Bread wheat response to saline stress across four experimental settings

Yield, growth and physiological response of bread wheat to saline stress in raised beds outdoors, pots outdoors, greenhouse and climate chamber.

All experiments were conducted on spring wheat (Triticum aestivum) cv. Tybalt (Limagrain, Saint Beauzire, France, https://www.limagrain.com/). Seeds were sown in wet compost in individual wells of a germination tray in a greenhouse and transplanted to the different settings a week after sowing.

The greenhouse pots (GH), outdoors pots (OP) and raised bed (RB) experiments took place from February to July 2022 in the experimental “Skardon” garden at the University of Plymouth, Plymouth, United Kingdom (50° 37’ N 4° 14’ W), all within 15 m of each other. The four raised beds and OP experiment were located next to each other (with the pots equally split in four groups at the edge of each raised bed under a rain out shelter in the form of a polytunnel that could be opened and closed manually. The shelter was open for the majority of the experiment, so the plants experienced natural outdoors conditions; it was closed shortly before heavy precipitation to avoid rainwater draining the salt from the pots or too deep in the soil of the raised beds. Caution was taken for the shelter to always be open under sunny weather to avoid creating a greenhouse effect artificially increasing temperature. The climate chamber (CC) experiment was conducted in a Sanyo growth chamber (photosynthetically active radiation 183 µmol m^-2 s^-1) after the other experiments were completed. Photoperiod and day and night temperatures were adjusted weekly to match the average of the conditions in the GH experiment.

For CC, GH and OP, three seedlings were transplanted in 3 L black plastic square pots (15 cm x 15 cm at the top and 20 cm high) containing 3 kg of a mix of 50% Devon farming soil, 40% soil from the experimental garden and 10% compost (Peat Free Multi-purpose with added John Innes, Westland Horticulture Ltd., Dugannon, Northern Ireland, UK). Mixing of the different components of the growth medium was carried out in the top 50 cm of the raised beds and pots were filled with that same medium. Pots were saturated with water and left to drain to establish full pot water capacity. For the RB experiment, seedlings were transplanted at a density of 220 plants m^-2, which is within the range of density for wheat in the UK.

Pots and raised beds were manually weeded as necessary. Fertilisation was provided in the form of a 0.75% dilution of Wilko® lawn feed (15-3-3 NPK with iron), two doses of 50 mL per pot or 2 L per raised bed (equivalent volume per surface) during the vegetative stage.

Stress treatment

In all experiments, the salt stress treatment started at the two tiller-stage, approximately a month after sowing, with the exact time varying slightly between experiments (between 28 and 36 days after sowing), as seedlings grew faster in GH and CC. Salt stress was applied using a solution of sea salts (Instant Ocean®, Blacksburg, VA, USA, https://www.instantocean.com/) that represents better the mix of ions found in brackish water and saline soils than pure NaCl, dissolved in rain water for RB, OP and GH, and in distilled water for CC; rain water and distilled water were also used for control watering. Each pot was watered with 50 mL of salt solution and each raised bed with 2 L of salt solution, which amounted to the same volume per surface three times a week. The concentration of salt was gradually increased by 2 g L^-1 increments over three weeks until the final concentration of 22 g L^-1 (27 dS m^-1) was reached. This was done to mimic the slow increase in soil solution salinity that takes place during the growing season in saline agricultural fields and avoid osmotic shock (Shavrukov, 2013). The salt stress treatment was maintained until there was no more green leaf tissue visible in stressed plants, which amounted to 92 days of treatment for GH, 94 days for CC and OP and 110 days for RB. Soil water content was maintained at 70% pot water capacity in both control and stressed pots from transplanting until all plants in a pot had no more green tissue visible. Comparable soil moisture was obtained in the RBs as in the pots using a moisture meter (HH2 ThetaProbe, Delta-T Devices Ltd, Cambridge, UK, www.delta-t.co.uk).

Soil salinity was measured at the end of the experiment at different depths for the salt treatments using the EC1:5 method (He et al., 2012). For pots, soil samples were taken from the first two top cm, 5 to 7 cm deep and 12 to 14 cm deep (bottom of the pot); for the raised beds, soil samples were taken from the first two top cm, 4 to 9 cm deep, 10 to 14 cm deep and 23 to 28 cm deep in at least two replicates for each setting. The baseline EC1:5 value for the soil used in the experiment was 0.35 dS m^-1. Similar results were obtained for the pots in all three settings, with EC1:5 values around 21 dS m^-1 for the top two cm of soil, 3.5 dS m^-1 at middle depth and 7.5 dS m^-1 at the bottom of the pot. For the raised beds, the top two cm had an EC1:5 of 6.8 dS m^-1, 2.4 dS m^-1 at 4 to 9 cm deep, 1.4 dS m^-1 at 10 to 14 cm deep and 1.2 dS m^-1 23 to 28 cm deep.

Plant sampling, growth and yield measurements

Plant height (from the soil to the end of the longest leaf) was recorded weekly from the start of the stress treatment for six randomly selected plants per treatment in GH and CC, for eight plants in GP and 16 plants in RB. More plants were measured for OP and RB to account for the added variability created by the split into two blocks per treatment and the highest variability observed for plant size in RB.

Four plants per setting and treatment were harvested at three different times during the experiment: after 23 to 25 days of stress (tillering stage for CC, OP and RB, stem elongation stage for GH), after 29 to 31 days of stress (tillering stage for CC, end of tillering stage for OP and RB, stem elongation stage for GH) and after 46 to 49 days of stress (tillering stage for CC, start of booting stage for OP and RB, end of booting stage for GH). Whole shoots were harvested, rinsed to remove salt from stress treatment watering, with the ears removed when they were present and the shoots without ears were frozen in liquid nitrogen. Sampling was always carried out around 9 am to avoid circadian variations in metabolites between sampling points. For CC, GH and OP only one plant per pot was harvested over the three sampling times for each pot, so at maturity there were two plants in each pot.

Harvesting was done once plants had reached maturity. For CC, GH and OP all remaining plants (18 per treatment, minus six for OP control and four for OP salt because of an undiagnosed disease that affected yield); for RB 30 plants (none from the edge of the plot and none with the same undiagnosed disease as some GP plants) per treatment were randomly harvested. Plants were dried in an oven overnight at 60°C to evaporate any remaining humidity. Shoot biomass was recorded, and for each plant, spikes and grains were counted and weighed.

Shoot tissue analyses

Whole shoots were freeze dried for at least 48 h, ground into a fine powder using a coffee grinder and stored in dry conditions.

Sodium and potassium were quantified using flame photometry. The method was adapted from (Munns et al., 2010). About 15 mg of tissue powder was digested in 2 mL of 0.5 M nitric acid for one hour at 80°C. After 10 min of centrifugation, the supernatant was collected and diluted 5 times (control samples only for sodium analysis) and 50 times. Diluted samples and standards were quantified using a flame photometer (Model 420, Sherwood, Cambridge, UK, sherwood-scientific.com).

Proline quantification was adapted from (Shabnam et al., 2016). Twelve to 15 mg of tissue powder were extracted with 40% ethanol, 10 mM ascorbate for 18 h at 4°C. After 10 min of centrifugation, 33 µL of the supernatant was mixed with 66 µL of 1.25% w/v ninhydrin in glacial acetic acid in 200 µL PCR tubes and incubated at 100°C for 30 min. This was replicated for each sample replacing the ninhydrin reagent with glacial acetic as a “sample blank” to assess the baseline colouration (independent of the ninhydrin reacting with proline) of each sample. The content of each tube was transferred to a 96 well plate and absorbance at 520 nm was measured in a plate reader (Spectramax Plus 384, Molecular Devices, San Jose, CA, USA, moleculardevices.com/). Proline concentration was calculated using the values for standards included in each set of reactions and by subtracting the absorbance value for the “sample blank”.

Chlorophyll extraction was modified from Sims and Gamon (2002). About 12 mg of dry leaf powder was extracted twice with 1.8 mL of 80:20 (v/v) acetone Tris buffer (pH 7.8), shaking for 5 min, followed by 5 min of centrifugation. Absorbance was measured at 537, 647 and 663 nm and chlorophyll concentration was calculated using the equations provided by Sims and Gamon (2002).