Carolyn Jensen, NPR News
Researcher Katy Payne looks out onto the elephant bai.
A Letter Home
This is the text of two letters Katy Payne sent to her family and friends while researching elephants at the bai.
February 5, 2002
Dear family and friends,
I am writing from Andrea Turkalo's field camp in the Dzanga-Sangha forest in the Central African Republic. It’s a quiet, beautifully built group of thatched huts not far from the Sangha River, a great, green, greasy river which flows into the Congo. It’s just before dawn. I have just said "bmbooma" to our pygmy helpers, and prepared coffee and oatmeal for our still sleeping research team, while beside me Melebu washed dishes and Matofi made preparations for the day's bread baking. Mbanda and Zu are talking softly down in their camp below ours, and a few small birds are starting to sing, but not so much in a dawn chorus as in little cadenzas lying against a long brassy chord made mostly by insect voices that will keep sounding all day as long as the sun shines.
From my cabin, I also hear the slow dripping of condensed mist on thatch, and the occasional croak of a hadeda ibis in the swamp below camp. Soon we will wade across that swamp and walk through dense forest to a large, muddy clearing where elephants are always to be found, digging and sucking up salty water and dust. A clearing like this is a "bai": Ours has a roughly rectangular shape, with its long axis about 500 meters long and a shallow stream trickling down the middle of it. On one edge stands a huge, high-roofed mirador or observation platform — that has been a boon to our project. It was built years ago by the World Wildlife Fund for tourists, but very few appear. We ourselves spend all day there, watching, recording, filming and counting elephants as they come in to drink with their sisters and cousins, to be reckoned up by dozens, sometimes more than a hundred at once. The later (and hotter) it gets, the more elephants come, and also the more sweat bees which crawl into our eyes and noses — that's the price we pay to witness this miracle.
At nightfall, the heat gives way abruptly to a sweet cool, and everything changes. We hurry home, in order to cross the swamp before the elephants arrive, knowing that this is where many of them will probably spend most of the night. We, bathed and fed and lying in our netted beds, will be hearing their screams and rumbles and speculating about what goes on down there. The swamp after dark is too dangerous for us to visit.
All of this is so much the same as before, it sometimes seems to the four of us who were also here in 2000 that we have been in Dzanga continuously since that spring. In fact Mya, Melissa and I have all migrated twice between Dzanga and Ithaca, and Andrea, whose home is always in Dzanga, spent a month with us in Ithaca last summer. And everything about this continuity — made possible by the continuity of our funding, for which I’m everlastingly grateful — is wonderful for the work. We are terribly lucky and we know it. Meanwhile it was no small thing that my daughter Holly gave birth to our Sofi, and that Mya and Zev got married, and that I met some of you during the months back at home. But with both ends of the journey familiar, such events just make the two home fires feel closer to one another, simplifying our desire to protect everything we care for.
That of course is the point, or as the French say, "le but" of our ongoing work in Dzanga. We're developing and testing an acoustic method for monitoring the size and well-being of forest elephant populations. These elephants belong to a unique species whose vulnerability is very difficult to assess because they are almost never seen except in this one place. Nonetheless they are as vocal as all elephants everywhere, and many of their calls are powerful enough to carry long distances through the forest. Thanks to Andrea’s knowledge, gained during 11 years as a Wildlife Conservation Society scientist studying these elephants as individuals, we begin to classify their typical calls to use as indicators of age, sex in an acoustic monitoring program. A scream with such-and-such a structure means that an infant is present; such-and-such a sequence of rumbles means that a female in breeding condition has found a mate. Calls like these would seem to be indicators of a healthy population, but we must be sure the call categories are real, for whatever interpretations we develop here will be applied to recordings made in forests where the calling animals are never seen but only heard.
Over the years, Andrea has noted most courtship and mating during the dry season — that explains the different timing of our two expeditions here. She seems to be right — and words will never do justice to the drama that unfolded on the bai in the late afternoon of January 25. We’d been watching a large-tusked, long-tailed bull with interest, as he was draining from his temporal glands and walking in a sort of swagger as he rapidly surveyed the other elephants in the bai. These are signs of musth — that male hormonal condition that increases the ferocity of a male's competition with other males and the intensity of his olfactory search for a fertile female. We had no idea that a pretty female called "Teardrop" was in estrus until Horsetail suddenly lifted his head out of a desirable drinking pit and started running. Teardrop ran ahead of him at full speed, and when only two elephants out of 50 are running, you know what sort of chase this is. She led him to a grassy meadow and then all at once stood still, while Horsetail caught up and mounted her. Farther behind, Teardrop's juvenile daughter, finding herself suddenly separated from her mother, ran into the meadow screaming.
Hardly a minute then passed — hardly had Horsetail descended from his almost-vertical posture and the youngster regained her mother — before the three of them were surrounded by a milling, circling, rumbling, roaring, screaming, urinating, defecating, trunk-waving crowd. Probably the first to arrive were Teardrop's relatives; soon they were joined by a family whose matriarch we recognized as Aurora. As Aurora's family left, others arrived, and then others, and the Auroras again, and yet others, and sub-adult and adult males joined the fray. Even long after Teardrop and Horsetail themselves had moved off, others were visiting, and each visit started with a wildly excited, noisy trunk-exploration of the grass at the site of the mating. When afterwards we listened to an audio recording we'd made at the time, we found an eight-minute unbroken series of overlapping rumbles, roars and screams. A rhythmic surge in intensity led me to believe that the loudest of the calls contributing to the chaos were estrous calls, announcing the female's condition. Excited calling continued throughout the next hour in bursts, as new elephants encountered the amazing-smelling spot.
Here I will break off to say that years ago, when Joyce Poole and I were first recording the calls of savanna elephants in Amboseli, Kenya, we saw and heard the same extravagant reaction to the mating of a young female. It was a phenomenon Joyce and Cynthia Moss referred to as the "mating pandemonium," and when I saw it, I decided that of all animals I’d ever seen, elephants are the most emotive. At the time I believed that the excitement was concentrated in close relatives of the estrous female. In the event we just watched, however, the excitement extended way beyond her ordinary associates. And in both cases, even after the mated pair had moved away and separated, what elicited the most emotion was smelling what had fallen onto the grass. Now the story continues, eerily similar in the two observations.
Half an hour into their union, Teardrop and Horsetail had drifted apart. Horsetail had gone back to his pit while Teardrop was wandering in the west meadow close to our observation site. All at once I noticed that she was not alone, but was being followed by Gonjala, a slender sub-adult male we'd also had our eye on, for a short stream of darkness on his cheeks revealed that he, too, was in musth — probably for the first time. Without a preliminary chase, Gonjala mounted Teardrop. All was, to our ears, silent, but instantly Horsetail arrived on the run from quite far away, and drove Gonjala away. No audible sounds accompanied the takeover, and no other elephants visited the site of this new mounting (which was, to all appearances, a full mating), or showed any excitement at all! But a transformation had occurred in Horsetail. From then on he guarded Teardrop extremely closely, for the first few minutes nudging and shoving her ahead of him with the flat of his long tusks, and then mounting her again. Again the surrounding elephants showed no interest, but for the rest of the afternoon, Horsetail did not allow any other male near Teardrop.
At this point, I can add a juicy morsel from my memory of the parallel event in Amboseli. There, too, the musth male's guarding did not start until a lesser male had sneaked a mounting. Once it started, however, his attention was undivided. But before all was said and done, that guarding male was replaced by a bigger one, who had apparently been called in from a distance by the estrous female's calls. I predicted a similar takeover in Dzanga, when we noticed, several hours after the pandemonium, an absolutely enormous musth male with tusks to the ground, smelling his way out of the northern forest and across the bai to the spot where Teardrop and Horsetail had first mated.
"The plot thickens!" I cried, but just then we heard the padding of bare feet on the mirador's stairs. Matofi and Mbanda had arrived to escort us back to camp. Hastily we packed up, needing to cross the swamp before dark.
So abruptly ended our observational and video records of a life-changing event. But a set of audio recordings continued, as they will for the next two months, from the array of autonomous recording units (ARUs) deployed around the perimeter of the bai. The development of these units, and automatic ways to analyze the massive recordings they accumulate, have occupied Chris Clark, Kurt Fristrup, Tom Calupca, Kathy Dunsmore, Harold Mills and others in the Bioacoustics Research Program of our mother institution, the Cornell Lab of Ornithology, continuously for several years, with the elephant project as a major testing ground and our excellent new engineer, Eric Spaulding, responsible for initial success out here. From the recordings made by our array — simpler, more comprehensive, less costly, better surveyed, and much easier to use than the version we had last year — we will later, back in the lab, be able to extrapolate broadly about the nature of many other elephant happenings.
But of course we are greedy to actually see more with our own eyes, for each event contains its sine qua non, and very often the greatest beauty lies in the uniqueness.
All this and much more is food for thought, and even if there were no human drama associated with this project, there would be an ongoing saga to relate and attempts to explain it in evolutionary terms. But there is, of course, human drama as well and plenty of it, plenty of technical drama, even more political drama, and all kinds of dramas of the heart. By the time I write again, Chris Clark will be here. He is the founder and director of the Bioacoustics Research Program, to which we owe the technical foundations of this project. And we are also old, old friends, having met in 1972 when we were both wet behind the ears and studying whales in Patagonia. This too is a blessing … but enough for today. What I have written I will now entrust to the satellites.
So I'm thinking of you all. I would prefer to send to each a separate message: I hope you will do me the favor of imagining a good one. I send to all my deep gratitude for the many kinds of support that sustain us here.
April 4, 2002
We've reached the end of a beautiful and fruitful field season in the Dzanga Elephant Listening Project. Every day this week Eric and Matofi have lowered recording units from the trees surrounding the elephant clearing, and solar panels from the roof of the mirador, our observation platform on the bai. Samba, Sakonga, Yaku and Matofi, our BaAka pygmy trackers, have carried these and 20 truck batteries up the mile-long trail and across the swamp to camp, where in a bittersweet, gladsome mood the rest of us have been packing up. Of the five American members of our research team, three — Eric Spaulding, Mya Thompson, and Melissa Groo — will arrive back in Ithaca on April 8. Andrea Turkalo will stay here and return to normal life after our three-month invasion of her home and work space. I will visit another branch of the project, in the Kakum National Park, Ghana, and will return home to Ithaca on April 22.
Our basic news as we wrap up the field season is that we have kept well, and our data look very good. The Dzanga elephant population seems full of health despite several poaching incidents. New babies have been born, and more (to all appearances) are on the way. We have collected all sorts of hairy stories to relate by the home fires. Friendships have deepened — how could they not? — and the members of the team, whom I knew would be compatible and effective, have surpassed all expectations. And as our commitment to this miraculous place has also deepened, we've learned more about what threatens it.
This has been a sobering experience, whose details must wait for another letter. The point of this one is that we are profoundly grateful to have been in beautiful, difficult Dzanga the second time around, with a more ambitious program and a clearer sense of reality. Beautiful and difficult! The vast buttressed forest twinkling internally with sunspots, crimson butterflies and huge and tiny falling flowers — the sudden elephant or gorilla in your face — but oh, the heat and the sweat bees, chiggers, filaria and tsetse flies, the rashes and bellyaches! Words can't touch the rainforest, it's so mixed with one's own sweat — nor could words justly describe the strengths and lovelinesses of my four companions, not to mention their talents and doggedness — we all knew we were lucky, very lucky to be together, in the challenge and the ordeal.
It's hard to break the habit of collecting data. Last night it felt uncomfortable to not be out on the mirador as the moon rose. Of course our absence was the fulfillment of a promise — to give the privacy of the forest back to the creatures it belongs to — but to give it back at a time when a fresh mystery was still unfolding! For a (not wholly unexpected) mystery had caught my attention late in the season, when on March 23, under a waxing moon I spent the night on the mirador. No sooner had the sun gone down than far more elephants than I've ever seen together emerged out of the forest. They entered from all directions as if they knew what they were doing, and most of them stayed on the bai till sunrise.
At midnight I counted 160 elephants as dark silhouettes against the moon-struck central pool in the half-kilometer-long clearing; dozens more were probably hidden in the shadowed forest edge. Only one-third that many had been visible at sundown. All night long they milled around, jousting and scaring one another and making a concentration and variety of sounds that was off the scale (just as were their numbers) from what we've found during daylight hours. The overlapping rumbles that signify meeting and greeting were unusually frequent, and interspersed between these were repeated steady rumbles, as if family members were announcing their locations at regular intervals. Coarse, abrupt growls came from wandering youngsters who suddenly lifted their trunks, smelled something unfamiliar — presumably strangers — and called out to their mothers, siblings, aunts.
These calls were often followed by overlapping rumbles that might have been responses, but so many calls were emanating from hither and yon I couldn't be sure which were associated with whom or what. Nor was I sure that the elephants themselves were sure, for rumbling, growling, trumpeting, and screaming are contagious behaviors in times of excitement, cause in themselves for further vocalizing. I heard harsh, horrid screams from preferred digging pit areas and felt that these probably signified the sort of hassles, displacements and tusk-stabbings that accompany competition for a limited resource. But in addition, there were screams from other spots, and snores, and half a dozen sorts of calls I didn't recognize at all. What were these about, and who were these incomers? Where had they come from? What explained the intense affect they were having on one another? And what did this imply about the real size and structure and welfare of the population?
Plenty of answers suggested themselves, and my mind leapt forward with probable details. I thought I recognized an exchange between Etta 2 and Etta 3, motherless young adult sisters who, during the 2000 field season used to call for one another pitifully when separated. This year they're transformed by the arrival of Etta 2's first calf, a wild pranky little female who nurses from both of them and splashes between the two at neck-breaking speed. I thought I recognized — but never mind! The point is that the night experience, now seven times repeated by one or another of us, brought home a vivid sense of the task we have committed ourselves to — a sense of what we must able to deduce, statistically, from impersonal counts and computer-managed call-sorting, if we are to become responsible interpreters of long, unmanned recordings made in forests where elephants are invisible.
Our desire to automate the interpretation of elephant sounds comes as an effort to help in the conservation of this little-known species. Forest animals with large home ranges are notoriously difficult to protect because they're almost impossible to survey and monitor: Photographs made from airplanes or satellites do not reveal whether the ecosystems under a dense forest canopy are thriving or being depleted. Dung count surveys give some basic information, but the conclusions that can be drawn from such data are fairly limited, and our colleague Richard Barnes, who designed that strategy, has long proposed that they be supplemented with acoustic information. Now it's good news that we are finding robust clues to the size and health of forest elephant populations imbedded in their calling behavior.
Hand in hand with this finding is the development, in the Bioacoustics Research Program of the Cornell Lab of Ornithology, of a set of tools to implement terrestrial acoustic monitoring. Christopher Clark and Tom Calupca have created a small, powerful data-gathering tool called the ARU — autonomous recording unit. This is a hard drive specialized for the continuous collection of large amounts of acoustic data, packaged in a watertight case with connections to a microphone and an external battery. ARUs can be left unmanned for months at a time, to record the acoustic environment — animal calls, wind, rain, falling trees, local footfalls, gunshots if they occur — and insofar as they do not need to be serviced, their presence does not disturb the situations being monitored.
Of course the biological interpretation of recordings collected in this way depends on baseline knowledge of the calling behavior of the animals of concern, and this is the purpose of our intense study of the Dzanga elephants. In this study, we have been collaborating with Andrea Turkalo, the Wildlife Conservation Society research scientist who has throughout the last 11 years been recording the comings and goings and associations of these forest elephants and knows them as individuals.
One friend has asked me to explain how we have gone about collecting this year's data. We have recorded continuously, day and night, from an array of six ARUs placed above the reach of elephant trunks in trees surrounding the bai. The necessity to elevate them was new to this field season and learned at some cost — which added drama to Chris Clark's three-week visit in February. (And the fact that Chris is part monkey saved the day.) Meanwhile during daylight hours, we have counted all the elephants on the bai at half-hour intervals. These two sets of data will enable us to correlate elephants' dry season calling rates with their numbers. At one-hour intervals, we have taken the story further by characterizing all the elephants on the bai by age and sex class. This enables us to ask, what kinds of calls and rates of calling reveal certain age/sex groups and not others? Then, once a day towards late afternoon when the bai is increasingly populated, Andrea has named all the elephants we've been counting and, using a directional rangefinder, located them on a map of the bai. This provides the individual dimension, so we can ask questions like does the age of a female's infant determine how often she calls to it? Does it influence what kinds of calls she makes, and how comparable are females with similar-aged infants in these respects? Finally (well, not quite finally, but finally for the purposes of this description), Eric hooks up a digital camera array of his invention, which will enable us to trace the movements of all visible elephants on the bai throughout each day. In this context we made two or three hour-long focal video recordings every day, following a consistent protocol. The product of all this will be the ability to compare the calling behaviors of a large sample of known elephants, and of elephants in known age and sex classes, in a variety of social and environmental circumstances. The configuration of the ARU array was such that any calls strong enough to be picked up on several of the microphones — including calls too low in frequency to be audible to human listeners — can be located on the map and attributed to the individuals who made them, as revealed in the simultaneous video footage.
So we created quite a data trap. The task of analyzing it all looms large as we and our associates in the Bioacoustics Research Program think about the next two years. Many moons will wax and wane while we are getting all of this together. The goal is to produce a simple system enabling local rangers to characterize elephant populations on the basis of autonomous recordings, without requiring visual confirmation or the attention of foreign experts.
We also hope for something else. The same recordings were also designed to answer a set of questions about elephants' view of the world and individual experiences. Some easily recognized elephant calls occur, for example, only during competitive interactions. By noting the occurrence of these we can relate elephants' aggressive behaviors to several factors which are, in humans, predictors of violence: to density (level of crowdedness); to resource limitation (for mineral pits, which are highly desirable during drought but become too diluted following rains); to the ratio of adult and sub-adult males to other age and sex classes present; to the presence of males in a heightened sexual condition called musth; and so forth. What emerges is a summary of the conditions that provoke and that relieve intense competition in this species.
At the same time we have documented certain male behaviors that run contrary to gender-based predictions. In certain large males we've seen tolerance of infants, occasional acts of apparent compassion, greetings with particular females (in two confirmed cases, their mothers), and other interactions which seem for all the world like evidence of social memory and friendship. The emergence of such social graces cautions me against believing, even as we find males less vocal and less cooperative than females, that they are less aware of the subtleties imbedded in individual relationships. If only we could know what elephants know and are thinking! But thought is the very hardest thing to demonstrate in any species.
Yet our studies move us in the direction of understanding, for studies of animal communication offer the best mode of inquiry we have into animals' thoughts and feelings. And such inquiry interfaces nicely with the conservation-oriented aspects of our work here. The better we get at interpreting elephants' calls and calling patterns, the closer we'll come to valid interpretation of recordings made without matching observations. So we face a double challenge: to document the vocal behavior of these particular individuals, long known by Andrea, with great care; and to extrapolate from the results of these observations to other forest elephant groups whose size, structure and circumstances are unknown.
Finally we hope that the development of acoustic protocol for monitoring forest elephant calls will shorten the task of doing the same for other endangered species. The acoustic monitoring of biodiversity has been for a long time the hue and cry of the Bioacoustics Research Program at the Cornell Laboratory of Ornithology. Chris, director of the program, is responsible for the development of methodology for surveying whales in the world's oceans. The lab's endorsement of that project, and of my elephant acoustic research since its inception in 1984, is evidence of its dedication to the conservation of all wild species. There is no way to express my gratitude for the broad-mindedness that welcomes our work into this environment in a laboratory whose very name expresses its primary commitment to birds.
There's also no adequate way to thank the supporters of the Elephant Listening Project for keeping our effort alive during a trio of years in which funds for conservation initiatives have become increasingly scarce. Our supporters have included the Wildlife Conservation Society, Conservation International, the International Fund for Animal Welfare, World Wildlife Fund, the U.S. Fish and Wildlife Service, the Park Foundation, and the Harry Frank Guggenheim Foundation. May we complete the work, and may it help elephants.
Dawn is arriving so I must consider this finished. In two hours we will be on the road heading for Bangui.
Good morning to all,