4.1 Crop response to irrigation management
During the 2019–2020 MRD drought, the Ministry of Agriculture and Rural Development (MARD) urged farmers to grow undefined drought tolerant crops (MARD, 2020) in an effort to limit irrigation and save water resources for household consumption. Restricting irrigation in this trial negatively affected cowpea with a 20% yield decrease occurring under water limited conditions (Table 2). However, daily water consumption for intermittently irrigated saline plants was almost half (544 mm) of constantly irrigated non-saline treatments (1000 mm) (Table 3). Cowpea grown in water stressed conditions can save over 3 times more water than well-irrigated plants but with a 50% yield reduction (Oumarou et al., 2015). Changes in water use due to stress are a result of the accumulation of proline in cells to assist with osmotic adjustment and maintenance of plant water contents (Furlan et al., 2020). Proline accumulation can occur in response to relatively small decreases in relative water content (RWC) (from > 90–85%) (Zegaoui et al., 2017) and accumulates linearly as soil dries (RWC of 65 to 75%) with an approximate 12-fold increase in proline concentration occurring. Intermittent irrigation induced a stress response in cowpea (Table 6) with the youngest leaves having significantly higher proline contents than older leaves, or those of the constantly irrigated plants. Therefore, cowpea is sensitive to inadequate irrigation and responded by decreasing water use at the expense of yield production (Table 2). Irrigating cowpea using Chameleon sensors before the red light (-50 kPa) would be an appropriate soil water indicator to help farmers in the MRD save water and minimise yield losses of this crop.
Quinoa yield was unaffected by irrigation treatments in this glasshouse trial (Table 2). The drought tolerance of quinoa means it can grow in as little as 200 mm of rainfall in sandy soils (Jacobsen et al., 2003) and has produced over 2 t ha-1 of grain yield in as little as 350 mm growing season rainfall (Telahigue et al., 2017). In this trial, the total water use of intermittently irrigated quinoa treatments was 420 mm pot-1 (Table 3). Farmers implementing intermittent irrigation of quinoa plants in the MRD could reduce water use without yield compromise and thus would be suitable for growth in areas affected by water scarcity.
Whilst soybean yield was unaffected by irrigation treatments and its total water requirement was similar to quinoa (Table 3), it grew for a shorter period and thus had an increased daily water use, particularly during reproductive phases. The drought stress that occurred at anthesis has been shown to cause up to an 82% yield reduction (Wei et al., 2018). Evidence of drought stress in this trial could be observed through a significant reduction in the K/Na ratio of new leaves in intermittent non-saline irrigation treatments (Fig. 3) which has been observed in other water limiting trials (Al-Hakimi, 2006). Potassium plays pivotal roles in many physiological processes involved with plant water regulation including water uptake, stomatal regulation and photosynthesis (Hasanuzzaman et al., 2018) and a deficiency in K+ can limit photosynthetic rates (Kanai et al., 2011). It will be critical if soybean is grown in the MRD to avoid stress at this time, which for a March planting, would occur in May towards the end of the dry season. An earlier sowing date where the risk of freshwater scarcity during reproductive phases is lower would be suitable for Soybean.
4.2 Salinity tolerance
Quinoa yield was affected by salinity treatments (Table 2), although its halophytic qualities that allow partitioning of Na+ from leaves into plant roots to tolerate higher levels of salinity (Cai and Gao, 2020) was partially observed in this trial. The maintenance of higher shoot K+ concentrations occurred in response to salinity (Fig. 3). Increasing salt concentration will generally increase both leaf and root Na+ and K+ concentrations, although roots will preferentially accumulate Na+ with higher concentrations of K+ occurring in leaves (Cai and Gao, 2020). Root mineral analysis did not occur in this trial and thus it cannot be determined if translocation of Na+ to plant roots happened in response to saline treatments. Leaf K/Na were calculated and showed that quinoa plants irrigated with salt had significantly lower ratios (< 5) than plants irrigated with non-saline water (> 65) (Fig. 3). These results are consistent with other studies where exposing quinoa to ~ 5.8 g L− 1 NaCl caused a leaf K/Na of less than 30 compared to non-saline treatments with a ratio over 120 (Cai and Gao, 2020). Hariadi et al. (2011) reported that in non-saline conditions the younger quinoa leaves have significantly higher K/Na (588) than old leaves (15.9) (Hariadi et al., 2011). Although, once the plants were exposed to saline conditions (> 17 g L− 1) they found this ratio changed and young leaves (1.9) recorded similar ratios to the old leaves (1.0). Despite salinity affecting the K/Na in this trial (Fig. 3) there were no differences in K/Na due to leaf age of quinoa. Significantly lower leaf K/Na for saline treatments are reflective of salinity tolerance mechanisms and indicate possible translocation of Na+ to roots (Cai and Gao, 2020).
Another salinity tolerance mechanism in quinoa is the production of organic solutes including the key osmolyte proline that assists plant osmotic adjustment under salinity stress (Adolf et al., 2013). Proline accumulation correlates with plant stress (Hayat et al., 2012) and stimulation of proline production can be triggered by salinity (Mini et al., 2019). Proline has been reported to be higher in saline affected quinoa (Cai and Gao, 2020, Maleki et al., 2018) although this is variety dependant. Whilst quinoa yield decreased under saline conditions (Table 2), proline concentrations were unaffected by salinity stress (Table 6). The quinoa variety used in this glasshouse trial was bred by Australian researchers (Dhammu, 2021) and salinity tolerance was not an original targeted breeding trait. The inability of the plant to increase proline may have caused salinity to impact yield in the variety used in the trial. Plant biomass, photosynthetic characteristics, and leaf Na+ content can differ between quinoa varieties of varying salt tolerance (Adolf et al., 2012). Saline tolerant varieties grown in northern Vietnam have been identified as having increased root length, leaf and branch numbers, and shoot dry weight (Nguyen Viet, 2016, Nguyen et al., 2020c). It is possible that growing Vietnamese varieties specifically bred for saline tolerance in the MRD may result in increased proline contents and further improve the species suitability for coastal regions affected by dry season saline intrusion.
Soybean yield decline in response to salinity was less than quinoa (Table 2) however it was only irrigated with 4 g L− 1 (~ 7.3 dS m− 1) for 28 days (Table 5). Soybean grew for a shorter duration, used more water daily (Table 3), and was flowering when quinoa was still in early vegetative stages (Fig. 2). Higher soybean growth rates, biomass production and yield consume more water as evapotranspiration rates increase to match plant growth (Purcell et al., 2007). The increased water use and consequential evapotranspiration potentially aided in the very negative soil solute potentials (-2 Mpa) for saline treatments (Fig. 2) and the advancement of leaf chlorosis and senescence when plants were irrigated with 4 g L− 1. Soybean has demonstrated sensitivity to Na+ with the greatest reductions in K/Na ratios occurring in the lamina with the application of NaCl (Le et al., 2021, Farhangi-Abriz and Torabian, 2018). Saline treatments caused a significant drop in leaf greenness at week 10 (Fig. 4), in addition to lower stomatal conductance rates (Table 4) which have been reported before in saline stressed soybean (He et al., 2016) as intracellular CO2 concentration decreases (Lu et al., 2009). Despite the reduced exposure to 4 g L− 1 of saline irrigation owing to the early senescence of soybean (Fig. 2), proline concentrations increased (Table 6) as plants attempt to limit oxidative damage (Hayat et al., 2012). Similar responses in proline concentrations have occurred for soybean plants under salinity stress at 3–6 g L− 1 (Sadak et al., 2019) although, proline production can be significantly influenced by soybean variety (Venkata Ramana et al., 2012). The sensitivity of soybean to salinity suggests that soybean is suitable for growth in areas of the MRD that are not affected by high levels of saline intrusion or soil salinisation.
Cowpea yield reductions under saline conditions (Table 2) were similar to soybean, however cowpea was exposed to maximum salinity (Table 5) for a longer duration than soybean. Cowpea is considered relatively salt tolerant and has developed salinity tolerance mechanisms. In this trial the leaf K/Na were unaffected by salinity treatments but differed between leaf age (Fig. 3). Saline treatments did not affect Na+ concentrations in the new leaves, however significantly higher Na+ concentrations occurred in older leaves of the saline treatment. Protection of young leaves by the transport of Na+ to older, sacrificial parts of the plant are possible mechanisms for tolerating salt exposure (Tester and Davenport, 2003). Consequently, older cowpea leaves often have higher concentrations of Na+ than younger leaves (Lacerda et al., 2006). The recent identification of genes that enable the coding for proteins involved in salt tolerance mechanisms in select lines of cowpea (Ravelombola et al., 2022) and the ability to osmotically adjust by accumulating Na+ in the roots (Le et al., 2021) highlights the suitability of cowpea for growth in saline affected areas of the MRD. Additionally, proline concentrations were significantly higher in saline treated cowpea leaves with the youngest leaves having nearly double the proline content of older leaves (Table 6). Proline accumulation in younger leaves is indicative of osmotic adjustment as potential transportation via the phloem occurs from older source leaves (Zegaoui et al., 2017). Proline accumulation in salt tolerant cowpea varieties can be nearly twice that of non-tolerant varieties (Mini et al., 2019). Whilst yield implications occurred for saline treated cowpea plants (Table 2), the evidence of Na+ compartmentalisation and proline accumulation suggest that cowpea may be suitable for growth in saline affected areas of the MRD.
4.3 Growth duration and tolerance during reproductive phases
The reproductive phases for quinoa and cowpea coincided with the application of saline water at ~ 7.3 dS m-1 (4 g L-1 NaCl) (Fig. 2) where soil EC1:5 exceeded 1.5 dS m-1 (Fig. 2) and soil solute potentials where more negative than − 1.5 MPa (Fig. 2). Whilst yield was negatively affected when both species were exposed to saline irrigation (Table 2), the total time spent growing in irrigation water with a salinity of ~ 7.3 dS m-1 (4 g L-1 NaCl) for quinoa and cowpea was 49 and 35 days respectively. Applying saline (10–20 dS m-1) irrigation during reproductive phases of quinoa can negatively impact yield but not terminate plant growth (Hussain et al., 2020). Similarly, cowpea has also demonstrated salinity tolerance during reproductive phases although varietal differences will determine the ability of cowpea to withstand saline conditions (Desire Taf et al., 2009). Düzdemir et al. (2009) reported no negative impact on cowpea yield when irrigated with saline water up to 4 dS m-1 but nearly 50% yield reduction occurred when water salinity increased to 7 dS m-1.
Soybean did not tolerate irrigation with ~ 7.3 dS m-1 (4 g L-1 NaCl) (Fig. 2), which it was only exposed to for 28 days until early death (Table 5). Premature senescence indicates ionic toxicity and results in loss of photosynthetic efficiency, consequently causing yield reductions from pod and flower abortion caused by toxic levels of Na+ (Otie et al., 2021). Salinity as a constant abiotic stress throughout the lifecycle of soybean plants will greatly affect biomass production and can even inhibit the ability of the plant to reach reproductive growth phases (Bustingorri and Lavado, 2011). Soybean flowering and pod setting is also the most sensitive growth stage to water stress with significant reductions in biomass and yield compared to water deficits occurring at seedling and branching stages (Wei et al., 2018).
Without other management intervention, short growth duration crop (< 50 days) will likely avoid the abiotic stresses seen in the later stage of the MRD dry season. All three of the investigated alternative crops in this trial had growth durations exceeding 50 days (Fig. 2). Consequently, salinity tolerance and/or greater water efficiency during reproductive phases are a priority in the identification of an alternative crop to be grown in a rice crop rotation in the MRD. Soil salinisation increases during the dry season as water evaporates and solutes accumulate in the topsoil (Apel et al., 2020, Hoa et al., 2019). Farmers are forced to irrigate with saline canal water as freshwater resources are depleted, further enhancing the salinisation process and the peak in soil salinity towards the end of the growing season could mean that plants are exposed to saline irrigation water exceeding 4 g L− 1 (Kaveney et al., 2023a). Consequently, the growth duration and tolerance of abiotic stresses during reproductive phases of alternative species must be considered when examining their suitability in the MRD. The ability of quinoa and cowpea to grow in saline conditions (Table 5) demonstrates their potential suitability for growth in areas of the MRD that could be exposed to these abiotic stresses, or alternatively as a crop option for sowing later in the dry-season. Soybeans requirement to avoid abiotic stresses during reproduction suggests its suitability for an early sown crop option or to be grown in areas were unaffected by salinity or water scarcity.
4.4 Locational suitability of alternative species in the MRD
The halophytic nature (Table 6) and water efficiency (Table 2) of quinoa make it suitable for areas of the MRD affected by salinity intrusion and freshwater scarcity. These areas are likely to be coastal provinces consistently exposed to dry season saline intrusion and where sluice gates are used to prevent further intrusion but limit the flow of freshwater. Due to regular failures of dry season rice crops, these areas have started to decrease from three to two rice crops a year with a fallow period in the dry season (Phung et al., 2020, MARD, 2020). Focusing on an early time of sowing to avoid the hottest parts of the dry season (Nguyen et al., 2020b), the ability to withstand possible waterlogging during that period and saline tolerant varietal selection will be critical to the successful growth of quinoa in the MRD.
Cowpea expressed some salinity tolerance mechanisms (Fig. 3 and Table 6) but is sensitive to intermittent irrigation (Table 2) and drought (Düzdemir et al., 2009), hence it is suitable for areas of the MRD that are saline affected but not water scarce. Inland provinces that are at risk of saline intrusion but not limited by the closing of sluice gates may be suitable for cowpea as an alternative crop, with multiple products of different markets; fodder, fresh peas, or grain. The opportunity of using drought tolerant cowpea genotypes with high nitrogen fixation potential (Yahaya, 2019) could also provide MRD farmers with future crop options that require lower fertiliser requirements and provide consequentially lower input costs for subsequent rice crops.
Soybean had mild tolerance to salinity (Fig. 3) and the highest daily water requirements of the crops studied (Table 3). Therefore, it may be suitable for areas further inland in the MRD, away from saline intrusion and with access to reliable quantities of irrigation water. Within these regions, there is potential to improve species performance by appropriate varietal selection (Lu et al., 2009, Nguyen et al., 2020c, Yahaya, 2019) and inclusion of associated management practices like soil moisture monitoring (Stirzaker et al., 2017) to improve WUE and yield.
4.5 The use of Chameleons
Chameleon soil moisture sensors were developed as a farmer friendly tool to monitor irrigation and soil moisture status (Stirzaker et al., 2017). Participatory learning of soil moisture status provides direct feedback via the colour coded Chameleon system (Fig. 1) which allows farmers to easily identify and understand crop irrigation requirements. An 80% change in smallholder farmer irrigation practice in Mozambique occurred in response to monitoring Chameleons to reduce the frequency or duration of irrigation (Chilundo et al., 2020). This increased crop production efficiency and decreased nutrient losses by leaching thereby resulting in both higher yield and farmer income. These changes, attributed to use of Chameleons, not only changed farmer irrigation practice but also facilitated the introduction of higher value crops in Tanzania (Mdemu et al., 2020).
Farmers in the MRD are experienced rice producers, with some farmers growing up to three rice crops a year (Nguyen et al., 2020b). Irrigation requirements between rice and upland alternative crops differ greatly and farmers’ unfamiliarity of growing alternative crops to rice may lead to over irrigation and reduced WUE (Kaveney et al., 2023b). This trial demonstrated a significant reduction (Table 3) in water use when Chameleons were used as the trigger for intermittent irrigation treatments rather than continuous irrigation practice. Cowpea was the only species to experience yield implications from intermittent irrigation (Table 2) owing to the sensitivity of cowpea to drier soil moisture conditions (Oumarou et al., 2015, Lima et al., 2019). Variable rainfall and freshwater scarcity during the dry season (CGIAR, 2016, Dang et al., 2020, IFRC, 2020) highlights the importance of maximising WUE through improved farmer irrigation schedules. Intermittent irrigation saved over 16% of the irrigation water used for quinoa and soybean (Table 3) without yield losses (Table 2). The combination of water efficient crops like quinoa (Algosaibi et al., 2017), and soil moisture monitoring instruments like Chameleons, could provide farmers in the MRD with suitable alternative crop options to rice for the drought affected regions during the dry season.
Limitations do exist with Chameleons as they are affected by salinity in excess of 4 dS m− 1 which causes calibration shifts and inaccurate readings of soil moisture (VIA, 2023). Soil EC exceeded 1.5 dS m− 1 for all three species when irrigated with saline water in this trial with solute potentials ranging between − 1 and − 2 MPa (Fig. 2). However, this occurred when irrigation water exceeded ~ 7.3 dS m− 1 (4 g L− 1) and although it did cause error in measurement of soil moisture, the Chameleons are programmed to identify errors likely to be due to excess salinity. The additional use of soil tensiometers negated experimental concerns, although tensiometers are not suitable in the MRD due to practicality and availability issues. When implementing the use of Chameleons in conjunction with alternative crops in MRD farming systems, caution should be taken towards the end of the dry season when soil and canal salinity levels are high (Phung et al., 2020). It is recommended that simple handheld EC meters should be used to test both the irrigation water and the soil EC (Rayment and Lyons, 2011) during that period to monitor soil salinity. The water saving benefits provided by monitoring soil moisture and irrigating effectively could result in sufficient water savings for irrigation at critical reproductive phases and potential avoidance of saline canal water. The yield and water saving benefits associated with Chameleon use in small holder farmers in Africa has been well documented (Svedberg, 2019, Moyo et al., 2020, Mdemu et al., 2020, Chilundo et al., 2020) and their implementation in the MRD as an irrigation monitoring tool should be strongly considered.
Intermittent irrigation and salinity affected the three species differently. Cowpea yield was negatively affected by intermittent irrigation, but displayed salinity tolerance mechanisms and completed reproductive growth stages when irrigation water was over 7 dS m− 1. Cowpea provides an alternative crop option for growth in the MRD in saline affected but not water limited areas, or if farmers are using a soil moisture device like Chameleon sensors to develop irrigation schedules, ensuring irrigation occurs before reaching − 50 kPa. Quinoa had the lowest daily consumption of water, was unaffected by intermittent irrigation and grew for the longest period with irrigation exceeding 7 dS m− 1. Quinoa could be suitable for growth in coastal areas affected by salinity and regions that experience freshwater scarcity. Soybean was also unaffected by intermittent irrigation however it grew for the shortest duration and required the highest daily water use, particularly during reproductive phases. Soybean also did not tolerate irrigation with 4 g L− 1 despite being irrigated with this salinity concentration for the shortest period at 28 days with leaf chlorosis and senescence occurring. Earlier sowing of soybean to avoid abiotic stresses during reproductive phases or utilising a soil moisture monitoring tool like Chameleon sensors may help farmers avoid yield losses or crop fatalities when growing soybean in the MRD. The use of Chameleon soil moisture sensors worked effectively to establish irrigation schedules that enabled water savings of over 16% without yield compromise. Limited by high levels of soil salinity, caution must be used when implementing Chameleons in highly saline affected areas of the MRD. This research demonstrates that knowledge of WUE, growth duration and ability to tolerate salinity can be used to assess the suitability of crops for the cropping calendar and conditions of the MRD Vietnam.