Skip to content
LazaarWorks
ENFRع

Research brief · Agriculture × AI · November 2026

Feeding more people with less water.

Farming already takes 71% of the freshwater the world withdraws. Yet 645 million people went hungry in 2025, up to 40% of crops are lost to pests and 13% of food is lost before it reaches a shop. This brief maps where the losses are — in Tunisia first, then in the countries that shape world food supply — and which AI and IoT tools have the evidence to cut them.1475

71%

of freshwater withdrawals go to agriculture (world, 2022)1

645M

people faced hunger in 2025 — 7.8% of humanity4

13.3%

of food is lost between harvest and retail (2023)5

346 m³

of renewable water per Tunisian per year — below the 500 m³ absolute-scarcity line1

Water stress (withdrawals ÷ available freshwater)1

98.1%Tunisia · 2022

Drag to turn the globe. Bars show the selected indicator for the ten benchmark countries; Tunisia is in orange. Tap a bar for its value.

01

Six pressures on the world’s food supply

Each one is measured by a UN agency or a national statistics office. Together they explain why producing more food is no longer a question of planting more.

Water is the binding constraint

Agriculture accounts for 71% of global freshwater withdrawals. Irrigated land is only 20% of the cultivated area but grows 40% of the world’s food, so every drop lost in a canal or a flooded furrow is food not grown. Surface irrigation delivers about 60% of the water to the crop; drip about 90%.1910

Share of applied water that reaches the crop, by irrigation method10

Surface (furrow, basin, flood)60%
Sprinkler75%
Drip90%

Food lost before it is sold

13.3% of food is lost after harvest — on farms, in storage, transport and processing — before it reaches retail. Then 1.05 billion tonnes, 19% of the food available to consumers, were wasted in shops, restaurants and homes in 2022, 60% of it by households. Perishables suffer most: fruits and vegetables lose 25.4%.56

Food lost between harvest and retail, by product group (2023)5

Fruits & vegetables25.4%
Meat & animal products14.0%
Roots, tubers & oil crops12.3%
Cereals & pulses8.4%

Sub-Saharan Africa 23.0% · Eastern & South-Eastern Asia 13.9% · Northern America & Europe 10.0%

Pests and diseases

FAO estimates that up to 40% of global crop production is lost to pests every year. Plant diseases cost the world economy more than US$220 billion a year and invasive insects at least US$70 billion — and warming is pushing pests into new regions.7

up to 40%

of crop production lost to pests each year

>US$220B

yearly cost of plant diseases

≥US$70B

yearly cost of invasive insects

Hunger is falling — slowly

645 million people (7.8%) faced chronic hunger in 2025, down from 8.6% in 2022 — the third straight year of decline. But 2.1 billion people live with moderate or severe food insecurity, 2.7 billion cannot afford a healthy diet, and Africa now has more hungry people (309 million) than Asia. At this pace 510–520 million will still be hungry in 2030.4

645M

people in chronic hunger (2025)

2.1B

moderately or severely food insecure

2.7B

cannot afford a healthy diet

A third of the world’s emissions

Agrifood systems emitted 16.5 Gt CO₂-eq in 2023, 32% of all human emissions. Half comes from the farm itself (livestock, fertilizer, rice, on-farm energy), a third from what happens before and after it, a fifth from clearing land. Cutting losses and over-fertilisation cuts emissions at the same time.8

Agrifood-system emissions by stage, 2023 (Gt CO₂-eq)8

  • Farm gate (crops, livestock) — 49%
  • Before & after the farm (energy, processing, transport, waste) — 32%
  • Land-use change (deforestation) — 19%

More mouths, thinner margins

World population reached 8.2 billion in 2025 and is projected to peak at about 10.3 billion in the mid-2080s. Most of the extra food will have to come from land and water already in use — through more output per hectare and per cubic metre.112

8.2B

people in 2025

≈10.3B

projected peak, mid-2080s

02

Critical products: who grows the world’s food

Eleven products that feed the world and matter for Tunisia. Pick one to see the leading producers in 2024, where Tunisia stands, and the AI lever that fits.

Olives — Leading producers, 2024 (tonnes)2

Spain8.3M t · 32.5%
Türkiye3.8M t · 14.7%
Italy2.3M t · 9.0%
Portugal1.3M t · 5.2%
Egypt1.3M t · 5.0%
Morocco1.1M t · 4.2%
Tunisia1.1M t · 4.1%
Algeria913.2K t · 3.6%
Show the data
CountrytonnesShare
Spain8,310,25032.5%
Türkiye3,750,00014.7%
Italy2,299,6609.0%
Portugal1,340,8205.2%
Egypt1,269,7965.0%
Morocco1,079,8744.2%
Tunisia1,056,2284.1%
Algeria913,2433.6%
World25,551,898100%

World production, 2024

25.6M t

Tunisia’s rank

#7

of 58 producing countries

Tunisia, 2012–2024 (tonnes)

Why it matters

Tunisia’s flagship crop: about 1.96 million ha and 102 million trees, mostly rain-fed — hence harvests that swing from 376,000 t to 2 Mt.217

Where AI and IoT help

Olive-fly camera traps, deficit-irrigation scheduling, harvest and oil-yield forecasting for exporters.

03

Tunisia: a farm economy running out of water

Agriculture is 10.3% of Tunisia’s GDP and 12.5% of its jobs, and it uses three quarters of the country’s water. Tunisia now withdraws 98% of its renewable freshwater every year, so each extra tonne of food has to be won through efficiency.1

10.3%

of GDP from agriculture, forestry & fishing (2025)1

12.5%

of jobs are in agriculture (2025)1

75.5%

of water withdrawals go to agriculture (2022)1

98.1%

of renewable freshwater withdrawn each year (2022)1

Water: below the absolute-scarcity line

Renewable internal freshwater per person, 2022 (m³/year)1

Egypt9 m³
Israel80 m³
Tunisia346 m³
Netherlands621 m³
Morocco777 m³
India1,014 m³
China1,992 m³
Spain2,327 m³

World average: 5,393 m³. Egypt, Israel and the Netherlands look low because they rely on water flowing in from abroad (the Nile, the Rhine) or on desalination.

National dam fill rate, as reported (%)2021

Total capacity is about 2.3 billion m³ (computed from 1,075 million m³ = 46%). ONAGRI stopped its daily dam bulletin in October 2025.

Show the data
Date%
20 Sept 202422.7%
20 Jan 202531.4%
28 May 202541.1%
25 Aug 202530.9%
10 Oct 202527.4%
23 Jun 202660.0%
23 Oct 202646.0%
  • Deep aquifers are pumped at 133% of their renewable rate (2020, up from 119% in 2015), and groundwater supplies 55% of the irrigated area.22
  • Irrigated perimeters cover 442,500 ha (2023) — about 8% of farmland — yet produce around 37% of the value of agricultural output.2324
  • ONAS treated 295.7 million m³ of wastewater in 2024, but only 11.3 million m³ (3.8%) went to agriculture. Israel uses about 86% of its treated wastewater for irrigation. Tunisia’s target: over 80% reuse by 2050.253126
  • Seawater desalination reached about 250,000 m³ a day when the Sousse plant opened in September 2026 (with Djerba, Zarat and Sfax) — roughly 91 million m³ a year.27

Cereals: a structural import dependence

In 2023 drought cut the wheat harvest to 470,000 t while imports reached 1.89 Mt — about 80% of the wheat supply. Harvests recovered to about 1.7 Mt of cereals in 2025 and an estimated 1.8 Mt in 2026, yet wheat import needs for 2026/27 are still forecast at about 1.8 Mt.2328

Wheat in Tunisia: domestic production vs imports (tonnes)23

  • Production
  • Imports

80% = imports ÷ (production + imports), 2023, computed. Imports for 2024 are not yet in FAOSTAT.

Show the data
YearProductionImports
20121,523,0002,114,000
2013975,0001,868,000
20141,513,0002,165,000
2015913,0002,475,000
2016927,0002,834,000
20171,104,0002,498,000
20181,075,0002,220,000
20191,455,0002,361,000
20201,042,0002,437,000
20211,193,0002,368,000
20221,135,0002,158,000
2023470,0001,889,000
20241,150,000—

Olive oil: world class, at the mercy of rain

Tunisia produced 340,000 t in 2024/25 (+55%, third in the world behind Spain and Türkiye), and about 450,000 t is forecast for 2025/26 — enough to rank second. Over ten months of 2025/26, exports reached 381,200 t worth 4.78 billion TND, 56.6% of it sold to the EU. Yet between 2010 and 2023 output swung between 112,000 and 400,000 t: on rain-fed groves, yield is weather.1314152

Olive-oil production in Tunisia, calendar years (tonnes)2

Show the data
Yeartonnes
2010174,600
2011112,400
2012192,600
2013188,500
2014179,700
2015301,400
2016140,800
2017188,600
2018278,300
2019239,500
2020399,900
2021144,200
2022235,200
2023210,000

Largest olive-oil producers, 2024/25 crop year (tonnes)13

Spain1.4M t · 40%
Türkiye505K t · 14%
Tunisia340K t · 10%
Greece250K t · 7%
Italy248K t · 7%
Portugal177K t · 5%

World total: 3.57 Mt (provisional).

Dates: growing, exporting — and exposed

Production doubled since 2012 to 400,000 t in 2024, and the 2025/26 harvest is estimated at about 404,000 t. Exports passed 1 billion TND for the first time in 2025/26. Two risks hang over the oases: aquifers pumped beyond their recharge (133% for deep aquifers nationally in 2020) and the red palm weevil, recorded in Tunisia since 2011.219182230

Dates in Tunisia: production and exports (tonnes)23

  • Production
  • Exports
Show the data
YearProductionExports
2012192,000101,000
2013195,000106,000
2014199,00087,000
2015223,000103,000
2016241,000114,000
2017260,000104,000
2018305,000126,000
2019289,000114,000
2020332,000119,000
2021345,000120,000
2022369,000130,000
2023387,150130,000
2024400,219—

04

Benchmark: what the leaders do differently

Ten countries — Tunisia, its neighbours, and the giants and pioneers of world agriculture — on the same World Bank indicators. Choose an indicator.

Water stress (withdrawals ÷ available freshwater) (%)1

SDG 6.4.2. Above 100% = drawing more than is renewably available (fossil groundwater, desalination, the Nile from upstream).

Egypt141.2%
Israel132.0%
Tunisia98.1%
India66.5%
Morocco50.8%
Spain43.3%
China41.5%
United States28.2%
Netherlands16.7%
Brazil1.5%
Show the data
CountryValueYear
Egypt141.2%2022
Israel132.0%2021
Tunisia98.1%2022
India66.5%2022
Morocco50.8%2022
Spain43.3%2022
China41.5%2022
United States28.2%2022
Netherlands16.7%2022
Brazil1.5%2022

One lesson from each country

Tunisia

Withdraws 98% of its renewable water each year; its cereal yield (1.9 t/ha) is less than half the world average (4.2 t/ha).1

Morocco

About 850,000 ha equipped with drip irrigation by 2024, with a target of 1 million ha by 2030 — yet agriculture still takes 88% of water withdrawals.331

Egypt

Withdraws 141% of its own renewable water — it depends on the Nile flowing in from upstream — and is the world’s second date producer (1.75 Mt).12

Spain

The world’s first olive-oil producer: 1.42 Mt in 2024/25, about 40% of world output.13

Israel

Uses about 86% of its treated wastewater for irrigation — the highest share in the world — and creates US$250 of GDP per m³ withdrawn, 20 times Tunisia.311

Netherlands

Exported €137.5 billion of agricultural goods in 2025 (€88.4 billion home-grown), while farming takes just 3% of the country’s water withdrawals.321

United States

Auto-steer guided 58% of corn acres by 2016; camera-guided See & Spray covered 5 million acres in 2025 with about 50% less herbicide.3536

Brazil

Grows 36% of the world’s soybeans and 23% of its oranges; food is 41% of the country’s merchandise exports.21

India

Pumps 251 km³ of groundwater a year — more than a quarter of the world total, 89% of it for farming — and agriculture employs 42% of workers.341

China

Grows 21% of the world’s cereals and a third of its tomatoes, using 393 kg of fertilizer per hectare — almost three times the world average.21

05

Eight AI + IoT solutions with evidence behind them

Placed by the impact they could have in Tunisia and how feasible they are for a software studio working with hardware partners. The placement is our assessment; every evidence line is sourced.
  • Proven at scale
  • Scaling
  • Pilot stage
Our assessment
Feasibility for us
Impact in Tunisia
Medium
High
Very high
Very high
Groundwater and dam digital twin
Smart irrigation scheduling
High
Pest early-warning networkYield and harvest forecasting
Cold-chain and storage monitoringFarmer assistant in Tunisian Arabic and French
Medium
Crop-disease diagnosis from a phone photoQuality grading and traceability
Proven at scaleImpact in Tunisia 5/5 · Feasibility for us 5/5 (Our assessment)

Smart irrigation scheduling

Problem
Farmers irrigate by habit or by water-turn schedule, not by what the soil and the crop need.
How it works
Soil-moisture probes and flow meters, local weather and satellite evapotranspiration (FAO WaPOR at 100 m; 20 m over Jendouba and Kairouan) feed a model that tells each farm when and how much to irrigate — by WhatsApp, or by opening the valve directly.
Evidence
Sensor-based systems saved 20–41.5% of water versus traditional irrigation in published trials; drip delivers about 90% of the water to the crop against about 60% for surface irrigation. Caveat: savings only stay in the aquifer if allocations are capped — otherwise farmers expand the irrigated area.38101145
Where in Tunisia
Irrigated olive groves (Sfax, Kairouan), date oases (Kébili, Tozeur), vegetable perimeters (Cap Bon, Sidi Bouzid).
What we would build
Farm dashboard and WhatsApp advisor, LoRaWAN sensor kit, irrigation model calibrated per crop.
Success metric
m³ per tonne harvested; pumping kWh per hectare
ScalingImpact in Tunisia 5/5 · Feasibility for us 3/5 (Our assessment)

Groundwater and dam digital twin

Problem
Deep aquifers are pumped at 133% of their recharge and the national dam bulletin stopped in October 2025: decisions are made without live data.
How it works
Connected meters on wells, piezometers and dam telemetry stream into one model that forecasts reserves, flags over-pumping and simulates allocation rules for regional agriculture offices (CRDA) and water-user associations (GDA).
Evidence
Tunisia’s dam fill swung from 22.7% (September 2024) to about 60% (June 2026) and back to 46% (October 2026); deep-aquifer exploitation rose from 119% (2015) to 133% (2020). FAO’s open WaPOR water data already covers the whole country at 100 m.20212245
Where in Tunisia
Kairouan, Sidi Bouzid and Kébili aquifers; the northern dam system.
What we would build
Data platform, forecasting models and a public dashboard, built with the Ministry and the CRDAs.
Success metric
Forecast error on reserves; volume of unmetered pumping detected
ScalingImpact in Tunisia 4/5 · Feasibility for us 4/5 (Our assessment)

Pest early-warning network

Problem
Up to 40% of crops are lost to pests, and sprays are applied by calendar rather than by actual pest pressure.
How it works
Camera traps count olive flies automatically; acoustic sensors hear red-palm-weevil larvae inside trunks; a model turns counts and weather into a treat / don’t-treat alert per zone.
Evidence
A low-cost electronic trap identified the olive fly with 89–94.5% accuracy; fibre-optic acoustic sensing detected weevil larvae under three weeks old with about 97% accuracy; camera-guided spot spraying cut herbicide use by about 50% across 5 million acres in 2025.4041367
Where in Tunisia
The olive belt (Sfax, Sousse, Mahdia); date oases; ornamental palms in Greater Tunis, where the weevil was first found in 2011.
What we would build
Trap and sensor network, image and audio models trained on Tunisian data, regional alert map.
Success metric
Treatments per season; share of fruit damaged at harvest
Pilot stageImpact in Tunisia 3/5 · Feasibility for us 4/5 (Our assessment)

Crop-disease diagnosis from a phone photo

Problem
Diseases are spotted late, and extension agents cannot visit every farm.
How it works
The farmer sends a photo; a vision model proposes the likely disease and the recommended action, and uncertain cases go to an agronomist.
Evidence
A deep-learning model reached 99.35% accuracy on lab images — but just above 31% on photos taken in other conditions. The lesson: a model is only as good as its local field data.39
Where in Tunisia
Tomato and potato perimeters, citrus orchards, olive groves.
What we would build
A Tunisian field-image dataset built with research institutes, a WhatsApp diagnosis bot and an agronomist review queue.
Success metric
Accuracy on Tunisian field photos; days from first symptom to treatment
ScalingImpact in Tunisia 4/5 · Feasibility for us 4/5 (Our assessment)

Yield and harvest forecasting

Problem
Import tenders and export contracts are decided before anyone knows the size of the harvest.
How it works
Sentinel-2 imagery (10 m, every 5 days), rainfall and soil moisture feed crop models that estimate cereal and olive output by region weeks before harvest.
Evidence
Tunisia’s cereal harvest went from 612,000 t (2023) to 1.5 Mt (2024) and about 1.8 Mt (2026 estimate), while wheat import needs stay near 1.8–2 Mt a year: every early, accurate estimate weighs on the import bill.22846
Where in Tunisia
The north-west cereal belt (Béja, Jendouba, Siliana, Le Kef); olive regions, for oil exporters.
What we would build
A regional forecast service with monthly reports for public buyers, mills and exporters.
Success metric
Forecast error 60 days before harvest
Proven at scaleImpact in Tunisia 4/5 · Feasibility for us 5/5 (Our assessment)

Cold-chain and storage monitoring

Problem
Fruits and vegetables lose 25.4% of their volume between harvest and retail; lack of refrigeration alone costs 12% of world food production.
How it works
Low-cost temperature and humidity loggers in stores and trucks; a model predicts remaining shelf life and alerts before a batch spoils.
Evidence
UNEP and FAO estimate that 526 million tonnes of food (12% of production) are lost for lack of refrigeration, and that developing countries could save 144 million tonnes a year with cold chains at developed-country levels.425
Where in Tunisia
Date packing houses, citrus and vegetable exporters, wholesale markets.
What we would build
A plug-and-play logger kit, a shelf-life model and alerts in the exporter’s app.
Success metric
Share of batches rejected; tonnes saved per season
ScalingImpact in Tunisia 3/5 · Feasibility for us 4/5 (Our assessment)

Quality grading and traceability

Problem
Spain alone bought 31.9% of Tunisia’s olive-oil exports in 2025/26: proving quality and origin is how more of the value stays in Tunisia.
How it works
Cameras on packing lines grade dates by size, colour and defects; lab and sensor data predict olive-oil quality; each lot gets a digital record from grove to buyer.
Evidence
Olive-oil exports reached 4.78 billion TND in ten months of 2025/26 and date exports passed 1 billion TND: each extra 1% of value on olive oil alone is worth about 48 million TND.151618
Where in Tunisia
Date packing houses (Kébili, Tozeur), olive mills and bottlers (Sfax, Sousse).
What we would build
A vision grading module for packing lines and a traceability portal for buyers.
Success metric
Share sold packaged or branded; price premium per kg
Pilot stageImpact in Tunisia 4/5 · Feasibility for us 5/5 (Our assessment)

Farmer assistant in Tunisian Arabic and French

Problem
75% of Tunisian farms are under 10 ha — far too many for agronomists to visit one by one.
How it works
A voice and text assistant on WhatsApp answers in Darija or French from validated sources (ministry guides, research institutes, our sensor data) and hands hard questions to a human expert.
Evidence
Digital Green’s Farmer.Chat reached 15,000+ farmers and 300,000+ questions in four countries; training on domain data raised fact recall from 26.2% to 50.3%. A generic chatbot is not enough — local validation is what works.434429
Where in Tunisia
Nationwide, starting with one value chain (olive or dates) and one region.
What we would build
A retrieval-augmented assistant with Darija speech, a curated knowledge base and an expert-in-the-loop dashboard.
Success metric
Share of answers validated by experts; weekly active farmers

06

Impact simulator: the water smart irrigation could free

Agriculture withdrew about 2.9 billion m³ in Tunisia in 2022 (75.5% of 3.86 billion m³). Move the sliders to see what an adoption rate and a saving per farm would mean, next to the country’s other water sources.1
30%
5%100%
20%
10%40%

Water freed each year

175 million m³

  • 1.9× the yearly output of all seawater desalination plants
  • 16% of what the dams held on 23 October 2026

Formula: 2,916 million m³ × share equipped × saving. Assumes equipped farms use the national average amount of water per hectare.1

Compared with other sources (million m³ per year)272521

Freed by smart irrigation175 million m³
Wastewater treated by ONAS (2024)295.7 million m³
All seawater desalination plants91 million m³
Treated wastewater used in farming (2024)11.3 million m³

The published range for sensor-based savings is 20–41.5%, so 20% is a cautious default. Savings only refill aquifers if allocations are capped at today’s level; otherwise the freed water tends to irrigate new land.3811

computed: 250,000 m³ × 365 = 91.3 million m³

07

The market: where the money and the openings are

Agriculture adds US$4.75 trillion to the world economy (4% of GDP), yet AI for agriculture is still a market of a few billion dollars — young, fragmented and growing fast.14849

US$16.2B

agrifood-tech venture funding in 2025, flat on 202447

US$9B · +7%

went to upstream startups (farms and food production)47

−12%

fewer deals than in 202447

AI-in-agriculture market forecasts (US$ billion)4849

  • 2024 estimate
  • 2030 forecast
BCC Research48CAGR 25.1%
2.208.50
TechSci Research49CAGR 22.7%
1.424.84
09 US$ billion

Two firms, two definitions: 2030 forecasts range from US$4.8 B to US$8.5 B. Read them as a direction, not a target.

What it means for us

  • Paying customers exist in Tunisia: olive-oil exporters (4.78 billion TND in ten months), date packers (over 1 billion TND), and the State, which still has to import about 1.8 Mt of wheat a year.151828
  • The data is open: satellite water data (WaPOR, 20 m over Jendouba and Kairouan), imagery (Sentinel-2, 10 m every 5 days) and World Bank / FAO statistics can all be built on.45461
  • Investors are favouring upstream, on-farm solutions (+7% in 2025 while overall funding stayed flat) — exactly where water and pest tools sit.47
  • Our role is the software layer — data platform, models, apps and WhatsApp channels in Arabic, French and English — with sensor makers, research institutes and regional agriculture offices covering the field.

08

Roadmap: start with water

Three phases, each one paying for the next. Water comes first because it is Tunisia’s binding constraint and the evidence for savings is strongest.

  1. 01 · 0–6 months

    Prove it on two farms

    • Smart-irrigation pilot on one irrigated olive grove and one date oasis
    • Sensor kit + WaPOR and Sentinel-2 data + WhatsApp advice
    • Measured baseline: m³ per tonne, pumping energy, yield

  2. 02 · 6–18 months

    Add the second layer

    • Olive-fly and red-palm-weevil early-warning network
    • Darija / French farmer assistant with expert validation
    • Cold-chain loggers with one date or citrus exporter

  3. 03 · 18–36 months

    Go regional

    • Groundwater and dam digital twin with the Ministry and CRDAs
    • Regional yield forecasts for cereal imports and oil exports
    • Quality grading and traceability for exporters

Who to talk to first

  • Ministry of Agriculture — water and irrigation departments
  • CRDAs (regional agriculture offices)
  • ONAGRI (national agriculture observatory)
  • INRAT and the Olive Tree Institute
  • ONH (national olive-oil board) and GIFruits
  • GDAs (water-user associations)
  • FAO Tunisia

09

Sources and method

Every figure links to its source. We used primary statistics first — the World Bank’s World Development Indicators (updated 8 October 2026) and FAOSTAT (release of 25 February 2026, data year 2024) — then UN agency reports, peer-reviewed studies and official Tunisian figures. Where ONAGRI or a ministry only published through the press, the article is cited and marked “press”.

Caveats: FAOSTAT counts calendar years while the International Olive Council counts crop years (October–September), so their olive-oil figures differ. Country indicators use each country’s latest available year (2021–2025). Market-size forecasts from research firms rely on different definitions and are shown as a range. The placement of solutions is LazaarWorks’ own assessment.

All sources

  1. 1World Development Indicators (API), indicators ER.H2O.FWAG.ZS, ER.H2O.FWST.ZS, ER.H2O.INTR.PC, ER.GDP.FWTL.M3.KD, ER.H2O.FWTL.K3, AG.YLD.CREL.KG, AG.CON.FERT.ZS, NV.AGR.TOTL.ZS, NV.AGR.TOTL.CD, SL.AGR.EMPL.ZS, SN.ITK.DEFC.ZS, TX/TM.VAL.FOOD.ZS.UNWorld Bank · Updated 2026-10-08 · Primary dataLatest year available per country (2021–2025). Water indicators come from FAO AQUASTAT; undernourishment from FAO, shown as “<2.5%” below the floor FAO reports.
  2. 2FAOSTAT — Production: Crops and livestock products (data year 2024), via Our World in DataFAO · Release 2026-02-25 · Primary dataCalendar-year volumes in tonnes. Olive oil is not in this release after 2023.
  3. 3FAOSTAT — Food Balances: imports and exports (to 2023), via Our World in DataFAO · Release 2026-02-25 · Primary data
  4. 4The State of Food Security and Nutrition in the World 2026 (SOFI)FAO, IFAD, UNICEF, WFP, WHO · July 2026 · Official report
  5. 5SDG indicator 12.3.1(a) — Global Food Loss IndexFAO · 2025 (data year 2023) · Primary data
  6. 6Food Waste Index Report 2024UNEP & WRAP · 2024-03-27 · Official report
  7. 7Climate change fans spread of pests and threatens plants and crops — scientific reviewFAO · 2021 · Official report
  8. 8FAOSTAT Analytical Brief 115 — Greenhouse gas emissions from agrifood systems, 2001–2023FAO · October 2025 · Official report
  9. 9The State of the World’s Land and Water Resources for Food and Agriculture 2021 (SOLAW) — facts and figuresFAO · 2021 · Official report
  10. 10Irrigation Water Management, Training Manual 4 — Annex I: Irrigation efficienciesFAO · 1989 · Official report
  11. 11Does improved irrigation technology save water? A review of the evidenceFAO · 2017 · Official report
  12. 12World Population Prospects 2024 — press releaseUN DESA · 2024-07-11 · Official report
  13. 13Olive sector statistics, December 2025, and forecastsInternational Olive Council · December 2025 · Primary dataCrop years (October–September). 2024/25 figures are provisional.
  14. 14Tunisia set to become the world’s second-largest olive oil producer in 2025/26CMB, citing the IOC · 2025 · Industry
  15. 15Huile d’olive : des recettes records pour la saison 2025/2026La Presse (ONAGRI data) · 2026-09-22 · Press (official figures)
  16. 16Huile d’olive : la Tunisie exporte vers plus de 82 paysLa Presse (ONAGRI data) · 2026-09-23 · Press (official figures)
  17. 17Indicateurs de l’huile d’olive en TunisieAPIA (DGPA data) · 2021 · Official report
  18. 18Dattes : les exportations tunisiennes dépassent pour la première fois le milliard de dinarsLa Presse · 2026-09-26 · Press (official figures)
  19. 19Tunisie | Récolte de dattes en hausse de 16,3 %Kapitalis · 2026-03-12 · Press (official figures)
  20. 20Dam fill rates: Managers 2024-09-20 · AllAfrica 2025-01-20 · Managers 2025-05-28 & 2025-08-26 · La Presse 2025-10-11 · Business News 2026-06-23Tunisian press (ONAGRI / Ministry of Agriculture data) · 2024–2026 · Press (official figures)ONAGRI stopped publishing its daily dam bulletin in October 2025, so recent points come from Ministry statements relayed by the press.
  21. 21Barrages : réserves d’environ 1,075 milliard de m³ au 23 octobre 2026Tuniscope (Ministry of Agriculture data) · 2026-10-24 · Press (official figures)
  22. 22Groundwater exploitation rates and irrigated area supplied by groundwater in TunisiaCahiers Agricultures (citing MARHP 2021) · 2024 · Research paper
  23. 23Déficit hydrique, agriculture irriguée : le recours aux eaux usées traitées est devenu une nécessitéTunisie Numérique (observatory data) · 2024 · Press (official figures)
  24. 24Irrigation project kicks off in TunisiaFruitnet · 2019 · Press (official figures)
  25. 25Tunisie | Vers une utilisation sûre des eaux usées traitées en agricultureKapitalis (Rapport national sur l’eau 2024) · 2026-06-23 · Press (official figures)
  26. 26Vers la réutilisation de plus de 80 % des eaux usées traitées à l’horizon 2050AllAfrica / La Presse · 2026-08-11 · Press (official figures)
  27. 27Une station de dessalement de 50 000 m³ par jour inaugurée à SousseLa Presse · 2026-10-16 · Press (official figures)Djerba (50,000 m³/day), Zarat (50,000) and Sfax (100,000) were already running; Sousse brings the total to about 250,000 m³/day.
  28. 28Country Brief — TunisiaFAO GIEWS · 2026-08-26 · Official report
  29. 29Fiche pays Tunisie — les politiques agricoles à travers le mondeFrench Ministry of Agriculture · 2014 · Official report
  30. 30Le charançon rouge tue en silence nos palmiersKapitalis · 2016-01-18 · Press (official figures)
  31. 31Israel bets on recycled water to meet its growing thirstNPR / WFDD (Israel Water Authority figures) · 2023 · Press (official figures)
  32. 32Value of agricultural exports up by over 8 percent in 2025CBS Statistics Netherlands & WUR · January 2026 · Official report
  33. 33Morocco to expand water-saving irrigation to one million hectares by 2030Morocco World News (Ministry of Agriculture figures) · 2024-07 · Press (official figures)
  34. 34UN World Water Development Report 2022 — Groundwater: making the invisible visibleUN-Water / UNESCO · 2022 · Official report
  35. 35Precision Agriculture in the Digital Era: Recent Adoption on U.S. Farms (EIB-248)USDA Economic Research Service · 2023 · Official report
  36. 36Customers use See & Spray technology across 5 million acres in 2025John Deere · 2025-11-04 · Industry
  37. 37Precision ag leads to reductions in fertilizer, herbicide and water useAEM (Association of Equipment Manufacturers), via Farms.com · 2025-09-10 · Industry
  38. 38Review of smart irrigation systems — water savings by control typeJournal of Water and Sustainable Development (Ferdowsi University of Mashhad) · 2024 · Research paper
  39. 39Using deep learning for image-based plant disease detectionMohanty, Hughes & Salathé — Frontiers in Plant Science · 2016 · Research paper
  40. 40Low-cost real-time electronic trap for the olive fruit fly (Bactrocera oleae)Molina-Rotger et al. — Frontiers in Plant Science · 2023 · Research paper
  41. 41CNN-aided optical fiber distributed acoustic sensing for early detection of red palm weevil: a field experimentAshry et al. — Sensors · 2022 · Research paper
  42. 42Sustainable Food Cold Chains: Opportunities, Challenges and the Way ForwardUNEP & FAO · 2022-11 · Official report
  43. 43Farmer.Chat: Scaling AI-powered agricultural services for smallholder farmersDigital Green — arXiv · 2024 · Research paper
  44. 44DG-EVAL: evaluating agricultural answers against expert ground truthDigital Green — arXiv · 2026-03 · Research paper
  45. 45WaPOR v3 — remote sensing for water productivity (300 m global · 100 m Africa & Near East · 20 m Jendouba, Kairouan)FAO · 2023–2025 · Primary data
  46. 46Sentinel-2 — 10 m multispectral imagery, 5-day revisitESA Copernicus · 2024 · Primary data
  47. 47Global AgriFoodTech Investment Report 2026AgFunder · 2026-03-18 · Industry
  48. 48AI in Agriculture: Global Markets (2025–2030)BCC Research · 2025 · Industry
  49. 49Global AI in Agriculture Market, 2020–2030TechSci Research · 2025 · Industry