One ape, two ape: why counting apes is so difficult — but crucial

It’s no secret: one of the biggest threats facing the world’s great ape populations is industrial agricultural, now rapidly extending throughout their native home ranges in Africa and Southeast Asia — threatening gorillas, orangutans and other primate species.

But as bleak as that future may seem, there are ways apes and industry can coexist.

However, central to gauging the impacts of industrial agriculture and the effectiveness of innovative new conservation solutions as they are tried out, is comparing how ape populations are responding by placing current threats in context with those of the past.

Which, it turns out, is easier said than done. The fact is, very little is conclusively known about how ape populations have evolved over time as they’ve been impacted by human development, due to a lack of long-term, comparable baseline data.

Lilian Pintea, Vice President of Conservation Science with the Jane Goodall Institute and a contributing author to the recent Arcus Foundation report, State of the Apes: Industrial Agriculture and Ape Conservation, says there aren’t many instances in the ape research community where the data can support long-term assessments of how populations fluctuate.

“I’ve been working with ape conservation now for more than a decade and I’ve come to find most research is not designed with a long-term perspective in mind,” he says. “Yet this is a crucial consideration for accurate scientific conclusions about how populations change over time and for setting conservation guidelines.”

Based on discussions with ape experts, it seems no species is spared from this fundamental research flaw. But there is some good news: researchers say that there are ways to lessen the negative impact of past oversights. And in many cases the tools to achieve this are arguably already available — they’re simply not being employed to full advantage.

Working without the data: Silvery gibbons in Java

Vincent Nijman, a primate conservationist from Oxford Brookes University works with the Silvery gibbon (Hylobates moloch). He told Mongabay that it was the rare Javan Hawk-eagle that first drew him to Java and Indonesia 22 years ago, but while trekking and camping in the forests surrounding Mount Halimun-Salak National Park, he encountered the region’s gibbons and was instantly enchanted.

“One morning I awoke to their calls, which are truly more like singing, then saw my first gibbon,” says Nijman. “Once I knew how to recognize them, I found quite a few of the animals, including in places not thought to have gibbon populations.”

Nijman assumed at the time that the local research community would have swallowed up the potential for major discoveries. But he learned otherwise. He found that population estimates for the species varied immensely, and some even claimed that the population had fractured into two genetically distinct groups — an idea since disproven by genetic evidence. Determined to get a better sense of the gibbon’s true numbers and range, Nijman returned to Java in 1994 to study them.

In 2004, he published the results of his work, backing up his earlier forest observations with data collected by fixed-point call counts and line transect walks conducted from 1994 to 2002 in Gede-Pangrango National Park and surrounding West Java, as well as the Dieng mountains in Central Java and locations in East Java.

Nijman found that previous estimates of populations had indeed been conservative. Past research had overlooked gibbon habitats and populations, especially in Central Java, an area the species wasn’t thought to inhabit. In 1994, the Javan Gibbon and Javan Langur (PHVA) Workshop estimated a total gibbon population of between 386 to 1,957 individuals, but Nijman’s moderate estimates more than doubled the number to between 4,000 and 4,500 animals.

“Prior estimates qualified the Javan gibbon as Critically Endangered, but luckily it turned out this was wrong,” says Nijman. In the Dieng Mountains alone he found about 800 animals — a region where the species was thought to be extinct.

“That doubled the total number of the gibbons we thought we had right there,” he says. “The more people look for apes, especially gibbons, so far, the more they find of them, not because there’s more animals but because people are using more ways to look for them.”

The apples-to-oranges problem

Nijman’s new, larger population estimates were encouraging, but he needed to know more to evaluate future conservation possibilities. He analyzed survey data gathered over the past 30 years, but found that nothing could be comparatively concluded about how populations had changed over time.

He could only create a vague sketch of the past.

Researchers had come and gone over the years, adding pieces to the gibbon population puzzle — but that work was spotty not continuous, limited by funding shortfalls or academic deadlines. And each survey relied on different assumptions of the gibbon’s range, preferred habitat, and group size — creating a challenging apples-to-oranges comparison problem.

Pintea explains that what Nijman encountered is not uncommon. “Most funding is short term and local in scope, using a diversity of definitions of vegetation types, suitable habitat, and so on,” he says. And conservation projects are often focused on very specific temporal or spatial goals. “This may provide a lengthy string of data, but often not one you can draw scientifically accurate conclusions from on how populations change.”

Ian Singleton, Director of the Sumatran Orangutan Conservation Programme told Mongabay that the data comparison problem is further complicated as survey capacities and data collection techniques improve over time. It is, he says, precisely because scientific research always aims to utilize the most recently available, best cutting edge tools — new technologies, methodologies and insights — that a natural disjointedness results when new, more reliable data is compared to less precise data collected over long periods of time, making the apples-to-oranges problem even more challenging.

“For past data we’re always stuck with what was possible then, which normally would be far less reliable and accurate due to fewer surveys, less experience, old methods and a lot of extrapolation between few sites,” writes Singleton. The Java gibbons are a case in point: their populations didn’t magically jump from 386-1,957 individuals to 4,000-4,500, Nijman simply used more advanced and thorough survey techniques.

Pictures of the gone world: the chimps of Gombe

There’s no way to totally and accurately map the past, but there are ways to learn from it. One promising place to look is at one of the very few existing long-term, long-running great ape research projects: the chimps of Tanzania’s Gombe Stream National Park, who first inspired Jane Goodall and have been continuously studied since the 1960s.

“The data we have from Gombe is on a totally different level,” says Pintea. “We’ve basically been following the chimps daily, recording every 15 minutes for the last 50 years, using very rigorous protocols and methodology standards to get just the right behavior, population and ecological data.”

Pintea and the Jane Goodall Institute team have used this massive database to make inferences about how chimp populations inside and outside Gombe have fared over the past 40 years, and their results, featured alongside Nijman’s work in the State of the Apes report, showcases the benefits of tracking populations long-term.

The institute team coupled chimp info with remote sensing data dating back to 1947, imagery from historical aerial photos and satellite sensors like SPOT and Landsat, plus field observations, and was able to evaluate Gombe population fluctuations and range changes continuously between 1972 and 2012. The findings were mostly good news.

While all chimp populations surveyed had suffered some losses, those inside the protected park’s boundaries had lost far fewer members than those whose range overlapped with areas outside the park. And within the park, some 58 percent of deaths were linked to disease, rather than poaching or other human disturbances. These findings clearly demonstrated the perks of protection, and showed that more consideration needs to be given to chimps whose populations’ fall outside the shelter of reserves and parks.

Pintea says the longevity and consistency of the Gombe data allowed the institute scientists to ask meaningful questions about how the populations have changed. But not all ape species, or even subsets of chimps, are so lucky. And in the end, there’s always room left to improve estimates, even in the case of Gombe.

“Gombe is very small, say 36 square kilometers (~14 square miles), where we’ve focused on only a few chimp communities intensely with the goal of understanding their social structure, behavior and individual characteristics,” says Pintea. “What we learn from all this is what it means to be a chimp, but this long-term data is also useful to understand how to conserve or protect them in Gombe and elsewhere.”

He notes that, on one hand, the research advances he’s seen over his long career have been amazing. For example, the Jane Goodall Institute in partnership with the University of Maryland (and funded by NASA) demonstrated that it’s now possible to annually map all chimp habitats in Africa using LandSat data that compliments data platforms like Global Forest Watch. But on the other hand, there’s still a tremendous amount of work to be done to address each chimp population throughout their expansive range. Pintea says that, researcher labor aside, the cost of securing this level of detail is unimaginable considering the limited budgets with which most primatologists work, when compared to the sustained effort at Gombe through the years.

“It wasn’t until 2006 that the Institute was even able to make the first potential suitability maps and population estimates about chimp abundance in Tanzania,” Pintea points out.

Clearly, there will always be unknowns: large past gaps in the great ape data continuum, and shortcomings even in current surveys and studies. So how much can be done to chart real conservation progress, improving comparisons of data sets over time? Quite a lot, it turns out, thanks to new technology and an increasingly global scientific community.

Empowering the data we have

A clear picture of the past may not be necessary to lessen the impact of data unknowns, and there may be practical ways to ensure that data sets can be used to ensure that future conservation plans are viable, effective and economical. The place to begin working, primatologists agree, is on the general lack of organization and communication within the great ape research community — the key stumbling block standing in the way of a more comprehensive knowledge base, and a rather sad admission in today’s Internet age.

That’s why big investments are currently being made in online databases — amassing all of what is globally known about a species and making it fully accessible to researchers. Hjalmar Kuehl is Project Manager of the IUCN SSC A.P.E.S. database, a virtual repository for chimp, bonobo, gorilla, orangutan and gibbon information. Kuehl says the project’s origins date back to 2005, and have been driven by the IUCN/SCC Primate Specialist Group’s desire to create an all-encompassing platform for ape data.

“People have always said the data is out there, it’s just never been in one central place,” says Kuehl.

In 2007, several other big name groups, including the Arcus Foundation, came onboard to help expand the database, hoping in particular to make it more user friendly. “The idea was to add visuals to the platform; at the time it was only really for experts,” says Kuehl.

To improve the user interface, a mapping feature was added, allowing for the layering of an array of data sets atop each other — including diverse factors and threats such as land cover, conflict zones, and infrastructure, plus great ape population info such as density, distribution, and abundance.

“Our data comes from NGOs, individual researchers, field groups, government agencies — bringing datasets together gives you insight impossible to gain from one point,” says Kuehl. “And [its ease of use] brings people into the conversation otherwise unable to join, or unaware they should be involved.”

Pintea praises the A.P.E.S. database, asserting that the overarching collection and organization of information on this scale is critical for great ape research going forward.

“All the time we hear how big data is changing conservation, we also need to understand how big data can be converted to better conservation decisions. Simply having the data is not enough,” he says. Decision makers can’t look at the numbers and hope to make good sense of it, let alone see nuances. “You can do a lot with the existing data to support some conservation questions, but we’re still putting the system in place to turn it into a powerful collective database, which is why we need to support long-term data alongside harmonizing efforts, like the A.P.E.S. database.”

Kuehl agrees that conservation planning could greatly benefit from databases that offer a broad assemblage of information and overlays concerning threats and populations, as A.P.E.S. does. Pintea and Kuehl say that, with the inclusion of a broad set of factors and the integration of all research available, the impact of errors and oversights can be buffered.

“Though more subtle threats like bush meat hunting or poaching can be just as damaging to a population as more obvious threats like habitat loss,” says Pintea. “And often these are the factors that really drive decisions, or impact the people who make and enforce them.”

Using such a multidimensional approach, Kuehl notes that no matter how scattered or incomplete the data may be, there’s almost always something that can be done with it. In the case of the Silvery gibbons, Nijman was able to calculate percent forest cover change based on Prasetyo et. al’s data. In 2005, he was able to conclude that Java’s Mount Halimun National Park had decreased in forested area by 2 percent annually between 1989 and 2004. This is useful knowledge for determining likely shrinkage of ape habitat, Nijman says, but not satisfactory grounds to make comprehensive conservation choices.

“Unfortunately, we can measure factors like forest loss with great precision now, but we still don’t know how this impacts gibbons,” says Nijman, and in the case of most gibbons, not enough is known about other major threats to counter this missing info. “Cracking that code still requires good old, on the ground, pricey, and somewhat tedious fieldwork, at least to begin with.”

After the data: the tricky task of prediction

As Kuehl and Pintea note, there’s a slew of contributing factors fuelling ape conservation decisions that have nothing to do with population data. And interpreting all that data to determine an effective conservation course of action is fraught with difficulty.

Conservation scientist Dr Susan Cheyne has been working with gibbons for more than 17 years, and she believes that for most Indonesian apes there are enough surveys and datasets now available to draw decent, albeit imperfect, population estimates. However, this Director of Gibbon and Felid Research and Conservation of the Orang-utan Tropical Peatland Project (OuTrop) also contends that outside pressures can put researchers in a bit of a lose-lose situation when making predictions used to drive conservation goals. Especially since once those decisions are made, there’s often an all-or-nothing approach, where each new revelation about population, be it an upward or downward trend, is used to challenge the overall status of a species or the protection plans already in place.

“There’s a general need to be able to make bold statements [to decision makers], to say things [definitively] like, ‘Only X number of animals will be left in X years, or X animal will go extinct in X years,’” says Cheyne. “But when these calculations turn out to be even the slightest bit wrong, which is nearly inevitable, you’re called out for either fear mongering or downplaying a crisis.”

Another worry for researchers is that information like population estimates, when taken in isolation and without historical context, can overshadow true species trends. Earlier this month, the Sumatran Orangutan Conservation Programme (SOCP) team published a paper in Science Advances doubling the estimated total number of Sumatran orangutans; the former assessment of 6,600 was axed and replaced by a more than doubling to 14,613. This didn’t indicate some sudden orangutan population explosion. As in the case of Nijman’s Silvery gibbon, SOCP had simply looked in places the species wasn’t thought to inhabit, namely at higher altitudes and in previously logged habitats, and found populations there. They also discovered orangutans in never surveyed regions.

Despite this seeming good news, the paper’s title offered little reason to celebrate: Land-cover changes predict steep declines for the Sumatran orangutan (Pongo abelii). The bolstered population estimate wasn’t the key finding, rather the study predicts one-third of Sumatra’s current orangutan population will be gone by 2030 due to habitat loss.

Yet most media outlets ran with the one-dimensional positive population numbers news, much to the team’s disappointment. “It is of serious concern to those working to conserve the species that the majority of media coverage has misrepresented the new findings in their headlines, leading readers to mistakenly believe that the wild population of this Critically Endangered species may have actually increased,” writes the team.

Another reason the new stats aren’t as exciting as they may seem is that they don’t come with much historical context. Singleton, part of the SOCP team, writes that he is often asked what the population of Sumatran orangutans used to be in the past as compared to today. But such assessments can only be inferred by apples-and-oranges comparisons of differing surveys, leaving considerable room for error. For past and present numbers to be useful, he notes, the exact same survey methodology would need to have been repeated at regular intervals — maybe every 10 years — using the same transect sites.

The team’s recent findings represent the most up to date, and by far the most thorough, comprehensive and systematic surveys ever done on Sumatran orangutans, writes Singleton. That should ensure the work fits in with future appraisals. But again, the more research conditions and procedures change in the future, the more relative data comparisons become.

“Who’s to say the methods available in 10 or 20 years won’t be extremely different, and hence not really comparable to the data we have now?” Singleton writes. That’s why research efforts must be as consistent and as rigorous as possible. It’s also why all invested groups need to be onboard with the same conservation plans.

Spotlight on the smaller apes

While a myriad of similar challenges face all great ape research and conservation initiatives — including the shared struggles of financing, data collection, interpretation, planning and coordination — those working with the smaller apes face additional, often even greater handicaps.

Primatologists pretty much agree: the lesser apes simply don’t get much notice — and the gibbons are among the most undeservedly ignored.

While there are five branches off the Hominidae ape family — bonobos, chimps, gorillas, orangutans and humans — the other ape family, Hylobatidae, is entirely made up of gibbons, divided into four genera, Nomascus, Hoolock, Hylobates, and Symphalangus, aka siamangs.

“Of all the apes, gibbons cover the most geographical ground, with 19 species found in ten countries,” Cheyne reveals. “They’re the singing, swinging apes. Truly, they’re brilliant and if you were to watch footage of gibbons flying through the canopy, I guarantee you’ll be impressed.”

Gibbons are found exclusively in Southeast Asia and unlike some of their massive cousins, only reach an average height of 3 feet (.91 meters) standing, and a weight of between 10 to 30 pounds (4.5 to 13.6 kilograms). They can walk bipedally, and do so when not using their primary form of locomotion — braciation — swinging from branch to branch at top speeds of 35 miles per hour (56 kilometers per hour), covering up to 20 feet (6 meters) per swing. Unlike the rest of the apes, gibbons have kept, and evolved, traits to favor this highly efficient mode of tree-travel, such as permanently curved fingers, super-strong rotator cuffs, long arms, short legs, and a slighter body frame.

Unfortunately for the remarkable gibbons, all 19 species are listed on the IUCN Red List as threatened with extinction. Eleven species are Endangered and another four are Critically Endangered. And almost every species suffers from a serious lack of long-term data and research.

Cheyne says that, even though the first wild studies of gibbons date back to 1947, there’s been far less funding for gibbon work over time compared to that going to gorillas, chimps or orangutans, possibly because the search for human origins has drawn researchers again and again to the great apes.

“Unfortunately gibbons fall in a funny category, being called lesser or small apes because they’re literally smaller than our own direct relatives,” says Cheyne. “And no one person has ever championed wild gibbons, as Goodall stood for chimps, or Fossey for gorillas.”

Kuehl reports that when it came to gibbons, the A.P.E.S. team didn’t feel it had the expertise or information to treat the lesser apes with the same sort of depth as with the great apes, but they’re currently redoing the gibbon database to reflect what is known.

“There’s a definitive lack of data when it comes to the smaller apes, like gibbons, but this is primarily because we haven’t yet dedicated the time and funds to support long-term surveys,” he says. And getting all this information and processing it takes a very long time. The data SOCP used to assess Sumatran orangutans took 6 months to analyze alone. “We’re getting to gibbons,” says Kuehl.

The Indonesian Country Director for the Aspinall Foundation, Made Wadana, seconded Kuehl’s views. He told Mongabay that there is still sparse regional research for gibbons in the wild compared with other ape species like orangutans.

According to Cheyne, part of the issue is that each of the smaller apes come with their own quirks, many of which researchers are still figuring out. Gibbons spend most of their time in trees, and must be counted by their morning calls. But not every member of the group will sing every day or at the precise same time. And there’s no guide yet to decipher what calls mean, or which members of a group make them.

Another disadvantage facing small ape surveyors is that most remaining populations are fragmented and far-flung geographically.

“One of the problems in Java is that [Silvery] gibbon populations are so isolated, and there may be only 4,000 or so of them left out there,” says Cheyne. “In the main site where I work in the Kalimatan peat-swamp forests of Indonesia, there are conservatively 19,000 White-bearded gibbons.”

Kuehl notes that this fragmentation of remote habitats makes it easy to miss persisting outlier populations that may endure or spring up. Adding to the problem, the smaller apes live in habitats that have undergone more deforestation than some of the great apes.

“Chimps, for example, still occur over a relatively large area,” he says. “And in Africa, despite more recent developments, there are still remaining huge blocks of forest, something mostly absent in large parts of Southeast Asia today.”

Given that the threats facing the smaller apes are so advanced, and so little is confidently known about them, more assumptions will likely have to be made when it comes to defining conservation goals. But progress is being made, as the smaller apes gain global attention. Last year was the Year of the Gibbon, though the event didn’t garner quite as much publicity as was hoped, something Wadana attributes to a lack of organization.

Nijman says he’s thrilled that the 2015 Arcus State of the Apes report made room for gibbons. “This is something that wouldn’t have happened even 10 years ago.”

Learning from the past, preparing for the future

As all the experts interviewed for this article point out, assessing the big picture is likely the only way to effectively conserve a species. And getting that picture entails a lot of long-term work: dedicated researchers, committed funds, and maybe most crucially, good ongoing communication.

“When it comes to defining population numbers over time, patterns, or assessing what different threats mean to population viability, you need respectable, long-term data to make conclusive statements,” Pintea says. “Unification of information is a big part of this.”

Kuehl adds that the scientific community is becoming much more open to the idea of sharing knowledge, though there are still some researchers who are hesitant about permitting the A.P.E.S. database to feature their work. “We have a strict policy that authors retain ownership of their work, and attitudes are changing, but it took a long time to settle all the fears people had,” he says. Kuehl is confident that the resources already exist to fill in many of the major gaps in ape data.

The A.P.E.S. database has aided in systematic surveys for entire countries — something rather unconceivable before the dawn of computing. It’s also helped establish conservation habitat priority areas for apes, elephants and other mammals in Equatorial New Guinea, as well as chimps in Liberia, Sierra Leone, and Tanzania. Kuehl says in some of these countries there was previously no established baseline data to work from.

Time to embrace the lesser apes

If any large conservation group today is “committed to updating the baseline data on the lesser apes and combining it in a platform like A.P.E.S., I believe it could be done,” Kuehl says optimistically. “It’d be huge, and it’s true the information wouldn’t be flawless, but yes, it could be done.”

Cheyne agrees that smaller apes can, and will, eventually be fully mapped in terms of their population density, abundance and distribution, but if this is going to happen, best practice guidelines need to be set in stone and strictly adhered to.

“There have to be assumptions made, even when you’re working with the best data and most advanced mathematical modeling tools, so there will always be some degree of error,” she says. “That’s why we need to perfect how to make good estimates and then commit to using these methods for the long haul.”

Luckily the motivation to establish these standards is growing as the general public and research community learn more about the plight of the smaller apes.

In 2011, the Gibbon Specialist Group was founded — known technically as the IUCN SSC Primate Specialist Group’s Section on Small Apes (SSA). And this Spring, the South Asian Primate Network is updating their section of the IUCN Red List. Cheyne notes that over the years the IUCN, with the help of the Section on Great Apes, has been working hard to gather together the very best science available. She hopes now the smaller apes will get their time in the spotlight.

Nijman says overall that he’s cautiously optimistic about the fate of Silvery gibbons. Some of his students are now working in Java, and there are numerous universities, conservation groups, and NGOs in the region who could set up long-term monitoring projects with relative ease when compared to the cost and logistics of foreign efforts.

“This is an amazing animal that lives in an amazing forest,” says Nijman. “Both are worth fighting for and there’s plenty of people in the position to help out.”

Wadana writes that the time to act is now, especially in Java where high rates of hunting and habitat change increase the need for continuous monitoring. Given Nijman’s surveys are now more than a decade old, Wadana recommends that it’s also time to revisit their findings. Nijman agrees.

“It’s good to draw attention to Silvery gibbons now — animals in a somewhat exceptional situation living on an island side-by-side with 120 million, or so, people,” he says. Everything considered, the gibbons seem to be doing better than predicted. But this is a case of good luck not good planning.

“Keeping up this fine balance will require a lot of careful consideration,” Nijman concludes. “This means people really need to start learning from each other, sharing data, and using consistent, long-term, tried and true methods. There’s just no other way around it.”

Citations

Geissmann, T. (2007). Status reassessment of the gibbons: Results of the Asian Primate Red List Workshop 2006. Gibbon Journal. (3), 5-15.

Lanjouw, A., Husson S. J., Morrogh-Bernard, H., Santiano, Ruwanto, A., Harsono, F., …Nijman, N. Chapter 7: Ape populations over time: Case studies from Gombe, Mount Halimun Salak, Sabangau and Wamba. State of the Apes: Industrial Agriculture and Ape Conservation. (2015). Arcus Foundation and Cambridge University Press, Vol. 2., 196-227.

Moyer, D., Plumptre, A. J., Pintea, L., Hernandez-Aguilar, A., Moore, J., Stewart, F., … & Mpunga, N. (2006). Surveys of chimpanzees and other biodiversity in Western Tanzania. United States Fish and Wildlife Service.

Nijman.,V. (2004). Conservation of the Javan gibbon Hylobates moloch: population estimates, local extinctions, and conservation priorities. The Raffles Bulletin of Zoology, 52(1), 271-280.

Pintea, L., Pusey, A., Wilson, M., Gilby, I., Collins, A., Kamenya, S., & Goodall, J. (2011). Long-term ecological changes affecting the chimpanzees of gombe national park, Tanzania. The Ecological Impact of Long-Term Changes in Africa’s Rift Valley. 227-247. Nova Science Publishers, Inc.

Pusey, AE., Pintea. L.,Wilson, ML., Kamenya, S., Goodall, J. (2007). The contribution of long-term research at Gombe National Park to chimpanzee conservation. Conserv Biol, 21(3), 523-534.

Surpiatna, Jatna. (2006). Conservation programs for the Endangered Javan Gibbon (Hylobates Moloch). Primate Conservation. (21), 155–162.

Supriatna, J., Tilson, R., Gurmaya, K. J., Manansang, J., Wardojo, W., Sriyanto, A., … Seal, U. (1994). Javan Gibbon and Javan Langur Population and Habitat Viability Analysis Report. Captive Breeding Specialist Group Species Survival Commission of the IUCN.

Wich, S. A., Singleton, I., Nowak, M. G., Utami Atmoko, S. S., Nisam, G., Arif, S. M., … Kühl, H. S. (2016). Land-cover changes predict steep declines for the Sumatran orangutan (Pongo abelii). Science Advances, 2(3), e1500789.

Article published by Glenn Scherer

Agriculture, Animals, Apes, Biodiversity, Chimpanzees, Conservation, Drivers Of Deforestation, Ecology, Endangered Species, Environment, Featured, Forests, Gibbons, Great Apes, Industrial Agriculture, Mammals, Orangutans, Palm Oil, Primates, Protected Areas, Rainforest People, Rainforests, Tropical Forests, Wildlife

Africa, Asia, Indonesia, Tanzania

Print