We studied the geographic distribution of the endangered black howler monkey, Alouatta pigra, in its southeastern range in eastern Guatemala along its putative contact zone with the mantled howler monkey, A. palliata. A distributional survey of both species was conducted through direct observations and interviews, and the data analyzed with GIS to detect potentially limiting geographic and ecological factors. Alouatta pigra was not associated with any specific vegetation type or altitudinal range, as it occurred in many forest types from the lowlands up to elevations of 2705 m a.s.l., in contrast to prior reports of it being restricted to elevations below 400 m a.s.l. and having a strong affinity for riverine forests. Alouatta palliata did not occur at the higher elevations. Distribution of both howler monkey species was unaffected by the presence of Ateles, by protection status of the survey site, or by hunting activities. Deforestation appears to have restricted both species in some areas. Throughout the southeastern range of A. pigra, we found no evidence for current sympatry with A. palliata, as their distributions are separated by a river barrier and unsuitable dry shrubland. Ecological preferences also seem to reinforce allopatry, with only A. pigra occurring in wet and cold montane forests of the Sierra de las Minas. Our data indicate that highland habitats in Mexico and Guatemala, previously considered unsuitable for A. pigra, may need to be evaluated for distribution assessments and management plans. Populations of A. pigra in the diverse protected ecosystems in eastern Guatemala may be an important component for conservation of the species.
Introduction
The black howler monkey, Alouatta pigra, is endemic to Guatemala, Belize and southern Mexico. Recently, it was categorized as endangered in the IUCN Red List of Threatened Species due to population size reduction based on habitat decline (Cuarón et al., 2003). Conservation assessments and action plans for the species mandate updating distribution maps and describing the status of wild populations (Rodríguez-Luna et al., 1996; Matamoros et al., 1997).
Limited knowledge of the geographic distribution of A. pigra is based on museum specimens (Smith, 1970; Hall, 1981) and broad field studies done in Belize and Mexico (Horwich and Johnson, 1986; Watts et al., 1986) and Guatemala (Curdts, 1993). Recent fieldwork in the Yucatan peninsula detailed the species' distribution and habitat preferences (Navarro et al., 2003; Serio-Silva et al., 2006), but other areas of Mexico and Guatemala need similar detailed studies. Of particular importance are potential contact zones with the mantled howler monkey, A. palliata, at the northwest and southeast range limits, where sympatry has been reported in Tabasco, Mexico (Smith, 1970; Cortés-Ortiz et al., 2003) and suspected in eastern Guatemala (Horwich and Johnson, 1986; Curdts, 1993). Contact zones are important for the preservation of primate biodiversity (Jones and Bicca-Marques, 2004) and essential for the study of genetics and ecology of speciation (Jiggins et al., 1996).
The studies in southern Belize and eastern Guatemala have resulted in ambiguous and inconclusive definition of ranges for the two species. For example, Horwich and Johnson (1986) identified sympatry at the Belize-Guatemala border around the Sarstún River, based on suggestions of palliata-like individuals on the Guatemala side of the river, where only A. pigra was expected. Curdts (1993) proposed large areas of sympatry in the mountains of Baja and Alta Verapaz and Sierra de las Minas, Guatemala, but provided no empirical evidence. The objective of the present study was to define the geographic distribution of A. pigra and A. palliata in their contact zone in eastern Guatemala at the southeastern range limit of A. pigra. In addition we hoped to identify ecological factors and geographic barriers that might affect both species' distributions.
Methods
Study area
The contact zone and putative area of sympatry of A. pigra and A. palliata lies within 88°–90° W longitude and 15°–16° N latitude, encompassing the eastern part of Guatemala, bounded by Belize, Honduras and the Atlantic Ocean (Fig. 1). Five Guatemalan departments are partly or totally within the area: Alta Verapaz, Baja Verapaz, Izabal, Zacapa and El Progreso. The area is topographically and ecologically heterogeneous with elevations ranging from sea level to 3,000 m a.s.l. (CONAP, 2005) and ecosystems varying from flooded coastal forest to montane cloud forest. The region is transversed by several mountain ranges and large rivers (Fig. 1). Continuous tracts of tropical forests are restricted primarily to protected areas. Forest conversion by traditional slash-and-burn maize cultivation is combined with intensive socioeconomic activities: coffee and cardamom plantations in the highlands, and cattle ranching, banana, African palm and rubber plantations in the lowlands. The Department of Izabal, which occupies a large part of the study area, has 20% of its total area in cattle pastures and crop monocultures (INAB, 2001).
Surveys
Following Brockelman and Ali (1987), we conducted a distributional field survey during the dry season, February to April 2005. Cartographic maps (1:25000), vegetation maps, and a Landsat TM satellite image (2003) were used to identify suitable monkey habitat, access routes, and survey sites. Survey sites were selected to test potential geographic barriers and to cover the altitudinal ranges and vegetation types of the region. Additionally, site selection was dependent on the presence of suitable forest cover and existence of key persons who would facilitate entrance to villages, contact appropriate interviewees, and serve as translators in K'ekchi villages. Survey sites were accessed using vehicle, boat, mule or by foot. At each survey site geographic location and altitude were recorded with a Garmin GPS 72.
Howler monkey occurrences were detected by two methods: interviews (Pinto and Rylands, 1997; Iwanaga and Ferrari, 2002) and broad forest surveys along trails and rivers (Brockelman and Ali, 1987). Interviews consisted of non-leading questioning about the primates in the area, as well as other questions related to the interviewee's occupation, residence time in the area, and hunting activities. Additionally, interviewees were asked to recognize photographs of possible local primate species (Alouatta, Ateles, Cebus) and South American Alouatta species, as well as playbacks of loud calls of A. pigra and A. palliata. People interviewed were mainly subsistence hunters, town elders with knowledge of the forest, park rangers, landowners, and field biologists. Forest surveys were made during 3-day visits to key sites in early mornings and late afternoons during peak hours of howler monkey vocal activity, in order to locate individuals and make visual and/or auditory identification.
Species presence/absence records obtained from valid interviews and/or direct observations at each site were incorporated into a Geographic Information System (ArcView version 3.3). Digital maps of the vegetation types of Guatemala (INAB, 2001; CCAD-WB, 2003), protected areas, and elevation curves were overlaid with the geographical coordinates of sites to create joined attribute tables and allow analyses of the distribution of howler monkeys in relation to physical and ecological features. Information collected through interviews on hunting pressure and on the occurrence of the spider monkey, Ateles geoffroyi, was also related to the presence/absence records of the howler monkeys. A search of collection databases of 18 natural history museums in North America and Europe and a review of literature were made to obtain any historic records of the occurrence of A. pigra and A. palliata in the region.
Statistical analyses
We used tests for homogeneity of proportions with a logistic analysis approach to detect habitat-specific distributional patterns. Specifically, we used a nominal model to test if vegetation type had an effect on the probability of occurrence of each species and an ordinal model to test for a trend in probability of occurrence along an altitudinal gradient. Also, tests of independence were used to assess howler monkey association with (a) the presence of Ateles and (b) the protection status of the site. All statistical tests were performed with SAS version 9.0.
Results
A total of 58 sites were visited (Fig. 1); interviews were conducted at 47 sites and direct observations at 23 sites. We conducted 97 interviews, 59% with ladino and 41% with K'ekchi interviewees. Five interviews were classified as invalid due to inconsistencies in descriptions or identifications of photographs. Playbacks proved useful only to identify A. palliata, as interviewees clearly distinguished the calls. In contrast, at A. pigra sites, interviewees could not distinguish between the two species' calls. Overall, Alouatta pigra was found at 26 sites: seven through both interviews and direct observations, seven through observations only, and 12 sites through interviews only (Table 1). Alouatta palliata was found at 12 sites: five through interviews and direct observation, three through observations only, and four through interviews only (Table 2).
Table 1.
Localities where Alouatta pigra occurs (see Fig. 1), protection status, and detection method.
Alouatta pigra was verified along both banks of the Río Sarstún, the Atlantic coast, the north shore of Lake Izabal-Río Dulce, both banks of the Río Polochic and inland in the Purulhá. mountains, Sierra Yalijux, Sierra Santa Cruz, and Sierra de las Minas (Fig. 1, Table 1). Alouatta palliata was verified along the Atlantic coast south of the Río Dulce and inland along the Río Motagua valley and in Sierra Caral (Fig. 1, Table 2).
Of five potential vegetation types, A. pigra was found in four and A. palliata in three (Table 3). Neither species occurred in “seasonal evergreen shrubland with mixed forest”. Only A. pigra occurred in “tropical evergreen and semievergreen mixed forest” typical of cloud forest. Alouatta palliata was most often in “agroproductive systems with significant portions of broad-leaved forest”, although the association was not statistically significant (test of homogeneity of proportions χ2 = 7.6, df = 4, p = 0.10). In contrast, A. pigra, was less frequent in this disturbed vegetation type and more commonly associated with flooded forests, mixed forests and broad-leaved forest (test of homogeneity of proportions χ2= 14.41, df = 4, p < 0.01). Survey sites were distributed across elevations from sea level to more than 2500 m a.s.l. (Table 4). Alouatta pigra was found in all altitudinal ranges (Table 4) and no ordinal association was detected (test of homogeneity of proportions, ordinal model χ2 = 1.61, df = 1, p = 0.20). The altimontane survey sites (> 2000 m a.s.l.) were all occupied by A. pigra, the highest being 2705 m a.s.l. in Sierra de las Minas. In contrast, A. palliata occurred significantly more frequently in lowland sites and was not recorded at elevations higher than 1000 m a.s.l. (test of homogeneity of proportions, ordinal model χ2 = 4.07, df = 1, p<0.05).
Table 2.
Localities where Alouatta palliata occurs (see Fig. 1), protection status, and detection method.
Table 3.
Vegetation types (UNESCO classifications following INAB, 2001) of the survey sites and number of sites with presence/absence records for each species.
Table 4.
Altitudinal ranges (following INAB, 2001) of the survey sites and number of sites with presence/absence records for each species.
Of the 10 major protected natural areas surveyed, 80% sustained howler monkey populations: Alouatta pigra was found in six and A. palliata in two (Tables 1 and 2). Of the total survey sites, 32 sites (55%) were in protected areas with enforcement and 26 sites (45%) were located outside protected areas or in protected areas without vigilance. Occurrence of howler monkeys was not associated with the protection status of the survey site (test of independence G2 = 0.33, df = 1, p = 0.59).
Of 97 interviewees, 76% denied hunting or ever hearing of anyone hunting howler monkeys. Most (19 of 22) of the remaining 24% responded that they had only heard of other people killing howler monkeys and did not consider them a valuable bush meat; two respondents used howler monkeys as bait for river shrimp, and one hunted these primates for meat and medicine (howler monkey broth was used as treatment for respiratory illness). Howler monkeys kept as pets were rare. The reports on hunting activity were distributed across the survey sites, unrelated to ethnic group of interviewees or geographic region. The only other primate species in the area reported in the interviews was the spider monkey Ateles geoffroyi. The white-faced capuchin monkey, Cebus capucinus, was not known to locals and probably does not extend into Guatemala. Records for Ateles geoffroyi presence were slightly lower than those for Alouatta species (34% and 40% of interview sites, respectively), and its range overlapped with both A. palliata and A. pigra. Ateles occurred in sites with and without howler monkeys and there was no evidence of association or dissociation with Alouatta (test of independence G2= 0.11, df = 1, p = 0.76).
No evidence was found for a current zone of sympatry between Alouatta pigra and A. palliata. Only one interviewee claimed to have seen both species in mixed troops, but this was disputed by other interviewees at the same site (site 19) and it conflicted with records from surrounding sites as well. The identified range limits for A. pigra are: Lake Izabal and Río Dulce in the east, unsuitable habitat of pine forest west of Purulhá in Baja Verapaz in the southwest and pine forest and dry shrubland in foothills of Sierra de las Minas in the south (Fig. 1). The southeastern range limit is not associated with any barrier. We found the most southeasterly populations of A. pigra in the tip of Sierra de las Minas (site 35), and historical records place the species farther east in Quirigua in the middle Motagua valley (Salvin and Goodman, 1879; museum specimen in the Smithsonian National Museum of Natural History, USNM 238704) (Fig. 1). The western range limit of A. palliata is dry deciduous shrubland. The heavily disturbed area to the south of Lake Izabal in the lower Motagua valley is the northwest limit and the Río Dulce the north limit.
Discussion
This study reports the distribution of the endangered black howler monkey, Alouatta pigra, in its southern geographic range in topographically and ecologically heterogeneous eastern Guatemala, providing needed information on detailed occurrence localities (Rodríguez-Luna et al., 1996; Matamoros et al., 1997) and new reports of its ecological tolerances. It also contributes to the limited knowledge of the status and distribution of the mantled howler monkey, A. palliata, in Guatemala. Alouatta pigra was widespread and occupied several different habitat types across altitudinal ranges from sea level up to 2705 m. Our results show no evidence of significant association with a specific vegetation type or altitude. In his preliminary study in Guatemala, Curdts (1993) also found A. pigra at high elevations above 2000 m a.s.l. These results contrast with the reports that A. pigra is restricted to elevations below 400 m a.s.l., with a strong affinity for riverine forest (Horwich and Johnson, 1986; Watts et al., 1986; Ostro et al., 2000). Models of the species' current range based on preference for lowland forests (Luecke, 2004) need to consider potential premontane and montane habitats to avoid underestimation of actual range.
Forest disturbance by humans influences the distribution patterns of A. palliata and A. pigra in the region mainly by eliminating suitable monkey habitat. This is especially the case along the Río Motagua Valley and south shores of Lake Izabal and Río Dulce. Here extensive monocultures and cattle pastures have replaced the original tropical forest (INAB, 2001), which may account for the high frequencies of A. palliata in disturbed vegetation types. The ability of A. palliata to survive in fragmented habitats is widely recognized by other authors (e.g., Estrada and Coates-Estrada, 1984; Silva López et al., 1988; Clarke et al., 2002). Deforestation in the range of A. pigra occurs mainly in the upper and middle Rio Polochic valley, where the species was once very common (Salvin and Godman, 1879), as well as north of Lake Izabal. The species was present in some disturbed sites, in accordance with other reports of A. pigra thriving in forest fragments (Horwich and Johnson, 1984; Baumgarten, 2000; Estrada et al., 2002).
The absence of A. palliata from the Cerro San Gil area (sites 44, 47, 48, 49) was unexpected, since the area has protected status and extensive forest cover with other wildlife, including the spider monkey. The premontane and montane elevations at these sites may act as a limiting factor for the species. Other studies have also found that A. palliata is absent from elevations above 700 m a.s.l. in mountainous ranges in Mexico (Silva López et al., 1988; Estrada and Coates-Estrada, 1996). The species is present at elevations up to 1500 m a.s.l., but only at lower latitudes in Central America (Timm et al., 1989). The red howler monkey, Alouatta seniculus, occurs at high elevations in the Colombian Andes (Hernández-Camacho and Cooper, 1976; Gaulin and Gaulin, 1982).
We found little evidence for hunting pressure on howler monkeys and no association between hunting and the distribution of the two species. Hunting of howler monkeys is also limited in Belize (Horwich and Johnson, 1984) but more common in Mexico (Estrada and Coates-Estrada, 1984; Horwich and Johnson, 1984; Silva López et al., 1988). In contrast, subsistence hunting has accounted for local extinctions of Alouatta in localities in the Brazilian Amazon (Peres, 1990) and the Peruvian and Bolivian Amazon (Freese et al., 1982). Alouatta occurrence may be influenced by competition with other primates in some habitats (Eisenberg, 1979), although not always (Peres, 1997; Iwanaga and Ferrari, 2002). We found no evidence for competitive exclusion between Ateles and Alouatta species. In Mexico the spider monkey Ateles is more widespread than Alouatta and sympatric with it in only a few places (Watts et al., 1986; Rodríguez-Luna et al., 1987; Silva López et al., 1988).
The study region has a high potential for the conservation of the endangered black howler monkey, A. pigra, since six of its protected areas harbor the species (Table 1), the largest being the Sierra de las Minas Biosphere Reserve (246,803 ha) (CONAP, 2005). Furthermore, the low hunting pressure and the occurrence of the species outside protected areas provides the opportunity for community-based conservation projects and sustainable resource programs, which have proven effective for howler monkey conservation in Belize and Mexico (Horwich, 1998). The occupied habitats are ecologically heterogeneous in comparison with the species' range in the Yucatan Peninsula, which is predominantly lowland rainforest (Horwich and Johnson, 1986; Watts et al., 1986; Navarro et al., 2003). The populations of A. pigra living in cloud forests at high elevations are an important component in conserving the species variability and its gene pool. Previously considered to be unsuitable habitats, highlands will have to be considered in distribution assessments and management plans for the species. The severe deforestation south of Lake Izabal-Río Dulce puts the populations of the mantled howler monkey A. palliata at risk in Guatemala. These are the most northerly populations of the subspecies A. palliata palliata. Important conservation actions here include conservation in the protected areas without enforcement, the legal declaration of the proposed Sierra Caral protected area, and the establishment of the Mesoamerican Biological Corridor project that connects Punta de Manabique Wildlife Reserve with protected areas in Honduras.
The results of this study are relevant to clarifying the distribution of Alouatta pigra and A. palliata in their contact zone in eastern Guatemala. We found no evidence for current range overlap or sympatry as well as no support for the previous proposed sympatric areas around Río Sarstun (Horwich and Johnson, 1986) or in the highlands of Baja Verapaz, Alta Verapaz, and Sierra de las Minas (Curdts, 1993). In the northwestern part of the putative contact zone, the Rio Dulce acts as a physical barrier separating both species. Curdts (1993) also reported A. pigra on the north bank and A. palliata on the south bank of this river. River boundaries are often limiting factors for the distribution of primates (Ayres and Clutton-Brock, 1992; Wallace et al., 1996). In the eastern part of the contact zone, A. pigra occurs on the southwest side of Lake Izabal and tip of Sierra de las Minas. South of Lake Izabal suitable monkey habitat has been removed and A. pigra is known only from historic records. Further east in the lower Motagua Valley we registered exclusively A. palliata. It is possible that a narrow sympatry existed south of the lake before forest loss. Further range overlap is not evident, probably explained by ecological differences. Our results suggest that A. palliata is associated with forests found at low elevations, precluding its expansion into the wet and cold habitats of Sierra de las Minas where A. pigra is found. Habitat preferences seem to act as a barrier between howler monkey species in other cases of near sympatry (Crockett, 1998) as reported for A. fusca and A. caraya in northern Argentina (Di Bitetti et al., 1994) and A. seniculus and A. caraya in southwestern Amazonia (Iwanaga and Ferrari, 2002). Further, A. pigra's lack of wide expansion into A. palliata's range in the lowlands may perhaps involve assortative mating or hybrid inviability.
In the extreme south of the potential contact zone, the ranges of both species are interrupted by dry deciduous shrubland south of the Sierra de las Minas. The semidesert characteristics of this vegetation clearly constitute an effective ecological barrier between the species. In contrast to the scenario in Guatemala, A. pigra and A. palliata in Mexico have a broad sympatric area that extends over the lowlands of the states of Tabasco and Campeche. Nevertheless, species introgression seems to be limited as Smith (1970) found no evidence for hybridization based on museum specimens from the same localities, although occasional interbreeding may occur in mixed troops of howler monkeys in the area (Cortés-Ortiz et al., 2003).
Acknowledgments
This study was supported by two Louisiana Governors Office of Environmental Education Research Grants. We are grateful to FUNDAECO, Fundación Defensores de la Naturaleza, FUNDARY, CECON-USAC, Parroquia El Estor, Finca Tijax, Finca Cálix and Finca Champas for technical support and research permission. Special thanks to all interviewees and enthusiastic guides and to Daniel Tenes, Carlos Avendaño, Haroldo García and José Moreira for assistance in the field. We thank Kellen Gilbert for comments on this manuscript. This study was carried out with the required permits from the Consejo Nacional de Areas Protegidas-Guatemala. The senior author was supported by a grant from the Fulbright Faculty Development Program Central America.