You are here

Captive Western Lowland and Virunga Mountain Gorilla Gestural and Vocal Dialect Comparisons

Lisa Wilding


Due to the tenuous existence of the endangered Virunga Mountain gorilla, studies of their communications compared with that of western lowland gorilla species can prove valuable.  In this research, visual and audio recordings of captive western lowland gorillas across the United States were compared with that of the endangered Virunga Mountain gorillas.  Analysis using Raven, ELAN, and Comparisonics software provides positive evidence that these two species could communicate, after a phylogenetic split occurring approximately 1.6 million years ago.  With socialization a key factor to their existence and no successful mountain gorilla breeding program in captivity, research in this area is helpful to determine if captive western lowland gorillas can, in some capacity, assist in the survival of young mountain gorillas rescued from the bushmeat trade.


Humans are not the only species to have acquired a complex language through evolution.  Humankind's closest relatives, the great apes, have actually learned one of our languages, American Sign Language, yet we fail to fully understand theirs.  Revealing a higher level of cognition, chimpanzees and gorillas can communicate in American Sign Language (Fouts, 1997; Patterson, 1978), and bonobos can communicate in a computer symbolic language called Kanzi (Savage-Rumbaugh & Lewin, 1996).  Communication in great apes is far more sophisticated than originally thought (Genty et al., 2009).  A connection between gesture and context has been found in all great apes (Call & Tomasello, 2007), with many gestures used in more than one context.  In both natural and captive environments, gorillas exchange information between themselves by means of olfactory signals, gesturing (both tactile and visual) and vocalization (Pika et al., 2003), in addition to developing a language variation through cultural learning, the way in which a group within a society passes on new information (Genty et al., 2009).  Just how complex is gorilla language and what advantage do we gain by researching it?

Cultural distinctions in language, called dialects, are the presence of cues in gesturing or vocalization that are cultural differences from the standard language, yet subtle enough to allow accurate communication across the species (Elfenbein et al., 2007).  Languages change over time and species that share the same but somewhat isolated geographic location can develop variations in communication (Elfenbein et al., 2007).  This is most easily illustrated in humans, where differences in words can include grammar, vocabulary, syntax, inflection, and common expressions.  The Southern dialects make up the largest accent group in the United States, with an area extending over 15 states.  There is not one Southern accent, but a number of sub-regional dialects.  In some Southern dialects of English, the "r" is quite commonly dropped after vowels unless following by another vowel, so "far" becomes "fah."  In this case, the actual word and meaning remain the same, just the pronunciation differs.  Furthermore, although at times difficult to understand, communication can transpire between humans living in the northern United States and those living in the southern United States.

Dialects have been seen and recorded in a number of other animal species.  Marine mammals, such as the killer whale (Orcinus orca) and sperm whale (Physeter macrocephlus), display dialects, consistent differences in the vocal repertoires primarily arising from social learning (Weilgart & Whitehead, 1997).  Significant differences in the alarm call of Gunnison's prairie dog (Cynomys gunnisoni) have verified dialect existence (Perla & Slobodchikoff, 2002).  Along with cultural differences in hunting and tool use (Gruber et al., 2009), chimpanzees have shown gesturing dialects in captivity (de Waal, 1989), in calls in the wild (Mitani et al., 1992), and also variations in gestures in the wild (McGrew & Tutin, 1978).  Studies have revealed acoustic variability in non-human primates and differences in vocalizations between interspecific and individual variation (Mitani et al., 1999).  Male chimpanzees living in separate geographic regions produced acoustically distinguishable pant hoot vocalizations (Mitani et al., 1999).  Wild populations of African great apes have been observed to have regional variations in communication and behavior referred to as cultural distinctions (Whiten et al., 1999).  Evidence of cultural differences in wild western lowland gorillas can be seen in two populations, Lop and Belinga, located approximately 250 km apart in Gabon (Tutin & Fernandez, 1992).  Although small, fungus-farming termites and weaver ants are common to both areas, the Lop population of western lowland gorillas will only eat the termites and the Belinga population will eat only the weaver ants.

  It is known that gesturing dialects do occur in captive lowland gorillas (Tanner & Byrne, 1999), most likely a result of the transfer of gorillas between zoos, as these transfers are directed by the Gorilla Species Survival Plan (SSP).  Formed in 1988, the Gorilla SSP ensures that genetic representation is spread throughout the population of captive gorillas and promotes the health and social well-being of captive troops.  Variations in communication learned in one geographic location with a particular social group will be viewed as a dialect in a new location.  As these distinct differences in communication are seen in captive gorillas, it attests to the adaptability of gorilla troops to accept new dialects that may veer from their ancestral roots.

Gorillas are endemic to Africa and they are commonly allocated to two species, western and eastern gorillas.  The western species is composed of the Cross River gorilla sub-species, Gorilla gorilla diehli, a population of 200 found on the Nigerian-Cameroon border, and the western lowland gorilla sub-species, Gorilla gorilla gorilla.  Occupying Africa's densest and most remote rainforests in Equatorial Guinea, Gabon, Angola, Cameroon, Central African Republic, and the Republic of Congo (Robbins, 2007), their population is difficult to estimate.  In 2007, researchers of the Wildlife Conservation Society discovered a large population of western lowland gorillas living in a vast area of swamp forest in northern Republic of Congo (Rainey et al., 2010).  It is the western lowland gorillas who are typically found in zoos across the world, as they have been able to adapt and reproduce in captivity.  Both western sub-species are listed as critically endangered (IUCN, 2010).  The eastern species is composed of the Grauer or eastern lowland gorilla sub-species, Gorilla beringei graueri, a population of 170 living in the eastern part of the Democratic Republic of Congo, and the mountain gorilla sub-species, Gorilla beringei beringei, a population of 680, occupying the volcanic Virunga Mountains at the intersection of Rwanda, Uganda, and the Democratic Republic of the Congo and the Bwindi Impenetrable National Park in Uganda (Robbins, 2007).  Both eastern sub-species are also listed as endangered  (IUCN, 2010).  This thesis focuses on the western lowland gorilla of the western species and the Virunga Mountain gorilla of the eastern species.

Gorillas are predominately vegetarian, although they do consume small amounts of ants and termites (Robbins, 2007). The Virunga Mountains, often cloudy, misty, and cold, is the home of Virunga Mountain gorillas, where they live at elevations between 8,500 and 13,000 feet (Fossey, 1972).  Due to this altitude, food resources are limited to celery, nettles, bamboo, and thistle; whereas, the western lowland gorillas, occupying lower altitudes, have access to a much wider variety of food items including a variety of fruits (Robbins, 2007).

Gorillas live in stable, cohesive social groups called troops led by the adult, dominant silverback male (> 12 years), who provides protection, mediates disputes, determines the home range, has exclusive breeding rights to the females, and basically regulates what time they wake up, eat, and go to sleep (Dian Fossey Gorilla Fund International, 2010b).  Other members of the troop can be several young males called blackbacks (8-12 years), adult females (> 8 years), subadults (6-8 years), juveniles (3.5-6 years), and infants (0-3.5 years) (Robbins, 2007).  The social structure is clearly defined in gorillas and seems to be an important factor in the stability of the troop.

A large body and a large brain require a long maturation and the average age for females to give birth is 10 years (gestation of 8.5 months), with an interbirth interval of 4 years (Robbins, 2007).  Newborn gorillas are completely dependent on their mothers for the first several years and thus there is a great maternal investment in rearing offspring (Robbins, 2007).  Infants are mainly cared for by their mother and protected by the dominant silverback (Robbins, 2007), but a troop of captive lowland gorillas seem to share responsibility in the infants' well-being, possibly providing an opportunity for other females to learn to mother.  Gorillas can live approximately 35 years in the wild and up to 54 years in captivity; however, commercial hunting, deforestation, and Ebola virus induced mortality jeopardize the life span for those living in the wild (Dian Fossey Gorilla Fund International, 2010b; Rainey et al., 2010).

DNA sequencing reveals that the human genome shares more than 98% homology with gorillas and chimpanzees (Toder et al., 2001), but despite being close relatives of humans, the Virunga Mountain gorillas are one of the most endangered species in the world.  Virunga National Park, home to nearly 200 mountain gorillas, has become a battleground for militia groups and the Congolese army.  Poachers hunt park gorillas for meat and sale, and forests are harvested for charcoal.   In 2007, a family of at least six gorillas was murdered "execution style" for political reasons attributed to charcoal production (Jenkins, 2008).  With the long maturation and interbirth interval, these losses are not sustainable for this population.  Survival rates of rescued gorilla infants, confiscated as orphans of the bushmeat "industry," are extremely low in captivity.  Currently, MGVP, Inc. (Mountain Gorilla Veterinary Project, 2010) is reporting only four surviving confiscated mountain gorilla orphans.  Unable to live in captivity within sanctuaries in the hope of being released to the wild, the future survival of mountain gorillas is uncertain.

Until recently, knowledge on the social organization and behavior of gorillas has been dominated by results focused on mountain gorillas, due largely to the work of Dr. George Schaller and Dr. Dian Fossey (Dian Fossey Gorilla Fund International, 2010a; Harcourt et al., 1993; Wallace, 2007).  Schaller is recognized by many as the world's preeminent field biologist, studying wildlife since the late 1950s throughout Africa, Asia, and South America.  Fossey was hand selected by paleontologist Louis Leakey to study mountain gorillas in the 1960s and spent close to two decades habituating, studying, and conserving the endangered mountain gorillas (Fossey, 2000).  During her research, she was able to identify and analyze sixteen distinct mountain gorilla vocalizations (Fossey, 1972). 

The gesturing ethograms compiled can vary greatly depending upon the finiteness of the observational study, the age group selected, the context and dynamics of the troop, and what particular behavior was focused on, such as sexual behaviors, mother-infant relationships, etc.  While gorilla gesturing has been studied, it is complicated and not fully understood (Kalan & Rainey, 2009), and very little research has been conducted on gorilla vocalizations.  To the best of this writer's knowledge, there has been no previous research done on gesturing and vocalization dialects together.

Genetics has revealed that western lowland gorillas and mountain gorillas physically split somewhere between 0.9 and 1.6 million years ago (Thalmann et al., 2007) and their present day ranges are separated by 1,000 km or approximately 621 miles (Thalmann et al., 2005).  It is quite likely that variations in communication or language would have developed between these two species over this time period.  What if these two species were brought together, could they communicate?  Could captive western lowland gorillas in sanctuaries thus provide a less stressful environment crucial to the survival of rescued mountain gorilla orphans?  By cataloguing current gestural and vocal dialects in captive western lowland gorillas and comparing them with Virunga Mountain gorilla communication, justification may be established that Virunga Mountain gorilla gesturing and vocalization should be considered yet another dialect, increasing the probabilities that these two sub-species, western lowland gorillas and mountain gorillas, could in fact communicate with each other. 



The James R. Davis Studbook (Gorilla Haven Org., 2010) was referenced to determine which zoos in the United States house gorillas.  Fifty-two zoos and/or sanctuaries in the United States currently have captive lowland gorillas (Gorilla Haven Org., 2010).  Limited to travel during semester breaks and the summer months, a subset of these zoos was considered for data collection.

Contacts were established through each zoo's webpage either by phone or by e-mail to gain approval for this observational study.  Each zoo had its own criteria for approving such a study.  Oftentimes, zoos have a select committee that meets to review such proposals.  While some required a detailed e-mail describing the purposes of this research and how it would be conducted, others required a written proposal, the zoo's application form, a curriculum vitae and/or a letter of support from Professor Covert, advising professor of this research. 

After any further questions regarding the research were answered, the zoo's decision to support or not support this research was received within a day or two up to a couple months.  When approvals came in, dates had to be negotiated to determine when would be the best time to collect data.  Gorilla transfers, enclosure construction, the health of gorillas, available zoo personnel, coordination of times with other zoos in the vicinity, and weather were considerations in making this decision.

This observational study required video equipment that was self-contained, as power sources were oftentimes not available, and enough memory to be able to hold the large video files that were created throughout each day.  The SANYO VPC-HD1 video recorder (SANYO North America Corporation, San Diego, CA) utilized in this study was able to film in poor lighting, through glass without glare, pick up fine detail on dark subjects, allow 10x digital zoom capabilities, six image quality settings, and could operate solely on small SANYO DB-L40 rechargeable batteries. This camcorder was limited to 2 GB SD memory cards, such that transferring data after the SD reached capacity onto a laptop was required periodically throughout each day.  Tv-SHQ video quality setting was selected, as it provided the highest quality film while still maximizing the memory usage.  Ten battery packs would generally be used in an 8-hour period of filming.  If more zooming was required, the batteries would drain faster.  A sturdy tripod with a joystick was used to provide more stable filming, even though this video recorder has image stabilization features.  A laptop was taken onsite to store data as it was collected after video SDs became full.  After successful recovery of the data, SDs were formatted in the camcorder and recording resumed.

The Olympus LS-10 linear PCM audio recorder (Olympus, America Corporate Headquarters, Center Valley, PA) was chosen based on the very positive recommendation by Dr. Bernie Krause, a professional bioacousitician, who records wildlife throughout the world, including the Virunga Mountain gorillas.  This audio recorder can be left unmanned running on two AA batteries or off a power source.  This was often placed on a tripod or on a shelf in a secluded area away from crowd noise, such as in the gorilla keepers' area, on the roof, or in an area accessed only by zoo personnel and non-intrusive to the gorillas.  As with most audio recorders, it was not possible to accurately record through glass, so careful placement had to be negotiated with the appropriate zoo personnel.  If left outside, windscreens were attached to each microphone to prevent wind distortion and the tripod was often secured in some way to avoid the pitfalls of falling into a dry moat or off the roof.  Due to a limit of 2 GB with built-in flash memory, data was always recorded to the secondary SD memory card.  Since this device was often left unattended, the record level was set to adjust automatically to ensure the least amount of distortion.  The record mode of MP3 256 kbps was selected, as this was a format that was recommended by Cornell University for their Raven software, an application for the acquisition, visualization, measurement, and analysis of acoustic signals, widely used in various research in animal acoustic communication (Rautio et al., 2009; Slocombe et al., 2010).  MP3 ranges are best suited for gorilla vocalizations, as ultrasonic ranges, frequencies greater than 20,000 Hz, are not needed.

Collecting Data

When developing a multi-institutional observation recording program, it is inevitable that numerous variables, including weather and facility layout will influence the recording somewhat.  Geissmann (2003) offers a thorough explanation of how to approach recording primate vocalizations and recommendations to be considered before beginning.  A minimum of three days was spent at each zoo.  Data sheets on each gorilla, including date of birth, gender, parents, and offspring, were printed from the James R. Davis Studbook (Gorilla Haven Org., 2010) and ethogram sheets were brought to record specific events.  Zoo personnel were contacted the first day to gain access to the zoo, identify gorillas, discuss personalities, and provide options for placement of equipment.  Audio was allowed to run throughout the night in some zoos to capture vocalizations in their off-exhibit enclosures.  Power generally needed to be provided in these cases, as the batteries could not sustain running more than eight hours.

With the exception of changing batteries and saving data off SD memory cards, continuous filming occurred from early morning until the gorillas were taken off exhibit and/or the zoo closed for the day.  While filming, focus was given to any interaction between gorillas, younger gorillas, and subjects who appeared to gesture and/or vocalize to oneself.  Juveniles seemed to interact more with others in the troop, so there was an emphasis to film them.  If a gesture and/or vocalization occurred, the time was recorded, along with the subjects involved and the circumstances surrounding the event onto the ethogram sheet.  These sheets could then be used to cross-reference during playback and analysis for accuracy as to the particular circumstances surrounding the interaction and/or vocalization.

At night, all data on SD memory cards and the laptop were backed up to an external drive to prevent data loss.  Also during the evening, a DB-L40 battery charger was used to sequentially recharge the video recorder battery packs that were used during the day, each charge requiring up to 1-1/2 hours.  In the morning, two fresh AA batteries were inserted and a formatted 16 GB SD memory card was installed in the audio recorder.

Analyzing Data

With the multitude of video and audio files that accumulated, it was important to devise a naming convention.  Each video file was named according to the zoo, the date, and the sequential number of that particular day.  For instance, the third video file recorded at Woodland Park Zoo on May 28th would be named "WPZ_28MAY_3.mp3."  As audio files were quite large and configuration requirements within the Raven software limited the reading of each file to under 700 MB, many audio files needed to be split.  Each audio file was named according to the zoo, the date, the sequential number of that particular day, and the split number.  For instance, the third audio file recorded at Woodland Park Zoo on May 28th and the second split would be named "wpz-28may-3-2.mp3" (the lowercase lettering indicating that the file was split).

Detailed instructions are included here to document exactly how audio files were split, as time was invested to find an approach that was reliable and cost effective.  Audio files were split using QuickTime Pro.  By opening the MP3 file in QuickTime Pro and waiting until the entire file was read in, moving the right bar at the bottom to the location of the split, Edit...Copy, File->New Player, Edit...Paste, File->Save As... the first half of the segmented audio file was then saved to a .mov file.  To convert it back to .mp3 so that Raven software can analyze it, iTunes was utilized.  Through iTunes, choose File->Add to Library, select the file and add to the library, highlight the file and Advanced->Create MP3 Version and the .mp3 file would be created.  The above was repeated for each subsequent split until the end of the file was reached.  Due to undocumented restrictions in iTunes, files needed to be less than 4 hours long to convert.

With data now in workable formats, ELAN (EUDICO Linguistic Annotator) software created complex annotations on the video and audio resources describing features observed in the media.  For each zoo, eighteen tiers were created for each gorilla (individuals up to 3 years old were referred to as infants) incorporating a range of broad gesturing and vocalization categories obtained from an extensive compilated ethogram.  This compilated ethogram included over 45 gorilla ethograms provided by The Gorilla Behavior Advisory Group of the Gorilla SSP (Gorilla Behavior Advisory Group, 1991), which also incorporates Dr. Dian Fossey's behavioral observation of the mountain gorillas.  Tiers could then be hierarchically interconnected.  Video files were viewed and for each group of frames, detailed annotations were noted using the tiers to indicate which gorillas were involved and what broad category of gesturing or vocalization was observed.  When all video files were cataloged in this way, multi-tier regression searches were done to determine unique communications.

Figure 1. ELAN (EUDICO Linguistic Annotator) software screen example showing an annotated video taken at the San Francisco Zoo.  The file name indicates that the video was filmed at the San Francisco Zoo on the second day of filming and it is the first video in a sequence taken that day.  In the center, is a bar running the width of the screen with small hash marks on it.  Each hash mark represents an annotation in this file (this particular file has 243 individual annotation entries).  On the left-hand side is a partial list of the tiers that were created for each gorilla in a range of behaviors.  The annotations are noted to the left of that, using the wording from the large compilation of ethograms.


Audio files of captive western lowland gorillas were converted from MP3 to WAV files to be read first by ELAN software, which is capable of importing video and audio and can display waveforms.  The Virunga Mountain gorilla vocalizations used in this study were recorded by Dr. Krause at Dr. Dian Fossey's Karisoke research site in 1987.  ELAN provides a fast scan and screen refresh capability that was useful when analyzing files that were up to four hours long.  Annotations were made in ELAN and the correlating waveform was found in the Raven software program.  It is in Raven that waveforms, including a marginal edge before and after a vocalization, were copied for input into the Comparisonics software.

Figure 2.  Raven Pro (Version 1.4) Interactive Sound Analysis software screen example showing several waveforms of gorilla vocalizations.  The filename indicates that the audio was recorded at Cheyenne Mountain Zoo on November 23rd.  Recorded in stereo, the upper screen displays the left channel and the lower screen displays the right channel.  The vertical height of the waveform indicates its amplitude and the timeline below can be used to calculate the length of each vocalization and indicates the number of hours, minutes, and seconds into the recording that these vocalizations occurred.

At a basic level, two aspects of any sound are loudness and pitch (Giancoli, 2005).  Each refers to a sensation in the consciousness of the listener and each has a physical measurable quantity (Giancoli, 2005).  The loudness or intensity of a sound is related to the amplitude squared of the wave (Giancoli, 2005).  The pitch of a sound, whether it is high like a piccolo or low like a bass drum, is determined by the frequency; the higher the frequency, the higher the pitch (Giancoli, 2005).  The audible range of frequencies for humans is roughly 20 Hz to 20,000 Hz (Giancoli, 2005).

ELAN and Raven can display waveforms, which is a common visual representation of audio showing amplitude (loudness) over time.  It indicates when the audio is loud or soft, but does not provide any information about how the audio sounds.  Comparisonics is an innovative program that utilizes both amplitude and frequency (pitch), and displays the waveform as other programs, but colorizes it to display frequency (20 Hz - 20 kHz) as well (Rice, 2005).  Shades of red are used for high-pitched sounds, greens and blues for mid-range sounds, and dark colors represent deep bass sounds (Rice, 2005).  Similar sounds are represented by similar colors and changes in sound are represented by changes in color (Rice, 2005). 

Using both Raven and Comparisonics applications, isolated waveforms of western lowland gorillas and Virunga Mountain gorillas were compared optically side-by-side to denote similarities and/or differences in basic waveform shape and color, which indicates frequency.  Measurements, using Raven software, were made for minimum amplitude, maximum amplitude, duration, energy, and frequency in comparing captive western lowland gorilla and the Virunga Mountain gorilla vocalizations.  Even though amplitude could not be controlled in this study, as subjects were either close or far away, it can enhance accuracy of the other measurements considered here.

Figure 3.  Comparisonics Audio Player (Version 2.1) screen example showing Raven software waveforms converted to display the same physical shaped waveform in amplitude (loudness), but also colorized to indicate frequency (pitch).  The filename indicates that this is vocalization from the Virunga Mountain gorillas.  The colors illustrate that the vocalization is higher-pitched and the extent of the height denotes loudness.


Fifty-one gorillas were observed in nine different zoos during this study.  Eighteen main categories of gesturing and vocalization repertoire were itemized for each gorilla, totaling 918 different tiers.  Through ELAN, 4,861 gestures and/or body movements were observed and annotated within sixteen hours and 31 minutes of video in total, which is approximately one-third of the video data actually recorded  (reference Table 1 below),.  Depending on the size of the troop filmed, 40 minutes of video took up to 6 hours to annotate.  The audio was captured separately and 416 distinct vocalizations (again, one-third of the data) were analyzed between the zoos and 133 distinct vocalizations were analyzed of the Virunga Mountain gorillas.

Gestural Communication

Chart 1.  This pie chart represents collectively the 45 combined behavioral ethograms provided by The Gorilla Behavior Advisory Group of the Gorilla SSP (Gorilla Behavior Advisory Group, 1991), the more-detailed gestures and/or body movements that have most likely been seen and documented in other ethograms, and the unique gestures and/or body movements observed in this study.

Table 1.  This is a sample segment of the large compilated ethogram incorporating the 45 smaller ethograms provided by The Gorilla Behavior Advisory Group of the Gorilla SSP (Gorilla Behavior Advisory Group, 1991) and the unique gestures recorded in this study.  Movements, gestures, and/or body positions with regard to "infants" (those still nursing), grooming, objects, mother-infant interaction, passive interaction with other troop members, and stationary positions.  Highlighted cells matching the color of the category are movements, gestures, and/or body positions that most likely have been noted in ethograms not included in this study.  Highlighted cells in yellow indicate unique movements, gestures, and/or body positions observed and not included in the large compilation of ethograms.  Numbers in parenthesis indicate the number of times this behavior was observed in this study.



Color and waveform shape was noted and measurements of duration were made on all vocalizations.  "Chestbeats," even though technically not a vocalization, were considered as both a gesture and as audible communication.  On multiple recordings of the same vocalization obtained for both the captive western lowland gorilla and the Virunga Mountain gorilla, namely "Chewbacca," "Chestbeat," "Lion," "Roar," "Grunt," "Growl," and "Growl that goes up and then down in pitch," a two-tailed t-test assuming unequal variances was run on the duration of the audio.  All vocalizations, except for the "Growl" with a p-value of 0.0479 (reference Chart 2 below), showed that the sounds were not significantly different in duration.

Chart 2.  Comparison of vocalization duration in seconds between 6 captive western lowland gorilla and Virunga Mountain gorilla "Growl" sounds.  The mean length for captive western lowland gorillas (Mean = 2.73 secs, Standard Error = 0.635) was significantly different (P(T<=) two-tail = 0.0479) than the mean length for Virunga Mountain gorillas (Mean = 1.04033 secs, Standard Error = 0.133) using the two-sample t-test for unequal variances. 


The purpose of this study is not to analyze the meaning behind the gestures, but to capture the visual and audio cues that are occurring between gorillas (or solitary communication) and increase the current ethograms to contain these unique gestures or dialects, both in gesturing and vocalization, to allow a determination of whether or not communication is possible between western lowland gorillas and Virunga Mountain gorillas.  All gestures and/or body movements or positions were included in this study so as not to make judgments on what is considered communication from that which is not.  The vocalizations quite often are expressed in a combination of vocalization sequences; however, within this research, audio sounds were purposely isolated and not considered in combination.

All video gestures and/or body movements were carefully analyzed and taken in context to events that happened before, during, and after the observation.  This is important, as many gestures were used in more than one context.  For example, the simple act of carrying vegetation, such as a branch, in the mouth can have an entirely different meaning if the avenue of transportation involved climbing a tree as compared to a blackback male carrying it as part of a display to another male.  Chestbeats can be used in a variety of different situations, including play and agnostic and solitary displays.  A yawn might indicate tiredness or might be a way for a silverback to display teeth size.  Context is important, as with many animals including humans, and must be considered when documenting gestures and/or actions.

The increase of the current ethogram by eleven percent, using less than one-third of the data collected, is compelling positive evidence that slight changes in communication are accepted in our captive western lowland gorilla population and shows a degree of flexibility, perhaps contributing to their success in adapting to captivity (reference Chart 1 above).  Distinctive dialects in this study were seen across the hierarchy of the troop, from the silverback to a subordinate or infant.  All members of the social group were included as part of the troop dynamics, even if their communication was slightly distinct.

Picture 1.  San Diego Zoo (28 Dec 2009) - "Ekuba" - male, born 2006 in San Diego Wild Animal Park, parent reared.  Gesture noted:  Clapping rapidly while standing bipedally.  Context of situation:  This gesture was only used by Ekuba and was usually directed at "Bouendje," his 3-year-old half brother, or Mandaazi, his 7-year-old brother.  Oftentimes, the clapping occurred while hunched over.  This immediately preceded Ekuba being chased by whomever the gesture was directed towards.


Creativity, along with flexibility, has been seen in gorillas' use of communication.  Koko, a western lowland gorilla living at the Gorilla Foundation, has been taught American Sign Language (Patterson & Linden, 1981).  When unable to communicate in signs that she knows, she invents words by using novel word combinations.  For instance, when shown a ring, she describes it as "finger" "bracelet," inventing a new sign from two signs she knows (Patterson & Linden, 1981), actually creating a dialect.  Bouendje and Ekuba, two 3-year-old males, from the San Diego Zoo were born only nine days apart.  Having different mothers, but the same father, they have grown up together.  Ekuba demonstrates a distinct gesture of clapping rapidly in a bipedal position directed toward another member of the troop (reference Picture 1 above).  This usually is preceded by a game of chase with Ekuba being chased.  Even though they have been together since birth, Bouendje does not use this gesture, but he definitely understands its intent.  Ekuba may have created this unique gesture and was able to obtain the desired results, thereby keeping it in his gesturing repertoire. 

Vocalizations were categorized into seven common audio articulations between both the western lowland gorillas and the Virunga Mountain gorillas.  Measurements of amplitude, energy, and frequency were noted, however these values are extremely variable and are strongly related to the subject's distance from the microphone, the environment the sound was recorded in, and the equipment used.  These measurements were deemed to not be conducive to comparing two species in entirely different habitats, as it would not be conclusive either way. The following colored waveforms (Images 1-2 below) depict one example of the recorded vocal repertoire of both the captive western lowland gorillas and the Virunga Mountain gorillas, with a side-by-side comparison of like vocalizations.  Not all known gorilla vocalizations are represented here, as some vocalizations did not occur during recording.  Although known gorilla vocalizations have commonly recognized names, some original names were used in this study to avoid the pitfall of mislabeling.

The vocalization classified as "Roar" is characterized by a sudden loud outburst of low-pitched sound.  Images 1 and 2 below are good depictions of the waveform, which is similar in both sub-species.  There is a buildup at the beginning and at the end.  The duration of the audible is not significantly different between both sub-species.

Image 1.  Sedgwick County Zoo (24 June 2010).  Vocalization:  Roar.


Image 2.  Virunga Mountain gorillas (1987).  Vocalization:  Roar.

Waveform shape was similar in all comparisons with distinct recognizable patterns.  Overall, the Virunga Mountain gorilla vocalizations were notably higher in pitch (lighter in color) compared to the captive western lowland gorillas.  Resonation of bass off zoo enclosure walls, even when recording outside, could account for the lower pitch of the lowland gorillas.   Consideration also must be made to individual body size and location of recording, such as heavily forested habitats, as both favor the production of low frequency calls (Mitani et al., 1999).  Recording equipment, ecological factors of the habitat acoustics, and morphological differences in body size can account for differences in sound recording.

Surrogate gorilla mothers have been used successfully in western lowland captive populations (Jendry, 1996).  First-time mothers, who may have been hand reared by humans, are often lacking the knowledge of rearing an infant.  Zoos today realize the importance of having young raised by their own species and seek surrogates for these young gorillas so they can learn social behavior from gorillas within a group environment (Jendry, 1996).  San Francisco Zoo's Hasani is a good example of success in this area.  After cleaning Hasani immediately after birth, his biological mother, Monifa, showed no interest in raising her newborn baby and abandoned him.  Zoo personnel hand reared him off-exhibit for several months, providing the nourishment he required, while introducing him to another female member of the troop, Bawang.  Bawang is now rearing Hasani and he has successfully joined the troop.  As with most western lowland gorilla troops, all females share in his well-being, including Monifa, and he has adjusted well to this troop and his new adoptive mother.

Cross-fostering within species is nothing new.  Young non-human orphans like Hasani, rejected by parents, can be adopted by adults of their social group (Blersch & Schmidt, 1992; Ellsworth & Andersen, 1997; Hamilton et al., 1982; Thierry & Anderson, 1986).  Unrelated females at the California Primate Research Center have routinely reared rhesus macaque (Macaca mulatta) infants; however, in 2005, there was a 67% success rate at cross-fostering between species, namely Japanese (Macaca fuscata) and rhesus (Macaca mulatta) macaques (Owren & Dieter, 1989).  Guerra et al. (1998) reported that cross-fostering in captivity between white tufted-ear (Callithrix jacchus) and black tufted-ear (Callithrix penicillata) marmosets was successful and advantageous over hand-rearing, as infants are not deprived of social interaction and receive comparable stimulations.  Developing techniques in cross-fostering can provide viable means, in times of need, of raising primate infants with members of a different species.

Virunga Mountain gorilla orphans witness their families being murdered before they are captured, and oftentimes are not fed or watered during transport through the animal trade.  After being rescued, orphans can be severely dehydrated and considerably stressed.  The Gorilla Foundation's gorilla, Michael, was able to give insight to what is experienced by gorilla orphans.  Michael came to the foundation in 1976 at the age of 3 from Cameroon, Africa, as a gorilla orphan.  He was taught American Sign Language, acquiring in excess of 600 signs (Patterson, 2010a).  When asked about his mother after he was an adult, he relayed a chilling account of how he witnessed the murder of his family and how they were killed (Patterson, 2010b).  He retold this story several times, unveiling the stress and significance of losing family members.  The Virunga Mountain gorilla orphans are experiencing the same horror.  They have just lost family members to humans, and if rescued are taken to sanctuaries where humans desperately work to save them.  At that age, orphans may only associate humans with the loss of family members, which may further exasperate their stressed condition.

The political unrest between militia groups, the Congolese army, and poachers is not likely to be solved in time to prevent extinction of these magnificent creatures.  While this study offers an explanatory analysis of gesture dialects and vocalization comparisons, it is a precursor to a more strategic study in gorilla language.  Both gestural and vocal sequencing, combinations of more than one cue, need to be researched in more depth to better understand the complexity in which gorillas communicate.  It is not this writer's intent to claim there is only one solution to this dilemma, but to illustrate that western lowland and Virunga Mountain gorillas can, most likely, communicate, making attempts to cross-foster or raise Virunga Mountain bushmeat orphans feasible, and to suggest pathways to rethink how we can prevent the further decimation of Virunga Mountain gorillas.


Blersch, B.H. & Schmidt, C.R. (1992). Adoption of an additional infant by a western lowland gorilla (Gorilla gorilla gorilla). Folia Primatology, 58(4), 190-196.

Call, J., & Tomasello, M. (2007). The gestural communication of apes and monkeys. Hillsdale: Lawrence Erlbaum Associates.

de Waal, F.B.M. (1989). Behavioral contrasts between bonobo and chimpanzee. In P.G. Heltne & L.A. Marquardt (Eds.), Understanding chimpanzees, (pp. 154-175). Cambridge, MA: Harvard University Press.

Dian Fossey Gorilla Fund International (2010a). Dian Fossey biography [Online]. Available:

Dian Fossey Gorilla Fund International (2010b). Learning about gorillas [Online]. Available:

Elfenbein, H.A., Beaupre, M., Levesque, M., & Hess, U. (2007). Toward a dialect theory: Cultural differences in the expression and recognition of posed facial expressions. Emotion, 7(1), 1311-146.

Ellsworth, J.A. & Andersen, C. (1997). Adoption by captive parturient rhesus macaques: Biological vs. adopted infants and the cost of being a "twin" and rearing "twins." American Journal of Primatology, 43(3), 259-264.

Fossey, D. (1972). Vocalizations of the mountain gorilla (Gorilla gorilla beringei). Animal Behaviour, 20(1), 36-53.

Fossey, D. (2000). Gorillas in the mist. New York: First Mariner Books.

Fouts, R. (1997). Next of Kin: My conversations with chimpanzees. New York: Avon Books, Inc.

Geissmann, T. (2003). Tape-recording primate vocalisations.  In J.M. Setchel & D.J. Curtis (Eds.), Field and laboratory methods for primatologists: A practical guide, (pp. 228-238). Cambridge:  Cambridge University Press.

Genty, E., Breuer, T., Hobaiter, C., & Byrne, R.W. (2009). Gestural communication of the gorilla (gorilla gorilla): Repertoire, intentionality and possible origins. Animal Cognition, 12, 527-546.

Giancoli, D.C. (2005). Physics: Principles with applications. Upper Saddle River, NJ: Pearson Education, Inc.

Gorilla Behavior Advisory Group (1991). Collection of Gorilla Ethograms [Online]. Available:

Gorilla Haven Org. (2010, September). Captive gorilla stud book [Online]. Available:

Gruber, T., Muller, M.N., Stimling, P., Wrangham, R., and Zuberbuhler, K. (2009). Wild chimpanzees rely on cultural knowledge to solve an experimental honey acquisition task. Current Biology, 19, 1806-1810.

Guerra, R.F., Takase, E., & Santos, C.V. (1998). Cross-fostering between two species of marmosets (Callithrix jacchus and Callithrix penicillata). Revista Brasileira de Biologia, 58(3), 665-669.

Hamilton, W.J., III, Busse, C., & Smith, K.S. (1982). Adoption of infant orphan chacma baboons. Animal Behaviour, 30, 29-34.

Harcourt, A.H., Stewart, K.J., & Hauser, M. (1993). Functions of wild gorilla 'close' calls. 1. repertoire, context, and interspecific comparison. Behavior, 124, 89-122.

IUCN (2010). IUCN Red List of Threatened Species. Version 2010.3. Available:

Jendry, C. (1996). Utilization of surrogates to integrate hand-reared infant gorillas into an age/sex diversified group of conspecifics. Applied Animal Behaviour Science, 48(3-4), 173-186.

Jenkins, M. (2008, July). Who murdered the Virunga gorillas? National Geographic.

Kalan, A. K. & Rainey, H. J. (2009). Hand-clapping as a communicative gesture by wild female swamp gorillas. Primates, 50(3), 273-275.

McGrew, W.C. & Tutin, C.E.G. (1978). Evidence for social custom in wild chimpanzees. Man, 13, 234-251.

Mitani, J.C., Hasegawa, T., Gros-Louis, J., Marler, P., & Byrne, R. (1992). Dialects in wild chimpanzees? American Journal of Primatology, 27, 233-243.

Mitani, J.C., Hunley, K.L., & Murdoch, M.E. (1999). Geographic variation in the calls of wild chimpanzees: A reassessment. American Journal of Primatology, 47, 133-151.

Mountain Gorilla Veterinary Project (2010). Providing medical care for confiscated primates (orphans) [Online]. Available:

Owren, M.J. & Dieter, J.A. (1989). Infant cross-fostering between Japanese (Macaca fuscata) and rhesus macaques (M. mulatta). American Journal of Primatology, 18, 245-250.

Patterson, F.G. (1978). The gestures of a gorilla: Language acquisition in another pongid.  Brain and Language, 5, 72-97.

Patterson, F. (2010a). Michael's Bio [Online]. Available:

Patterson, F. (2010b). Michael's Story [Online]. Available:

Patterson, F. and Linden, E. (1981). The education of Koko. New York: Holt, Rinehart and Winston.

Perla, B.S., & Slobodchikoff, C.N. (2002). Habitat structure and alarm call dialects in Gunnison's prairie dog (Cynomys gunnisoni). Behavioral Ecology, 13(6), 844-850.

Pika, S., Liebal, K., & Tomasello, M. (2003). Gestural communication in young gorillas (Gorilla gorilla): Gestural repertoire, learning, and use. American Journal of Primatology, 60, 95-111.

Rainey, H.J., Iyenguet, F.C., Malanda, G.-A.F., Madzoke, B., Dos Santos, D., Stokes, E.J., Maisels, F., & Strindberg, S. (2010). Survey of Raphia swamp forest, Republic of Congo, indicates high densities of critically endangered western lowland gorillas Gorilla gorilla gorilla. Oryx, 44, 124-132.

Rautio A., Niemi, M., Kunnasranta, M., Holopainen, I.J., Hyvarinen, H. (2009). Vocal repertoire of the Saimaa ringed seal (Phoca hispida saimensis) during the breeding season. Marine Mammal Science, 25(4), 920-930.

Rice, S.V. (2005, October). Frequency-based coloring of the waveform display to facilitate audio editing and retrieval. Paper presented at the 119th Audio Engineering Society Convention, New York, New York, USA.

Robbins, M. M. (2007). Gorillas. In C. J. Campbell, et. al. (Eds.), Primates in perspective (pp. 305-321). New York: Oxford University Press.

Savage-Rumbaugh, E.S., & Lewin, R. (1996). Kanzi: The ape at the brink of the human mind. New York: John Wiley & Sons, Inc.

Slocombe, K.E., Kaller, T., Call, J., & Kuberbuhler, K. (2010). Chimpanzees extract social information form agonistic screams. PLoS ONE 5(7), e11473.

Tanner, J. E., & Byrne, R. W. (1999). The development of spontaneous gestural communication in a group of zoo-living lowland gorillas. In S. T. Parker, et. al. (Eds.), The mentalities of gorillas and orangutans: Comparative perspectives (pp. 211-239). Cambridge, UK: Cambridge University Press.

Thalmann, O. H., Fischer, A. H., Lankester, F. H., Paabo, S. H., & Vigilant, L. H. (2007). The complex evolutionary history of gorillas: Insights from genomic data. Molecular Biology and Evolution, 24(1), 146-158.

Thalmann, O. H., Serre, D., Hofreiter, M., Lukas, D., Eriksson, J., & Vigilant, L. H. (2005). Nuclear insertions help and hinder inference of the evolutionary history of gorilla mtDNA. Molecular Ecology, 14, 179-188.

Thierry, B. & Anderson, J.R. (1986). Adoption in anthropoid primates. International Journal of Primatology, 7(2), 191-216.

Toder, R., Grutzner, F., Haaf, T., & Bausch, E. (2001). Species-specific evolution of repeated DNA sequences in great apes. Chromosome Research, 9(6), 431-435.

Tutin, C.E.G, & Fernandez, M. (1992). Insect-eating by sympatric lowland gorillas (Gorilla g. gorilla) and chimpanzees (Pan t. troglodytes) in the Lope Reserve, Gabon. American Journal of Primatology, 28, 29-40.

Wallace, S. (2007, October). Lifetime achievement: Biologist George Schaller. National Geographic.

Weilgart, L., & Whitehead, H. (1997). Group-specific dialects and geographical variation in coda repertoire in South Pacific sperm whales. Behavioral Ecology and Sociobiology, 40(5), 277-285.

Whiten, A., Goodall, J., McGrew, W.C., Nishida, T., Reynolds, V., Sugiyama, Y., Tutin, C.E.G., Wrangham, R.W., & Boesch, C. (1999). Cultures in chimpanzees. Nature, 399, 682-685.