Tonlé Sap: The River That Reversed Less

The Mekong River Commission's own booklet still prints the textbook figures: the lake grows sixfold, from 2,500 to 15,000 square kilometres, and from under 1.5 cubic kilometres of water to between sixty and seventy. In November 2025 a paper in Nature Sustainability found the inflow that does this had fallen 40 to 50 per cent since 1998 — and that only about eight per cent of that loss is attributable to upstream dams.

Silence Index — D-143

Popular, with quiet hours

Ranks 85 of 99 scored dossiers — quieter than 14% of the others we have scored. A low score is not a warning; it describes who else will be standing next to you.

How the total is weighted. These are the fixed weights of the four inputs, the same on every dossier — not this place’s scores. We do not publish per-input bars, because the per-input values were never recorded article by article and inventing four plausible numbers that sum to the total would be a fabrication dressed as a chart.

Footfall pressure — visitors measured against the ground that absorbs them30%

Access friction — permits, caps, seasons, boats, weather windows30%

Acoustic floor — how close the nearest engine is20%

Night darkness — how much sky is left20%

The index is a desk instrument scored by this publication, not a field measurement. How it is calculated, and where it fails.

This is a desk-researched dossier. Nobody from this publication has been on the Tonlé Sap. It is built from the Mekong River Commission's booklets, weekly flood and drought situation reports and fisheries monitoring pages; peer-reviewed hydrology in Hydrology and Earth System Sciences, Nature Sustainability, Hydrology and Water; the Stimson Center's Mekong Dam Monitor; UNESCO's Man and the Biosphere programme; SEAFDEC and Open Development Cambodia; residential care studies published by Cambodia's Ministry of Social Affairs, Veterans and Youth Rehabilitation with UNICEF; the ChildSafe Movement; the Statelessness & Citizenship Review; and Cambodian reporting where it carries an official figure. Where two sources disagree we have printed both.

What the reversal actually is

For most of the year the Tonlé Sap river runs south-east out of the lake and joins the Mekong at Phnom Penh. Then the monsoon arrives, the Mekong rises faster than the lake can drain into it, and the river turns around and runs north-west. The lake fills. Months later it drains again.

The MRC's own management booklet gives the classical numbers: area from 2,500 to 15,000 square kilometres, a sixfold increase; mean depth from one metre to six or nine depending on the year; and volume from under 1.5 cubic kilometres to between sixty and seventy. Those are the figures the institution itself publishes, and they are the ones we would use as the textbook baseline.

The season is measured, not estimated. The MRC's weekly wet-season situation report for 16–22 September 2025 records that the inflow to the lake began on 29 May 2025 and that accumulated reverse flow stood at 26.15 cubic kilometres by 22 September. It also notes that August 2025 came in at 98.32 per cent of the long-term average and above every year since 2019. Prek Kdam, on the Tonlé Sap river in Kandal province, read 7.33 metres on 16 September and 7.45 on 22 September. Kampong Luong, on the lake in Pursat, is the other station.

That 2025 detail matters for a reason people writing about the Tonlé Sap frequently miss: a good year does not disprove a bad trend, and a bad year does not prove one. 2025 was a near-average season. The decline is a thirty-year signal, not a description of last summer.

The decline, measured

Chua and colleagues published the reference study in Hydrology and Earth System Sciences in 2022, using the Prek Kdam and Kampong Luong gauges. Their headline pair: average annual reverse flow from the Mekong into the lake of 48.7 cubic kilometres in 1962–1972, against 31.7 in 2010–2018. The duration of the reverse phase fell from about 125 days to about 112.

We have to flag something here. The abstract as we read it describes that change as a decrease of 56.5 per cent. Forty-eight point seven to thirty-one point seven is a fall of roughly thirty-five per cent. The two statements do not reconcile on their face, we could not establish which pairing the 56.5 figure refers to, and rather than quietly picking the number that suited the sentence we are printing both and saying so.

The same paper reports what the change does to the lake. Comparing 2010–2019 with 1996–2009, maximum wet-season area fell 20.6 per cent and maximum wet-season volume fell 23.4 per cent, a loss of 12.1 cubic kilometres. The dry-season minimum barely moved: area down 3.1 per cent, from 2,080 to 2,010 square kilometres. And at the gauges, minimum water levels rose by up to 1.55 metres at Kratie while maximum levels fell by up to 0.79 metres at Prek Kdam. The flood is being flattened from both ends — higher lows, lower highs — which is precisely the signature of a river being regulated.

Flattening a flood pulse is not a cosmetic change to a lake. The Tonlé Sap's productivity is a function of how much floodplain gets inundated for how long, because that is where the fish feed and breed. Less amplitude means less floodplain, and less floodplain means less fish.

The attribution changed in November 2025

For fifteen years the argument about the Tonlé Sap has been dams versus drought. China's Lancang cascade upstream, the Stimson Center's Mekong Dam Monitor tracking restriction and release from satellite, the MRC being more cautious about attribution, Chinese operators disputing the whole framing. It is a genuine dispute among people acting in good faith and it is not settled.

But on 10 November 2025 a paper in Nature Sustainability moved the ground under it. Quan Le, Christopher Hackney and colleagues found that between 1998 and 2018 the reverse flow fell by 40 to 50 per cent — 47 per cent for medium-flow years — and attributed it not to dam operations but to the bed of the Mekong physically dropping, driven by sand mining and by sediment being trapped behind upstream reservoirs. Their attribution sentence is the striking one: only about eight per cent of the decline can be put down to changes in the upstream hydrograph. In medium water-flux years, total inflow to the lake falls from 45.1 cubic kilometres under 1998 conditions to 24.0 under 2018 conditions, and their projection for 2038 is 13.9 — a 69 per cent decline against 1998.

Read that carefully, because it cuts both ways. It is partly exculpatory for dam operations as operations. It is not exculpatory for dams as structures, because sediment trapping behind reservoirs is one of the two mechanisms named. And it moves a large share of responsibility to an industry that is much closer to home: dredgers working the Mekong mainstem in Cambodia and Vietnam, supplying construction sand for cities including Phnom Penh.

The Stimson Center's monitoring remains the best near-real-time record of what the dams are actually doing — sixty large dams tracked, an estimated 54.5 billion cubic metres of active storage in that set, accumulated reverse flow of 22.60 cubic kilometres in 2023 — and its own annual report is careful to say that 2023's shortfall was primarily a weak monsoon rather than dam operations alone. Nobody serious in this argument is claiming a single cause. What changed in November 2025 is which cause is largest.

The fishery, and the reform that was meant to fix it

SEAFDEC's 2025 Cambodia country profile puts roughly 70 per cent of the animal protein consumed in Cambodia as freshwater fish, and states that the Tonlé Sap supplies almost 71 per cent of the fish in Cambodian daily protein. Cambodia's total fisheries production it gives as 866,250 tonnes in 2023, down from 908,508 in 2019. The Mekong system as a whole yields around 2.1 to 2.3 million tonnes a year on MRC figures. Whichever number you take, this is not a scenic lake with a fishing industry attached. It is a national food supply that happens to be beautiful.

In 2012 Cambodia abolished the commercial fishing lot system. Open Development Cambodia records 158 lots covering 953,861 hectares cancelled, with about 98,000 hectares redesignated as conservation areas, and reports fishers netting 718,000 tonnes in 2013 against 360,000 in 2012.

That doubling should be treated with suspicion, and the peer-reviewed work does treat it that way. Chan, Ngor, Hogan, So, Brosse and Lek analysed the Dai fishery — the stationary trawl fishery on the Tonlé Sap river that the MRC and the Cambodian Fisheries Administration have monitored since 1995 — comparing 1995–2000 with 2012–2015. They found stable catch trends before the lots were abolished and decreasing catch trends after, with weaker seasonal variation. More telling still, the composition shifted: away from large-bodied, migratory and predatory species such as Pangasianodon hypophthalmus and Chitala ornata, and toward small-bodied floodplain species. That is the classic signature of a fishery being worked down the food chain.

So: abolishing the lots removed a rent-extraction system that had excluded ordinary fishers, which was a real and defensible reform. It did not stop the decline, and the reported tonnage jump and the scientific catch series point in opposite directions. Print both.

What popular coverage gets wrong

Three claims, all checkable.

"The lake grows from 2,500 to 16,000 square kilometres." The 2,500 is right. The upper figure moves depending on who you ask and, more importantly, on when. The MRC prints 15,000. SEAFDEC prints 13,000. Chua and colleagues measured the maximum wet-season extent falling by 20.6 per cent between 1996–2009 and 2010–2019, which means any single upper figure is a snapshot of a shrinking quantity presented as a constant. The 16,000 that circulates in travel writing is at the top of a range nobody has measured recently. There is a second confusion worth naming: Siev and colleagues, publishing in Hydrology in 2016, measured the Tonlé Sap river and its floodplain — a different thing from the lake — and found 123.8 to 3,251.2 square kilometres. Those numbers get quoted as lake figures and they are not.

"It is the only river in the world that reverses." It is not, and the claim is not necessary. Backwater-driven flow reversal is a documented phenomenon in large river systems: Meade and colleagues, in Environmental Geology and Water Sciences, described backwater effects in the Amazon basin propagating 300 to 800 kilometres up tributaries and inverting the normal relationship between stage and discharge on the Negro, Madeira and Purús. Tidal reversal on lower river reaches is ordinary. What is genuinely unusual about the Tonlé Sap is not the direction change but the scale and the consequence: a reversal that runs for around four months, moves tens of cubic kilometres of water, expands a lake sixfold, and underwrites the protein supply of a country. That is a better sentence than the false one, and it survives checking.

"The dams did it." As of November 2025 the best available attribution says otherwise, and we would rather change our mind in public than repeat the version everyone already believes. The dams are implicated through sediment trapping; the operations, on this analysis, account for about eight per cent. The dredger is closer to the crime scene than the turbine. This is a live scientific question and the finding is nine months old; treat it as the current best estimate rather than the last word.

The people who live on it

Around 1.38 million people live inside the Tonlé Sap Biosphere Reserve, which UNESCO designated in 1997 and which covers 1,483,339 hectares. The stilted and floating settlements are not exhibits. They are ordinary Cambodian communes with schools, pagodas, shops, clinics, debt, mobile phones and local politics, in which a substantial number of households are ethnic Vietnamese with contested legal status.

That last point is documented and it is not a footnote. Christoph Sperfeldt, writing in the Statelessness & Citizenship Review in 2021, records that from 2017 Cambodian authorities confiscated the documents of roughly 70,000 ethnic Vietnamese residents — old immigration cards, family books, in some cases Cambodian identity cards — and that a permanent-resident card introduced in 2015 costs 250,000 riel, over sixty US dollars, renewable every two years, and identifies the holder as an immigrant alien. Cambodia reported 57,444 stateless people to UNHCR by 2019. Population estimates for ethnic Vietnamese in Cambodia range from 78,090 Vietnamese speakers in the 2019 census to 400,000–700,000 in NGO estimates, which is itself a measure of how contested the question is. He also records over 10,000 people from floating villages facing relocation, beginning in Kampong Chhnang, and that relocated households cannot own the land they are moved to.

That process is continuing. On 10 July 2025 the Cambodian government established a commission chaired by the Minister of Interior to oversee relocation of floating houses and monitoring of fish farms across the five lake provinces — Battambang, Siem Reap, Kampong Thom, Pursat and Kampong Chhnang — citing pollution, biodiversity and the lake's ecology, and promising social land concessions, infrastructure, financial assistance and vocational training at the relocation sites. Whether that is delivered is the thing worth following. If you visit a floating village in 2026 you may be looking at a settlement that is scheduled to stop existing.

Boats, money, and the two things not to do

Chong Kneas is fifteen kilometres from Siem Reap at the foot of Phnom Krom and takes the great majority of the visitors. Kampong Phluk is further east and stilted rather than floating. Kampong Khleang, on the north-east shore, is the largest of the three communities and the least visited. Prek Toal, on the western side in Battambang province, is the bird colony and the serious naturalist's destination.

On price, we are going to disappoint you honestly. We could not find a published government tariff for boat hire at any of these landings. Operator quotes vary by a factor of three for what is described as the same trip, boats are priced per vessel rather than per person at some landings and per person at others, and Chong Kneas in particular has a long and well-aired reputation for pricing that changes once you are on the water. We are not going to print a figure we cannot source. Agree the total, in writing if possible, before you leave the landing, and understand that the difference between a fifteen-dollar trip and a fifty-dollar one is usually not the boat.

Two specific things. First, do not visit an orphanage or a residential care institution, on the lake or anywhere in Cambodia, and do not accept a boat detour to one. This is not a fringe position: the Ministry of Social Affairs, Veterans and Youth Rehabilitation, with UNICEF and USAID, mapped 639 residential care facilities holding 35,374 children and young people in February 2017, with 80 of those facilities in Siem Reap province alone. The same body of work found the large majority of those children have at least one living parent — a 2013 inspection put it at about 77 per cent — and states plainly that many centres turn to orphanage tourism to attract donors, fuelling a system that exposes children to risk. Cambodia's revised Alternative Care Policy of July 2024 prioritises family-based care, and UNICEF put the figure at roughly 10,000 children in about 300 facilities as of 2022, which is real progress. The ChildSafe Movement's position is one sentence: children living or studying in schools and orphanages should not be exposed to tourist visits.

Second, be sceptical of the on-water solicitation. The floating-school donation stop, the sack of rice sold at several times its market price to be given to a family, the child brought alongside your boat — these are the retail end of the same economy. Buying lunch at a village restaurant, paying a local guide properly, and paying the agreed boat price without haggling it down puts more money into the community than any of it.

Why the Silence Index is 44

Low, and it should be low. This is the most inhabited landscape in the archive. Access friction is minimal — a sealed road from Siem Reap and a boat that leaves when you pay for it, no permit, no cap, no season closed to visitors. Footfall pressure is genuinely heavy at Chong Kneas and moderate at Kampong Phluk, though it falls away fast at Kampong Khleang and Prek Toal. The acoustic floor is the real driver: long-tail diesels, outboards, generators and the ordinary noise of a working village of several thousand people carry a long way over flat water. Night darkness is the only input that scores well, and even that is qualified by Siem Reap's glow to the north. The Tonlé Sap is extraordinary and it is not quiet, and conflating those two things is how travel writing ends up describing somebody's home as an empty place.

The Traveler's Panel · checked 14 August 2026

Changing information. Prices, permits, seasons and road access change, sometimes at short notice. Everything here is a starting point for your own confirmation — not a quote, and not a guarantee. Verify with the operator and the relevant official body before you book.

The reversal	Roughly June to October. In 2025 it began on 29 May
Lake size	2,500 km² dry to 15,000 km² wet on MRC figures; measured maxima are falling
Best months to see high water	September and October, at the end of the inflow
Chong Kneas	15 km from Siem Reap. Busiest, and the most complaints about pricing
Kampong Khleang	About 50 km east. Largest community, fewest visitors
Boat cost	No official tariff published. Agree the total before leaving the landing
Do not	Visit orphanages or residential care institutions. This is the settled advice

Getting there — Siem Reap → Phnom Krom landing (Chong Kneas) · or road east to Kampong Phluk / Kampong Khleang

1. Chong Kneas is the closest landing to Siem Reap, about 15 km south at the foot of Phnom Krom, and is reached in twenty to thirty minutes by road. It is also where almost all the pricing complaints originate. If you have half a day, go further.
2. Kampong Phluk is roughly 30 km south-east of Siem Reap and is a stilt village rather than a floating one — the houses stand on poles six metres or more above the dry-season ground, which is the clearest demonstration of the flood pulse you will see anywhere.
3. Kampong Khleang, about 50 km east, is the largest of the lakeside communities and takes a fraction of the traffic. It is a working commune, not a viewpoint.
4. Prek Toal, on the western shore in Battambang province, is the bird sanctuary — reached by boat from Chong Kneas or from Battambang, and best in the dry season when waterbirds concentrate. It is a half-day boat journey, not a stop.
5. Water level dictates everything. Late in the dry season, roughly March to May, boats cannot reach some landings and the channels are shallow and dusty. Late in the wet season the same villages sit in open water.

Indicative costs — verify before booking

Boat hire	No published government tariff found. Quotes vary by a factor of three
What to do instead	Agree the total, per boat or per person, in writing, before departure
Guide	A licensed local guide is worth paying properly for and is the best value here
Rice sacks and donations	Solicited on the water at several times market price. Decline
Orphanage visits	Do not pay for one at any price. See the ChildSafe guidance

What to pack

• Sun protection for open water. There is no shade on a boat and the reflection doubles it.
• Small denominations of riel and US dollars, and enough of them that you are not negotiating change on the water.
• A dry bag. Boats take spray and the lake is not clean.
• Binoculars if you are going to Prek Toal. It is a bird colony, not a boat ride.
• A camera you are willing to keep pointed at the landscape rather than through people's doorways.

Where to stay

Almost everyone bases in Siem Reap, which has the full range from hostels to resorts and is twenty minutes from the Chong Kneas landing. A small number of community-based homestays operate at Kampong Khleang and are the only way to see the lake at dawn and dusk rather than in the middle of the day with everyone else; book through a Cambodian operator rather than an international platform. Kampong Chhnang and Pursat serve the southern and western shores and see almost no foreign visitors. Battambang is the base for Prek Toal from the west. There is no visitor accommodation on the water itself and the floating settlements are, as of July 2025, subject to a government relocation programme.

Safety & responsible travel

• Do not visit orphanages or residential care institutions and do not accept a detour to one. Cambodia's own Ministry of Social Affairs, with UNICEF, mapped 639 facilities holding 35,374 children in 2017, most of whom have a living parent, and documented orphanage tourism as a driver of the sector.
• Decline the rice-sack and floating-school donation stops. Buy lunch at a village restaurant and pay a local guide instead — the money stays and the transaction is honest.
• Photograph the landscape freely and people only with permission. These are homes with doors, not exhibits, and the boat passes within a couple of metres of them.
• Pay the agreed price without grinding it down. The pricing problem at Chong Kneas is real, but the answer is agreeing a fair total in advance, not underpaying a boat driver.
• If you want to help the lake itself, the leverage is upstream. As of November 2025 the largest measured driver of the shrinking flood pulse is riverbed lowering from sand mining in the Mekong mainstem, not anything that happens on the water in front of you.

Nearby, and quieter

Angkor is thirty minutes north of the Chong Kneas landing and is the reason almost everyone is in Siem Reap in the first place; Phnom Krom, the hill above the landing, carries a late-ninth-century temple with a view over the lake. Beng Mealea and Koh Ker lie north-east. Battambang, on the western side, is the base for Prek Toal's bird colony and for the Sangkae river. Kampong Chhnang, at the southern end, has the pottery villages and the least-visited stretch of the lake shore in the country.

Getting there, in order

• 1Siem Reap
• ↓
• 2Phnom Krom landing (Chong Kneas) · or road east to Kampong Phluk / Kampong Khleang

Each leg is a separate booking and a separate chance to be weathered out. Build spare days into the chain, not onto the end of it.

Where it is

Tonlé Sap lake and floodplain, Cambodia · 12.9000°N 104.1000°E · Open in Google Maps · OpenStreetMap

Key locations

• Phnom Krom — the hill above the Chong Kneas boat landing, 15 km south of Siem Reap
  13.2857°N 103.8138°E
• Kampong Phluk — stilt village where the houses stand six metres above the dry-season ground
  13.2121°N 103.9743°E
• Kampong Khleang — the largest lakeside community and the least visited of the three
  13.0489°N 104.1403°E
• Prek Toal — the western bird colony, a half-day boat journey from Chong Kneas or Battambang
  13.2312°N 103.6600°E
• Prek Kdam — the Tonlé Sap river gauge station whose record defines the reversal
  11.8149°N 104.8076°E
• Siem Reap — the gateway city and the base for every northern-shore trip
  13.3618°N 103.8606°E

Coordinates are for orientation and are accurate to roughly the width of the feature named, not to a doorway. Several of these points are inside protected areas or need a permit, a guide or a boat — read the access notes above before you plan a route to any of them.

Sources & further reading — checked 14 August 2026

1. Mekong River Commission — Management Information Booklet Series No. 2: The Flow of the Mekong. The MRC's own baseline figures for the reversal: lake area increasing sixfold from 2,500 km² to 15,000 km², mean depth from 1 m to 6–9 m, and volume from under 1.5 km³ to between 60 and 70 km³
2. Mekong River Commission — Weekly flood and drought situation report, 16–22 September 2025. That the 2025 reverse flow began on 29 May 2025 and had accumulated 26.15 km³ by 22 September; Prek Kdam water levels of 7.33 m and 7.45 m on those dates; and August 2025 at 98.32% of the long-term average, above every year since 2019
3. Mekong River Commission — Fisheries. Lower Mekong Basin inland capture fisheries of about 2.3 million tonnes and US$11 billion a year, total production of 4.4 million tonnes including aquaculture, and a basin population of about 70 million for whom fish is a staple
4. Mekong River Commission — Fisheries monitoring. That the MRC and Cambodia's Fisheries Administration have monitored the Tonlé Sap Dai fishery since 1995, tracking small mud carps migrating out of the lake from October to March, and that it is Cambodia's only inland industrial fishery, with over 140 years of history
5. Mekong River Commission — Mekong River Basin 2023 State of the Basin Report. That the MRC's periodic State of the Basin assessment exists and is the institution's own consolidated account of basin condition — used here to establish the publication rather than a figure, since we could not open the report file itself
6. Chua et al. — Drastic decline of flood pulse in the Cambodian floodplains (Hydrology and Earth System Sciences 26, 2022). Average annual reverse flow of 48.7 km³ in 1962–1972 against 31.7 km³ in 2010–2018; reverse phase duration falling from about 125 to about 112 days; maximum wet-season area down 20.6% and volume down 23.4% (12.1 km³) between 1996–2009 and 2010–2019; dry-season minimum area from 2,080 to 2,010 km²; and gauge changes of +1.55 m minimum at Kratie and −0.79 m maximum at Prek Kdam
7. Quan Le, Hackney et al. — Sand-mining-driven reduction in Tonle Sap Lake's critical flood pulse (Nature Sustainability, November 2025). That riverbed lowering from sand mining and upstream sediment trapping reduced reverse flow by 40–50% between 1998 and 2018 (47% for medium flows); that only about 8% of the decline is attributable to upstream hydrograph changes; and inflow of 45.1 km³ under 1998 conditions, 24.0 km³ under 2018 conditions and a projected 13.9 km³ by 2038
8. Newcastle University Press Office — Sand mining reduces Tonle Sap Lake's flood pulse (10 November 2025). The publication date of 10 November 2025, the author attribution and the researchers' own framing of the finding, including the link between global construction-sand demand and Mekong dredging
9. Siev, Paringit, Yoshimura & Hul — Seasonal Changes in the Inundation Area and Water Volume of the Tonle Sap River and Its Floodplain (Hydrology 3(4), 2016). Inundation area of the Tonlé Sap river and floodplain — not the lake — ranging from 123.8 to 3,251.2 km² with a mean of 1,028.5 km², the figures most often misquoted as lake extent
10. Chan, Ngor, Hogan, So, Brosse & Lek — Temporal dynamics of fish assemblages as a reflection of policy shift from fishing concession to co-management (Water 12(11), 2020). Dai fishery catch data for 1995–2000 and 2012–2015 showing stable catch trends before the fishing lots were abolished and decreasing trends afterwards, with a shift from large-bodied migratory and predatory species to small-bodied floodplain species
11. Stimson Center — Mekong Dam Monitor Annual Report 2023–2024. Sixty large dams tracked with about 54.5 billion m³ of active storage; accumulated Tonlé Sap reverse flow of 22.60 km³ in 2023; and the report's own caution that 2023's shortfall was primarily a weak monsoon rather than dam operations alone
12. Stimson Center — Mekong Dam Monitor project. How the Monitor works: a Stimson and Eyes on Earth partnership using Sentinel-1 virtual gauges with about ±1 m error and natural-flow modelling against MRC gauges at Chiang Saen and Vientiane to infer upstream dam manipulation
13. Stimson Center — Mekong Dam Monitor dashboard. The current near-real-time reporting format, including weekly restriction and release figures at named reservoirs and basin wetness conditions
14. UNESCO Man and the Biosphere Programme — Tonlé Sap Biosphere Reserve. Designation in 1997, total area of 1,483,339 hectares, a resident population of 1,380,615, the Stung Sen core area and the Boeung Chhmar Ramsar site
15. SEAFDEC — Fisheries Country Profile: Cambodia (2025). Around 70% of animal protein consumed in Cambodia coming from freshwater fish; the Tonlé Sap supplying almost 71% of daily fish protein; total fisheries production of 866,250 t in 2023 against 908,508 t in 2019; and the lake given as 2,500 km² dry to 13,000 km² wet
16. Open Development Cambodia — Wild capture commercial fishing and natural fisheries. The 2012 cancellation of 158 commercial fishing lots covering 953,861 ha, the redesignation of about 98,000 ha as conservation areas, and the reported jump from 360,000 t in 2012 to 718,000 t in 2013
17. MoSVY and UNICEF Cambodia — With the Best Intentions: a study of attitudes towards residential care in Cambodia (2011). 11,945 children in 269 residential care facilities in 2010, a 75% rise since 2005; that 44% were placed there by parents or extended family; and that centres turn to tourism to attract donors, including children performing for visitors
18. MoSVY, UNICEF and USAID — Mapping of residential care facilities in the capital and 24 provinces of the Kingdom of Cambodia (February 2017). 639 residential care facilities holding 35,374 children and young people; 80 facilities and 2,148 children in Siem Reap province; about 77% of children with at least one living parent on a 2013 inspection; and the statement that many centres turn to orphanage tourism, fuelling a system that exposes children to risk
19. UNICEF Cambodia — Revised Alternative Care Policy prioritises family-based care (29 July 2024). The revised policy making institutional care a last resort, and roughly 10,000 children and young people in about 300 residential care facilities as of 2022
20. ChildSafe Movement — Children are not tourist attractions. The campaign's core instruction that children living or studying in schools and orphanages should not be exposed to tourist visits, and the finding that 80% of the eight million children in institutions worldwide are not orphans
21. ChildSafe Movement — Media and resources. The catalogue of studies and campaigns underpinning the Cambodian orphanage-tourism evidence base, including the 2011 MoSVY/UNICEF attitudes study and the 2017 residential care mapping
22. UNICEF Cambodia — Living and learning on the Tonlé Sap Lake. Schooling conditions in the lake's fishing communities in Pursat province: floating primary schools reached by boat, secondary schools on the mainland, and early drop-out before the end of primary
23. Sperfeldt — Legal identity and minority statelessness in Cambodia (Statelessness & Citizenship Review 3(2), 2021). The confiscation from 2017 of documents held by about 70,000 ethnic Vietnamese residents; the 2015 permanent resident card at 250,000 riel renewable every two years and marking holders as immigrant aliens; 57,444 stateless people reported to UNHCR by 2019; population estimates from 78,090 Vietnamese speakers in the 2019 census to 400,000–700,000; and over 10,000 people from floating villages facing relocation without land ownership at the destination
24. Phnom Penh Post — Commission established to relocate floating houses from Tonlé Sap lake. The commission established on 10 July 2025 under the Minister of Interior covering Battambang, Siem Reap, Kampong Thom, Pursat and Kampong Chhnang, its stated environmental rationale, and the promised social land concessions, infrastructure and livelihood support
25. Meade, Rayol, Da Conceicao & Natividade — Backwater effects in the Amazon River basin of Brazil (Environmental Geology and Water Sciences, 1991), via USGS. Documented backwater effects propagating 300–800 km up Amazon tributaries and inverting the stage–discharge relationship on the Negro, Madeira and Purús — the evidence that annual flow reversal is not unique to the Tonlé Sap
26. US Department of State — Cambodia Travel Advisory. Level 2 as of 25 July 2025, and the specific warnings relevant to travel in Siem Reap and Kampong Thom provinces
27. GeoNames gazetteer — Tonlé Sap place records. Decimal coordinates for Kampong Khleang, Kampong Phluk, Prek Toal, Phnom Krom and Prek Kdam, used for the key-location points on the map panel

Official and scientific bodies are the authority for safety, regulatory and scientific claims. Commercial operators are used only for practical detail such as tour length, meeting points and indicative pricing — never for a safety or scientific statement.