Best Rice Varieties for Water-Scarce Regions: 2026 Guide

Rice drinks more water than almost anything else on your plate. Producing one kilogram of milled rice can take anywhere from 800 to 5,000 liters of water, and the average sits close to 2,500 liters. Now hold that number next to what millions of growers deal with every season: canals that open two weeks late, tubewells reaching deeper every year, and a monsoon that no longer follows the calendar it used to.

That collision is exactly why choosing the right rice varieties for water-scarce regions has quietly become the highest return decision on a rice farm. Not a bigger pump. Not more urea. The seed you put in the ground.

Here is the encouraging part. Over the past twenty years, breeders at the International Rice Research Institute, the Indian Council of Agricultural Research, Africa Rice and dozens of national programs have released rice that holds yield on a fraction of the usual water. Some of it matures in 95 days. Some of it survives a three week dry spell that wipes out the field next door. And a couple of the newest lines were built with gene editing tools that did not exist commercially when today’s mid career agronomists were in school.

This guide walks through the varieties that actually perform, the science behind why they work, and how to match one to your own water situation.

Why Rice and Water Scarcity Keep Colliding

Rice is not naturally thirsty. Flooded paddy management is thirsty. The crop itself transpires about as much as wheat or maize per unit of biomass. What burns the water is standing water: seepage, percolation, evaporation from a bare water surface, and the enormous volume needed to puddle a field before transplanting.

The scale is hard to ignore. Rice accounts for roughly 34 to 43 percent of the world’s irrigation water and draws down something like a quarter to a third of total freshwater withdrawals. In Asia, irrigation consumes more than 80 percent of available freshwater, and rice alone takes over half of that share.

Meanwhile, about 38 percent of global rice area is rainfed, and that area contributes only around 21 percent of production. The gap between those two numbers is the drought penalty, written in plain arithmetic.

In eastern India alone, roughly 4.28 million hectares of rice land face frequent drought. Similar patterns show up in northeast Thailand, the Cambodian lowlands, upland West Africa, and increasingly in irrigation commands that used to be considered safe, from Punjab in South Asia to parts of the Sacramento Valley.

Figure 1. Seasonal water input by rice production system. Moving off continuous flooding is where the biggest savings live.

What Actually Makes a Rice Variety Water Smart

Farmers often ask for “a drought variety” as if drought tolerance were one switch. It isn’t. Plant scientists split it into three separate strategies, and the best modern varieties combine at least two of them.

Figure 2. The three survival strategies breeders stack into a modern water efficient rice variety.

1. Drought Escape: Finish Before the Water Runs Out

The simplest trick in the book. A variety that matures in 105 days instead of 145 days sidesteps the late season dry spell entirely. It also needs fewer irrigations across the whole cycle, because there are fewer days to irrigate.

Short duration has a second benefit that farmers rate even higher than the water saving. It frees the field early, so you can plant a rabi crop such as wheat, chickpea or mustard on residual moisture. In eastern India, that shift has let some households move from two crops a year to three.

2. Drought Avoidance: Reach the Water That Is Still There

Soil moisture rarely disappears all at once. It retreats downward. Varieties with deep, thick, well branched roots keep drawing on moisture at 40 to 60 centimeters when shallow rooted irrigated types are already rolling their leaves.

Root architecture is heritable, and it is one of the traits breeders now select for directly. Waxy leaf cuticles, faster stomatal closure and leaf rolling all belong in this bucket too. The plant is not toughing out dehydration. It’s dodging it.

3. Drought Tolerance: Keep Working Under Stress

This is the hardest trait to breed and the most valuable. A truly tolerant plant keeps its cells functioning at low water potential, keeps pollen viable through flowering, and keeps filling grain when a susceptible plant has already given up.

Progress here was slow for decades because drought tolerance is controlled by many genes at once. The breakthrough came with quantitative trait loci known as the qDTY series, which allow marker assisted selection. Breeders can now screen seedlings in a lab instead of waiting for a natural drought to reveal winners in the field.

Traits Worth Checking Before You Buy Seed

• Days to maturity. Under 115 days is a meaningful advantage in an unreliable season.
• Yield under stress, not just potential yield. Ask for the drought trial number, not the irrigated one.
• Seedling vigor for direct seeding. Water saving usually means dropping transplanting, so emergence matters.
• Lodging resistance. Non puddled soils and mechanical harvest punish weak stems.
• Grain quality and market acceptance. A resilient variety nobody will buy is not a solution.

The Best Rice Varieties for Water-Scarce Regions

These are the lines with published field data and real farmer adoption behind them. Availability varies by country, so treat the list as a shortlist to take to your national seed agency or extension office rather than a shopping cart.

Sahbhagi Dhan: The Benchmark for Rainfed Lowlands

If one variety changed the conversation about drought rice, it’s this one. Developed as breeding line IR74371-70-1-1 and released in India in 2010, the same line went out as Sukha Dhan 3 in Nepal and BRRI dhan 56 in Bangladesh.

The numbers explain the adoption. In a normal season it delivers 4 to 5 tonnes per hectare against roughly 2.5 tonnes for the traditional varieties it replaced. Under severe drought it still brings in 1 to 2 tonnes per hectare in seasons where high yielding irrigated types produce nothing at all. It matures in about 105 days rather than 120 to 150, and it needs roughly two irrigations where older varieties such as Sarju 55 and Samba Mahsuri wanted four.

One household survey of 2,710 farms in eastern India found that adopters held yield losses in drought years to about 13.6 percent compared with their own normal season output. That is remarkable stability for a rainfed crop.

Figure 3. Sahbhagi Dhan holds a harvest in seasons when susceptible varieties fail completely.

DRR Dhan 42, 43 and 44: The Next Generation from ICAR

Bred at the Indian Institute of Rice Research and released for drought prone tracts, these lines were selected specifically using qDTY markers. Field trials place the average yield advantage of drought tolerant varieties over susceptible checks at 1.0 to 1.5 tonnes per hectare when drought hits.

DRR Dhan 42 and 44 are the ones you will most often see promoted for West Bengal, Odisha and Jharkhand. They pair well with direct seeding, which matters because that is where the second layer of water saving comes from.

Sahod Ulan Series: The Philippine Answer

The Sahod Ulan lines were bred for rainfed lowland conditions in the Philippines and released in a numbered series. They carry the same qDTY logic as the South Asian releases but were selected under local pest pressure and photoperiod conditions. If you farm in Southeast Asia, start here rather than importing an Indian variety that has never been tested against your blast races.

Sukha Dhan and Hardinath Lines: Nepal’s Rainfed Toolkit

Nepal has released the Sukha Dhan series alongside newer options such as Hardinath 5 and Hardinath 6 and the Ganga Sagar lines. These matter because much of Nepal’s rainfed area was historically planted with Sabitri, a variety bred for irrigated conditions that collapses under drought. Swapping in a drought bred line of similar duration is one of the cheapest resilience gains available in that system.

BRRI dhan 56 and BRRI dhan 66: Bangladesh Releases

BRRI dhan 56 is the Bangladeshi identity of Sahbhagi Dhan and suits the drought prone aman season in the northwest. BRRI dhan 66 followed as another short duration drought tolerant option. Both are aimed at the same problem: a late monsoon withdrawal that leaves the crop short of water precisely at flowering.

Apo (IR55423-01): The Aerobic Workhorse

Apo is an improved upland inbred that became the reference variety for aerobic rice research at IRRI. Grown in non flooded, unsaturated soil with flush irrigation, Apo reached 5 to 6 tonnes per hectare in dry season trials, which is unusually high for a non flooded system.

Be honest about the trade off, though. Across multiple seasons and varieties, aerobic systems cut water input sharply but average yields ran about 32 percent below flooded fields in the dry season and 22 percent below in the wet season. Aerobic rice is a water productivity play, not a yield maximization play.

MAS 946-1 and MAS 26: India’s First Aerobic Releases

The University of Agricultural Sciences in Bangalore released MAS 946-1 in 2007 as India’s first drought tolerant aerobic rice variety, followed by MAS 26 in 2008. Grown properly, aerobic cultivation can cut irrigation by roughly 40 to 50 percent compared with transplanted rice. The catch is weeds. Without a standing water layer suppressing them, weed management becomes the single biggest operational challenge, and it is where most first year aerobic growers lose their margin.

Han Dao Varieties: China’s Aerobic Program

China Agricultural University developed a family of aerobic types marketed as Han Dao, including HD297, HD277 and HD502. Reported water inputs of roughly 470 to 650 millimeters run more than 50 percent below lowland rice, with water productivity 64 to 88 percent higher. They are widely grown in northern China where water tables have been falling for decades.

NERICA: Built for African Uplands

New Rice for Africa is a genuinely different genetic idea. AfricaRice crossed Asian rice, Oryza sativa, with African rice, Oryza glaberrima, to combine Asian yield potential with African stress adaptation. There are now 82 NERICA varieties in total: 18 upland, 60 rainfed lowland and 4 irrigated.

Upland NERICAs mature in 75 to 100 days and tolerate short term drought, poor soils and weed pressure. Yield potential runs 2 to 3 tonnes per hectare under water limited conditions and up to 4 to 5 tonnes in better years. NERICA 4 is the most widely adopted upland line and is grown in more than ten African countries, partly because it also tolerates low phosphorus soils.

A detail farmers appreciate: a NERICA panicle typically carries 300 to 400 grains against 75 to 100 for many unimproved local varieties, and the crop ripens in about three months instead of six.

DRR Dhan 100 (Kamala) and Pusa DST Rice 1: The Gene Edited Arrivals

On 4 May 2025, ICAR released the world’s first two genome edited rice varieties. Both were made with CRISPR based site directed nuclease methods that alter the plant’s own DNA without inserting foreign genes, which is why India regulates them differently from conventional GM crops.

• DRR Dhan 100 (Kamala). Built on Samba Mahsuri by editing the CKX2 (Gn1a) gene to raise grain number per panicle. It yields about 19 percent more than its parent, averaging 5.37 tonnes per hectare against 4.5, and matures roughly 20 days earlier at about 130 days. That shorter cycle is where the water saving comes from. ICAR estimates a saving on the order of 7,500 million cubic meters of irrigation water if adoption reaches the recommended zones, along with lower methane emissions.
• Pusa DST Rice 1. Built on MTU 1010 by editing the DST gene, which improves tolerance to both drought and salinity. Trials showed yield gains of roughly 9.66 to 30.4 percent under saline and alkaline conditions while keeping the parent’s grain quality.

If your problem is water scarcity combined with rising salinity, which is the reality in a lot of coastal and over pumped inland areas, Pusa DST Rice 1 is the more interesting of the two.

Other Options Worth Asking Your Extension Office About

• CR Dhan 201, CR Dhan 204 and related upland lines from the National Rice Research Institute
• Swarna Shreya, a drought tolerant improvement on the hugely popular Swarna
• DRR Dhan 50, DRR Dhan 64, CR Dhan 80 and CR Dhan 802 for stress prone tracts
• PR 126, a short duration line widely promoted in northwest India to cut pumping
• Drought tolerant versions of Sabitri developed through marker assisted backcrossing

Water Efficient Rice Varieties Compared at a Glance

Variety	Best suited to	Maturity	Yield under stress	Standout trait
Sahbhagi Dhan / BRRI dhan 56 / Sukha Dhan 3	Rainfed lowland and direct seeded upland	About 105 days	1 to 2 t/ha in severe drought	Proven across three countries
DRR Dhan 42 and 44	Drought prone eastern India	Medium short	1.0 to 1.5 t/ha advantage over checks	qDTY marker bred
Sahod Ulan series	Philippine rainfed lowland	Short to medium	Stable under intermittent drought	Locally adapted pest profile
Apo (IR55423-01)	Irrigated aerobic systems	Medium	5 to 6 t/ha under aerobic dry season	Highest aerobic yield ceiling
MAS 946-1 and MAS 26	Aerobic cultivation in India	About 112 days	Holds yield on 40 to 50 percent less irrigation	First Indian aerobic releases
Han Dao HD297, HD277, HD502	Northern China, falling water tables	Medium	Water input 470 to 650 mm	Water productivity 64 to 88 percent higher
NERICA 1 to 18 (upland)	West and East African uplands	75 to 100 days	2 to 3 t/ha water limited	Weed competitive, low input
DRR Dhan 100 (Kamala)	Zones III, V and VII in India	About 130 days	5.37 t/ha average, 19 percent gain	Genome edited, matures 20 days early
Pusa DST Rice 1	Drought plus salinity or alkalinity	Medium	9.66 to 30.4 percent gain under stress	Dual drought and salt tolerance

Figure 4. Every 15 days you shave off maturity typically removes two to three irrigations from the season.

How to Choose the Right Variety for Your Field

Skip the temptation to pick the variety with the biggest yield number. Start with your water reality and work backward.

Figure 5. A simple decision path from water availability to variety shortlist to agronomy.

1. Classify your ecosystem honestly. Rainfed upland with zero irrigation, rainfed lowland with one or two emergency irrigations, and irrigated but declining supply are three genuinely different problems requiring three different variety groups.
2. Map when your drought usually arrives. Early season drought hurts establishment and calls for strong seedling vigor. Terminal drought at flowering and grain filling calls for short duration and reproductive stage tolerance. They are not the same fix.
3. Set a duration ceiling. Count the reliable rainfall or canal days you actually get, subtract a safety margin, and refuse any variety that needs longer than that window.
4. Check the local release list, not the internet. Varieties are released by zone for good reason. A line that shines in Odisha may be blast susceptible in your district.
5. Test on a strip before you commit the farm. Plant your candidate beside your usual variety on the same field for one season. Nothing beats your own soil as a trial site.
6. Buy certified seed and plan for renewal. Drought tolerance decays with poor seed management. Refresh certified seed every three to four seasons.

Water Saving Practices That Multiply the Varietal Gain

A drought tolerant variety grown with old habits gives you maybe half of what it is capable of. The variety and the agronomy work as a pair.

Alternate Wetting and Drying

AWD is the highest value practice on this list and it costs almost nothing to start. Instead of keeping the field flooded all season, you let the water level fall to about 15 centimeters below the soil surface, then reflood. A cheap perforated tube pushed into the soil lets you see the water table without guessing.

Figure 6. Alternate wetting and drying with a field water tube, the cheapest water saving tool in rice.

A global meta analysis covering 437 studies and 93 percent of world rice production found AWD cut irrigation water use by about 33.9 percent while raising irrigation water use efficiency by roughly 47.6 percent. Methane emissions typically fall by 30 to 80 percent as well, averaging around half.

One caution worth taking seriously. Pushing water savings past 20 percent raises the risk of a 15 to 25 percent yield penalty, and past 50 percent the penalty risk climbs above 25 percent. Safe AWD, where you never let the water table drop below 15 centimeters, is the version to run if you cannot monitor daily.

Direct Seeded Rice

Puddling and transplanting can consume a quarter of the season’s total water before the crop is even established. Direct seeding skips both. It works best with varieties bred for seedling vigor and anaerobic germination, which is precisely why the newer drought lines were selected for those traits. Budget for a stronger weed program in year one.

Laser Land Levelling

An unlevel field forces you to over irrigate the high spots to wet the low ones. Levelling to within a couple of centimeters commonly saves 10 to 25 percent of applied water and improves crop uniformity at the same time. It is a one time contractor cost that pays back within two to three seasons in most commands.

Timing and Bund Discipline

• Repair and compact bunds before the season so water stays where you put it
• Irrigate at the critical stages first: panicle initiation, flowering and early grain filling
• Shift sowing dates so flowering avoids the hottest, driest fortnight in your calendar
• Mulch or retain residue in aerobic and upland systems to slow surface evaporation

Five Mistakes That Waste Water and Yield

• Treating drought tolerance as drought proofing. These varieties reduce loss. They do not guarantee a normal harvest in a failed season.
• Running AWD during flowering. Keep the field wet through the reproductive window. Dry down before and after, not during.
• Cutting nitrogen because water is short. Aerobic and AWD systems still need proper nutrition. Underfeeding a stressed crop compounds the loss.
• Ignoring weeds in non flooded systems. Standing water was doing free weed control. Remove it and you have to replace that function.
• Saving grain from last year as seed indefinitely. Genetic purity drifts, and the drought traits you paid for erode with it.

What Is Coming Next in Drought Rice Breeding

The pipeline is moving faster than it has in decades. Three things are worth watching.

Genome editing has crossed from lab to field. India’s 2025 releases proved that a regulator will approve a non transgenic edited crop, and other rice producing countries are drafting similar frameworks. Expect edited lines targeting root depth, stomatal control and nitrogen use efficiency within a few seasons.

Trait stacking is the second shift. The interesting varieties now combine drought tolerance with submergence tolerance from the SUB1 gene, because many rainfed areas get both a dry spell and a flash flood in the same season. Swarna Sub1 and Samba Sub1 opened that door and drought plus submergence combinations are following.

Finally, water productivity is becoming the metric that matters, replacing yield per hectare as the headline number. When a government prices or rations irrigation water, the variety that produces the most grain per cubic meter wins, even if a thirstier line would top it on a per hectare basis.

Frequently Asked Questions

What is the best rice variety for water-scarce regions?

There isn’t a single global winner, but Sahbhagi Dhan is the most widely proven choice for rainfed lowlands across South Asia. It yields 4 to 5 tonnes per hectare in a normal season, still returns 1 to 2 tonnes under severe drought, matures in about 105 days and needs roughly two irrigations instead of four. For African uplands the equivalent answer is NERICA 4, and for irrigated systems facing shrinking supply it is DRR Dhan 100 (Kamala) or Pusa DST Rice 1.

How much water does rice actually need?

Seasonal water requirement ranges from about 660 to 5,280 millimeters depending on soil, climate and season, with most fields falling between 1,000 and 2,000 millimeters. Aerobic systems have delivered crops on 470 to 650 millimeters. On a product basis, one kilogram of rice takes 800 to 5,000 liters, averaging near 2,500.

Can rice grow without flooding?

Yes. Rice is not an aquatic plant by necessity. Aerobic rice systems, direct seeded rice and alternate wetting and drying all grow rice in non flooded or intermittently flooded soil. You need a variety bred for those conditions, a real weed plan, and acceptance that yields may run 20 to 30 percent below a fully flooded field while water use drops much further.

Do drought tolerant rice varieties yield less in a good year?

Modern ones generally don’t. That was the old trade off, and it is largely gone. Sahbhagi Dhan out yields the traditional varieties it replaced in normal seasons, and DRR Dhan 100 delivers 19 percent above its parent. The breeding goal for the past fifteen years has been stress tolerance without a yield penalty in favorable years.

What is alternate wetting and drying and how much water does it save?

AWD means letting the flood water recede to about 15 centimeters below the soil surface before reflooding, monitored with a simple perforated field tube. A meta analysis across 437 studies found average irrigation savings near 33.9 percent with irrigation water use efficiency up about 47.6 percent, plus a large cut in methane emissions.

Are genome edited rice varieties safe and legal?

The two Indian releases, DRR Dhan 100 (Kamala) and Pusa DST Rice 1, contain no foreign DNA. The edits change the plant’s own genes in ways considered equivalent to natural mutations, so India exempted them from strict GM biosafety rules under the Environment (Protection) Act, 1986. Regulatory treatment varies by country, so check your national framework before importing seed.

Which rice variety is best for saline plus dry conditions?

Pusa DST Rice 1 is currently the most targeted option, since the edited DST gene improves tolerance to drought and salinity together. Trials recorded yield gains of 9.66 to 30.4 percent in saline and alkaline soils. Combine it with gypsum amendment and careful leaching where drainage allows.

How much can short duration varieties really save?

Each two week reduction in maturity typically removes two to three irrigations. ICAR estimated that DRR Dhan 100 alone, maturing 20 days earlier than Samba Mahsuri, could save on the order of 7,500 million cubic meters of irrigation water across its recommended zones.

The Bottom Line

Water scarcity is not a problem you solve once. It is a constraint you design around, season after season. The best rice varieties for water-scarce regions give you the biggest single lever, because a better seed costs about the same as a worse one and changes the outcome of the entire year.

Start with your water reality. Pick a variety bred for that reality, released for your zone, and short enough to finish inside your reliable window. Then stack the agronomy on top: level the land, seed it directly, and manage the water table with a tube instead of a guess. Growers who do all four routinely cut irrigation by a third or more while holding yield steady.

Your next step: contact your national rice research institute or district extension office, ask for the current drought tolerant release list for your zone, and put one candidate variety on a trial strip this coming season. One strip, one season, and you will know more about what works on your farm than any article can tell you.