirect (global warming) consequences of human activities may have profoundly affected life of an insular species like Procolobus kirkii. We also examined whether the skull morphologies of the pooled sample of Procolobus kirkii were significantly different from those of congeneric mainland populations. We used permutation tests to assess pairwise the significance of differences in shape and size between allochronic populations of Procolobus kirkii and between P. kirkii and other species. We tested differences in sample means via a nonparametric analysis of variance (ANOVA) wherein we compared the sum-ofsquares explained by group membership in the data with that for random permutations of group membership (Fontaneto et al. 2004). We set the significance level of a difference between 2 samples via the frequency with which a random permutation of the group affiliation of specimens explained as much or more variation between pseudosamples than that observed in the original samples. The advantage of using permutation tests instead of standard parametric statistics is that they are unaffected by the large number of shape variables relative to sample size.

Evolution in Zanzibar Red Colobus

1321

We used a principal component analysis (PCA) of shape variables, which identifies the axes of greatest variation in a sample, to illustrate spatial (similarity) relationships among study species. We drew scatterplots of species mean shapes together with their bootstrap confidence ellipses (Cardini and Elton 2008b) to show the relative position of the means and the variation around their estimates along the major axes of shape variance. We summarized size variation via box plots. We measured shape variance in each sample as the sum of variances of all shape variables and computed standard bootstrap confidence intervals (Manly 1997, p. 35). To test for significance of differences in variance between allochronic populations of Procolobus kirkii and between P. kirkii and all other species, which might indicate population bottlenecks in the recent history of P. kirkii, we performed a series of Levene’s tests on the Procrustes distances. The Levene’s test requires calculating the absolute value of the deviation of each individual from the sample mean (Van Valen 1978), which is satisfied via Procrustes distances to the mean shape or absolute differences to the mean size. One then compares the deviations via ANOVA. Though researchers generally consider the test relatively robust to departures from normality, we chose to perform a randomization version of the test that compares the observed F-statistic with the distribution obtained by randomly reassigning deviations from sample means to the samples. We estimated variation in size within each population sample by the variance of centroid size. We used resampling statistics (via the same procedure as for shape) to estimate confidence intervals and to test differences in variance of size between samples. The absence of a well-supported phylogeny of Procolobus makes it difficult to undertake tests of evolutionary rates and consequently the extent to which insularity of P. kirkii has accelerated its morphological evolution. However, we obtained clues to this aspect of morphological evolution by studying the contribution of Procolobus kirkii to the diversity of form observed among its living relatives. We measured the relative contribution of each population to shape variation of study species via partial disparities (Zelditch et al. 2004). Partial disparity (PD) of the ith species is given by: . PDi ¼ ðDi Þ2 ðN
1Þ wherein D is the shape (Procrustes) distance between the mean of the ith species and the grand mean of all species and N is the sample size. For variance, we computed standard bootstrap confidence intervals by bootstrapping original samples and repeating analyses to estimate bootstrap standard deviations which we used to calculate confidence intervals. We measured partial disparity of size via the same equation as for shape, with D now being equal to the difference between the mean size of the ith species and the grand mean of all species. We also estimated the relative contribution of species to shape variation via a measure that we called mean distinctiveness (MD), which is related but not identical to PD. MD is computed as the sum of squared differences (pairwise Procrustes distances or size differences) between the mean of a species and those of others relative to the total sum of pairwise squared differences between all means. PD measures the amount of variation accounted for by each species relative to the grand mean of all species. MD, instead, measures the amount of variation of each species relative to all others in the study group. MD is strictly related to PD but, by using

1322

K. Nowak et al.

pairwise differences between species means instead of differences to the grand mean, does not allow 1 or a few species further apart from the grand mean to exhaust most of the variation in the sample. We explored the use of this measure in a previous study (Cardini and Elton 2008b) as an alternative to PD, to reduce the influence of strong outliers that may bias the grand mean of the species. Resampling Statistics in Small Samples Owing to the paucity of specimens of Procolobus in museum collections, we computed tests and confidence intervals based on resampling statistics for very small samples (n≤5). However, analyses based on very few specimens are problematic and their outcome needs to be confirmed by further studies on larger samples. For instance, when computing confidence intervals or ellipses for mean shapes, the actual number of bootstrap samples is <1000 in all samples of ≤5 specimens because the maximum number of independent bootstrap samples is given by (2N −1)!/N! (N −1) (Zelditch et al. 2004) and thus is 3 with N=2, 10 with N=3, 70 with N=4, 756 with N = 5. Also, repeated randomized selection experiments to build progressively smaller samples from an original data set of ca. 400 vervet skulls (Cardini and Elton 2007) indicated that the variation around estimates of parameters such as shape variance increases as sample size decreases. Thus, one expects bootstrap confidence intervals to become larger in smaller samples. Mean size, standard deviation of size, and variance of shape are fairly accurate even in relatively small samples. In contrast, mean shapes are strongly affected by sampling error. Thus, one needs to interpret results involving samples of few individuals with caution.

Results Differences in Means Between Allochronic Populations of Procolobus kirkii Three of four comparisons of cranial form between allochronic populations of Procolobus kirkii are not significant. Mean size is virtually identical in the field (female, 224.8 mm, n=17; male, 235.7 mm, n=6) and museum (female, 227.2 mm, n=16; male, 231.4 mm, n=4) samples (variance explained via group differences and corresponding p values: females, 3.9%, p=0.2750; males, 13.0%, p=0.2843). In contrast, mean shapes differed significantly in females (PRD intergroup mean shapes, 0.0218; variance explained via group differences, 4.9%; p=0.0060) but not in males (PRD intergroup mean shapes, 0.0348; variance explained via group differences, 13.1%; p=0.2260). However, the Procrustes shape distance between means of allochronic populations was, in both sexes, ca. two-thirds of the smallest distance between Procolobus kirkii and any other species, and percentages of variance explained via group differences were generally much smaller than those of interspecific comparisons. Thus, a small shape difference between allochronic populations of Procolobus kirkii might be present, but intraspecific variation is smaller than interspecific differences, a conclusion supported by our observation that the majority of specimens of Procolobus kirkii grouped together to the exclusion of other species in a cluster

Evolution in Zanzibar Red Colobus

1323

analysis on the matrix of Procrustes shape distances among all individuals (separate sexes; results not shown). We thus performed all further analyses on pooled samples of Procolobus kirkii. Differences in Means Between Procolobus kirkii and Continental Species Procolobus kirkii is significantly different from virtually all other species in both shape and size (Table III). Within Procolobus, variance explained via interspecific differences ranged from 10% to 30% for shape and 30–90% for size. Procolobus rufomitratus was the only species that did not differ significantly in size from P. kirkii, though shape differences are highly significant. Results did not change after

Table III Test of mean differences in shape and size of Procolobus kirkii and other taxa Sex

Females

Males

a

Taxona

Procolobus badius badius P. b. temminckii P. b. waldroni P. gordonorum* P. pennantii preussi P. rufomitratus* P. sp. ellioti P. sp. foai* P. sp. oustaleti P. sp. tephrosceles P. sp. tholloni* P. verus Colobus angolensis* C. guereza C. polykomos P. badius badius P. b. temminckii* P. b. waldroni P. pennantii preussi P. sp. ellioti P. sp. foai P. sp. oustaleti* P. sp. tephrosceles P. sp. tholloni* P. verus C. angolensis* C. guereza C. polykomos

Shape

Size b

N

PRD

expl. %

36 11 15 4 31 5 16 3 9 7 4 20 4 20 11 23 4 6 10 18 6 5 16 2 7 4 19 7

0.0409 0.0458 0.0420 0.0344 0.0562 0.0519 0.0440 0.0505 0.0465 0.0436 0.0597 0.0570 0.0902 0.0867 0.0735 0.0528 0.0596 0.0545 0.0592 0.0547 0.0699 0.0684 0.0619 0.0865 0.0533 0.1097 0.1218 0.0993

13.6 13.6 13.3 4.7 25.3 12.2 14.6 7.7 13.5 10.3 12.5 23.8 25.3 41.6 29.8 18.3 25.2 22.7 27.1 21.5 33.8 31.1 26.3 33.3 23.5 52.6 58.0 51.1

c

d

p

mean CSe

expl. %

pd

0.0001 0.0001 0.0001 0.0043 0.0001 0.0001 0.0002 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0006 0.0001 0.0001 0.0001 0.0006 0.0004 0.0001 0.0163 0.0001 0.0015 0.0001 0.0002

240.7 239.7 235.5 250.4 261.1 229.3 246.8 245.2 251.3 245.0 254.0 209.2 263.6 261.4 265.5 256.0 248.4 249.5 266.3 270.0 276.8 277.6 271.8 277.2 216.3 288.4 291.3 289.5

52.8 38.6 30.8 56.8 86.7 3.1 59.0 39.5 66.8 49.0 61.9 56.8 73.4 82.7 85.7 62.3 61.0 41.7 83.9 89.8 93.5 88.5 83.0 90.0 70.8 91.9 87.5 93.5

0.0001 0.0001 0.0002 0.0001 0.0001 0.2972 0.0001 0.0003 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0011 0.0047 0.0001 0.0001 0.0004 0.0002 0.0001 0.0161 0.0002 0.0010 0.0001 0.0001

Asterisks (*) in this and other tables and figures are used to indicate smallest samples (N≤5). Procrustes shape distance between mean shapes. c Percentage of variance explained via mean differences. d Significance estimated using 10,000 random permutations; all significant differences with a type I error probability α=0.05 are also significant after a sequential Bonferroni correction for multiple comparisons via Holm’s method (Howell 2002). e Mean of centroid size (mm); female and male P. kirkii means were respectively: 225.8–234.0 mm. b

1324

K. Nowak et al.

we introduced a sequential Bonferroni correction for multiple comparisons via Holm’s method (Howell 2002). Scatterplots (Figs. 2 and 3) of specific mean shapes (first 2 principal components) suggested that Procolobus kirkii is somewhat intermediate between all other members of the subgenus Piliocolobus and the one representative of the subgenus Procolobus, P. verus. The most evident apomorphism of Procolobus kirkii was a considerably shorter face relative to those of all continental species except P. verus

Fig. 2 Female species mean shapes. Scatterplots of the first principal components (PCs) of shape variables, which together account for ca. 50% of total variance. Shape changes at extremes of the axes are illustrated using surface rendering. For each species except those with the smallest samples, variation around the mean is illustrated with 95% bootstrap confidence ellipses. Groups of taxonomic and biogeographic relevance (Colyn 1991; Grubb et al. 2003; Ting et al. pers. comm.) are below the scatterplot and emphasized using different styles for lines of confidence ellipses.

Evolution in Zanzibar Red Colobus

1325

Fig. 3 Male species mean shapes. Scatterplots of the first 2 PCs of shape variables (ca. 56% of total variance). See Fig. 2 for legend.

(Figs. 2 and 3 with surface rendering for shapes in the lower left corner of the scatterplot, corresponding to the region of the shape space occupied by Procolobus kirkii and P. verus). We confirmed the expansion of the neurocranium relative to the face of Procolobus kirkii by pairwise comparisons of mean shapes via Morpheus, and also found it in P. gordonorum besides P. verus (results not shown). The scatterplot of female mean shapes also indicates strong similarities between Procolobus kirkii and P. gordonorum. Besides clearly showing that Procolobus kirkii is the smallest colobus monkey in the analysis after P. verus, boxplots of size (Fig. 4) support results of permutation tests that suggested strong similarities in size between females of P. kirkii and P. rufomitratus.

1326

K. Nowak et al.

Fig. 4 Box-plot of female and male centroid size of species samples.

The range of sexual dimorphism (male size/female size) in Procolobus is 1.03 (P. kirkii) to 1.13 (P. sp. foai). Thus, male Procolobus kirkii are approximately the same size as females and sexual dimorphism is generally smaller than in other populations (Groves 2007). Differences in Population Variances Between Allochronic Populations of Procolobus kirkii or Between P. kirkii and Continental Species Shape and size variances do not differ significantly between allochronic populations of Procolobus kirkii in both females and males (Table IV). Observed variances are virtually identical in 3 of 4 comparisons. In the interspecific comparisons, none of the tests for differences in shape or size variances (Table IV) is significant after a sequential Bonferroni correction for multiple tests (Howell 2002). Exceptions are represented by females of Procolobus badius badius and P. rufomitratus, which had, respectively, larger and smaller shape

Evolution in Zanzibar Red Colobus

1327

Table IV Levene’s test of differences in shape and size variance within (W) Procolobus kirkii or between (B) P. kirkii and other taxa Variables

Taxon

Females VARa

Males 99% CIb Lower

Shape W Procolobus kirkii (museum) P. kirkii (field) B P. kirkii P. badius badius P. b. temminckii*M P. b. waldroni P. gordonorum*F P. pennantii preussi P. rufomitratus*F P. sp. ellioti P. sp. foai*F P. sp. oustaleti*M P. sp. tephrosceles P. sp. tholloni* P. verus Colobus angolensis* C. guereza C. polykomos Size W P. kirkii (museum) P. kirkii (field) B P. kirkii P. badius badius P. b. temminckii*M P. b. waldroni P. gordonorum*F P. pennantii preussi P. rufomitratus*F P. sp. ellioti P. sp. foai*F P. sp. oustaleti*M P. sp. tephrosceles P. sp. tholloni* P. verus Colobus angolensis* C. guereza C. polykomos a

pc

VAR

Upper

99% CI Lower

p Upper

0.00233 0.00185 0.00281 —

0.00257 0.00055 0.00458 —

0.00246 0.00245 0.00291 0.00316 0.00283 0.00256 0.00235 0.00129 0.00290 0.00257 0.00242 0.00283 0.00343 0.00265 0.00239 0.00276 0.00263 48.3 50.0 49.6 52.4 95.5 44.2 15.7 49.3 12.2 115.3 0.2 87.5 103.1 54.6 53.2 85.4 90.5 58.9

0.00232 0.00247 0.00292 0.00263 0.00311 — 0.00278 — 0.00281 0.00272 0.00304 0.00290 0.00275 0.00265 0.00287 0.00267 0.00276 58.7 24.1 37.9 77.9 12.3 185.0 — 73.2 — 35.3 27.1 120.1 97.1 4.5 32.3 136.5 152.0 89.9

0.00202 0.00214 0.00258 0.00269 0.00245 0.00133 0.00215 >0 0.00252 0.00065 0.00185 0.00226 0.00151 0.00235 0.00121 0.00240 0.00214 >0 12.9 15.4 21.0 >0 9.2 >0 24.8 0.6 23.6 0.0 11.0 >0 >0 13.4 >0 >0 >0

0.00289 0.00277 0.00325 0.00364 0.00321 0.00380 0.00255 0.00424 0.00328 0.00449 0.00298 0.00339 0.00536 0.00295 0.00357 0.00311 0.00311 100.9 87.2 83.8 83.8 195.9 79.2 36.2 73.8 23.8 207.0 0.5 163.9 231.4 116.5 92.9 188.5 136.2 138.2

0.6267 — 0.0068 0.0167 0.1326 0.1578 0.6362 0.0002 0.0563 0.0613 0.3538 0.7962 0.6135 0.2915 0.0684 0.1423 0.8904 — 0.6341 — 0.6542 0.3842 0.8939 0.3210 0.4468 0.2458 0.0700 0.0610 0.2155 0.4060 0.8549 0.7421 0.6117 0.0238 0.8848

0.00162 0.00188 0.00246 0.00117 0.00212 — 0.00224 — 0.00246 0.00182 0.00188 0.00243 0.05762 0.00189 0.00143 0.00235 0.00196 6.1 >0 11.0 19.8 >0 20.1 — >0 — 3.3 >0 >0 37.8 >0 >0 >0 53.2 12.5

0.00302 0.00305 0.00337 0.00409 0.00409 — 0.00331 — 0.00316 0.00362 0.00420 0.00337 0.00777 0.00341 0.00431 0.00298 0.00357 111.3 61.3 64.8 136 27 350 — 185 — 67 56 244 156 10 67 279 251 167

0.9140 — 0.0585 0.4908 0.2293 — 0.2341 — 0.0412 0.7771 0.4550 0.0499 0.0489 0.8274 0.8807 0.1432 0.6114 — 0.2958 — 0.3771 0.1487 0.0318 — 0.8423 — 0.6329 0.4237 0.1832 0.0693 0.1518 0.6212 0.0579 0.0122 0.1615

Shape or size (mm) variance. Standard 99% bootstrap confidence intervals estimated using 1000 bootstraps for all species samples except those with N≤5 (indicated by asterisks followed by F or M, if the asterisk refers to either females alone or males alone); lower limits were set to >0 whenever small observed variances and large bootstrap standard deviations led to negative values. c Significance estimated using 10000 random permutations; among the few differences significant with a type I error probability α=0.05, only those for shape variance between females of Procolobus kirkii and either P. badius badius or P. rufomitratus are still significant after a sequential Bonferroni correction for multiple comparisons using Holm’s method (Howell 2002). b

1328

K. Nowak et al.

variance than P. kirkii. Also, standard bootstrap confidence intervals largely overlapped and, as expected, tended to be larger in smaller samples. Form Distinctiveness Results of PD and MD analyses are in Table V. Though PD tended, as expected, to emphasize the few species that were further apart from the grand mean, both PD and MD analyses show the same patterns. Since the purpose of this analysis was to see if our data provide clues about whether cranial morphology in Procolobus kirkii has evolved faster than expected for a species with a relatively short evolutionary history, we focus on the results of the MD analysis. This is because the MD analysis spreads the variation among species means more evenly, and it is thus somewhat more conservative in its estimates of relative differences. A species which looks highly distinctive in the MD analysis will look even more distinctive in the PD analysis. Figures 5 and 6 contain MDs of all species and their standard bootstrap confidence intervals via profile plots. The percentage of shape variation accounted for by Procolobus kirkii is comparable to that of other species in the female sample. Also, confidence intervals for MDs largely overlapped among representatives of the subgenus Piliocolobus. In contrast, in the male sample, MD of Procolobus kirkii was large compared to its closest relatives (second only to that of P. sp. tholloni within Piliocolobus), and its confidence interval only partly overlap with those of other species. However, if shape MDs generally were not very large, results of MD analyses of size in both sexes left little doubt about the unusual divergence of Procolobus kirkii. Procolobus kirkii has the third (females) or second (male) largest MD among all species (including the outgroup) and very little or no overlap in confidence intervals with other species of the subgenus.

Discussion Size and Shape Divergence Morphological divergence in Zanzibar red colobus over the last century appears to have been negligible. However, at least in shape, there are small differences between allochronic populations that may be related to environmental changes that caused small shifts in gene frequencies or, more likely on such a short temporal scale, promoted plastic responses in the phenotype. With the exception of Procolobus verus, P. kirkii has the smallest cranium of all the species in our sample. In line with previous studies, cranial size within clades can be a proxy for overall body size (Cardini et al. 2007a; Fooden and Albrecht 1993). Therefore, our results are consistent with the expectation that large mammals, including primates, may become smaller on islands (Bromham and Cardillo 2007; Foster 1964; Lomolino 1985; Van Valen 1973). Though this island rule does not generalize to all taxa (Meiri et al. 2008), Goldman and Walsh (1997) and Kingdon (1990) noted a trend toward smaller size for Zanzibar leopards (Panthera pardus adersi). Thus, at least the few data that are available for mammals endemic to the Zanzibar Archipelago support the island rule.

Procolobus kirkii P. badius badius P. b. temminckii*M P. b. waldroni P. gordonorum*F P. pennantii preussi P. rufomitratus*F P. sp. ellioti P. sp. foai*F P. sp. oustaleti*M P. sp. tephrosceles P. sp. tholloni* P. verus Colobus angolensis* C. guereza C. polykomos P. kirkii P. badius badius P. b. temminckii*M P. b. waldroni P. gordonorum*F P. pennantii preussi P. rufomitratus*F P. sp. ellioti P. sp. foai*F P. sp. oustaleti*M P. sp. tephrosceles P. sp. tholloni* P. verus C. angolensis* C. guereza C. polykomos

Shape

5.2 4.6 4.3 4.4 6.3 5.8 6.8 4.3 5.5 4.2 4.6 5.8 10.3 10.9 9.5 7.4 8.6 3.4 3.6 4.5 3.5 6.7 6.8 3.2 3.1 3.6 3.1 4.2 21.8 7.9 6.9 9.0

MD

Females

4.8 4.2 3.5 3.9 5.1 5.3 6.1 3.8 3.8 3.5 3.8 4.3 9.1 9.3 8.2 6.6 6.8 3.0 2.5 3.4 2.8 5.2 5.2 2.7 3.1 2.4 2.6 2.1 18.4 3.1 4.6 5.9

Lower

99% CI

a

5.7 4.9 5.0 5.0 7.6 6.3 7.5 4.8 7.2 4.8 5.4 7.4 11.5 12.5 10.9 8.1 10.3 3.8 4.7 5.6 4.2 8.2 8.4 3.6 3.1 4.9 3.7 6.4 25.2 12.8 9.1 12.1

Upper

Standard 99% bootstrap confidence intervals (note 2 in Table III).

Size

Species

Variables

4.2 2.9 2.3 2.6 6.4 5.3 7.3 2.4 4.8 2.1 3.0 5.5 14.4 15.6 12.8 8.5 10.9 0.6 0.9 2.7 0.8 7.2 7.3 0.1 0.0 1.0 0.0 2.2 37.4 9.6 7.5 11.7

PD

Table V Mean distinctiveness (MD) and partial disparity (PD) of focal species

3.3 2.2 0.8 1.6 3.9 4.3 5.9 1.3 1.4 0.8 1.4 2.4 12.0 12.4 10.1 6.9 7.4 >0 >0 0.5 >0 4.2 4.1 >0 0.0 >0 >0 >0 30.6 >0 2.9 5.6

Lower

99% CI

5.1 3.6 3.8 3.7 9.0 6.3 8.7 3.4 8.1 3.4 4.6 8.6 16.8 18.7 15.5 10.0 14.4 1.4 3.1 5.0 2.1 10.2 10.6 0.9 0.0 3.5 1.1 6.5 44.2 19.4 12.0 17.9

Upper 7.3 4.8 5.9 5.9 — 6.0 — 4.5 5.3 4.9 4.7 8.0 11.0 11.3 12.0 8.4 11.3 4.2 5.8 5.5 — 3.6 — 3.8 4.6 4.8 3.9 4.7 22.6 7.8 9.0 8.3

MD

Males

6.2 4.4 4.7 4.6 — 5.4 — 4.1 4.2 3.9 4.4 5.0 9.5 9.8 10.5 7.1 9.0 3.6 4.8 2.5 — 3.3 — 3.5 3.7 2.7 3.3 4.0 19.2 3.8 6.3 5.4

Lower

99% CI

8.5 5.3 7.1 7.1 — 6.7 — 4.9 6.4 5.9 5.0 11.0 12.6 12.8 13.4 9.7 13.6 4.9 6.9 8.6 — 3.8 — 4.1 5.6 6.9 4.6 5.4 26.0 11.9 11.7 11.1

Upper 7.5 2.5 4.6 4.6 — 5.0 — 1.8 3.5 2.6 2.3 8.9 15.0 15.4 16.8 9.6 15.5 1.3 4.5 3.9 — 0.0 — 0.4 2.1 2.5 0.7 2.3 38.1 8.6 10.8 9.3

PD

5.2 1.6 2.2 2.1 — 3.7 — 1.0 1.3 0.7 1.6 3.0 12.0 12.4 13.9 7.0 10.8 >0 2.4 >0 — >0 — >0 0.2 >0 >0 0.9 31.2 0.6 5.4 3.6

Lower

99% CIb

9.8 3.4 7.0 7.1 — 6.2 — 2.6 5.7 4.6 2.9 14.9 17.9 18.3 19.6 12.2 20.2 2.7 6.7 9.9 — 0.5 — 1.0 4.1 6.7 2.0 3.7 45.0 16.6 16.2 15.1

Upper

Evolution in Zanzibar Red Colobus 1329

1330

K. Nowak et al.

Fig. 5 Female species shape and size MD profile plots. Solid black line, observed value; dotted gray lines, 99% standard bootstrap confidence intervals.

Various ecological and behavioral factors might have promoted a reduction in size in Procolobus kirkii. Though a thorough examination of these is beyond the scope of our present study, we hope that researchers will examine further the ideas presented here in future work. Predation pressure is one potential influence on primate body size (Palkovacs 2003). Zanzibar red colobus spend a significant proportion (≤23%) of the day on the ground (Siex 2003). The extent of terrestriality, unusual in most continental Colobinae, indicates that predation pressure on Procolobus kirkii is relatively low (Struhsaker 2000), an observation supported by its fission-fusion social system (Nowak 2007; Siex 2003), which tends to be associated with an absence of predators (McFarland 1986). Two well-known predators of red colobus,

Evolution in Zanzibar Red Colobus

1331

Fig. 6 Male species shape and size MD profile plots. Solid black line, observed value; dotted gray lines, 99% standard bootstrap confidence intervals.

crowned hawk-eagles (Stephanoaetus coronatus) and chimpanzees (Pan troglodytes), do not occur in Zanzibar. However, the relatively recent presence of another key primate predator, leopards, on the island suggests that Zanzibar red colobus would have probably experienced some nonhuman predation in the past. Work at Täi indicates that leopard predation selects against large body size, whereas larger size

1332

K. Nowak et al.

might be beneficial in the face of eagle predation (Zuberbühler and Jenny 2002). Thus, the lack of crowned hawk eagles in combination with the presence of leopards might have made decreases in body size beneficial for Procolobus kirkii. It is also possible that limited resources, linked to decrease of body size on islands (Lomolino 1985; Palkovacs 2003), may have contributed to the evolution of the smaller size of Procolobus kirkii compared to that of other red colobus species. Procolobus kirkii demonstrates greater flexibility in habitat use and dietary choice than that observed in many other red colobus taxa (Nowak 2007; Siex 2003) and it is possible that decreased resource availability in a restricted range acted as a selective pressure for increased dietary flexibility and reduction in size. In addition to size reduction in Procolobus kirkii, there may have been a concomitant decrease in size sexual dimorphism. Overall body size is often argued to be a major correlate of the relative magnitude of size sexual dimorphism within a clade (Fairbairn 1997). In other words, as body size decreases, sexual dimorphism also becomes smaller, per Rensch’s rule (Rensch 1959; Smith and Cheverud 2002). That sexual dimorphism in the small Procolobus kirkii is less than in other Procolobus spp. is consistent with predictions of Rensch’s rule, and our study indicates that sexual dimorphism in Piliocolobus tends to be proportional to size variation both within and across species. The observation is congruent with similar findings from guenons (Cardini and Elton 2008c). An interesting side-effect of the very modest sexual dimorphism in the cranium of Zanzibar red colobus is that this trend appears to have resulted in relatively smaller males with paedomorphic traits —short face, large orbits, enlarged neurocranium— reminiscent of those in another small but less closely related member of the same genus, olive colobus (Procolobus verus). Groves (2007) even suggested the possibility that female Procolobus kirkii exceed males in size. Besides differences in size, there are significant shape differences in the cranium of Procolobus kirkii compared to virtually all other members of Procolobus. Procolobus kirkii has a considerably shorter face, with a relatively expanded neurocranium that is also evident in the slightly smaller P. verus. This is a good example of a common allometric trend, occurring in many primates, and probably other mammals, in which small-bodied representatives of a group of closely related species tend to have shorter faces and larger braincases than those of larger species (Cardini and Elton 2008b). However, contrary to expectations from this widespread allometric pattern, a short face also occurs in large-bodied Procolobus gordonorum, the putative sister species of P. kirkii (Strushaker 1981; Ting 2008; Ting et al. 2006) and the geographically closest continental population of Procolobus. This observation raises an interesting question about where and when paedomorphic traits of Procolobus kirkii might have evolved, a question that researchers are unlikely to answer convincingly without detailed palaeontological data from the Zanzibar Archipelago. Phenotypic Variance and Island Endemicity Isolated populations, including those on islands, may experience population bottlenecks during initial colonization, reduction of effective population size due to range restriction, and fixation of neutral alleles (Woolfit and Bromham 2005). The observation that population bottlenecks and genetic drift may be common after island colonization was an important element in Mayr’s model of speciation in small

Evolution in Zanzibar Red Colobus

1333

peripheral isolates (1963) and his often strongly criticized (Coyne and Orr 2004), hypothesis of accelerated morphological evolution in insular species. Thus, if strong bottlenecks occurred after isolation in a restricted portion of the original geographic range, we might expect smaller variance in the gene pool of a young insular population. Using phenotypic variance as a proxy for genetic variance, there is no evidence for such a reduction in our data. Again, the virtual absence of palaeontological and genetic data hampers any historical reconstruction of the evolutionary history of Procolobus kirkii and any statement is inevitably speculative. For instance, it is possible that Procolobus kirkii colonized the island several times, which would maintain variance. However, phenotypic variance reflects not only genetic but also environmental effects and possibly interactions between genes and environments. Thus, a reduction in genetic variance may have been masked by more substantial environmental influences on morphology. Indeed, after its isolation in the Zanzibar Archipelago, Procolobus kirkii is likely to have experienced strong selective pressures in a rapidly changing environment owing to climate change and human activities. The pressures may have varied across the forest patches of the fragmented habitat inhabited by the Zanzibar red colobus. Thus, either adaptation to local differences in habitats or developmental plasticity under differential environmental conditions may have promoted phenotypic variation across the archipelago. Accelerated Evolution Results of the partial disparity (PD) and mean distinctiveness (MD) analyses indicate that Procolobus kirkii has a highly distinctive form vs. that of its congeners. PD and MD measure the relative disparity or distinctiveness of a species regardless of its time of origin. The approach is appropriate given the lack of paleontological data for red colobus and, at the time of our study, the absence of a molecular clock for the red colobus radiation. Even if one takes the more conservative estimate and considers only MD, the disparity in the size of Procolobus kirkii is on average more than twice as large as that of other populations, a finding that also holds when comparing P. kirkii to its sister species, P. gordonorum. Such a disparity factor corresponds with Millien’s (2006, p. 1863) estimates that “rates of morphological evolution are significantly greater – up to a factor of 3.1 – for islands than for mainland mammal populations.” Ting’s molecular data indicate that Procolobus kirkii diverged from its sister taxon P. gordonorum ca. 600 Ka (Ting 2008). Researchers previously assumed that the Zanzibar red colobus was a much younger species, dating from the end of the last glaciation. The revised estimate of a longer evolutionary age of Procolobus kirkii does not weaken our observation of acceleration in its rate of morphological evolution. Acceleration is a relative concept and dwarfism in Procolobus kirkii is so pronounced that it stands out as a clear outlier relative to other populations of red, olive, and black-and-white colobus even with divergence times based on Ting (2008), and size differences scaled accordingly. Because the observed morphological change is unlikely to have occurred at a constant rate over 600,000 yr of evolution of Procolobus kirkii, our explorative analysis of the rate of morphological evolution in the species is probably conservative. Thus, it is unlikely to have produced a false positive, unless we make the much less parsimonious assumption that dwarfism was ancestral to the entire clade of central and east African red colobus.

1334

K. Nowak et al.

Our findings of size distinctiveness in Procolobus kirkii are consistent with observations of morphological distinctiveness in close relatives of several other insular mammal species (Millien 2006). The acceleration in morphological evolution on islands is likely to be related to significant environmental pressures (Millien 2006), related to predation, competition, habitat, or resource availability. However, the ways in which the changes shape the genetic structure of the founding population, as in Mayr’s model and subsequent elaborations of it (Carson and Templeton 1984; Templeton 1980), are less clear (Coyne and Orr 2004). Evidence for Speciation in Procolobus kirkii Helbig et al. (2002, p. 519) defined species as “evolutionary lineages that maintain their integrity…through time and space”. Helbig et al. (2002) suggested that multiple lines of evidence (qualitative and quantitative morphological traits, behavior and ecology, and genetics) are needed to diagnose a species. One should also base specific rank on the likelihood of future retention of genetic and phenotypic integrity, and the characters chosen for diagnosis should be functionally independent and have a low environmental component to exclude differences caused simply by environmental factors such as nutrition or temperature (Helbig et al. 2002). For Procolobus kirkii, we found large and significant differences in cranial size and shape. That they are purely responses to local environmental factors seems unlikely because the magnitude of differences is about as large as among other putative species of red colobus and, for males, almost as large as between red and olive colobus. In addition, their distinctive cranial form co-occurs with differences in vocalizations and external morphology, specifically pelage color. Molecular analyses have also indicated that Procolobus kirkii is fairly closely related to P. gordonorum but more distantly related to other red colobus populations (Ting 2008; Ting et al. 2006). Thus, several largely independent lines of evidence suggest that the diagnosability criterion of Helbig et al. (2002) is met by Zanzibar red colobus. The second criterion, the retention of genotypic and phenotypic integrity, is extremely hard to assess, particularly for allopatric populations. Of course, it is impossible to observe reproductive isolation directly in geographically separated taxa such as Procolobus kirkii. However, differences in vocalization between Procolobus kirkii and mainland red colobus (Struhsaker 1981) might be a clue to past reproductive isolation, given the importance of calls in determining the mating structures of forest primates. It is possible that the red colobus group is a superspecies that comprises diagnosable, geographically separated taxonomic units — allospecies (sensu Helbig et al. 2002)— one of which is Procolobus kirkii. Grubb (2006) adopted a similar approach, suggesting that groups such as the red colobus should be termed geospecies.

Conclusion Our observation of morphological distinctiveness as a result of evolutionary acceleration suggests that insularity has significantly influenced the morphology of Procolobus kirkii. Accordingly, one can easily distinguish them from other members

Evolution in Zanzibar Red Colobus

1335

of the genus on the basis of overall cranial morphology and via features including cranial sutures, coat color and pattern, and acoustics of male calls (Kingdon 1997; Napier 1985; Struhsaker 1975, 1981; Verheyen 1962). This supports the consensus (Grubb et al. 2003) that Procolobus kirkii is a unique taxon among Procolobus and needs to be actively conserved (IUCN 2000). Given that each Procolobus species is judged to be either Vulnerable (Procolobus gordonorum), Endangered (P. kirkii, P. badius, P. pennantii), or Critically Endangered (P. rufomitratus) and that ca. 40% of Procolobus taxa are threatened with extinction (Struhsaker 2005), further clarification of the phylogeny of the red colobus group, in part via detailed morphological analyses of the type presented here for P. kirkii, would reinforce the establishment of conservation targets and inform management decisions and priorities. Further work on the morphology of the understudied red colobus group may also shed light on the evolutionary and ecological processes that have acted to differentiate species and subspecies. Researchers have undertaken relatively little work on primate isolates, but appreciating the effects of insularity and studying peripheral primate populations might be especially important as habitats become increasingly fragmented and, in essence, become islands (Channel and Lomolino 2000). Kingdon (1990, p. 244) described Africa as “a pattern of islands” (p. 244) and warned that “Africa’s natural communities are in the process of being cut up into separate islands and it will become increasingly difficult to distinguish man-made island communities from those that were once naturally isolated” (p. 234). Endemic species such as Procolobus kirkii, as well as other Procolobus spp., all of which are at risk of becoming increasingly relict and disjunctly distributed within remaining forests, may eventually become extinct unless they adapt or continue to adapt to drier and less forested conditions (Burgess et al. 1998) or find refugia in alternative and less human-accessible habitats such as mangroves (Galat-Luong and Galat 2005; Nowak 2008). Carefully designed studies that integrate specific aspects of morphology, behavior, and ecology, undertaken by collaborative multidisciplinary teams, will help to identify the limits of such adaptation and adaptability, and will facilitate conservation planning. Acknowledgments We dedicate this paper to the memory of Marco Corti (1950–2007) in recognition of his great contribution to the development and application of geometric morphometrics to the study of systematics and the mechanisms of speciation in mammals. An eclectic scientist, Marco may have been the first to use the term geometric morphometrics. With his studies, he was a pioneer, clearing the path and serving as an inspiration for many mammalogists. His papers were a model to follow for many of us and his advice was invaluable for many young zoologists who were struggling to learn geometric morphometrics and multivariate statistics. We are deeply grateful to all museum curators and collection managers who allowed and helped us to study their collections. Among them, we especially thank HansWalter Mittmann (Staatliches Museum für Naturkunde, Karlsruhe) for sending us specimens on loan during our visit at the Museum für Naturkunde in Berlin. Wim Wendelen (Royal Museum for Central Africa, Tervuren) and Olav Olav Röhrer-Ertl (formerly at Staatliche Naturwissenschaftliche Sammlugen Bayerns, Munich) provided invaluable help with specimen identification and advice on collections of Procolobus, and Cristina Murari (University of Modena and Reggio Emilia) provided crucial support for running computer analyses on Linux work stations. Claudio Gentilini, Maria Teresa Martinelli, Roberta Cantaroni, Costantino Crescimanno and Andrea Ghidoni (all of them at the University of Modena and Reggio Emilia) also were of great help to solving computer and network problems. For their support of our study, we sincerely thank Robert J. Asher and the University Museum of Zoology in Cambridge, where the specimens of Procolobus kirkii collected by K. Nowak are held. Craig Ludwig (National

1336

K. Nowak et al.

Museum of Natural History, Washington), Emiliano Bruner and Paolo Colangelo (University of Rome), Damiano Preatoni and Adriano Martinoli (University of Insubria), and Andrew Marshall (University of York) were of great help during various stages of this study. We also thank Zanzibar authorities in the Department of Commercial Crops, Fruits and Forestry (DCCFF), especially Director Dr. Bakari Asseid, for logistical support in situ. We particularly thank a reviewer of one of our previous papers for their advice on and suggestions about how to recalculate the partial disparity analysis so that the clumped means did not overly influence the analysis. Finally, we thank Nelson Ting (City University of New York) and his coauthors for sharing with us the preliminary results of their study of the molecular systematics of Piliocolobus, and Colin Groves (Australian National University) for his always invaluable advice on primate taxonomy and his careful review which improved a previous version of this manuscript. Grants from the Leverhulme Trust and the Ruggles-Gates Fund for Biological Anthropology provided funding for the study.

References Adams, D. C., Slice, D. E., & Rohlf, F. J. (2004). Geometric morphometrics: ten years of progress following the ‘revolution’. Italian Journal of Zoology, 71, 5–16. Bookstein, F. L. (1991). Morphometric Tools for Landmark Data. Cambridge,UK: Cambridge University Press. Bromham, L., & Cardillo, M. (2007). Primates follow the ‘island rule’: Implications for interpreting Homo floresiensis. Biology Letters, 3, 398–400. Brown, P., Sutikna, T., Morwood, M. J., Soejono, R. P., Jatmiko, A., Wayhu Saptomo, E., & Awedue, R. (2004). A new small-bodied hominin from the Late Pleistocene of Flores, Indonesia. Nature, 431, 1055–1061. Burgess, N. D., Clarke, G. P., & Rodgers, W. A. (1998). Coastal forests of eastern Africa: Status, endemism patterns and their potential causes. Biological Journal of the Linnaean Society of London, 64, 337–367. Camperio Ciani, A., Palentini, L., & Finotto, E. (2001). Survival of a small translocated Procolobus kirkii population on Pemba Island. Biodiversity and Conservation, 24, 15–18. Cardini, A., & Elton, S. (2007). Sample size and sampling error in geometric morphometric studies of size and shape. Zoomorphology, 126, 121–134. Cardini, A., & Elton, S. (2008a). Does the skull carry a phylogenetic signal? Evolution and modularity in the guenons. Biological Journal of the Linnean Society of London, 93, 813–834. Cardini, A., & Elton, S. (2008b). Variation in guenon skulls I: species divergence, ecological and genetic differences. Journal of Human Evolution, 54, 615–637. Cardini, A., & Elton, S. (2008c). Variation in guenon skulls II: sexual dimorphism. Journal of Human Evolution, 54, 638–647. Cardini, A., & O’Higgins, P. (2004). Patterns of morphological evolution in Marmota (Rodentia, Sciuridae): Geometric morphometrics of the cranium in the context of marmot phylogeny, ecology and conservation. Biological Journal of the Linnaean Society, 82(3), 385–407. Cardini, A., & Thorington Jr., R. W. (2006). Post-natal ontogeny of the marmot (Rodentia, Sciuridae) cranium: Allometric trajectories and species divergence. Journal of Mammalogy, 87, 201–216. Cardini, A., Jansson, A.-U., & Elton, S. (2007a). Ecomorphology of vervet monkeys: A geometric morphometric approach to the study of clinal variation. Journal of Biogeography, 34, 1663–1678. Cardini, A., Thorington Jr., R. W., & Polly, P. D. (2007b). Evolutionary acceleration in the most endangered mammal of Canada: Phylogenetic signal and cranial divergence in the Vancouver Island marmot (Rodentia, Sciuridae). Journal of Evolutionary Biology, 20, 1833–1846. Carson, H. L., & Templeton, A. R. (1984). Genetic revolutions in relation to speciation phenomena: The founding of new populations. Annual Review of Ecological Systems, 15, 97–131. Channell, R., & Lomolino, M. V. (2000). Dynamic biogeography and conservation of endangered species. Nature, 403, 84–86. Colyn, M. M. (1991). L’importance zoog’eographique du Bassin du Fleuve Zaire pour la speciation: le cas des Primates simiens. Annalen Zoologische Wetenschappen, Koninklijk Museum Voor Midden-Afrika Tervuren, Belgie, 264, 1–250. Coyne, J. A., & Orr, H. A. (2004). Speciation. Sunderland, MA: Sinauer Associates.

Evolution in Zanzibar Red Colobus

1337

Delson, E. (1994). Evolutionary history of the colobine monkeys in paleoenvironmental perspective. In A. G. Davies, & J. F. Oates (Eds.), Colobine monkeys: Their ecology, behavior and evolution (pp. 11–43). Cambridge, UK: Cambridge University Press. Dryden, I. L., & Mardia, K. V. (1998). Statistical Shape Analysis. New York: John Wiley & Sons. Fairbairn, D. J. (1997). Allometry for sexual size dimorphism: Pattern and process in the coevolution of body size in males and females. Annual Review of Ecological Systems, 28, 659–687. Fontaneto, D., Melone, G., & Cardini, A. (2004). Geometric morphometrics study of the jaws in microscopic aquatic pseudocoelomates: Shape diversity in the trophy of different species of Rotaria (Rotifera, Bdelloidea). Italian Journal of Zoology, 71, 63–72. Fooden, J., & Albrecht, G. H. (1993). Latitudinal and insular variation of skull size in crab-eating macaques (primates, Cercopithecidae: Macaca fascicularis). American Journal of Physical Anthropology, 92, 521–538. Foster, J. B. (1964). Evolution of mammals on islands. Nature, 202, 234–235. Galat-Luong, A., & Galat, G. (2005). Conservation and survival adaptations of Temminck’s red colobus (Procolobus badius temminckii) in Senegal. International Journal of Primatology, 26, 585–603. Goldman, H. V., & Walsh, M. T. (1997). A leopard in jeopardy: An anthropological survey of practices and beliefs which threaten the survival of the Zanzibar leopard (Panthera pardus adersi). Unpublished report to the Department of Commercial Crops, Fruits, and Forestry, Zanzibar. Goltsman, M., Kruchenkova, E. P., Sergeev, S., Volodin, I., & Macdonald, D. W. (2005). “Island syndrome” in a population of Arctic foxes (Alopex lagopus) from Mednyi Island. Journal of Zoology, 267, 405–418. Groves, C. P. (2001). Primate Taxoxomy. Washington, DC: Smithsonian Institution Press. Groves, C. P. (2007). The taxonomic diversity of the Colobinae of Africa. Journal of Anthropological Science, 85, 7–34. Grubb, P. (2006). Geospecies and superspecies in the African primate fauna. Primate Conservation, 20, 75–78. Grubb, P., Butynski, T. M., Oates, J. F., Bearder, S. K., Disotell, T. R., Groves, C. P., & Struhsaker, T. T. (2003). Assessment of the diversity of African primates. International Journal of Primatology, 24, 1301–1357. Helbig, A. J., Knox, A. G., Parkin, D. T., Sangster, G., & Collinson, M. (2002). Guidelines for assigning species rank. Ibis, 144, 518–525. Hilton-Taylor, C. (2000). 2000 IUCN Red List of Threatened Species. Cambridge, UK: IUCN. Howell, D. C. (2002). Statistical Methods for Psychology (pp. 384–387, 5th ed.). Pacific Grove, CA: Duxbury Thompson Learning. IUCN (2000). 2000 IUCN Red List of Threatened Species. Cambridge, UK: IUCN. Kingdon, J. (1981). Where have the colonists come from? A zoogeographical examination of some mammalian isolates in eastern Africa. African Journal of Ecology, 19, 115–124. Kingdon, J. (1990). Island Africa: The evolution of Africa’s rare animals and plants. London: William Collins Sons. Kingdon, J. (1997). The kingdon field guide to African mammals. Princeton: Princeton University Press. Klingenberg, C. P., & Monteiro, L. R. (2005). Distances and directions in multidimensional shape spaces: Implications for morphometric applications. Systems Biology, 54, 678–688. Lomolino, M. V. (1985). Body size of mammals on islands: the island rule re-examined. American Naturalist, 125, 310–316. Lomolino, M. V., Sax, D. F., Riddle, B. R., & Brown, J. H. (2006). The island rule and a research agenda for studying ecogeographical patterns. Journal of Biogeography, 33, 1503–1510. Manly, B. (1997). Randomization, bootstrap, and Monte Carlo methods in biology (2nd ed.). London: Chapman and Hall. Marcus, L. F., Hingst-Zaher, E., & Zaher, H. (2000). Application of landmark morphometrics to skulls representing the orders of living mammals. Hystrix, 11, 27–48. Mayr, E. (1963). Animal Species and Evolution. Cambridge, MA: Harvard University Press. McFarland, M. J. (1986). Ecological determinants of fission-fusion sociality in Ateles and Pan. In J. G. Else, & P. C. Lee (Eds.), Primate ecology and conservation, Vol. 2 (pp. 181–190). Cambridge, UK: Cambridge University Press. Meiri, S., Cooper, N., & Purvis, A. (2008). The island rule: made to be broken? Proceedings of the Royal Society of London B – Biological Science, 275, 141–148. Millien, V. (2006). Morphological evolution is accelerated among island mammals. PLoS Biology, 4, 1863–1868.

1338

K. Nowak et al.

Millien, V. (2007). Spurious or island effect? A response to J. A. Pérez-Claros and J. C. Aledo’s comment on Morphological evolution is accelerated among island mammals. PLoS Biology, 5(7), e176. doi:10.1371/journal.pbio.0050176. Mturi, F. A. (1991). The Feeding Ecology and Behavior of the Red Colobus Monkey (Colobus badius kirkii). Ph.D. dissertation, University of Dar es Salaam, Dar es Salaam, Tanzania. Napier, P. H. (1985). Catalogue of Primates in the British Museum (Natural History) and Elsewhere in the British Isles. Part III: Family Cercopithecidae, subfamily Colobinae. London: British Museum (Natural History). Nowak, K. (2008). Frequent water drinking by Zanzibar red colobus (Procolobus kirkii) in a mangrove forest refuge. American Journal of Primatology, 70, 1081–1092. Nowak, K. (2007). Behavioral Flexibility and Demography of Procolobus kirkii across Floristic and Disturbance Gradients. Ph.D. dissertation, University of Cambridge, Cambridge, UK. Oates, J. F., & Davies, A. G. (1994). What are the colobines? In A. G. Davies, & J. F. Oates (Eds.), Colobine monkeys: Their ecology, behavior and evolution (pp. 1–9). Cambridge, UK: Cambridge University Press. O’Higgins, P., Jones, N. (2006). Tools for statistical shape analysis, Hull York Medical School. http:// www.york.ac.uk/res/fme/resources/software.htm. Palkovacs, E. P. (2003). Explaining adaptive shifts in body size on islands: A life history approach. Oikos, 103, 37–44. Perez-Claros, J. A., & Aledo, J. C. (2007). Comment on “Morphological evolution is accelerated among island mammals”. PLoS Biology, 5(7), e180. Rensch, B. (1959). Evolution above the species level. London: Methuen. Rodgers, W. A. (1981). The distribution and conservation status of colobus monkeys in Tanzania. Primates, 22, 33–45. Rodgers, W. A., Owen, C. F., & Homewood, K. M. (1982). Biogeography of East African forest mammals. Journal of Biogeography, 9, 41–54. Rohlf, F. J. (1998). On applications of geometric morphometrics to study of ontogeny and phylogeny. Systems Biology, 47, 147–158. Rohlf, F. J. (2006a). TpsSmall. Department of Ecology and Evolution, State University of New York, Stony Brook, New York. Http://life.bio.sunysb.edu/morph/. Rohlf, F. J. (2006b). NTSYSpc, version 2.20L. Setauket, New York: Exeter Software. Rohlf, F. J., & Marcus, L. F. (1993). A revolution in morphometrics. Trends in Ecological Evolution, 8, 129–132. Rohlf, F. J., & Slice, D. E. (1990). Extensions of the Procrustes method for the optimal superimposition of landmarks. Systematic Zoology, 39, 40–59. Schultz, A. H. (1957). Cranial and dental variability in colobus monkeys. Proceedings of the Zoological Society of London, 130, 79–105. Siex, K. S. (2003). Effects of population compression on the demography, ecology, and behavior of the Zanzibar Red Colobus Monkey (Procolobus kirkii). Ph.D. Dissertation, Duke University, Durham, NC. Siex, K. S., & Struhsaker, T. T. (1999). Ecology of the Zanzibar red colobus monkey: Demographic variability and habitat stability. International Journal of Primatology, 20, 163–191. Silkiluwasha, F. (1981). The distribution and conservation status of the Zanzibar red colobus. African Journal of Ecology, 19, 187–194. Slice, D. E. (1999). Morpheus (beta version). Stony Brook, New York: Department of Ecology and Evolution, State University of New York. Smith, R. J., & Cheverud, J. M. (2002). Scaling of sexual dimorphism in body mass: A phylogenetic analysis of Rensch’s rule in primates. International Journal of Primatology, 23, 1095–1135. Struhsaker, T. T. (1975). The Red Colobus Monkey. Chicago: University of Chicago Press. Struhsaker, T. T. (1981). Vocalizations, phylogeny and palaeogeography of red colobus monkeys (Colobus badius). African Journal of Ecology, 19, 265–283. Struhsaker, T. T. (2000). The effects of predation and habitat quality on the socioecology of African monkeys: Lessons from the islands of Bioko and Zanzibar. In P. F. Whitehead, & C. J. Jolly (Eds.), Old World Monkeys (pp. 393–430). Cambridge, UK: Cambridge University Press. Struhsaker, T. T. (2005). Conservation of red colobus and their habitats. International Journal of Primatology, 26, 525–538. Struhsaker, T. T., & Leland, L. (1980). Observations on two rare and endangered populations of red colobus monkeys in East Africa: Colobus badius gordonorum and Colobus badius kirkii. African Journal of Ecology, 18, 191–216.

Evolution in Zanzibar Red Colobus

1339

Struhsaker, T. T., & Siex, K. S. (1998). Translocation and introduction of the Zanzibar red colobus monkey: Success and failure with an endangered island endemic. Oryx, 32, 277–284. Tappen, N. C. (1960). Problems of distribution and adaptation of the African monkeys. Current Anthropology, 1, 91–120. Templeton, A. R. (1980). The theory of speciation via the founder principle. Genetics, 93, 1011–1038. Ting, N. (2008). Mitochondrial relationships and divergence dates of the African colobines: evidence of Miocene origins for the living colobus monkeys. Journal of Human Evolution, xx, 1–14. Ting, N., Delson, E., Oates, J. F., & Disotell, T. R. (2006). Molecular systematics of red colobus monkeys (Procolobus [Piliocolobus]). International Journal of Primatology, 27(Suppl 1), Abstract 96. Van Valen, L. M. (1973). Pattern and the balance of nature. Evolutionary Theory, 1, 31–49. Van Valen, L. M. (1978). The statistics of variation. Evolutionary Theory, 4, 33–43. Verheyen, W. N. (1962). Contribution a la craniologie comparee des Primates: les genres Colobus Illiger 1811 et Cercopithecus Linne 1758. Musee Royal de l’Afrique Central Tervuren, Belgique Annales Sciences Zoologiques, 105, 1–255. Woolfit, M., & Bromham, L. (2005). Population size and molecular evolution on islands. Proceeding of the Royal Society of Biology B, 272, 2277–2282. Zelditch, M. L., Swiderski, D. L., Sheets, H. D., & Fink, W. L. (2004). Geometric Morphometrics for Biologists: A Primer pp. 291–320. London: Elsevier Academic Press. Zuberbuehler, K., & Jenny, D. (2002). Leopard predation and primate evolution. Journal of Human Evolution, 43, 873–886.