The Arithmetic at Dzanga Bai
From 1990 a study at one forest clearing in the south-west of the Central African Republic did something nobody had managed for forest elephants: it watched the same individually known animals for long enough to work out how fast they breed. The answer — a first calf at around twenty-three, then one more every five or six years — is why a poached population needs the better part of a century to come back.
The platform at Dzanga Bai is a wooden deck at the edge of the trees, and you get to it by walking forty-five minutes over swamp on fallen logs. Below it the forest simply stops. A few hundred metres of open mud and mineral springs, water standing in the holes elephants have dug, and elephants in the holes: family groups arriving out of the wall of green on one side and leaving through it on the other, all afternoon, sometimes fifteen of them and sometimes more than a hundred. The trackers sitting beside you know a great many of them by name, which is not sentimentality. It is the method.
Forest elephants are almost impossible to study. They live under a canopy that hides them from aircraft, they move in small groups, and they are counted across Central Africa not by looking at them but by counting their dung along survey transects. Dzanga Bai is the exception, because the salts in the mud bring the animals out into the open on their own schedule and the same individuals keep coming back. From 1990 the American biologist Andrea Turkalo sat on that platform and wrote down who arrived. Ear tears, tusk shape and angle, scars, notches — the same identification technique used on savanna elephants for decades, applied for the first time to an animal that had never held still long enough for it.
What a clearing can measure that a transect cannot
A dung survey gives you a number and a trend. It cannot tell you how old a female is when she has her first calf, how long she waits before the next one, or how many of them live. Those three quantities are the entire demography of a species: put them together and you get the rate at which a population can replace losses, which is the only figure that matters when you are trying to work out whether an animal being shot at can outrun the shooting. Nobody had them for Loxodonta cyclotis, and the working assumption, for want of anything better, was that forest elephants behaved roughly like savanna elephants, which had been studied at Amboseli since 1972 and were reasonably well understood. Getting the real numbers required an absurd commitment of time. An elephant that is a calf when your study begins does not tell you her age at first birth until more than two decades later; there is no shortcut, no model that substitutes for it, and no funding cycle shaped like it. The Dzanga observations ran for well over twenty years before the demographic paper could be written at all, and they ran through a country that had repeated changes of government and successive armed rebellions in the same period.
The answer, and what it meant
The result, published by Turkalo with Peter Wrege and George Wittemyer in the Journal of Applied Ecology in 2017, was worse than anyone had assumed. Females at Dzanga did not begin breeding until around twenty-three years old — roughly a decade later than savanna elephants at Amboseli — and the interval between calves was five to six years. Put through the arithmetic, that gives an intrinsic growth rate about a third of a savanna population's. A forest elephant population knocked down by poaching would need something close to ninety years to climb back to where it started, assuming the poaching stopped entirely, which it had not.
You cannot measure how fast an animal recovers without watching it for longer than it takes to recover. That is the trap forest elephant science was in, and the only way out of it was to sit at one clearing for thirty years.
A species, separated
The finding landed in the middle of an argument that had been running since the 1990s about whether forest and savanna elephants were one species or two. Genetic work had been pointing at two for years; a 2010 genomic study put the split between the lineages at somewhere in the range of two and a half to five and a half million years, comparable to the gap between ourselves and chimpanzees. Meanwhile the survey data was catastrophic. A continent-wide analysis published in 2013 found forest elephant numbers had fallen by about 62 per cent between 2002 and 2011, with their range contracting by roughly 30 per cent — a decade in which the ivory price ran and Central African law enforcement mostly did not. On 25 March 2021 the IUCN assessed the two as separate species for the first time. The savanna elephant was listed as Endangered. The forest elephant, whose numbers had dropped by more than 86 per cent over thirty-one years, was listed as Critically Endangered — the last category before extinct in the wild. The Dzanga demography is what made that assessment possible to defend, because a Red List category depends on the rate of decline measured over three generations, and the generation length of a forest elephant was one of the things nobody had known until the clearing produced it.
May 2013
Two months after the Séléka coalition took Bangui in March 2013, a group of armed men — reported at the time as seventeen — reached Dzanga Bai and occupied the observation platform. They shot twenty-six elephants at the clearing. The staff had already been evacuated. It is the single worst thing that has happened at the site, and it happened in the place where the animals are most exposed by design: a bai draws elephants into the open, which is exactly what makes it useful to a researcher and to anybody else.
What is less often told is what happened next, which is that the reserve did not close. Guard patrols were rebuilt, the observation programme resumed, and Dzanga-Sangha spent the following years as one of the few institutions in that part of the country still paying a payroll and still functioning as an employer — funded from outside, certainly, but staffed and run by people from Bayanga and the villages around it. The BaAka trackers who habituated the gorilla groups from 1997 onwards, and who read the ground on the walk out to the bai, are the operating instrument of everything here, scientific and commercial alike.
The number that is hard to argue with
Conservation arguments usually run on charisma and they age badly. The value of what came out of this clearing is that it produced a figure instead — a slow, dull, checkable figure about breeding intervals that turned out to be the most damaging fact about the species. It changed the category, and the category changes what governments and CITES parties are obliged to consider. It also sets a hard limit on what any amount of good intention can achieve: if the last twenty years of losses were reversed tomorrow, the population still would not be back within the working life of anybody now employed to protect it. That is not a reason for despair and the people at Bayanga do not treat it as one. It is a reason to be exact about what protection buys. Ninety years is a long time to hold a line in a country that has struggled to hold anything, and the argument at the reserve office is not about whether the elephants will recover quickly, because everyone there now knows the answer to that. It is about whether the institution outlasts the arithmetic.