The Nam Et-Phou Louey National Protected Area

Rasphone: Laos – or Lao PDR, Lao People’s Democratic Republic – is located in Southeast Asia in the Chinese region. And it’s surrounded by China, Vietnam, Cambodia, Thailand and Myanmar. So it’s a landlocked country and it’s quite mountainous. The capital city is called Vientiane, and the population of Laos is about 7 million people. So Lao has 18 provinces, and is very mountainous. It has a very mountainous landscape in the north of Laos. And we have over 24 protected areas – National Protected Areas, not counting the provincial protected areas or district protected areas.

Most of the large protected areas are along the border between Laos and Vietnam. The protected areas of Laos, it’s spread out from the north to the south. And in NEPL – or the full name is called Nam Et-Phou Louey National Protected Area – is one largest of the protected areas in Laos. It has the area of about 5,900 square kilometers, covering three provinces known as Houaphan, Luang Prabang and Xieng Khouang. It is located in the northern highlands of Laos and it shares the border in the northern part with Vietnam the elevation range from 400 to about 2,200 meters. And the habitat is dominated by dry evergreen and semi-evergreen forest. So the area itself is divided into two zones. So they have a Core Zone – or Total Protected Zone – and the Managed Use Zone, where livelihood activities are permitted following the National Protected Area regulations. And the characteristic of this National Protected Area in Lao is quite different from a lot of other places in the way that there’s still some people living inside the protected area. So the livelihood of the villages in the area includes gathering forest products, shifting cultivation, livestock raising, and also hunting – but subsistence hunting using, like, traditional tools; traditional methods.

[ Back to topics ]

Snares and poachers

Watkins: Snares have been used throughout the ages to trap wild animals. Traditionally, subsistence hunters in the NEPL have made their snares from rattan and other natural products that are prone to decomposing relatively quickly. On the other hand, contemporary commercial poachers more often use the brake cables of motorbikes and bicycles … which, obviously, remain dangerous for much, much longer. So Doug and I were eager for Akchou to help us understand what drives this illegal commerce.

Rasphone: Illegal hunting mainly is done by snaring. Snaring is very easy to set [up]. It’s very cheap. They can set many out in the forest. So one of the things is that they don’t discriminate any type of species. So any mammals going through the snares would be caught. And so it’s quite deadly for a lot of species. And sometimes we got them on our trip. If we found some, we would take them out. My teams would take them out. And sometimes they may target larger species, like tigers. But even that type of snare caught other type of species, other types of mammals. So poachers will put their hands on every single piece of every single animal they can get from the area. Because the bush meat demand is very high, especially in China, it would still be a challenge for any species population in this protected area, as well as any other habitats in Laos. If it’s small mammals, we get to see sometimes in certain markets. But large cats, we don’t see it being sold in Laos. It’s usually get sent to Vietnam, and probably the end market is in China, because of this high demand in tiger parts, and also for other cat species as traditional medicine.

[ Back to topics ]

Stopping snaring and poaching

Leigh: The NEPL has been widely regarded for its diverse community of carnivores. For instance, just a decade ago, it was noted as being an important site for tiger conservation in Southeast Asia. But – so long as poaching remains financially lucrative – there’s only so much that legislation and law enforcement can do to quash it. So Ryan and I were interested in learning what Akchou thinks might be the key to solving the region’s problem with snaring and poaching.

Rasphone: This particular protected area is co-managed by [the] Wildlife Conservation Society and the government counterpart, which is we have the Protected Area Management Unit. But there’s always some turnover in law enforcement staff members. Some would get trained and then moved on, and then we have another one. So the turnover rate is quite high. And then the change in law enforcement strategy is also defined by the availability of financial resources. There’s certain years that funding has become very limited, so that we have to adjust the enforcement strategy to reflect the funding situation. But primarily, law enforcement is targeting to keep people out of Core Zone area. That area should not have anyone going in there. And whether that’s going to be poachers or any other kind of illegal activities. So that’s the main purpose.

[ Back to topics ]

Field research in the NEPL

Watkins: The NEPL is a remote region in Northern Laos, and the terrain there is quite rugged. So Doug and I were curious to hear how Akchou and her team went about accessing the region for this research.

Rasphone: NEPL has an office in a small district in Houaphan province. So when I did my research, my collaboration, I collaborated with the Wildlife Conservation Society and they have office is in Vientiane. So I moved from the capital city to get to Houaphan, which involved driving 12 hours to get to that Hiam District, it’s called. And there’s an office there, a National Protected Area office there. So we would go there and arrange everything – equipment and everything – there, and then we would set off to go to the field. Which, we would have a car arranged to drop us off – depending on which area we’re going in. Either it’s going to be the nearest village or the nearest entry to the park. And then we just started walking.

But I feel quite safe in there. It’s not very scary at all. Being attacked by animals is the last thing I would think about. I was more worried about running into poachers, illegal poachers, because the type of research that I did it’s involving camera trapping. Normally we would go out for 15 days to 18 days. We would pack everything with us: food, camera traps, camping gear. We sleep on hammocks. Me and my team’s 16 guys, we would go out and then split up into four teams. And each of us would be in there for 15 days or so. So if the camera trap locations where we want to go and set up is very nearby to the spot where we picked for camping, then we would stay at the camp for two days, maybe, and then move. But sometimes we had to move our camp every single day.

[ Back to topics ]

Biodiversity as richness and evenness

Leigh: According to Robert Peet [then at] Cornell University, biodiversity has two components: the richness of species, and their evenness. As she and her team employed this distinction in interpreting their findings, Ryan and I asked Akchou to tell us more about this idea.

Rasphone: When we talk about species richness, we looking at the type of species; so different species: how many different species are there in the region, or in the area? But when you talk about species evenness, we talked about the size of the populations or the distribution of the species population. So we used the Dynamic Community Model., which was developed by Dr. Marc Kéry. He’s one of the co-authors of my paper. So the Dynamic Community Model is a form of occupancy modeling [in] which we look at, like, the percentage of the species that occupy the area. So it looks at the occupancy rate, looking at the rate of which the species occupy the area.

So this model is also known as the Multi-species Occupancy Model or Dynamic Multi-species Occupancy Model, which is written in [the] Bayesian statistical framework. So the classic statistics record frequentist statistics [and] look at probabilities, or hypothesis testing. But there’s another statistical framework, which is [the] Bayesian statistical framework. And with this one – with [the] Bayesian statistical framework – the advantage of this is that when you have a small pool of data, the Bayesian statistical framework is of advantage.

[ Back to topics ]

Documenting population density

Watkins: Archeologists may use material culture as their evidence; biologists, assays; and social psychologists, questionnaires. But in Akchou’s research, she and her team used photographic evidence to document the population density of identifiable species present in the NEPL, as she explains after this short break.

ad: Altmetric‘s podcast, available on Soundcloud, Apple Podcasts & Google Podcasts

Watkins: Here again is Akchou Rasphone.

Rasphone: For this research we use infrared camera traps, which is automatic cameras. So with infrared camera traps, they would trigger when there’s movement and there’s heat that passed in front of the cameras. Depending on the model of the camera – so we used, I think, three different models of cameras – and at each location we would put two camera traps at one location, so one station. And that’s because we want to study population density of identifiable species, meaning the species that we can identify individuals based on their patterns. For example clouded leopards or tigers. So they have distinctive patterns, so you can actually tell them apart. You can tell if this one is different from the other one. So with two camera traps, you can get both sides of the animal. For example, you have a trail and then you have one camera on the left side and then one camera on the right side of the trail. But they would be a little bit … not directly pointing at each other, but a little bit off. That’s also to ensure that if one camera missed., then the other one would still have a chance to to get the photo. Also to avoid if there’s a flash – if both cameras [are] going off at the same time – then the flash might have affected the photos. You know, you would have this white area and that will decrease the quality of the image.

[ Back to topics ]

Using infrared cameras to collect data

Leigh: As she just described, Akchou and her team collected their data through camera traps. Ryan and I followed up by asking her how she went about determining where to locate the cameras so that they stood the best chance of capturing the nocturnal movements of the last Laotian tigers.

Rasphone: The camera trap space [is] design[ed] systemically, so we have a specific spacing between each location. So, normally when we determine the sampling block, we look at the previous records of the species in the area … of mainly carnivores, because I was then targeting clouded leopards and medium-sized carnivores. And I looked at the threats level, which I used the patrol data, or law enforcement data, to inform the threats density within the area. And then the cameras will be spread out evenly, systematically, like evenly. So when I say systemic, it’s when I say “Okay, within this space, I want 80 camera trap locations. And I want each location to be apart from the other location [by] 1.5 kilometers.” So we have like 80 locations at 1.5 kilometers spacing. So these [are] used as a guidance to get to the location. So you don’t really know what it’s there in reality yet when you have this sampling block planned out. So when we actually go there to the – let’s say location number one – we use our GPS to guide us to get there, and when [we] get closer we can actually assess whether or not that location is appropriate to put the Camera traps. Is there a place to put the camera? So we searched around [the] location. It has to be within 200 [to] 400 meters of the proposed location. So if it’s like on the cliff, then we have to move it a little bit.

[ Back to topics ]

Theft of automatic cameras

Watkins: Each of Akchou’s automated photographic surveys lasted for a minimum of 50 days. But despite being in a national protected area, Doug and I figured that at least some of those cameras would have been stolen or destroyed … maybe by poachers, maybe by locals, or maybe even by the wildlife that they were deployed to document. We asked Akchou just how prevalent this kind of thing is when collecting data remotely in the forests of Laos.

Rasphone: So the setting up would take maybe 10 [to] 15 days, or maybe more. And then the camera traps will be left in the forest for up to 50 days, or a little bit more, or little bit less. But, ideally, our target was to leave the cameras in for 50 days for certain assumptions. So, you know, [with] population modeling, we have a closure assumption, where we assume that there’s no population movements happening within our survey period. So that’s why we have to limit the number of days when the cameras should be out. And so we would leave the cameras out for 50 days, and then we would go back and pick them up. But not all of them were there. We would expect at least 10 cameras to be stolen or broken. So in some areas there are very small number of cameras stolen, but in some of the blocks a lot of them got stolen. This is quite common in this type of research, especially in the area[s] where there are people living in there, and they’re also a lot of poachers and illegal activities inside.

[ Back to topics ]

False positives in the dataset

Leigh: If you’ve got a video doorbell, you’re probably familiar with the many false positives that they alert you to. Mine, for instance, has a special affinity for seeing a human face in a cluster of tree branches just beyond the front door. So, given the many images that must come from the thousands of days that Akchou’s cameras spent collectively gathering data, Ryan and I figured that there must have been many such images to wade through, and wondered how she went out doing so.

Rasphone: So after we retrieve the camera traps, we would download all the photos. And then, usually in rainy season, we would go through the images and identify which species we see in the photo[s]. We also look at how many see in the photo, and then we organize them for our analysis later. So, apart from looking at identifying the species, we also identify individuals for those species that have distinctive patterns. So it’s ah … you’re looking at some months of work. That’s probably the hardest part; not the hard … it’s not physically hard, but it’s the tedious part of the research. It’s interesting, also, right? You get to see all these beautiful images, but at the same time it’s also a lot of work to go through all those images. And sometimes you look through all this downloaded photos, and sometimes your camera just triggered because of the wind and then there’s like thousands of just … of nothing. But you just … you can’t skip them. Because you never know. Maybe within a split second something actually went past the front of the camera. So you might miss something. So you kind of have to stare at those images for a long, long time. For some images that I’m not very sure about the species – you know, there are only certain parts of the body that got taken, or it’s a bit blurry – I would send them out to some experts to help me identify them.

[ Back to topics ]

Modeling the prevalence of species in the NEPL

Watkins: With 120 to 160 cameras taking pictures every time they were triggered for 50 days or more, Akchou and her team did indeed end up with lots of pictures: about 10,000 labeled images from this study alone. So Doug and I were interested in hearing how she and her team turned all this raw data into information that they could use in modeling the prevalence of the various species in the NEPL.

Rasphone: So even though I have 10,000 images, those 10,000 images divided into species, different species, right? And then sometimes with the images, we have to break them up into independent events. So, meaning that the shot I took – let’s say four species that you cannot identify individuals so like the barking deer, for example. So with this species it’s very hard to tell if the picture that I take now is different from the other one I take 5 minutes later. So, when we prepare the data, we try to define independent event[s] based on the group of images that are [taken] 30 minutes apart. So if they are 30 minutes apart … let me give you this example. At my first camera trap location on the 1st of March, for example. I get the images of of a wild pig, and then I would have, like, maybe five images – right? – within that 30 minutes of [the] wild pig. And that is just one independent event. But if they’re a different species walking passed within the first 30 minutes, that’s another independent event, because it’s not the same species. So that means even if you have 10,000 images, you still one data point … like, you still have fewer data points when you break them up like that.

[ Back to topics ]

The extirpation of Laotian tigers

Leigh: Across the 32,000 nights that they spent outdoors in the field, Akchou’s cameras captured nearly 10,000 independent encounters of wildlife. And from these records, they detected 43 different species of animals. But neither tigers nor leopards were photographed after the 14 times they were captured in 2013, the first year of their survey. This led Akchou to conclude that the species has effectively been “extirpated” in Laos, which is a scientific way of saying that they’d been completely wiped out … as she explains next.

Rasphone: So the 14 events, they’re are only from two tigers, right? We would get them in a few different locations, but maybe in one location there are few shots of a single tiger. So when we actually look at the patterns, we were able to tell that they’re two different tigers, one male and one female. Then the following years, we didn’t get any. It was a huge reduction. So [our] assumption was that if they are still in the area – and I mean these species, if there are there they are ready to be photographed. So these two tigers were the same individuals that were trapped in 2009, but I was not able to compare with any other photographs that were recorded during Johnson’s time. So Dr. Eileen Johnson, she was at the time the country director of [the] Wildlife Conservation Society, Loas program. And I started working with [the] Wildlife Conservation Society in late 2005. So that was the time [when] she was the director. She started the systematic camera trapping to look at the tiger population in the area. But, definitely, there were a lot more records then compared to when I started my research.

[ Back to topics ]

Hope for tigers in the future

Watkins: Despite all of this, Doug and I were curious what hope Akchou might have for the future of big cats in the Nam Et-Phou Louey National National Protected Area.

Rasphone: The pattern here – as much as I like to look at it in the ecological context – it’s also influenced by human factors, which is the poaching. But, we found that some of the prey species are still quite widely distributed, but tigers are not there. So prey depletion, especially the large prey, may have played a role in the disappearing of tigers, but the major issue here is illegal hunting for trade. So these findings from my studies should be taken as lessons learned, and also as information or guidance to where we are at this stage in terms of our biodiversity. And despite the big cats are likely to be extirpated from the area, we still have a lot of other species that represent the diversity of carnivores and mammals and other prey species in this area. And those species should not be neglected. And I believe that if we are able to conserve or to protect this area to be free from threats, [then] we can bring the condition of the area to be back to where it was 10 [or] 20 years ago. And would still be a good site for supporting carnivore population. And perhaps, who knows, maybe one day tigers [will be] wandering back there from somewhere. And it could also be a suitable habitats. Although we don’t have tigers there [now], but that doesn’t mean that it’s not possible to bring up the population back to the area.

[ Back to topics ]

Links to manuscript, bonus audio and other materials

Leigh: That was Akchou Rasphone, discussing her article “Documenting the demise of tiger and leopard, and the status of other carnivores and prey, in Lao PDR’s most prized protected area: Nam Et – Phou Louey,” published in October 2019 with Marc Kéry, Jan Kamler, and David Macdonald in the journal Global Ecology and Conservation. You’ll find a link to their open-access paper at parsingscience.org/e72, along with transcripts, bonus audio clips, and other materials that we discussed during the episode.

Watkins: We hope that Parsing Science helps you hear what you might not have the time to read. And if you’re new to the show – or just missed a few of our recent episodes – then head over to parsingscience.org to check out our entire catalogue. There, you’ll find our conversation with our guest from the previous episode, Verónica Sevillano, who spoke with us about why it is that we treat different animal species so differently … as well as the episode before that, in which we spoke with Jeremy Gunawardena about that idea that even prehistoric single-celled organisms can be taught to learn new tricks.

[ Back to topics ]

Preview of next episode

Leigh: Next time, in episode 73 of Parsing Science, we’ll hear from Courtney Coughenour and Jennifer Pharr, from the University of Nevada, Las Vegas, about their research into what differentiates the likelihood that drivers – in their cars – will yield to pedestrians of different races and genders.

Pharr: Looking at the make and model of the car from the video, and then using Kelly Blue Book to get the estimated value of the car … and then, from there, we did the statistical analysis to see if in fact it was true: if people who drove more expensive cars were less likely to yield for pedestrians who were attempting to cross the street.

DL: We hope that you’ll join us again.

[ Back to topics ]