Revisiting the taxonomic status and ecological partitioning of night monkeys genus Aotus in western Colombia, with notes on Aotus zonalis Goldman, 1914

Revisando el estatus taxonómico y la partición ecológica de los monos nocturnos del género Aotus en el noroccidente de Colombia, con notas sobre Aotus zonalis Goldman, 1914

Hugo Mantilla-Meluk1, Alex Mauricio Jiménez-Ortega2

1 Department of Biological Sciences, Texas Tech University, Lubbock, TX, USA. e-mail: hugo.mantilla@ttu.edu
2 Universidad Tecnológica del Chocó, Quibdó, Colombia. e-mail: almajior@hotmail.com
    Fecha recepción: Octubre 5, 2010        Fecha aprobación: Febrero 4, 2011

Abstract

As part of the project Programa Regional de Biodiversidad, subprograma Línea Base de Conocimiento de la Biodiversidad del Chocó, designed to document the mammalian fauna of the Biogeographic Chocó, museum voucher specimens from the Colombian Chocoan Region deposited in American institutions have been taxonomically assessed in the past three years. Herein, we review the morphological variation, distribution patterns, and taxonomy of night monkeys in the genus Aotus from northwestern Colombia deposited at the Field Museum of Natural History (FMNH), and GIS based analyses and Maxent modeling are used to define the geographic extent and ecological limiting factors of the analyzed taxa. Our Principal Component Analysis showed high variation in skull morphology among Aotus from northwestern Colombia. However, differences observed in both discrete and morphometric analyses of the interorbital region of A. zonalis (more depressed than that in the grographically adjacent A. griseimembra and A. lemurinus) appear as good diagnostic characters for this taxon. Our analyses on the ecological variation associated with collecting localities of Aotus specimens support the geographic subdivision previously proposed based on karyotypic data. Based on obtained models of potential distribution we define the location and extent of potential contact zones among species of Aotus from northwestern Colombia.

Keywords: Aotus; A. zonalis;  Biogeographic Chocó;  Ecological partitioning; Skull morphology; Taxonomy. 

Resumen

Como producto del proyecto Programa Regional de Biodiversidad, subprograma Línea Base de Conocimiento de la Biodiversidad del Chocó, que busca documentar la biodiversidad del Chocó biogeográfico colombiano, se han revisado en los últimos tres años, especímenes museológicos depositados en colecciones americanas. En este trabajo  se evalúa la variación en la morfología craneal de especímenes de monos nocturnos género  Aotus, del noroccidente de Colombia depositados en el Field Museum of Natural History (FMNH), y se utiliza modelamiento basado en sistemas de información geográfica (SIG) y Maxent en la definición de los límites geográficos y ecológicos de los taxa analizados. El presente análisis de componentes principales mostró un alto grado de variabilidad en morfología craneal entre las poblaciones de Aotus del noroccidente colombiano. Sin embargo, diferencias observadas en análisis de caracteres discretos y morfométricos de la región interorbital de A. zonalis (caracterizada por ser más deprimida con respecto a las geográficamente adyacentes A. griseimembra y A. lemurinus), se presentan como buenos caracteres diagnósticos para este taxón. Nuestro análisis de la variación ambiental asociada con las localidades de colecta de los especímenes de Aotus analizados, apoya la subdivisión geográfica previamente sugerida por datos cariotípicos. Basados en los modelos de distribución obtenidos se define también  la ubica-
ción y extensión geográfica de potenciales zonas de contacto entre las especies de Aotus del noroccidente colombiano.

Palabras clave: Aotus; A. zonalis; Chocó biogeográfico;  Partición ecológica; Morfología craneal; Taxonomía. 

Introduction

With the aim of creating a database of the mammalian species of the Biogeographic Chocó, specimens deposited in American Institutions have been documented and their systematic status has been revisited in the last three years. Our work has been focused on those groups of mammals characterized by controversial taxonomic histories such Aotus zonalis, Goldman, 1914 described from the Canal Zone, Gatún, Panama, which constitutes the only currently recognized Chocoan night monkey (Defler, 2003, 2004; Defler and Bueno, 2007). Difficulties in interpreting the morphological and coat color variation among night monkeys in the genus Aotus has resulted in controversy on the designation of appropriate taxonomic status for night monkey populations from the western versant of the Andes in Colombia (Defler and Bueno, 2007). Recent molecular and karyotipic analyses have provided important insights on the systematic relationships among geographically distinctive Aotus variants, supporting a greater diversity within the genus than previously documented (Defler and Bueno, 2007; Groves, 2001, 2005). Based on karyotypic data, Defler and Bueno (2007) recognized six species of Aotus for the Colombian territory with a suggested geographic structure (A. brumbacki, A. griseimembra, A. jorgehernandezi, A. lemurinus, A. trivirgatus, A. vociferans, and A. zonalis). Three of these species occur on the western versant of the Andean System distributed as follows: A. griseimembra (Elliot, 1912) (2N=52-54), distributed along the eastern bank of the Sinú River east across the lowlands to high elevations of the Caribbean Region up to west of the Gulf of Maracaibo in Venezuela (Defler, 2004); A. lemurinus (I. Geoffroy-St. Hilaire, 1843) (2N=55-56), distributed along the Andean range above 1000 m; and A. zonalis (Goldman, 1914) (2n=58), distributed across the lowlands of the Biogeographic Chocó north up to Costa Rica in Central America (Ford, 1994).
 
Herein, we revisit Chocoan specimens of A. zonalis deposited at the Field Museum of Natural History and present a preliminary analysis on the macroecological component of the observed morphological variation among Aotus species from western Colombia: A. griseimembra, A. lemurinus, and A. zonalis.  In addition, models of potential distribution created for these three taxa are used to analyze the location and extent of potential zones of contact among species of Aotus from northwestern Colombia.

Materials and methods

Group designation. Craniodental and external characters in the descriptions of A. lemurinus (I. Geoffroy-St.Hilaire, 1843), A. griseimembra (Elliot, 1912), and A. zonalis (Goldman, 1914) were evaluated for the analyzed individuals. Age of the specimens was estimated based upon totally erupted and functional dentition, as well as totally fused spheno-occipital and/or ethmoid sutures.

Morphological variation. In order to assess the morphological variation among Aotus populations from the western versant of the Colombian Andes, a Principal Component Analysis (PCA) was performed on 30 craniodental variables (Appendix I) of 32 Colombian Aotus (29 adults and three juveniles (FMNH 68859, 68861, 70679, that were removed after a preliminary analysis) deposited at the scientific collections of the Field Museum of Natural History (FMNH) (Appendix I), including A. griseimembra (N=11), A. lemurinus (N=18), and A. zonalis (N=3) (Table 1) in the statistical package PAST, available at http://www.nhm.uio.no/norges/past/download.html. In addition, the variation of discrete skull characters included in the description of A. zonalis (Goldman, 1914) was also investigated for the three analyzed taxa.


Environmental variation among Aotus sampling localities. To determine the degree of ecological differentiation among analyzed sampling localities of A. griseimembra, A, lemurinus and A. zonalis (including their type localities) (N=12, Appendix II), a PCA was performed on four environmental variables (Elevation, Mean Annual Maximum Temperature, Mean Annual Minimum Temperature, and Mean Annual Precipitation). Environmental data were derived from the Bioclim data set available at http://www.worldclim. org/bioclim
Finally, to test for ecological similarities among analyzed sampling localities of Aotus, an ecological niche model was created for each one of the analyzed taxon by applying the Maxent algorithm (program available at http://www.cs.princeton. edu/~schapire/maxent/) following the procedures described in (Phillips et al., 2006).  Overlap among obtained outputs was used to infer the location and extent of potential contact zones among species of Aotus from the northwestern range of Colombia.

Results 

Principal components analyses. Our PCA showed a relatively low degree of morphological differentiation among the three analyzed taxa in Aotus, with A. lemurinus having the greatest range of skull variation. In our PCA’s morphospace, samples of A. lemurinus partially overlapped A. griseimembra and A. zonalis (Figure 1). Most of the observed skull variation was explained by the first two components (42.4 and 9.10%, respectively) (Table 2). Samples of A.griseimembra and A. zonalis were primarily located on the III quadrant of the morphospace of our PCA, with negative loadings suggesting a smaller size for these two taxa when compared with analyzed samples of A. lemurinus (Figure 1).  Our PCA on measurements of the interorbital region is presented in Figure 2.






Discriminant function analysis.
In our DFA only  55.2% of the samples identified as A. griseimembra and A. lemurinus were correctly assigned to predicted groups (Table 3). However, all samples of A. zonalis (N=3) were correctly assigned in our DFA.






Discrete characters in A. zonalis.
Discrete characters observed in A. zonalis are commented in the Discussion section of this work.

Ecological differentiation among Aotus from northwestern Colombia. In our PCA most of the observed environmental variation among analyzed sampling localities was explained by the first component (98%) with elevation having positive loading among A. lemurinus localities and precipitation having positive loadings for A. zonalis localities at the department of Chocó (Figure 3). As product of our analysis we identified three clearly differentiated geographic groups as follows:


Ecological overlap and potential contact zones among Aotus species from northwestern Colombia.
In our Maxent models A. griseimem-bra appear as the most resilient species with the widest distribution among northwestern Colombian Aotus. When our Maxent models constructed for A. griseimembra and A. lemurinus were evaluated at a predictive value of 50%, A. griseimembra distribution overlapped 100% of the predicted area for A. lemurinus. At probability values greater than 80% there was low overlap among the three analyzed species of Aotus (Figure 4). Potential areas of overlapping among Aotus species from northwestern Colombia are shown in Figure 4 and commented in the Discussion section of this work.


Discussion

Morphometric analyses. In a revision of the taxonomy and the distribution of night monkeys, genus Aotus, Ford (1994) carried out multivariate analyses of craniodental measurements in combination with the analysis of pelage patterns and color, chromosomal data and blood protein variations. Ford (1994) concluded that there was «good support» for A. brumbacki, A. lemurinus, A. griseimembra and A. zonalis distributed in northern Colombia and the colombian Andes and Panamá. Our PCA and DFA performed on the whole set of skull measurements failed in producing a clear discrimination among analyzed population of Aotus from northwestern Colombia. However, most of the analyzed samples of A. lemurinus had positive loadings on the first component, suggesting differences in size between this taxon from the Andean region, and A. griseimembra and A. zonalis from the lowlands of the Caribbean and the Chocoan Region of Colombia respectively (Figure 1). Aotus lemurinus proved to have longer skulls (with greater loadings associated with condylobasal length) and broader interorbital regions (associated with greater loadings for interorbital constriction width) when compared with the two analyzed lowland forms A. griseimembra and A. zonalis (Table 1 and Figure 5). We analyzed measurements associated with the interorbital region in a separate PCA obtaining a better differentiation between A. lemurinus and A. zonalis with no overlap between these two forms in the morphospace. The morphometric gap between A. lemurinus and A. zonalis was filled by A. griseimembra (Figure 2).


Due to our low sample size, males and females were treated together in both PCA and DFA. The low differentiation in our morphometric analyses can be associated with our small sample size and skull differences associated with sexual dimorphism. In addition, the observed variation in terms developmental stages in our sampling, prevented a better differentiation among groups. Our knowledge on the allometry of members of the genus Aotus is still poor and to differentiate between immature and mature individuals is difficult based on skull morphology alone.

Significance of discrete characters, the case of interorbital and nasal features in A. zonalis. As Hershkovitz asserts, the interorbital region may have been narrow in primitive platyrrhines (Hershkovitz, 1977). Among living cebids, the interorbital region of Aotus and Saimiri agree with Callicebus but their orbits are wider relative to cranial breadth (Hershkovitz, 1977). Major orbital expansion in Aotus has been lateral but median compression is also evident in the narrowness of its interorbital region (Hershkovitz, 1977). As documented in callithichids (Hershkovitz, 1977), the narrower interorbital septum of small species of platyrrhines (Aotus, Callicebus, and Saimiri) is due to relative larger eyes in smaller heads. Among higher primates the nasoturbinal is best developed in platyrrhines particularly in prehensile-tailed cebids, but most notably in pithecines. The ethmoturbinal II, as well of the ectoturbinal I are well developed in Aotus contrasting other platyrrhines (Hershkovitz, 1977). Although not well documented among platyrrhines, enlargements of the interorbital region and their associated turbinal systems have been linked with smell functions. Goldman (1914) mentioned that although skull of A. zonalis is in general broader to that of A. griseimembra, with the greater breath more noticeable in the braincase, the interorbital region of A. zonalis is «more depressed materially altering the facial angle» (Goldman, 1914; p. 6) (Figure 1). This character was consistent among the three adult A. zonalis specimens analyzed in this work. In the scatter plot of our PCA on three nasal and interorbital measurements, individuals of A. zonalis clustered together (Figure 2). However, a depressed interorbital region was a variable character among individuals of A. griseimembra and A. lemurinus. Juvenile and sub-adult individuals of A. griseimembra and A. lemurinus presented a more depressed interorbital region when compared with adult individuals, suggesting that the presence of a depressed interorbital region in A. zonalis might be a neotenic characteristic in this taxon.

Distribution patterns of Aotus in northwestern Colombia. Hernández-Camacho and Cooper (1976) restricted A. lemurinus to the Colombian Andes (1000-1500 m up to 3000-3200 m) and A. griseimembra to the northern lowlands of the Colombian Caribbean Region, Santa Marta mountains, west to río Sinú, río San Jorge, lower río Cauca and lowlands of middle and upper río Magdalena. Same authors recognize the form A. zonalis as the night monkey of north-western Colombia, in the department of Chocó, extending its distribution up to Panamá. In spite of the controversies on the taxonomic status of Aotus populations in northwestern Colombia, the geographic subdivision proposed by Hernández-Camacho and Cooper (1976) still accepted (Defler and Bueno, 2007). Hershkovitz (1983) recognized lemurinus and griseimembra as distinct, but considered them to be subspecies of A. lemurinus; he made no mention of A. zonalis, but as he ascribed Central American night monkeys to A. lemurinus lemurinus, by implication he was regarding it as a synonym of this latter form. Groves (2001) followed Hernández-Camacho and Cooper (1976) in recognizing A. zonalis as the typical form in Panama, and listed it as a subspecies of A. lemurinus along with A. griseimembra and A. brumbacki. Defler et al. (2001) concluded that the karyotype of A. hershkovitzi Ramírez-Cerquera, 1983 (from the upper río Cusiana, Boyacá, Colombia; 2n=58) was in fact that of true lemurinus, and they also mentioned that karyotypes considered by Hershkovitz (1983) as lemurinus in the northwestern distribution of the genus, were in fact of A. zonalis. Defler et al. (2001) and others (Defler, 2003, 2004; Defler and Bueno, 2007) concluded that Aotus lemurinus of Hershkovitz (1983) represent three karyotypically well-defined species of night monkeys as follows: 1) lowlands of Panamá and the Chocó region of Colombia represent A. zonalis, 2) Magdalena valley, A. griseimembra, and 3) those from above 1500 m A. lemurinus.

Potential contact zones among Aotus from northwestern Colombia. As shown in our Maxent model outputs (Figure 4), even at low predictive values (80%) it is possible to identify several areas of potential contact among A. griseimembra, A. lemu-rinus, and A. zonalis. Aotus griseimembra appears as the most ecologically resilient species of Aotus partially overlapping the distribution of A. lemurinus, and the area of the model obtained for A. zonalis. Intermingle between A. griseimembra and A. lemurinus is most likely to occur along the northern slopes of the Central and Western Cordilleras, as well as along the slopes of the Serranía del Perijá and the Sierra Nevada de Santa Marta (Figure 4). In the same way, A. zonalis and A. griseimembra overlap in their ecological requirements at the northwestern most corner of Colombia in the area of the Gulf of Urabá and the Colombian Darién. Although our Maxent model    identified this area as part of the predictive output of A. lemurinus, probability values associated with this portion of the model were lower than 50%; in addition, based on morphological trends observed in the present work we consider unlikely for A. l. lemurinus, the typical form of the Andean region, to be present in the lowlands of the Chocó.

Conclusion

Despite of the low resolution of our PCA and DFA analyses, we agree with Defler and Bueno (2007) in the recognition of A. griseimembra, A. lemurinus, and A. zonalis as valid species. However, we disagree with Defler and Bueno (2007) on their argument on the validity of morphological characters to differentiate among the so called subspecies in the A. lemurinus group sensu Hershkovitz (1983).  Taking into account the observed ecological overlapping and low morphological differentiation among Aotus from northwestern Colombia the potential presence of hybrid populations is high. The presence of hybrids and the lack of strong mechanisms of sexual isolation do not necessarily imply the absence of autopomorphic characters among karyologically distinguishable populations of Aotus from northern Colombia. In our work, the three adult individuals of A. zonalis analyzed have a more depressed interorbital region contrasting the broad interorbital region of A. griseimembra and A. lemurinus. Based on the analysis of our niche model we agree with Hernández-Camacho and Cooper (1976) geographic subdivision among Aotus from northwestern Colombia.

Aknowledgements

This work would not have been possible without the generous cooperation of B. D. Patterson, L. R. Heaney, B. Starck, N. Upham, and the staff of the Field Museum of Natural History. We thank all of them for their support and assistance. We recognize the generous contribution of Dr. Miguel Angel Medina Rivas (Program Coordinator: Cooperación Gestión Ambiental Local y Cadenas Productivas UTCH-NUFFIC). Dr. Medina Rivas commitment to the understanding of the megadiversity enclosed by the Chocoan region has been inspirational for several generations of young Chocoan biologists. Financial support was provided by the project: Programa Regional de Biodiversidad, subprograma Línea Base de Conocimiento de la Biodiversidad del Chocó; the Short term visitors program of the Field Museum of Natural History; and the Department of Biological Sciences of Texas Tech University.

Literature cited



Appendix I

Craniodental measurements analyzed: Brain case length (BCL); external interorbital width (IOC); anterorbital constriction (AOC); orbit height (OH); nasal length (NL); nasal width (NW); tooth row (TR); greatest length of the skull (GLS); Condyobasal length (CB); palatal length (PAL); zygomatic (ZYG); braincase width (BCW); mastoid breadth (MB); brain case height (BCH); orbital height ext (OHE); molars row (MR); premolars row (PR); first molar width (M1W); canine to canine distance (C-C); internal width across molars (AIM); maximum distance of the nasals (MDN); temporal width (TW); basioccipital length (BOL); molar to molar length (M-M); mandibular length (ML); ramus height (RM); mandibular length 1 (MnL1); mandibular length 2 (MnL2); mandibular toothrow (MTR); and distance across mandibular canines (C-CM).

Appendix II

Collecting localities associated with analyzed specimens of Aotus: A. griseimembra­.-  (N=11) COLOMBIA: Department of Córdoba, Catival, upper río San Jorge, 120 m (FMNH 68859 f, 68861 f, 68860 f, 68862 f); Department of Sucre, Colosó, Las Campanas, 175 m (FMNH 68850 f, 68851 f, 86652 f, 68853 f, 68855 f, 68856 f); Undetermined locality (FMNH); A. lemurinus. (N=18) COLOMBIA, Department of Antioquia, Bellavista (FMNH 6960 f, 69608 m, 69609 f, 69610 f); Urabá (FMNH 69612 f, 69613 f); Department of Cauca, río Mechengue (FMNH 88471 f); Munchique (FMNH 88470 m); Department of Huila, Acevedo, Aguas Claras (FMNH 70672 m, 70673 m, 70674 m, 70675 m, 70676 f);  San Agustín, San Antonio 2,300 m (FMNH 70677 f, 70678 m, 70679 m, 70680 f); Undetermined locality (FMNH); A. zonalis. (N=3) COLOMBIA, Deparment of Chocó, río Baudó, río Sandó (FMNH 90322 f, 90323 f, 90324 f).

f = female     m = male