Thailand's planted durian area has more than doubled in a decade. This review asks whether that capacity is translating into dependable output, where the 2024 disruption and 2025 recovery actually occurred, and whether Vietnam's growth in China is concentrated in Thailand's own supply window.
Thailand's production base is growing, but the data does not support a simple capacity-equals-output story. The 2024 decline was driven by a broad fall in output per bearing rai. The 2025 recovery was real, but still partly dependent on additional area entering production rather than on orchards performing better.
In China, Vietnam's share gain was commercially material during Thailand's main window, and was not uniquely concentrated there. That distinction matters more than it first appears.
Standing area rose from 805,000 rai in 2015 to 1.887 million in 2025. Bearing area doubled over the same period, yet the gap between the two widened to 607,326 rai. That gap is orchard area planted but not yet producing.
The obvious reading is a supply wave arriving in four years. The flow data says otherwise. In 2025 bearing area rose 138,160 rai, but standing area rose 126,753 rai at the same time. New planting offset 91.7% of the movement into bearing, so the gap fell by only 11,407 rai.
On observed net flows the pipeline takes decades to close, not years. That is a different and more consequential story: sustained expansion rather than a single wave that passes.
| Assumption | Annual gap change | Years to close |
|---|---|---|
| Observed 2025 net flows continue | −11,407 rai | 53.2 |
| Net standing additions halve | −74,784 rai | 8.1 |
| Net standing additions stop | −138,160 rai | 4.4 |
Constant-flow sensitivities, not forecasts. Gross planting, removals, abandonment and reclassification are not separately observed.
Production rose every year from 2016 to 2023, then fell 193,556 tonnes in 2024 despite bearing area growing by 81,339 rai. Output per bearing rai dropped 19.26%, the largest fall in the selected series, and every OAE region recorded a negative output-per-rai effect even while all four had positive area effects.
2025 recovered 22.2% to 1.574 million tonnes. Splitting that change symmetrically, roughly 162,956 tonnes came from additional bearing area and 122,452 tonnes from improved output per rai. Output per rai remains 11.94% below the 2019–2023 average.
The benchmark is not biologically clean. The whole 2015–2025 series sits inside an expansion, and no orchard-age distribution is published, so the shortfall's direction is more secure than its exact size.
Production increased in 61 of 67 matched provinces, which reads as broad. By volume it was not. Chanthaburi alone generated 38.7% of gross gains; with Rayong and Chumphon, 70.7%. Five of the six declining provinces were in the South.
High province counts can overstate diversification. Anyone treating the 2025 recovery as a national return to normal is describing a small number of places.
The area and output-per-rai split is an exact accounting identity. It shows where the arithmetic sits, not why output changed. Weather, disease, tree health, management and orchard age are absent from the data and remain competing explanations.
China issued phytosanitary requirements for Vietnamese fresh durian in July 2022, so 2021 and most of 2022 are pre-access. Their near-zero recorded shares are a market-access baseline, not competitive change, and do not mean zero physical trade, since unofficial border movements would not appear in customs records.
The defensible comparison is 2023 against 2024, both fully post-protocol. Vietnam's full-year quantity share rose 12.62 percentage points. Inside April to August, Thailand's own peak window, it rose 13.43 points. The difference is 0.81 points.
The market itself grew. China's recorded import quantity rose 9.39% between 2023 and 2024. Thailand's recorded quantity fell about 12.84% while Vietnam's rose about 49.31%. That is descriptive displacement inside an expanding market, not proof that Vietnam caused Thailand's decline.
The test uses China reporter-side imports rather than Thai exports, because the question is share within one destination market. That places Thailand and Vietnam on the same customs basis.
The project set a rule in advance: a model must beat a simple benchmark in a rolling-origin backtest before anything is published. The monthly series that would make a credible forecast possible, Thai Customs exports and OAE Monthong prices, were identified but not loaded and validated within this scope. The model-ready evidence held 11 comparable annual observations and two post-protocol years of monthly import data.
That is a scope decision, not a finding that the data does not exist. Fitting a line to eleven annual points would have added the appearance of sophistication without the validation to support it.
| The data supports | The data does not establish |
|---|---|
| The non-bearing pipeline remains large and was barely reduced in 2025 | Future planting, removals or exact absorption timing |
| Output per bearing rai stayed below benchmark in 2025 | Whether lower productivity or non-producing classified area caused it |
| The 2024 output-per-rai effect was negative in all four regions | A shared causal shock |
| Vietnam gained share broadly across the calendar after access | Tonne-for-tonne displacement, or a stable long-run trend |
The public sources contain no costs, margins, contracts or firm-level transactions. The analysis cannot speak to profitability, company market share or investment returns.
ASA builds commercial intelligence from official data, with the validation record and the limitations stated openly.