Quinoa as an alternative dry season climate resilient crop in the Mekong Delta Vietnam

Saline intrusion, drought, and water shortages are causing rice production losses in the Mekong River Delta dry season. Quinoa is a drought tolerant halophyte that could be an alternative crop option to rice or fallow in the dry season. The novelty of growing quinoa in the delta requires an understanding of variety selection and crop optimisation in conditions experienced in the dry season. A range of greenhouse and field trials screened five quinoa varieties (2-want, Sluga, Titicaca, 42- test, and Atlas) for salinity tolerance, fertiliser nitrogen response, and sowing density. The higher yielding varieties, Titicaca and 2-want, had lower biomass and improved water use efficiency in saline and water limited conditions. Nitrogen fertiliser rates of 120 kg N/ha produced the greatest biomass and yield, however lodging occurred, and the 2-want variety recorded negligible yield. Sowing densities between 15 plants/m 2 and 50 plants/m 2 had no significant effect on lodging rate and an uncharacteristic heatwave during anthesis causing no measurable yield. Though these experiments identify Titicaca as suitable variety of quinoa based on tolerance to salinity and agronomic performance, sensitivity to temperature has been identified as a risk factor. Manipulating the growing calendar so that saline and water limited tolerant varieties like Titicaca avoid peaks in temperature during anthesis may provide farmers with a suitable dry season alternative crop to rice.

The Mekong River Delta (MRD) is a significant agricultural region of Vietnam producing over 50% of the country’s rice (GSO, 2021). Extensive investment in irrigation infrastructure and the heightening of dikes in 2000 allowed triple rice production to dominate land use and contribute to Vietnam’s shift to net rice exportation (Tran et al., 2018; Sakamoto et al., 2009). However, the MRD is ranked highly vulnerable to climate change (WorldBank, 2021) with dry season saline intrusion and water shortages continuously threatening rice production (Nghia et al., 2024; Loc et al., 2021; Dang et al., 2020). The land use policies of the peri coastal regions of the MRD had designated some at risk areas as triple rice (three rice crops per year; Nguyen, Q. H. et al., 2020). In these areas, salinity intrusion events in 2016, 2020 resulted damage to rice in more than 200,000 and 40,000 ha respectively (WorldBank, 2017; Nghia et al., 2024). The severely affected province of Sóc Trăng has experienced rice yield decline of up to 4 t/ha due to salinity stress (Minh et al., 2021). In response to multiple severe droughts and salinity intrusion events, there was a focus on management and policy changes to mitigate these hazardous events (Nghia et al., 2024). A changed land use policy in 2000 identified the need to adapt to changing environmental conditions via diversification of agricultural production from intensive rice production in order to improve food security (Nguyen, Q. H. et al., 2020). Simultaneously, sluice gates were constructed as control mechanisms for saline intrusion. Despite these efforts, dry season saline intrusion and drought has continued to cause significant production losses in parts of the lower delta making rice unsuitable for growth during this period (Loc et al., 2021). Local authorities have advised farmers to change crop systems, shifting from three to two rice crops per year, and combine a rotation of alternative crops during the dry season to suit the changing conditions and improve land use efficiency (Nguyen et al., 2020). Dry season conditions mean alternative crops could be exposed to saline irrigation water exceeding 4 g L −1 which is the trigger for sluice gate closure (Loc et al., 2021) however these concentrations are often exceeded with reports of 10 g L −1 occurring 70 km deep into the network of canals in the 2020 drought (MARD, 2020). Concentrations exceeding 15 g L −1 have been consistently recorded in locations closer to the ocean (Eslami et al., 2019). The dry season experiences only 10% of the yearly rainfall and the increased occurrence of drought and changes to upstream river flows mean alternative crops may also experience water scarcity stress in addition to salinity (Dang et al., 2020). Therefore, there is an urgent need to find upland crop species displaying high salt tolerance and water use efficiency to cope with the changing climate in the MRD.

Quinoa (Chenopodium quinoa Willd.) is a pseudocereal that originated from the Andean regions of South America and since domestication has undergone many morphological changes (Adolf et al., 2013). It is cultivated for its seeds with the grain of quinoa providing a highly nutritional human food source (Angeli et al., 2020). Quinoa can tolerate a range of adverse abiotic growing conditions allowing it to be grown across many geographical regions (Bazile et al., 2016). Quinoa has a positive yield response to nitrogen application with rates of 120 kg N/ha increasing grain yield by up to 94% compared to plants not fertilised with nitrogen (Schulte auf’m Erley et al., 2005). Nitrogen application will increase panicle number, plant height, and seed weight (AbdElgalil et al., 2023), however excess nitrogen can cause lodging and delayed maturity (Wang et al., 2014). Plant sowing density can influence lodging and affect plant light interception, nutrient competition, and canopy structure, with optimal plant density levels varying by genotype and environment (Minh et al., 2021). High density may increase land-use efficiency but reduce plant biomass and seed size, whilst low density boosts per-plant yield but lowers total productivity. A Northern Vietnam study by Minh et al. (2021) identified an optimal sowing density of 8 plants/m 2 under the conditions tested; however, these environments did not experience the abiotic salinity stresses characteristic of the Southern Vietnamese Mekong River Delta (MRD).

Quinoa is not currently cultivated in the MRD and remains an uncommon crop in Vietnam, yet its halophytic characteristics (Hussain et al., 2020) and low water requirements (Telahigue et al., 2017) suggest strong potential as an alternative dry-season crop for MRD farmers (Kaveney et al., 2023). Recent research conducted under the cooler growing conditions of northern and central Vietnam has examined quinoa’s agronomic performance, including responses to nutrition (Hoang et al., 2021; Nguyen and Chuyen, 2023), plant density (Minh et al., 2021), soil type (Nguyen and Văn Minh, 2022; Văn Minh et al., 2022), and salinity (Long, 2016). However, no studies have evaluated quinoa varieties specifically suited to the saline and drought-affected soils of the MRD, where environmental conditions differ substantially from those of northern and central regions.

Knowledge of quinoa performance and its agronomic optimisation in the Mekong River Delta (MRD) is lacking. This research therefore aimed to identify quinoa varieties tolerant of salinity and water-limited conditions and optimise their production through nitrogen fertiliser and sowing density management to support climate-resilient farming systems in the MRD.

Experiments were conducted to investigate the suitability of quinoa for growth in the MRD where varieties were first screened for salinity and irrigation tolerance in a greenhouse pot trial, before selected varieties were tested in two field trials exploring different nitrogen fertiliser rates and sowing densities.

Salinity and irrigation screening pot trial

The soil used for the pot trial was collected from the 0 cm–20 cm topsoil layer of a rice field in Long An Province, Vietnam. The soil was air dried, sieved (<10 mm), and packed into pots measuring 22 diameter x 25 cm, with each pot containing approximately 9 kg of soil to give a 20 cm soil depth. Soil properties are reported in Table 1.

The randomised block pot experiment with four replications was designed to screen five quinoa varieties (2-Want, Sluga, Titicaca, 42-Test, and Atlas) under four different salinity and two soil moisture conditions. 10 seeds of each variety were sown per pot at a depth of approximately 1.5 cm in bare soil without mulch. 2 weeks after sowing, seedlings were thinned to two plants per pot. Each pot was treated as an experimental unit. All pots were watered with tap water (0 g/L NaCl) for the first two weeks before salinity and irrigation treatments commenced. Saline water was introduced starting from the third week after sowing, simulating the point when irrigation with salinised canal water would occur in the Soc Trang region (Ratering Arntz, 2018). Four salinity concentrations were applied (0, 2, 4, and 6 g/L NaCl) by adding water with incremental increases each week of 0.5 g/L until the designated application concentration was reached, after which pots were irrigated with solutions of that concentration as needed. The 0 g/L treatment was irrigated with deionised water throughout the experiment.

Two soil moisture conditions (−22 kPa and −50 kPa) were implemented using Chameleon soil moisture sensors placed at a depth of 10 cm–15 cm to trigger irrigation events. The Chameleon sensors indicate soil tension, so readings are independent of soil type (Stirzaker et al., 2017). Chameleons emit a (indicating a moisture content of −20 kPa) or red (indicating a moisture content of −50 kPa), water was applied until the light turned blue (0 kPa). Applied water quantity for each pot was recorded every irrigation event which was used to calculate water use efficiency.

A basal fertiliser, granular diammonium phosphate (DAP), was applied to all pots at sowing at a rate equivalent to 60 kg/ha. The experiment was conducted under greenhouse conditions at the Institute of Agricultural Sciences for Southern Vietnam from December 2022 to March 2023.

Harvest occurred at maturity, with Titicaca harvested at 75 days after sowing (DAS), and the remaining varieties harvested at approximately 80 DAS. Harvest was carried out intermittently, once every 1 day–2 days, and completed within 10 days. Aboveground plant material of the two combined plants per pot was removed for biomass measurements and air dried prior to recording dry mass of stem and leaves per pot. Roots were washed from the soil, dried until a constant weight was achieved, and then weighed. Individual yield was determined as the dry weight of harvested seeds per plant.

The effect of nitrogen fertiliser rate in the field

A field experiment was implemented on seasonally saline-affected rice land in Lieu Tu Commune, Tran De District, Soc Trang Province (9°29′13.02″N, 106°6′28.39″E) in the 2022–23 dry season. The soil properties of the site are shown in Table 2.

A randomised block experiment of two factors, quinoa variety and urea nitrogen fertiliser rate, was applied to four replicates. The two best performing varieties from the pot trial screened for salinity tolerance (Section 2.1) were selected for field testing: 2-Want and Titicaca. 4 application rates of urea were applied (0, 40, 80, and 120 kg N/ha) in even split applications occurring 10, 25, and 40 DAS. Super phosphate (90 kg P/ha) and potassium chloride (75 kg K/ha) were used as a source of phosphorus and potassium fertiliser and were applied to all treatments at sowing. Plots 4 m × 5 m were formed by creating three raised beds of 0.9 m width, and 30 cm high. Bunds of earth (0.2 m high and 0.3 m wide) separated the plots to ensure soluble nutrients could not move from each treated plot.

Quinoa varieties were direct sown at a density of 8 plants/m 2. Each raised bed had 3 rows (30 cm apart) with 25 cm between plants. All beds were mulched with rice straw at 1 kg/m. Sowing commenced at the end of the rainy season on the October 31, 2022.

Plots were irrigated with canal water in irrigation events triggered when the Chameleon soil moisture sensor lights indicated red (<50 kPa) and occurred until the sensor light turned blue (>−22 kPa). The irrigation cycle depended on weather conditions. All plots were irrigated simultaneously. Soil moisture was monitored at a depth of 0 cm–15 cm, with the Chameleon sensor installed at approximately 15 cm depth. Rainfall, humidity and temperature were recorded hourly using a weather station.

Plants were harvested manually on a single day (January 14, 2023). 6 plants per plot were harvested to determine grain yield and biomass. Each plot consisted of three rows with the samples collected following a zigzag pattern. Dry biomass was collected from the stems, leaves, and roots (after being washed and dried until a constant weight was achieved) and then weighed. Individual yield was determined as the dry weight of harvested seeds per plant.

The effect of sowing density in the field

A second field trial was conducted in the 2023–2024 dry season on the same site as the 2022–2023 fertiliser field trial following two subsequent rice crops. The nitrogen rate field trial (Section 2.2) demonstrated Titicaca showed promising characteristics for growth in the MRD and thus was chosen for further investigation in a plant sowing density trial as means to decrease lodging. Densities of 15, 25, and 50 plants/m 2 were chosen based on previous studies by Taaime et al. (2023) and Nguyen Van Minh et al. (2020). Three sowing dates were also investigated in an early (February 27, 2024) mid (March 12, 2024) and late (March 26, 2024) sowing based off the previous rice harvest schedule.

Each experimental plot had an area of 12 m 2 (4 m x 3 m) with treatments replicated four times. Drainage ditches were placed between the plots. Phosphorus (super phosphate) was applied to all plots at a rate of 90 kg P 2 O 5 /ha. Nitrogen (urea) and potassium (potassium chloride, KCl) were applied in three separate applications (Table 3). Superphosphate (90 kg P 2 O 5 /ha) and potassium chloride (90 kg K 2 O/ha) were uniformly applied across all treatments. Urea was applied at a rate of 0.64 g/plant and varied based on sowing density.

Dry biomass was collected on June 3, 2024 from the stems, leaves, and roots (after being washed and dried until a constant weight was achieved) from six plants per plot and then weighed. Individual yield was determined as the dry weight of harvested seeds per plant. The lodging rate was determined for each treatment plot based on the number of lodged plants per plot.

Statistical analysis

Statistical analyses were performed using Genstat version 22 with two-way ANOVA of salinity and variety, and irrigation and variety for yield, biomass, root mass, and water use efficiency (WUE). Genstat was applied to the experimental content of experiment 1 (Section 2.1) and experiment 2 (Section 2.2). Minitab (Minitab, LLC, Pennsylvania, US) was used for one-way ANOVAs of sowing density, examining biomass, root mass, and lodging rate. Minitab was applied to the content of experiment 3 (Section 2.3). Water use efficiency was calculated as yield per pot divided by the volume of water applied per pot. Results reported as significant were done so with the use of least significant difference (LSD) calculated at p = 0.05 with significant main effects reported if no significant interactions between treatments occurred.

Salinity and irrigation screening pot trial

Salinity

Atlas had the greatest shoot and root biomass but was significantly influenced by the presence of salinity (Figures 1A,B). Titicaca had the smallest shoot biomass compared to all other varieties. It also had the smallest root mass but was comparable to saline affected Sluga, and 42-test varieties. The grain yield of all varieties was significantly affected by the presence of increasing salinity. Again, the low biomass production of Titicaca was associated with significantly higher yield than any other variety for the same salinity. Though the yield of Titicaca was sensitive to salinity (Figure 1C), the yield at the highest salinity (6 g/L) was greater than all other varieties with the exception of 2-want, Sluga and 42-test in the absence of salinity. The WUE of varieties decreased with increasing salinity (Figure 1D). Titicaca resulted in the greatest WUE in the presence of salinity.

A strong negative correlation between WUE and shoot biomass existed in non-saline conditions (Figure 2). As plant biomass increased, its WUE decreased with varieties including Atlas and 2-want demonstrating the lowest WUE. As salt concentrations increased, the correlation coefficient became weaker with biomass impacting WUE the least at 6 g/L.

A negative correlation between yield and shoot biomass existed across all salt concentrations but was strongest at the lowest salt concentrations (Figure 3). As salt concentrations increased, the correlation strength became weaker. Yield implications occurred for plants like Atlas that had a higher biomass, whilst Titicaca was a variety that had low shoot biomass and higher yields.

Irrigation

There were no significant main effects due to irrigation for any measured parameter. Significant interactions of irrigation treatment and variety existed for plant biomass and yield (Figure 4). The varieties that had more shoot biomass also had more root biomass (2-want and Atlas) with irrigation at 22 kPa having significantly higher shoot and root mass for these treatments (Figure 4). Root mass was also greater at 22 kPa for Sluga and 42-test, though there were no differences in the shoot biomass of those variety due to irrigation. Interestingly, Titicaca had the lowest above and below ground biomass but was unaffected by irrigation. Titicaca returned the highest grain yield and 50 kPa was significantly greater than 22 kPa (Figure 4). This resulted in Titicaca having the greatest WUE and 50 kPa being significantly better than 22 kPa for Titicaca (Figure 4).

The effect of fertiliser rates in the field

Yield

There was a clear difference in yield between the two quinoa varieties under field conditions (Figure 5). Titicaca demonstrated the potential for achieving harvestable yields when grown in MRD dry season conditions. Increasing the nitrogen fertiliser application rate significantly contributed to higher yields with 120 kg N/ha resulting in the highest yield. In contrast, the 2-Want variety did not produce measurable yields under the same planting time and experimental conditions.

Biomass

The results showed that the addition of nitrogen fertiliser positively impacted the biomass of both varieties (Figure 5). As nitrogen concentration increased, plant biomass also increased. In the treatment without nitrogen supplementation, biomass was significantly lower compared to the other treatments for both varieties. Despite 2-Want not producing harvestable yield, this varieties biomass was the same as Titicaca for all nitrogen rates.

The effect of sowing density in the field

Yield and biomass

Titicaca plant shoot biomass decreased as sowing density increased (Table 4), although this was not significant. A sowing density of 15 plants/m 2 resulted in the highest biomass compared to the other two densities (25 plants/m 2 and 50 plants/m 2). At harvest, there was an occurrence of shrivelled seeds and failure in seed formation which resulted in negligible seed yield. The early sowing date was the only successful trial where plants reached maturity and produced harvestable yield. Plants from the mid and late sowing dates failed to grow and produce seeds.

Root mass

The results indicated an increase in root biomass with higher planting densities, although this was not significant (Table 4). A planting density of 50 plants m 2 produced the highest root biomass compared to the other two densities (15 plants/m 2 and 25 plants/m 2).

Lodging rate

Sowing density had no significant impact on quinoa lodging rate (Figure 6). A trend of higher planting densities (25 and 50 plants/m 2) having increased lodging rates existed, but it was not statistically different.

Temperature during the field trial

During the flowering stage, a heat wave occurred from the beginning of April that resulted in daytime temperatures consistently exceeding 34 °C and nighttime temperatures above 25 °C (Figure 7). The heat wave period where daily temperatures exceeded 34 °C lasted for 21 days. This was followed by patterns of peaks and troughs where daytime temperatures reached 34 °C–36 °C and dipped to 31 °C every 6–7 days. The pattern of these temperatures was considered atypical for the MRD dry season.

The importance of variety selection to suit MRD conditions

Saline intrusion (Loc et al., 2021), droughts (Lee et al., 2019) and water scarcity (Anh et al., 2019) are prominent dry season constraints to production in the MRD. Selecting quinoa varieties that can tolerate these conditions will increase their likelihood of success as an alternative crop to rice. The greenhouse results showed increasing salt concentrations negatively affected the yield of all quinoa varieties, though the degree of salinity tolerance was variety dependent (Figure 1). Lowest yielding varieties (Atlas and 42-test) had the greatest biomass (Figure 1A) whilst Titicaca produced the highest yield (Figure 1C) across all salt concentrations and recorded the lowest plant biomass. Quinoa varieties have been screened for salinity tolerance before (Adolf et al., 2012; Cai and Gao, 2020; Hussain et al., 2020), including in temperate northern Vietnam (Long, 2016; Nguyen Viet, 2016). However, quinoa has not been evaluated under the tropical, saline, and drought affected conditions characteristic of southern Vietnam, highlighting a critical research gap. The negative correlation between yield and shoot biomass (Figure 3) may be due to transpiration rates. In saline conditions, quinoa plants reduce transpiration rates by decreasing stomatal density and leaf area (Turcios et al., 2021). Varieties more efficient at reducing leaf area and consequently biomass, save water by transpiring at lower rates and display a greater salinity tolerance (Agirresarobe et al., 2022). Additionally, Titicaca recorded the lowest root weight compared to other varieties (Figure 1B). Decreased root growth in saline tolerant varieties reduces the excessive uptake of toxic Na + and Cl − (Cai and Gao, 2020).

High water use efficiency (WUE) is a desirable trait when assessing the suitability of varieties for development in the MRD. In conditions of water scarcity (Anh et al., 2019), crop varieties that can achieve high yields with less irrigation water are more suitable, especially during the dry season. In this experiment, Titicaca had significantly higher yield and WUE at both −22 kPa and −50 kPa than any other variety, and the lowest root and biomass (Figure 4). The variety Atlas that yielded significantly less than Titicaca, recorded much higher root mass and lower WUE (Figures 1B,D). It was also found through the correlation that WUE improved with lower biomass varieties (Figure 2). Drought tolerant quinoa varieties will decrease root diameter in order to increase root length and promote water uptake (Nguyen et al., 2021). In the present study, Titicaca was the only variety to have a significant yield difference between irrigation treatments, with drier conditions interestingly promoting higher yields. Irrigating quinoa using alternative root-zone drying techniques actually improved quinoa plant growth and WUE compared to full and deficit irrigation (Yang et al., 2016). The varieties Atlas and 2-Want have previously exhibited drought tolerant phenotypes when grown in northern Vietnam (Nguyen et al., 2021), and thus the difference between −22 kPa and −50 kPa in the current study may not have been enough to induce drought like conditions. Decreased yields may be attributed to experimental factors other than drought.

Whilst salinity tolerance and WUE are important for quinoa growth in the MRD, temperature also strongly influences varietal performance. Atlas and 2-Want have previously demonstrated high yield potential under saline conditions in northern Vietnam (Long, 2016; Nguyen, L. V. et al., 2020; Hoang et al., 2021), yet Atlas performed poorly under MRD greenhouse conditions and 2-Want failed to produce harvestable grain in the field (Figures 1, 5). High temperatures during flowering and seed development are amongst the greatest constraints to quinoa yield (Yang et al., 2016; Hinojosa et al., 2019). Although moderate temperatures (25/20 °C day/night) promote growth, temperatures above 33 °C–35 °C reduce yield (Peterson and Murphy, 2015; Walters et al., 2016; Alvar-Beltrán et al., 2019), whilst temperatures approaching 40 °C can reduce pollen viability by 30%–70%, causing poor seed set (Hinojosa et al., 2019). These conditions are increasingly common in the MRD, where dry season temperatures regularly reach 27 °C–34 °C (Kontgis et al., 2019) and climate change is increasing the frequency of droughts and heatwaves (Phan et al., 2020; WorldBank, 2021; Nghia et al., 2024). During the 2024 field trial, temperatures exceeded 34 °C during the day and 25 °C at night for up to 21 consecutive days (Figure 7), coinciding with anthesis and likely causing sterility and grain failure (Figure 5), as reported elsewhere (Taaime et al., 2023). Similar interactions between salinity, drought, and high temperature have reduced quinoa yields by up to 50% (Abbas et al., 2023), whereas little effect was observed when maximum temperatures remained below 28 °C (Becker et al., 2017). Amongst the varieties evaluated, Titicaca showed the greatest tolerance to combined salinity, drought, and heat stress, suggesting that aligning its reproductive stage to avoid peak dry season temperatures could improve its suitability for MRD production.

Optimising quinoa growth in the MRD

Quinoa's tolerance of abiotic stresses highlights its possibility as an alternative dry season crop under managed conditions in the MRD, although optimising agronomic management is essential to maximise yields and reduce crop failure. Nitrogen fertiliser substantially increased Titicaca yield, with 40, 80, and 120 kg N/ha increasing grain yield by approximately 20, 60, and 130%, respectively, compared with the control (Figure 5). Although biomass also increased in ‘2-Want,’ no grain was produced, likely due to temperature-induced sterility at anthesis (Hinojosa et al., 2019). Previous studies have reported variable quinoa responses to nitrogen depending on environmental conditions, with optimum rates of 40 kg N/ha under semi-arid, water-limited conditions (Taaime et al., 2023) and 90 kg N/ha under saline conditions, beyond which yield declined (Hoang et al., 2021). In contrast, the linear yield response observed in this study (Figure 5) suggests that rates above 120 kg N/ha may further increase yield potential in the MRD. Similar responses have been reported under favourable conditions, where applications exceeding 75 kg N/ha improved growth and yield (Basra et al., 2014) whilst quinoa grown under mediterranean conditions yielded highest at 150 kg N/ha (Geren, 2015). As nitrogen application does not alter quinoa growth duration (Hoang et al., 2021) adjusting planting dates may be necessary to avoid high temperatures during sensitive reproductive stages such as anthesis.

Saline intrusion, water scarcity, and temperatures peak towards the end of the MRD dry season during April and May (Eslami et al., 2021; Phan et al., 2020). Quinoa was planted in the field in late February and March 2024 with anthesis occurring in the period of peak abiotic stresses. Aforementioned, temperature stress during anthesis likely caused grain sterility (Hinojosa et al., 2019) in 2-Want (Figure 5) and Titicaca. Manipulating the cropping calendar to sow quinoa earlier in the dry season may reduce the risk of abiotic stresses impacting yield. Quinoas susceptibility to waterlogging (González et al., 2009) may present alternative issues to seedling establishment if sown too soon after the second rice crop, yet management techniques like raised beds may mitigate these risks (Bakker et al., 2010).

Plant lodging occurred during the 2023 field trial evaluating nitrogen fertiliser effects on quinoa yield (Figure 5). However, the 2024 field trial found that sowing density had no significant effect on lodging (Figure 6). Previous studies in the cooler climate of Vietnam's Central Highlands (~25 °C) identified optimum planting densities between 66,000 and 133,333 plants/ha for maximising biomass and yield (Minh et al., 2021; Nguyen and Chuyen, 2023; Nguyen and Văn Minh, 2022). In contrast, higher plant densities can increase stem lodging due to greater wind susceptibility, although reducing density may also reduce yield (Wang et al., 2021). Well-irrigated conditions can further increase lodging by promoting plant height. In the present study, sowing density did not significantly affect biomass or root mass (Table 4), and yield effects could not be measured due to temperature stresses. Selecting shorted statured varieties with thicker stems may reduce lodging risk (Tang et al., 2024) but further research is needed to optimise plant density and variety selection for quinoa growth in the MRD.

Given the increasingly complex climate change situation in the Mekong River Delta, crop diversification is crucial for future adaptation and development. Growing quinoa in saline and water limited areas of the MRD requires a variety that can tolerate these conditions and produce harvestable yield. Titicaca displayed the greatest saline tolerance through increased yield despite being exposed to 6 g/L salt concentration and −50 kPa soil moisture irrigation conditions. However, field trials highlighted significant challenges associated with the MRD environment, particularly high temperatures during reproductive development, which substantially reduced yield. Although plant yield increased with nitrogen application, lodging still occurred across a range of sowing densities. Further research is therefore required to identify suitable agronomic solutions, including the management implication of manipulating the cropping calendar to avoid periods of peak temperature during anthesis and reduce abiotic stress. By timing sensitive reproductive phases to cooler conditions, varieties such as Titicaca could provide a promising climate-resilient alternative under managed conditions in the MRD dry season.