Squirrel Monkeys in Costa Rica: Drifting to Extinction

Squirrel monkeys in Costa Rica: drifting to extinction

Sue Boinski, Katharine Jack, Craig Lamarsh and Jessica A. Coltrane

Central American squirrel monkeys Saimiri oerstedii are limited to Costa Rica and Panama, and have never been abundant. The Costa Rican population is now decimated. Based on our survey of squirrel monkeys throughout a large portion of the Costa Rican range of this species in 1996, we can confirm that at least 1246 squirrel monkeys remain in 26 spatially dispersed localities. Despite probable undercounts of squirrel monkeys within sites and potentially missed localities, the total population size should be considered Jar below a size that would provide long term genetic viability. Deforestation and habitat fragmentation spurred by agricultural and tourism development are the familiar culprits contributing to this decline. Our strong recommendation is that future conservation efforts be targeted at the level of each specific locality, perhaps recruiting local, national and international sponsors. A focused strategy would allow management efforts to be tailored to the circumstances specific to each site and thus allocate scarce resources more efficiently.

Introduction subspecies were most recently estimated at c. 3000 and 500-1000, respectively (Boinski and 'Drifting' in the sense 'to move or be carried in Sirot, 1997). Concomitant with this steep de a casual, unimpeded manner' is, unfortu cline in the estimated population of squirrel nately, an appropriate term to describe the monkeys was a massive increase in the rate progression of Central American squirrel and absolute amount of deforestation and for monkeys in Costa Rica, Saimiri oerstedii oerste est fragmentation in Costa Rica (Butterfield, dii and S. o. citrinellus, towards extinction. 1994). Squirrel monkeys in Panama (only the Historical records indicate that the distri S. o. oerstedii subspecies) have fared even bution of this species has always been re worse. For more than a decade Panamanian stricted to the Pacific wet lowlands of Costa wildlife biologists thought that squirrel Rica and Panama, an area of less than 8000 sq monkeys in their country were extinct as a re km, and 600-1000 km from the nearest South sult of massive deforestation (Boinski and American squirrel monkeys (Hershkovitz, Sirot, 1997). In 1996, however, a small popu 1984). The urgent need for conservation lation (< 25 individuals) was discovered in a measures has been emphasized relatively re 70-ha reserve on the Pacific coast at La cently (Boinski, 1985, 1987a; Wong et al., 1994). ' Barqueta (8°23'N, 82°36'E; R. Samudio, pers. In 1983 the number of squirrel monkeys in comm.). Costa Rica was estimated at 200,000 based on One of the notable features of squirrel informant interviews (Vaughan, 1983). monkey conservation in Costa Rica has been Currently Saimiri oerstedii oerstedii (Figure 1) is the lack of detailed documentation of the sur listed as Endangered while S. o. citrinellus viving populations. The few surveys reported (Figure 2) is listed as Critically Endangered thus far have either relied exclusively on inter (Rylands et al., 1995; Rodriguez-Luna et al., views with local inhabitants (Vaughan, 1983), 1996, IUCN, 1996). Populations of these two censuses focused on highly restricted © 1998 FFI, Oryx, 32 (1), 45-58 45 S. BOINSKI ET AL.

Figure 1. Three juvenile male S. o. oerstedii, each approximately 2.5 years old (Craig Lamarsh) . geographic areas (Wong, 1990a; Arauz, 1993), countries of Central and South America with or fieldwork in restricted areas combined with the greatest conservation sophistication information provided by biologists with (Fundaci6n Ambio, 1993; Reid et al., 1994), and knowledge of other regions (Boinski, 1985, the largest proportion of land reasonably pro 1987a; Boinski and Sirot, 1997). The lack of ac tected in national parks, reserves, and wildlife curate data has probably hindered restoration refuges (11 per cent; Carrillo and Vaughan, or stabilization of squirrel monkey numbers. 1993). Furthermore, tourism, much of it gener For example, two recent international work ated by Costa Rica's scenery and wildlife, is its shops devising management strategies for S. o. single largest source of foreign currency. The citrinellus relied heavily on VORTEX Instituto Costarricense de Tourismo (ICT) esti Population Viability Analysis (PV A) models mates that in 1996 alone over $US1 billion was (Wong et al., 1994; Matamoros et al., 1996). Yet generated by 800,000 foreign tourists in Costa the accuracy of PVA models is constrained by Rica {ICT, unpubl. data). If an appealing the quality of the data on which they are based monkey can become extinct or closely ap (Lacy, 1993), and the population estimates proach extinction here, what are the realistic used in these PV As are inflated relative to prospects for effective species conservation more recent surveys (Boinski and Sirot, 1997). elsewhere? In 1996 we conducted the most extensive survey to date of squirrel monkeys in Costa Rica, encompassing the range of S. o. oerstedii Natural history and much of the range of S. o. citrinellus. The purpose of this report is to present distri Owing to arguments that squirrel monkeys bution data for these two subspecies, evaluate were probably first introduced into Central current threats to their survival, and recom America from South America by prehispanic mend appropriate management strategies. In Amerind traders (Hershkovitz, 1984), the addition, we suggest that S. oerstedii provides merit of conservation efforts has been some a useful case study of the effectiveness of con what clouded and suspect. Why bother with servation efforts confronting pressures for de an 'introduced' species when so many in velopment. Costa Rica is arguably among the digenous species are at risk? Convincing 46 © 1998 FFI, Oryx, 32 (1), 45-58 SQUIRREL MONKEYS IN COSTA RICA

Figure 2. A fully mature adult male S. o. citrinellus (Sue Boinski) . evidence now exists that S. oerstedii was pres social organization of any primate taxon with ent in Central America prior to the arrival of typically large troops; neither males nor fe humans, and is distinguished from squirrel males can be described as 'dominant' over the monkeys in South America in numerous re other sex. spects. First, cladistic analyses based on exten Distinctive features of the natural history of sive mtDNA sequence data from hair follicle Central American squirrel monkeys enhance collections from squirrel monkeys of known their susceptiblility to local extinction (Boinski geographic origin in Costa Rica (S. o. oerstedii 1987b, 1988, 1989; Boinski and Fowler, 1989; and S. o. citrinellus), Guyana and Suriname Boinski and Sirot, 1997; Wong, 1990b). (Saimiri sciureus sciureus), and Peru (S. s. bo 1 In continuous forest, Central American liviensis) found that approximately the same squirrel monkeys have large troop sizes degree of sequence divergence exists among (40-70) and large annual ranges (> 2 sq km). the three clades, and that the divergence is Long-term monitoring of isolated populations likely to be ancient (> 500,000 years; Boinski indicates that local extinction is almost certain and Jacobs, 1997). Second, detailed compari when a group contains fewer than 15 mem sons of the morphology and enzyme banding bers and less than c. 30 ha available habitat patterns of S. oerstedii and South American (Boinski and Sirot, 1997). squirrel monkeys indicate greater differences 2 Travel routes, and thus the habitat areas between the two geographic areas than are that can be exploited, have stringent require found within South American populations ments and are readily abandoned. Central (Costello et al., 1993). Lastly, the social organ American squirrel monkeys travel and forage ization of squirrel monkeys S. oerstedii in almost exclusively on branches between 1-2 Costa Rica is markedly distinct (Boinski, in an in diameter and seldom make horizontal press). Squirrel monkey troops in South leaps greater than 2 m. Moreover, travel or America always exhibit strong dominance hi any activity on the ground is extremely rare. erarchies and high levels of aggression. Costa As a consequence, road cuts, clearings for tele Rican squirrel monkeys, however, arguably phone and electric power lines, or even pri exhibit the most egalitarian, least aggressive mary forests with gaps between tree canopies © 1998 FFI , Oryx, 32 (1), 45-58 47 S. BOINSKI ET AL.

Figure 3. A typical example of the early successional habitat on which Costa Rican squirrel monkeys are now heavily dependent: an overgrown pasture adjacent to a heavily disturbed secondary forest on the Burica Peninsula (Katharine Jack). and little understorey can severely circum forest, is seldom beneficial because squirrel scribe troop movements. monkeys exploiting these areas become 3 The diet of Central American squirrel agricultural pests and vulnerable to a panoply monkeys consists largely of phytophagous of risks unique to this habitat (see below). arthropods (Orthoptera such as grasshoppers 4 Central American squirrel monkeys are and katydids, and lepidopterous caterpillars), documented to suffer severe food stress dur and small berry-like fruits (and cultivated ing the annual rainy season. The Pacific wet guava, banana, mango and cacao when little lowland region of Costa Rica has extremely else is available). Disturbed and early succes pronounced seasonal fluctuations in rainfall, sional habitats are clearly preferred to primary with negligible rain from January to March forest. This is because regenerating forest, but an intense rainy season from August to compared with primary forest, generally has a November, when arthropod abundance de greater abundance of phytophagous arthro clines dramatically. Moreover, the availability pods, fruits and flowers (Figure 3). Therefore, of the small berry-like fruits that squirrel squirrel monkeys are extremely vulnerable to monkeys specialize in harvesting is also at the habitat succession. Parts of Costa Rican pro annual nadir. As a result, squirrel monkeys tected areas, such as the Llorona plateau of greatly expand their ranging areas and virtu Parque Nacional Corcovado, in which sec ally every moment of diurnal activity is allo ondary habitat was allowed to follow its typi cated to searching for food. Not surprisingly, cal course of regeneration, have experienced the physical condition of squirrel monkeys local extinction of squirrel monkeys. Yet suc visibly declines during the rainy season. The cessional habitat created by human settle minimum habitat required to sustain a troop ments, although more abundant than primary over the long term is undoubtedly best

CARIBBEAN SEA

PACIFIC OCEAN COSTA RICA

9·00·

...... ·········...... ( ...... f \.. PANAMA ·· · · · ········. ~ . - " . . . . ~ f 8"30' ( '\:" } Survey Area

PACIFIC OCEAN

54· 00· 83"30'

Figure 4. Area of the Pacific wet lowlands of Costa Rka surveyed for this report represented by the extent of its habitat needs Sirot, 1997). This collation determined the during the rainy season. maximum limits of the surveyed area and the areas targeted as likely to contain squirrel monkeys but for which no recent and definite Methods proof was available. Most sites were visited by three to four fieldworkers, the exceptions The primary objective of our study was to being Parque Nacional Corcovado and adjac document Costa Rican localities where squir ent sites, which were visited by one field rel monkey troops still survived. Figure 4 indi worker. In the course of our travels in the cates the the areas surveyed for this report. Pacific wet lowlands of Costa Rica we also Published information on definite or possible searched for any unreported forested habitat populations, in addition to extensive inter that might contain squirrel monkeys. The only views with knowledgeable informants were area we excluded from our survey was a zone collated immediately prior to the initiation of of approximately 150 sq km, between Quepos the survey (see also Map 1 in Boinski and (9°24'N 84°10E) and Parrita (9°32N 84°l8'E), © 1998 FFI , Oryx, 32 (1), 45-58 49 S. BOINSKI ET AL.

CARIBBEAN SEA

l.._ . _... ·

COSTA RICA PAC/AC OCEAN

9·00· Saimiri oerstedi citrinellus- letters Saimiri oerstedi oerstedi-numbers <... J i... PANAMA " "\

8"30' J .r ·..r _/

PACIFIC OCEAN

N o.__~.____..10~...... ~-20~~.__~JOKm A

84 '00' 83' 30'

Figure 5. A map of the confirmed localities for remaining populations of both subspecies of Costa Rican squirrel monkeys. Historical records indicate that the Rio Grande de Terraba was formerly the boundary separating S. o. citrinellus to the north of the river and S. o. oerstedii to the south.

representing the most northern area in which areas (trail construction being neither feasible squirrel monkeys (S. o. citrinellus) have oc nor desirable), as well as exploiting stream curred in historical knowledge. Wildlife biolo beds, extant trails and roads. Approximately 4 gists from the Programa Regional en Manejo h per day were spent surveying in the field, de Vida Silvestre of the Universidad Nacional concentrated in the morning and late after in Heredia, Costa Rica were completing a sur noon, when squirrel monkeys are most active vey for 5. o. citrinellus in that zone just as our and readily detected (Boinski and Sirot, 1997). survey began (Altrichter et al., 1996). Other daylight hours were allocated to travel Once a known or potential squirrel monkey ling between sites and interviewing infor locality was reached several techniques were mants. Localities and areas surveyed were employed to track and locate squirrel considered 'unlikely' or 'impossible' to con monkeys in site transects. For transects we fol tain squirrel monkeys when they were effec lowed compass bearings through forested tively denuded of trees and shrubs or 50 © 1998 FFI, Oryx, 32 (1), 45-58 SQUIRREL MONKEYS IN COSTA RICA consisted of large, continuous plantations of Nacional Corcovado, 41,789 ha, with a mini trees, such as oil palm Elais guaianensis, known mum of 200 S. o. oerstedii, and (iii) Refugio not to provide significant food resources to Nacional de Fauna Silvestre, 2830 ha, with at squirrel monkeys. When squirrel monkeys least one troop of 58 S. o. oerstedii. Much of the were encountered, as much information as habitat in all three protected areas is primary possible was recorded on troop size, age-sex or high-grade forest and thus rarely exploited composition, foraging behaviour, habitat use, by squirrel monkeys. All other reported lo social and vocal behaviour. Whether or not calities are on private and otherwise unpro squirrel monkeys were found, we investigated tected lands. The largest surviving and least current and former threats to this species at disturbed breeding population of either sub each site. species is undoubtedly that in Parque Nacional Corcovado. It is possible that current numbers may be higher than the 200 monkeys Confirmed localities we counted because the population in this park in 1984 was estimated to be 400-500 Here we report only those Costa Rican lo based on 3 years of field work resulting in 10 calities in which the presence of squirrel individually distinguished troops (Boinski, monkeys was visually confirmed. Our 1996 1985; Boinski and Sirot, 1997). survey established that squirrel monkeys were Our survey almost certainly missed some present at a minimum of 26 localities in Costa sites with squirrel monkeys and, with the ex Rica, six with S. o. citrinellus and 20 with S. o. ception of the 105 S. o. citrinellus in Parque oerstedii (Table 1, Figure 5). These localities are Nacional Manuel Antonio, probably under often so distant from the nearest other locality counted the number of individuals at some of (mean 3.6 ± 1.1 SE km, n = 26), compounded the sites where squirrel monkeys were found. by the absence of intervening forest corridors, However, we are reasonably confident that no that dispersal of squirrel monkeys between lo locality with large numbers(> 100) of squirrel calities is extremely unlikely. Even troops monkeys remains unidentified in the area we within 1 km of the range of another troop surveyed. The most likely portion of Costa commonly face a gauntlet of challenges to dis Rica in which undocumented troops may re persal (see below). We found 284 S. o. citrinel main in our survey area are valley bottoms in lus and 962 S. o. oerstedii individuals. Twenty the vicinity of the village of Coto Brus (8°58'N, five squirrel monkeys per troop was used as a 83°06'E; Table 1, Figure 5, key label 1). There conservative estimate of troop size for 10 we found a single troop in an isolated patch of troops of S. o. oerstedii in which exact counts riverine forest, and squirrel monkeys may also were not achieved. For example, only a cluster inhabit other relict forest patches in this exten of three adult males was sighted at both sive, although heavily deforested, valley sys Puerto Escondito and Finca San Gerardo tem. Several inaccessible patches of forest in (Table 1; Figure 5, key labels 13 and 19). the southernmost tip of the Burica Peninsula Because adult males in poorly habituated also warrant further scrutiny . Saimiri oerstedii troops often approach human It should be noted that approximately 20 observers and other potentially dangerous in isolated groups totalling 445 individual (397 truders while other troop members flee, an ap adult) S. o. citrinellus were identified by the parent defense strategy (Boinski, unpubl. data; Universidad Nacional survey between Boinski and Mitchell, 1994), these small Quepos and Parrita (Altrichter et al., 1996). groups of males indicated the presence of a troop at the site. A small proportion of the squirrel monkeys Current threats recorded was located within three protected areas: (i) Parque Nacional Manuel Antonio, The situation in Panama foreshadows the fu 683 ha, with 105 S. o. citrinellus; (ii) Parque ture of Costa Rican squirrel monkeys. Squirrel © 1998 FFI, Oryx, 32 (1), 45-58 51 S. BOINSKI ET AL.

Table 1. Confirmed squirrel monkey localities in Costa Rica

Location, key Sub- Min. count Latitude/ Major label and name species* of troop(s) longitude threatst Comments

A Parque Nacional Soc 37 9°23.5', 84° 08' DT,F Hurricane in 1992 levelled Manuel Antonio 40 c. 25 per cent of 5-sq-km 28 reserve. These areas are now In June 1996 prime squirrel monkey only these 3 habitat. This absolute troops (n=105) population count is less than rangedin one half of some previous this reserve. estimates. B Reserva Biologica, Soc 30 9'25', 84'09' DA, DT, Manuel Antonio F c Mariposa Hotel, Soc 45 9°24', 84°10' DT,F Three S. o. oerstedii adult Manuel Antonio females were photographed carrying oerstedii x citrinellus hybrid infants. D Jardin Gaia, Soc 24 9°24.5', DA, DT, Two S. o. oerstedii adult Manuel Antonio 20 84° 08.5' F females were photographed carrying oerstedii x citronellus hybrid infants. This and sightings above imply that at least five adult females of S. o. oerstedii were released in vicinity of Manuel Antonio in previous years. E Boca de! Ro Soc 25 9' 22.5', 84°06' DA,F,P Monkeys travel through Naranjo palm plantations to reach forest patches. F La Botanica, Soc 35 9°30', 84°03' D,F A planned dam may Londres jeopardize remaining troops. 1 Coto Brus Soo 25 s·5s·, 83°06' D,F Monkeys in this valley system rely on relict large trees adjacent to the rivers as travel routes. The high elevations of valley walls restrict group movements out of the valleys. 2 Refugio Nacional Soo 58 8°39', 83°09.5' D, F,H This reserve is cut off from de Fauna Silvestre, populations of monkeys to Golfito the south by major roads. 3 Esperanza Soo 31 8°34', 83'06' DA,F Habitat fragmented by roads, rivers, agriculture. 4 Bajo de Coto Soo 28 8°34', 83' 07' DA,F Habitat fragmented by roads, rivers, agriculture. 5 Bajo Grapa (Finca Soo 25 8°34.5', 83'03' DA, F Habitat fragmented by de David roads, rivers, agriculture. Olovorado Zomora) 6 Tiskita Lodge, Soo 35 8°23.5', DA,DT, Tourism developing. Palm Punta Banco 25 83° 08.5' P, S and Gmelina plantations encroaching from the north.

Table I. Confirmed squirrel monkey localities in Costa Rica (continued)

Location, key Sub- Min. count Latitude/ Major label and name species* of troop(s) longitude threats+ Comments

7 Pavones (Finca de Soo 25 8°24', 83°07.5' DA,DT, Tourism developing. Palm Laura Alvarado 30 P, S and Gmelina plantations Gomez) 30 encroaching from the north. 8 Punta Banco, Soo (2 troops) 8°21·, 83°06.5' CH,D Burica Peninsula 50§ 9 La Penita, Burica Soo (1 troop) 8°19', 83"04' CH,D, Peninsula 25§ F 10 Clarita, Burica Soo 38 8°16', 82"58' CH,D, Peninsula F 11 Playa Zancudo Soo (2 troops) 8°33', 83°08' CH, DA, Gmelina plantations most 50§ DT,F, S immediate threat. 12 Matapalo, Bosque Soo 20 8°22', 83°18' DA,F del Cabo, Osa 22 Peninsula 13 Puerto Escondito, Soo (3 males 8°40.5', 83"27 DA,F Osa Peninsula probably defensive sub-group of troop) 25§ 14 Lapas Rios, Osa Soo (1 troop) 8°24', 83°17' D,F Peninsula 25§ 15 Between Agua Soo 2 troops 8°26', 83°25' D,F Buena and Carate totalling approx. lOOinds. 16 Carate, Osa Soo (1 troop) 8°27', 83°27' D,F Peninsula 25§ 17 Agua Buena, Osa Soo (1 troop) 8°26', 83°24' Peninsula 25§ 18 Parque Nacional Soo Min. of Between H Corcovado, Osa 5 troops, 8°27' and 8°32' Peninsula each with and 40 members, 83°31' and N;,;200 83°39' 19 FincaSan Soo (3males, 8°34.5', 82°53' D,DA, F Installation of power lines Gerardo (Sr probably have destroyed and Ramon Cabezas) defensive fragmented the limited subgroup remaining habitat. of troop) 25§ 20 Sierpe Soo 20 8°52.5', 83°29' DA,F, Banana, palm and Gmelina P,S plantations pose increasing threats. Toxicity from extensive pesticide use.

* Soc, Saimiri oerstedii citrinellus; Soo, S. o. oerstedii. t CH, capture for pets and hunting as crop pests; D, deforestation; DA, deforestation due to agriculture; DT, deforestation due to tourist development; F, habitat fragmentation; H, habitat succession; P, banana and oil palm plantations; S, silviculture. §Estimated minimum group sizes (25 animals per troop) are given in instances when accurate counts could not be obtained.

© 1998 FFI, Oryx, 32 (1), 45-58 53 S. BOINSKI ET AL. monkey populations in Costa Rica are deci travel routes when thin arboreal branches are mated relative to historic levels. An optimistic unavailable and at many sites power lines are estimate, based simply on doubling the ap the only nonterrestrial connections among dis proximate cumulative number of squirrel continuous ranging areas (Figure 6). Despite monkeys found by our survey and that of the this apparently positive feature, insulation on Universidad Nacional team (Altrichter et al., the power lines is poor and the electrocution 1996), is that a total of 1500 S. o. citrinellus and of squirrel monkeys and other arboreal mam 2000 S. o. oerstedii survives in the wild or as mals is common (Figure 7; Boinski and Sirot, near commensals to humans. In any case, the 1997). current effective population size of squirrel Plantations and silviculture are threats monkeys in Costa Rica is undoubtedly below meriting mention apart from agricultural de the 5000 individuals that Lande (1995) calcu velopment because of the huge expanse of for lates are needed to avoid inbreeding de mer squirrel monkey habitat they now pression and maintain long-term genetic occupy. As far as we could determine, effec viability in a species. This pessimistic prog tively all of the lowland coastal area (550 sq nosis is only exacerbated when the genetic km) between Savegre and Palmar Norte fragmentation of the remaining squirrel (Figure 4), formerly prime S. o. citrinellus habi monkeys in Costa Rica is considered; Lande' s tat (Boinski 1987b), is now converted to oil estimation is premised on an assumption of palm plantations and human settlements. random mating across a population. Our edu Patchier, but still extensive oil palm and ba cated guess is that if the current development nana plantations are also a significant land regime in Costa Rica continues unimpeded, in scape component in the range of S. o. oerstedii 15 years squirrel monkeys will survive at no to the south. more than half, and probably fewer, of the lo A recent incursion on S. o. oerstedii habitat calities this survey documented. In 50 years, around the Golfo Dulce and Burica Peninsula the distribution map of Costa Rican squirrel is the establishment in the past decade of plan monkeys may have only two localities indi tations totalling at least 14,000 ha of Gmelina cated, Parques Nacionales Corcovado and arborea, a fast-growing tree native to India, by Manuel Antonio. Stone Container Corporation of the United Deforestation and habitat fragmentation are States. Gmelina is planted in pastures and the primary causes of the present critical other previously cleared areas as monoculture status of Costa Rican squirrel monkeys. stands, which are regularly sprayed with Agriculture continues to erode squirrel pesticide, cleared of undergrowth and har monkey habitat at the level of both agri vested for wood pulp after 8 years. Squirrel business and the individual small farmer. monkeys do not eat its fruit and have not yet Development for tourism is perhaps a less ap been observed to forage for whatever arthro preciated, but no less important, threat, es pod fauna might survive the pesticide appli pecially for S. o. citrinellus in the vicinity of cations. From the perspective of squirrel Quepos and Parque N acional Manuel monkeys this nominally 'empty' land was cer Antonio. Tourists and squirrel monkeys alike tainly either already prime habitat or would have strong habitat preferences for beachfront have quickly succeeded to the early success and lowland forests. With tourists come roads, ional habitat preferred by squirrel monkeys. restaurants, hotels, stores, and electric and On the other hand, plantations of Gmelina do telephone lines, all of which fragment and iso provide transient benefits in that they allow late squirrel monkey populations. The clearing squirrel monkey troops to travel between of forest and scrub along the routes of power otherwise isolated patches of exploitable forest. lines has led to electrocution being arguably Other notable threats, albeit of reduced im the major source of direct mortality for S. o. pact compared with deforestation and habitat citrinellus, and a significant one for S. o. oerste fragmentation, include habitat succession, the dii. Squirrel monkeys use power lines for pet trade, and hunting as crop pests.

Figure 6. Squirrel monkey habitat within Parque Nacional Manuel Antonio is bisected by a road paralleled by a wide clearance for powerlines. S. o. citrinellus, such as these females with 2-month old infants, depend on the power lines as travel routes along and across the road (Sue Boinski).

1 Squirrel monkeys are clearly at risk of los ing further numbers in Parque Nacional Corcovado and. the Refugio Nacional de Fauna Silvestre in Golfito as a result of succession of formerly extensive disturbed habitats. Fortuitously for the squirrel monkeys in Parque Nacional Manuel Antonio, a 1992 hurricane toppled approximately 25 per cent of the forest of this small protected area, and the areas of regenerating forest now provide more prime habitat for squirrel monkeys than was present before the hurricane. 2 A pet trade still exists where Boinski and Sirot (1997) predicted none. Ten pet squirrel monkeys were encountered during our sur vey, and five more were illegally on exhibit at a private zoological garden catering for tourists. At least 15 of the squirrel monkeys currently held in the two legal captive colonies Gardin Gaia, a wildlife rehabilitation center in Manuel Antonio that is open to visi tors, and a private colony on the outskirts of San Jose owned by Dr Ludwig Muller) are for mer pets. More illicit captives undoubtedly exist. Numerous informants independently re ported that (a) the current price for a pet squirrel monkey was $US30-50, (b) pet squir rel monkeys were relatively easy to obtain, Figure 7. A dead adult male S. o. citrinellus and, (c) animals were not infrequently smug electrocuted while traversing a power line about 2 gled out of the country. One restaurant be km outside the Parque Nacional Manuel Antonio tween Quepos and Parque Nacional Manuel (Katharine Jack) . © 1998 FFI, Oryx, 32 (1), 45-58 55 S. BOINSKI ET AL.

Antonio has a commensal troop that forages The basic arsenal of conservation measures on the patrons' tables and is also given food for squirrel monkeys is not large, and each by restaurant staff to maintain the troop as a should be fitted to the circumstances of the tourist attraction. We strongly suspect that the specific locality. five adult female S. o. oerstedii we closely ob 1 Available forest must be protected from fur served in the Quepos tourist developments ther deforestation and fragmentation. All surrounding Parque Nacional Manuel parties making land-use decisions must be in Antonio (72 km up the Pacific coast from the formed of the economic and ecological bene closest point of this subspecies's documented fits of maintaining a local squirrel monkey range) were captives released in 1995. population. Owners of land for vacation re 3 Although we did not see hunting or car treats and tourist lodges, in addition to casses, numerous local informants, especially tourism industry associations, should be par in the vicinity of Quepos and on the Burica ticularly targeted for enlistment in conser Peninsula, readily and independently de vation projects. scribed hunting as an effective deterrent to 2 Second-growth and early successional habi crop raiding. This is consistent with an up tat must be maintained. Nationally protected surge of complaints of crop damage caused by areas and privately owned forests should be squirrel monkeys harvesting garden produce managed, not merely 'protected'. Squirrel compared with previous surveys (Boinski monkey habitat must contain a large compo 1985, 1987a; Boinski and Sirot, 1997). As with nent of secondary forest. Many of the plant the pet trade, hunting was previously thought genera that define early successional growth to be a negligible risk (Boinski and Sirot, in Pacific lowland Costa Rica, including 1997). Cecrapia, Piper, Miconia, Palicourea and Lycianthes, provide most of the fruit and arthropod components of the squirrel monkey Suggested management strategies diet. Food sources can be further augmented and stabilized across the annual cycle by Each confirmed squirrel monkey locality rep planting and maintaining patches of guava resents a distinctive situation regarding the trees Psidium guajava. During the wet season, immediate conservation threats, the mix of in the fruits of feral and cultivated guava trees terested government agencies, non-govern probably are crucial sources of food in many mental organizations, private landowners, of the remaining squirrel monkey localities. economic interests of the local community and 3 Squirrel monkeys use existing corridors for potentially useful remedies. Under these con crossing roads or moving between separated ditions, a simple, comprehensive prescription forest patches when they are available. In for squirrel monkey conservation suitable for areas inhabitated by squirrel monkeys safe all sites is not attainable. In any case, broad travel routes must be maintained or estab conservation attempts for either all Costa lished. The removal of branches overhanging Rican squirrel monkeys or one of the sub roads should be discouraged, as should un species based largely on public education in the necessary brush and tree clearance from road national media have been notably unsuccess sides, pastures and underneath powerlines. ful thus far. Therefore, we strongly advocate Along commonly used travel routes, ropes or that future conservation efforts by concerned other branch substitutes should be provided individuals, local groups and non-governmen across roads and other relatively narrow clear tal organizations should target a single site or ings, and the junctures of powerlines should tight cluster of sites to secure and maintain a be insulated where electrocution is a risk. specific local population of squirrel monkeys. 4 Squirrel monkey numbers should be moni Perhaps a separate advocacy and support tored within each locality at frequent inter group could be recruited locally, nationally vals, at least four times annually. This is much and internationally for each locality. easier than it appears at first glance. Costa 56 © 1998 FFI, Oryx, 32 (1), 45-58 SQUIRREL MONKEYS IN COSTA RICA

Rican squirrel monkeys are highly conserva References tive in their use of sleeping trees. Probably one Altrichter, M., Fernandez, M.T., Gomez, L.G., or two alternative trees account for 90 per cent Gonzalez, J., Fernandez, C., Herrera, H. et al. 1996. or more of trees used as sleeping sites. Once Evaluaci6n de la fragmentaci6n del hlibitat y estado del these sleeping sites are identified (sleeping mono titf (Saimiri oerstedi citrinellus) en el Pacifico trees are reasonably predictable as to location Central, Costa Rica . Programa Regional en Manejo and other characteristics; Boinski, 1987b; de Vida Silvestre, Universidad Nacional, Heredia, Costa Rica. Boinski and Sirot, 1997), observers can readily Arauz, J. 1993. Estado de conserva{:i6n de! mono tit{ document local population levels by monitor (Saimiri oerstedi citrinellus) en su area de distribu ing the trees at predawn and dusk. cion original. Thesis de Grado, Escuela de Ciencias 5 Based on criteria presented in Kleiman Ambientales, Universidad Nacional, Heredia, (1996), reintroduction or translocation of Costa Rica. squirrel monkey troops into areas in which Boinski, S. 1985. Status of squirrel monkeys in Costa they were previously extirpated is not war Rica. Primate Conservation, 6, 15-16. Boinski, S. 1987a. The status of Saimiri oerstedi cit ranted at this stage of Costa Rican squirrel rinellus in Costa Rica. Primate Conservation, 8, monkey conservation. Furthermore, we argue 69- 72. that resources are much more effectively spent Boinski, S. 1987b. Habitat use in squirrel monkeys maintaining squirrel monkeys in localities (Saimiri oerstedi) in Costa Rica. Folia primatologia, where they are already present rather than in 49, 151- 167. attempts to establish new populations in areas Boinski, S. 1988. Sex differences in the foraging be in which long term habitat suitability is uncer havior of squirrel monkeys in a seasonal habitat. Behavioral Ecology and Sociobiology, 23, 177- 186. tain. Boinski, S. 1989. The positional behaviour and sub 6 Once population and habitat losses are con strate use of squirrel monkeys: ecological implica trolled within each locality, attention should tions. Journal of Human Evolution, 18, 659-677. be given to maintaining genetic diversity. Boinski, S. In press. Stress responses in primates: Female dispersal predominates among Costa proximate mechanisms in the evolution of social Rican squirrel monkeys, and females readily organization. In Geographic Variation in Behavior: An Evolutionary Perspective (eds S. A. Foster and J. change troops with no aggression from either A. Endler). Oxford University Press, Oxford. resident male or female troop members Boinski, S. and Fowler, N.L. 1989. Seasonal patterns (Boinski and Mitchell, 1994). Wildlife popu in a tropical lowland forest. Biotropica, 21, lation geneticists should be consulted about 223--234. the minimum transfer rate of females needed Boinski, S. and Jacobs, 5.C. 1997. Behavioral and between isolated localities (Lande, 1995). mtDNA divergence in squirrel monkeys. American Journal of Physical Anthropology Supplement, 24, 78--79. Acknowledgements Boinski, S. and Mitchell, C.L. 1994. Male dispersal and association patterns in Costa Rican squirrel Ludwig Muller, Dario Castelfranco and Anne monkeys (Saimiri oerstedi). American Journal of Simpson de Gamboa are foremost among the many Primatology, 34, 157- 170. who rendered patient assistance with the planning Boinski, Sand Sirot, L. 1997. Uncertain conservation and implementation of this survey. Many Costa status of squirrel monkeys in Costa Rica, Saimiri Ricans generously gave us their time, advice and oerstedi oerstedi and Saimiri oerstedi citrinellus. Folia permission to enter their properties. We thank the primatologica, 68, 181-193. Direcci6n General de Vida Silvestre for permission Butterfield, RP. 1994. Forestry in Costa Rica: status, to work in Costa Rican parks and preserves. A grant research priorities, and the role of the La Selva from the National Geographic Society to SB pro Biological Station. In La Selva: Ecology and Natural vided essential financial support for this survey. History of a Neotropical Rainforest (eds L.A. Colin Chapman, Margaret Kinnaird and Gary Steck McDade, K. S. Bawa, H. A. Hespenheid and G. 5. thoughtfully commented on earlier versions of the Hartshorn), pp. 317-328. University of Chicago manuscript. Press, Chicago. Carrillo, E. and Vaughan, C. 1993. Costa Rica. In La Vida Silvestre de Mesoamerica: Diagnostico y Estrategia para su Conservacion (eds E. Carrillo and © 1998 FFI, Oryx, ai (1), 45-58 57 S. BOINSKI ET AL.

C. Vaughan), pp. 237-263. Editorial de la primates (Platyrrhini): distribution by country, Universidad Nacional (EUNA), Heredia, Costa endemism, and conservation status according to Rica. the Mace-Lande system. Neotropical Primates, 3 Costello, R.K., Dickinson, C., Rosenberger, A.L., (suppl.), 113-164. Boinski, S. and Szalay, F.S. 1993. A multidiscipli Vaughan, C. 1983. A report on dense forest habitat for nary approach to squirrel monkey (genus Saimiri) endangered wildlife species in Costa Rica. species taxonomy. In Species, Species Concepts, and Environmental Sciences School, National Primate Evolution (eds W. Kimbel and L. Martin), University, Heredia, Costa Rica. pp.177-237. Plenum Press, New York. Wong, G. 1990a. Uso del habitat, estimaci6n de compos Fundaci6n Ambio. 1993. Diversidad Biol6gica y icci6n y densidad poblacional del mono tit( (Saimiri Dessarrollo Sostenible. Euroamericana Ediciones oerstedi citrinellus) en la zona de Manuel Antonio, (CR) SA, San Jos~. Quepos, Costa Rica. Thesis de Grado, Escuela de Hershkovitz, P. 1984. Taxonomy of squirrel Ciencias Ambientales, Universidad Nacional, monkeys genus Saimiri (Cebidae, Platyrrhini): a Heredia, Costa Rica. preliminary report with a description of a hitherto Wong, G. 1990b. Ecolog(a del mono tit(: Saimiri oer unnamed form. American Journal of Primatology, 6, stedi citrinellus en el Parque Nacional Manuel 257-312. Antonio, Costa Rica. Thesis de Grado, Escuela de IUCN. 1996. 1996 IUCN Red List of Endangered Ciencias Ambientales, Universidad Nacional, Animals. IUCN, Gland, Switzerland. Heredia, Costa Rica. Kleiman, D.G. 1996. Reintroduction programs. In Wong, G., Matamoros, Y. and Seal, U. 1994. Wild Mammals in Captivity (eds D. G. Kleiman, M. Population and Habitat Viability Assessment E. Allen, K. V. Thompson, S. Lumpkin and H. Workshop for Saimiri oerstedi citrinellus. Final Harris), pp. 297-305. University of Chicago Press, Report. Conservation Breeding Specialist Group Chicago. (IUCN/SSC), Apple Valley, MN, USA. Lacy, R.C. 1993. VORTEX: a computer simulation model for population viability analysis. Wildlife Sue Boinski, Department of Anthropology and Research, 20, 45-65. Division of Comparative Medicine, 1350 Turlington Lande, R. 1995. Mutation and conservation. Hall, PO Box 117305, University of Florida, Conservation Biology, 9, 782-791. Gainesville, FL 32611-7305, USA. E-mail: Matamoros, Y. Wong, G. and Seal, U. (eds) 1996. [email protected] Population and Habitat Viability Assessment Workshop for Saimiri oerstedi citrinellus. Final Katharine Jack, Department of Anthropology, Report. Conservation Breeding Specialist Group University of Alberta, Edmonton, Alberta, Canada (IUCN /SSC), Apple Valley, MN, USA. T6G2H4. Reid, W.V., Laird, S.A., Gamez, R., Sittenfeld, A., Janzen, D.H., Gollin, M.A. and Juma, C. (eds). Craig Lamarsh, Department of Anthropology, 1994. La Prospecci6n de la Biodiversidad: El Uso de los University of Alberta, Edmonton, Alberta, Canada Recursos Geneticos para el Desarrollo Sostenible. T6G2H4. Instituto Nacional de Biodiversidad, San Jo~. Rodriguez-Luna, E., Mittermeier, R., Rylands, A., Jessica A. Coltrane, Tropical Conservation and Wong-Reyes, G., Carrillo, E., Matamoros, Y. et al. Development Program, Department of Wildlife and 1996. Hada un plan de acci6n para los primates Range Sciences, University of Florida, Gainesville, mesoamericanos. Neotropical Primates, 4 (suppl.), FL 32611, USA. 119-132. Rylands, A.B., Mittermeier, R.A. and Rodriguez Received 27 February 1997 Luna, E. 1995. A species list for the New W-Oi:J.d Accepte(fl3 O,aober 1997