Yellow-bellied spider monkey
(Ateles belzebuth)
Individuals
Three juvenile individuals at the center were considered apt to undergo the reintroduction process. These individuals are named Watson, Corny, and Kasha. Watson is a female that was confiscated by the Ministry of the Environment of Ecuador (MAATE) in July 2021 after being held as an illegal pet. She was an infant upon arrival. Watson was raised at amaZOOnico until September 2022 when she was moved to Makisapa Alpa to begin the reintroduction process. Corny is a male that was born at the center in 2018. His parents, Guima and Johan, are long term residents at amaZOOnico who cannot be reintroduced because of their psychological dispositions. Corny was raised by his mother at amaZOOnico until he was moved to Makisapa Alpa in September 2022. Kasha is a female that was born at amaZOOnico in 2017. She is the daughter of Johan and Mea (also a long-term resident at the center). Kasha was initially introduced into the group of free-living spider monkeys in 2019 (without an extensive rehabilitation period). In 2022, Kasha went missing from the group and was discovered weeks later at a cacao plantation several kilometers from amaZOOnico. Kasha was returned to the center and brought to Makisapa Alpa in September 2022 to join Watson and Corny for further rehabilitation.
Housing and care
Corny, Watson, and Kasha underwent rehabilitation at Makisapa Alpa for 17 months between September 2022 and January 2024. Efforts during this period focused on three behavioral aspects: 1) reducing comfort with humans, 2) food selection, and 3) reducing time spent on the ground. The Makisapa Alpa site is designed to reduce the number of humans that the monkeys will have contact with. For this reason, it is located 2 km from amaZOOnico and is inaccessible to the general public. Two local staff members were hired to care for the monkeys during this period, thus limiting regular human contact. To reduce the monkey’s desire for cultivated fruits (e.g. papayas, bananas…) they were provided a special diet consisting of large proportions of wild fruits (e.g. ungurahua and morete) and leaves. Furthermore, several enrichment activities designed to promote the use of the upper strata of the rehabilitation cage were implemented. This included placing food on top of the enclosure, hanging food in puzzle containers, and hanging special platforms for feeding.
Release site
The monkeys were released from the Makisapa Alpa enclosure on February 4, 2024. This date corresponded to a period when fruit availability was high in the forest surrounding the Makisapa Alpa site. A soft-release approach was taken whereby two feeding platforms were hung in the canopies of nearby trees and used to offer daily supplemental feeding to the monkeys, thereby preventing immediate long-distance dispersal and allowing post-release visual observation to monitor the health and well-being of the individuals.
Pre-release adaptations
To monitor post-release movement and behavior patterns, Kasha and Corny, were fitted with iridium-GPS collars made by Advanced Telemetry Systems. The collars were fitted to the individuals two weeks prior to release by a veterinary staff from the Fundacion Condor Andino. The individuals were briefly anesthetized, and tissue and blood samples were taken opportunistically and sent for laboratory testing to ensure the monkeys were in good health prior to the release. The collars were programmed to record four locations per individual per day covering the span of the typical active hours for yellow-bellied spider monkeys (i.e. 6 am, 10 am, 2 pm, and 6 pm). IridiumGPS technology allows locational points to be sent automatically to a central server every 24 hours, after which they can be downloaded using the company’s web-browser application at any time. Accordingly, the monkeys’ locations could be followed with a maximum delay of 24 hours. The collars feature an electronic drop-off system which was programmed to disengage in January 2025. To estimate the latent behavioral states of Corny and Kasha throughout the post-release period, GPS fixes were analyzed using non-parametric Bayesian movement models. These calculations were performed using the opensource Bayesmove package in R v4.1.1 (Cullen et al., 2022). The number of discreet latent behavioral states was set to those that could explain 95% of the GPS fixes. Classification of further latent behavioral states beyond this level is typically redundant. A total of 2578 behavioral points were taken of the three monkeys in the pre-release period between May 2023 and January 2024. The time-series of human-attention behavior showed a strong habituation effect. During the early months of observation, human attention increased for several months, before decreasing sharply beginning in October 2023. This habituating effect is common in primate observation studies, and important to ensure that the presence of human observers does not alter natural behaviors once the individuals are released. Another positive behavioral indicator was the increase in upper strata use for all three individuals. A particularly large increase in upper strata use correlated with an increased focus on enrichments such as alternative methods of hanging food items in the enclosure starting in October 2023. Patterns of social development between the individuals throughout the pre-release period were less clear. Social behaviors for Corny and Kasha were relatively consistent throughout the period but increased slightly for Watson. A likely explanation may be that Corny and Kasha had many years of familiarity with each other, whereas Watson is a relatively new introduction to the group. Watson, being a juvenile, would also be expected to initiate more playful social behaviors compared to the other individuals, which was evident in the final months of the pre-release period.
Post-release
A total of 1283 GPS fixes were registered for Corny and 1075 for Kasha during the postrelease observation period from February 4, 2024, to December 31, 2024. Both collars remained attached to the individuals throughout the planned observation period. Given perfect functionality, the GPS collars would have been expected to record 1328 locations for each individual (i.e. four fixes per day). Accordingly, the data represents a completeness of 96.6% for Corny and 80.9% for Kasha. The GPS collar of Kasha performed well until roughly August 2024, when it consistently began acquiring less than the programmed four GPS fixes per day (Figure 4). Despite this drop in performance, the collar still acquired at least one GPS fix on all but 13 days. Horizontal dilution of precision (HDOP; a measure of location accuracy where a lower value indicates better performance) averaged 1.32 for Corny and 1.72 for Kasha indicating excellent locational accuracy for both collars. HDOP for Kasha’s collar indicated a drop in performance beginning in August 2023, however, the average HDOP between August and December 2023 was 2.26, and thus, still of sufficient quality for locations to be classified as highly accurate. Corny’s collar recorded a slight decrease in HDOP performance in June and July corresponding to a period where he was kept for observation in the infirmary building at Amazoonico, and thus, inhibiting the collar’s ability to connect with satellites. The collar then regained optimal performance when Corny was rereleased in late July.
For roughly the first month postrelease, the three individuals occupied the site near Makisapa Alpa, returning to sleep near the enclosure each night. In early March, Kasha and Watson dispersed in a northeastern direction, while Corny remained alone near Makisapa Alpa. One week later, Corny dispersed in a northwestern direction. Kasha and Watson continued to occupy sites near the Arajuno River until late April when Kasha appeared near the perimeter of amaZOOnico. Females within the free-living spider monkey group around amaZOOnico showed signs of aggressive behavior towards Kasha, preventing her from joining the group or entering far into the limits of the center. She was recaptured by amaZOOnico staff on May 2 and translocated back to the Makisapa Alpa enclosure for re-release. Watson was recaptured on the perimeter of amaZOOnico a few days later and re-released at the Makisapa Alpa site with Kasha on May 13. At this time three juvenile spider monkeys were translocated to the Makisapa Alpa enclosure to begin rehabilitation for future release. It was hoped that their presence in the enclosure would cause Kasha and Watson to remain close to the release site. After Corny’s initial dispersal in March he began to move west towards amaZOOnico. On April 24 he came within 100 m of the center’s limits, but avoided entering, and quickly dispersed further to the southwest along the Arajuno River. In early May, GPS data indicated that Corny was approaching agricultural fields and a rural bus stop, a potentially dangerous situation. AmaZOOnico staff alerted the owner of the agricultural fields, and a few days later local residents confirmed his presence in the area. Corny was recaptured at the site on June 19 with the purpose of translocating him back to the Makisapa Alpa site for another release attempt. During the process, however, it was noted that Corny had a deep and badly infected wound in the right scapular region that required medical intervention. Corny was thus translocated to the infirmary building at amaZOOnico where he received treatment until July 24. He was then translocated to the Makisapa Alpa site and re-released a few days later. Kasha and Watson were still present at Makisapa Alpa when Corny was re-released, having occupied the site continuously following their re-release in May. The three individuals occupied the site for roughly one week following Corny’s re-release before Kasha and Watson dispersed to the northeast, leaving Corny alone at the Makisapa Alpa site for a second time. Kasha’s GPS data indicates that since this dispersal event she has occupied a site roughly 800 m to the northeast of Makisapa Alpa. Dispersion of GPS fixes indicates normal activity and suggests that Kasha was alive as of December 31 when the collars were programmed to disengage. Despite attempts to locate Kasha and Watson for visual observation during this period, efforts have failed. Accordingly, it cannot be said definitively that Watson remains alive or has been with Kasha throughout this period, however, there is also no evidence to suggest the contrary. Following the dispersal of the females in early August, Corny remained alone at the Makisapa Alpa site (although having contact with the three juveniles in the enclosure), until early December when he dispersed to the east. According to GPS fixes, Corny approached but did not enter agricultural fields outside the northeastern border of Selva Viva. Corny reached a maximum distance of 5965 m from Makisapa Alpa on December 12, outside the limits of the protected forest, but still within mature forest habitat. After this, he returned within the limits of the eastern section of Selva Viva to a distance of 3950 m from Makisapa Alpa at the time his GPS collar was programmed to disengage.
​
Bayesian movement models estimated that three latent behavioral states account for 95.9% of GPS fixes between February 4 and September 30. Based on average displacement, step length, and turning angle between GPS fixes these behavioral states were classified as 1) intense foraging (low displacement and step length with high turning angles), 2) broad foraging (intermediate displacement and step length with high turning angles), and 3) exploration (high displacement and step length with low turning angles). The amount of time spent in each behavioral state was similar for Corny and Kasha, with Corny exhibiting slightly more exploratory behavior.
Discussion
Pre-release ethograms suggest that enrichment activities can positively reinforce natural behaviors in captively raised and rescued yellow-bellied spider monkeys. Enrichments that presented food on high hanging structures seemed to drive more use of the upper strata of the enclosure. This contrasts with traditional feeding tables from which food often falls to the ground, causing monkeys to spend most of their foraging and consumption time in lower strata. A large mortality risk for released arboreal primates is attack by ground dwelling predators. Selva Viva contains both jaguars and pumas, which present a large risk to primates that spend unnaturally large amounts of time on the ground. Accordingly, the three individuals in this case exhibited good preparedness in their comfort of using high vertical strata. Ethograms did not suggest high pre-release social cohesion amongst the three individuals, and this was evident in the post-release period as the females twice dispersed away from the male, failing to form a cohesive social unit. In the wild, it is necessary for spider monkeys to form social units as this is thought to guard against predation as monkeys consistently warn other members of the group when dangers are present. Accordingly, a split group containing one and two individuals respectively is not ideal as it puts the individuals at higher risk of predation and precludes all reproductive potential unless the individuals eventually locate and are accepted into wild spider monkey troops. Spider monkeys naturally exhibit fission-fusion social dynamics whereby larger groups split into smaller subgroups for foraging and then reunite after a short period. It is possible the group is exhibiting this style of social dynamic, however, with such small group sizes this is not ideal, and the group has shown no signs of reuniting after quite extended periods apart. Iridium GPS technology used to track post-release progress exceeded expectations by recording 88.7% of all possible GPS fixes throughout the monitoring period. Because this technology requires a clear signal to satellites, it is often not recommended for use in closed canopy forest conditions, such as those in the Amazon Rainforest. Our results, however, confirm that this technology can be used to acquire highly accurate locational data on spider monkeys in this habitat. We also demonstrate that this data can be combined with non-parametric Bayesian models to estimate latent behavioral states. To our knowledge, this is the first study to combine iridium-GPS technology with these mathematical models to estimate primate behavior. Visual observation of GPS fix data correlated well with the estimated classifications indicating that these methods can be used to reliably estimate latent behavioral states in primates, an important metric for both ecological studies and conservation purposes. As these methods are used more regularly to monitor behaviors of both wild and released spider monkeys, we will gain a better understanding of how to focus pre-release behavioral conditioning efforts and when post-release intervention becomes necessary to prevent project failure.
​
​Our case study also demonstrates that iridium-GPS technology can be an important tool to prevent human-wildlife conflict in wildlife release programs. Because we were able to monitor the location of the monkeys with high temporal accuracy, we could alert members of the local community when Corny was approaching and likely raiding farms. People will often retaliate against wildlife that they perceive as a threat to their livelihoods, presenting obvious dangers to humans and wildlife alike. By alerting locals before conflict occurred, we were able to enlist their assistance in safely recapturing Corny and were able to attempt a second release. Despite the separation of the group at least two of the individuals, and possibly all three, remain alive at the end of the post-release monitoring period, indicating moderate success. To prevent a similar lack of social cohesion in future release efforts we recommend a stronger focus on enrichment activities that reinforce social connection. For example, constructions that require multiple monkeys to work together to gain access to food. One possible cause of group splitting, although speculative, may be that a group size of three is perceived as too small by the monkeys themselves, and therefore, they are employing an exploratory strategy to split off and seek larger groups to join. Accordingly, it may be necessary to release larger groups to guard against potential splitting in future efforts. Currently, another group of three spider monkeys is undergoing rehabilitation for release at Makisapa Alpa, while two infants that are potential candidates for release remain at amaZOOnico. It would likely be beneficial for the project to release all five individuals together when the infants are of the appropriate age and have undergone sufficient rehabilitation to establish social cohesion with the three older individuals. A final positive aspect of the release attempt is the apparent buffer effect of having a pre-established spider monkey group surrounding the amaZOOnico territory. This group of seven monkeys is considered only semi-wild in that they still receive supplemental feeding at amaZOOnico to prevent conflict with local communities as they are prone to seek humans for food given their captive history. The objective of the release program is to establish fully wild spider monkey troops within Selva Viva, and therefore, it would not be considered optimal if current and future released monkeys simply join the established amaZOOnico group. In the current case, this did not happen. Kasha and Watson were actively pushed away from the group by aggressive females, and Corny approached but did not attempt to enter the territory. This is positive as it increases the chances that groups establish new territories in Selva Viva and confirms that there is a positive conservation purpose for maintaining the semi-wild group with supplementary feedings at amaZOOnico.
​
​
Other releases by Merazonia
