Final Report 2 --

Efficacy of Ac Saltlander in Saline Soils by Ryan Hildebrand Student# 7799489

Advised by Steve Sager-AAFC Soil Resource Specialist

A final report submitted to the Department of Environment and Geography, University of Manitoba, In partial fulfillment of the requirements for course ENVR 4500 (Honours Thesis Project)

I

Abstract Agriculture and Agri-Food Canada trials for AC Saltlander Green Wheatgrass (ACS) (Elymus hoffmanii) in southern Manitoba have tested biomass, weed suppression and forage quality in various salinity conditions. Salinity was measured in millisiemens per meter using an Em38 ground conductivity meter and calibrated using electrical conductivity results from soil samples. Subsequently, vegetation samples were collected to determine biomass, and forage quality. Weed suppression data was collected using qualitative field assessment reports. The average yield from all treatments collected in first cut (June 19th) is 6,690 Kg/ha. The second cut (September 6th) biomass had an average total biomass of 4965 kg/ha. Weed suppression was rated as high in 2017 and was reduced to low rating in 2018. Forage quality testing parameters show ACS has a protein content was ranked as supreme, premium TDN in high salinity/ good TDN in low salinity, fair ADF, and utility NDF.

II

Acknowledgements I would like to thank Agriculture and Agri-Food Canada for employing me as a coop student and giving me the experience needed to succeed in the workplace. A thank you to my boss and advisor Steve Sager for teaching and guiding the direction of this report. I would like to thank my colleagues in the Soil & Water program for assistance at various stages of the project. Additionally, the work could not be done without the help of Mae Elsinger and her team from Brandon collecting, weighing, and organizing the data collection. Lastly, I would like to thank Bill Houston for providing me the necessary data and reference reports needed to complete this project.

III

Table of Contents Abstract .............................................................................................................................................1 Acknowledgements............................................................................................................................2 List of Figures.....................................................................................................................................4 List of Tables ......................................................................................................................................4 Introduction ......................................................................................................................................5 Background ............................................................................................................................................... 5 Cause of Salinization ................................................................................................................................. 6 Research Purpose and Objectives .......................................................................................................7 Background Literature........................................................................................................................8 Biomass ..................................................................................................................................................... 8 Weed Suppression .................................................................................................................................... 9 Forage Quality ........................................................................................................................................... 9 Methods .......................................................................................................................................... 11 Site Description ....................................................................................................................................... 11 Local Geology .......................................................................................................................................... 11 Biomass ................................................................................................................................................... 11 Weed Suppression .................................................................................................................................. 15 Forage Quality ......................................................................................................................................... 15 Management........................................................................................................................................... 16 Results............................................................................................................................................. 16 Biomass ................................................................................................................................................... 16 Weed Suppression .................................................................................................................................. 17 Forage Quality ......................................................................................................................................... 18 Management........................................................................................................................................... 20 Discussion........................................................................................................................................ 20 Biomass ................................................................................................................................................... 20 Weed Suppression .................................................................................................................................. 20 Forage Quality ......................................................................................................................................... 20 Management........................................................................................................................................... 21

IV Appendix 1.0- Hypothesis testing ..................................................................................................... 22 References....................................................................................................................................... 25

List of Figures Figure 1 yield salinity curve ................................................................................................................................... 8 Figure 2 comparison of fields with and without ACS ............................................................................................ 9 Figure 3 salinity maps ......................................................................................................................................... 12 Figure 4 regression of em38 data and soil sample data ..................................................................................... 13 Figure 5 ec maps ................................................................................................................................................. 14 Figure 6 average biomass ................................................................................................................................... 17 Figure 7 relationship between salinity and yield................................................................................................. 17 Figure 8 forage quality results ............................................................................................................................ 19

List of Tables Table 1 mean forage quality swiftcurrent .......................................................................................................... 10 Table 2 forge quality guidelines .......................................................................................................................... 15 Table 3 forage quality mean with std error ........................................................................................................ 19

V

Introduction Background Salinization has been a challenge in agriculture production for centuries. In ancient Mesopotamia (226-637ce), massive irrigation channels were made to irrigate low lying fields. Early summer flow from melted mountain snow would often flood the land. Producers in Mesopotamia noticed that when the water evaporated the soil was white and unproductive. The poor soil conditions would lead to food shortages in Mesopotamia and inevitably urban decline in the southern regions (Irrigation and Soil Salinization, 2015). The Mesopotamian’s were experiencing the process of salinization, defined as the increase in levels of soluble salts in soil. When there is an increase in salinity plants ability to uptake water is decreased, contributing to the plant using more energy which reduces yields. Today the issue remains, and every year $226 million is lost due to salinization impacting 4 million hectares in Canada annually (Wiebe, 2007). Furthermore, salinization is an issue locally, as a large proportion of southern Manitoba is in the Chernozemic soil zone. Chernozemic soil has properties that are prone to moderate to severe salinity issues. Therefore, effective solutions are needed to keep up with growing food demands (Amacher, 2000). Agriculture and Agri-Food Canada is conducting trials for AC Saltlander Green Wheatgrass (ACS) in different salinity ranges at Morden Research and Development Centre (MRDC). ACS was developed in 2004 by Dr. H. Steppuhn at the Semiarid Prairie Agricultural Research Center (SPARC) in Swift Current, Saskatchewan. ACS is a naturally occurring hybrid wheatgrass between Pseudoregneria spp. and Elymus repens. Chosen for its deep roots, high

VI yields, nutrient content, and salinity tolerance. Research trials testing the biomass, forage quality, weed suppression and management methods are taking place in Manitoba, Saskatchewan and Alberta research facilities. (Kayter, 2017).

Cause of Salinization Salts are unbound when precipitation infiltrates soil weathering soluble salts from minerals. Lateral flow in groundwater transports soluble salts causing accumulation of salts in depressions. Fluctuations in the water table leaves salts in soil, when temperatures increase the water evaporates leaving behind high salt concentrations (Kayter, 2017). Physical properties in soil such as soil texture, soil structure, and clay type impact the movement and retention of salts in soil. The texture of clay soils is more susceptible to salinization due to the small particle size and large surface area, resulting in salts and nutrients being held. Conversely, sandy soils have a large particle size and a small surface area, resulting in salts and nutrients being easily flushed out of the soil. Additionally, the three clay types montmorilonite, illite, and kaolinite have different structures impacting clay swelling and dispersion. Montmorilonite soil types are the most heavily impacted by salinization, due to having a weakly bound 2:1 phyllosilicate structure allowing water to enter. When the water that enters the clay dissipates it leaves behind dissolved salts. Kaolinite clays are a 1:1 phyllosilicate, thus, they are less likely to retain water and are less prone to undergo salinization (Pearson, 2003). Climate change is creating an increase in the amount of arid and semi-arid zones. Therefore, producers will become more heavily reliant on irrigation to produce enough food for a growing population (USGS, 2017). Heavy irrigation can make the water-table rises increasing the likelihood of surface salinity. Additionally, 55% of irrigation water is sourced from groundwater that is high in salinity. This can be highly problematic if proper water testing is not completed before water is used for irrigation (Yeo, 1998).

VII

Research Purpose and Objectives 1) How does each treatment compare in biomass? •

Changes in biomass from first cut to second cut?

•

How does salinity impact yield?

2) How does ACS perform supressing weeds? •

Is foxtail barely, Pigweed, Lambsquarters and Japanese Brome reduced in the site?

3) What is the forage quality of ACS? •

How does ACS forage quality change from the first cut to the second cut?

•

How does ACS quality compare to Slender wheatgrass (SWG) in different salinity conditions?

•

Does salinity level affect ACS forage quality?

4) How does different management practices impact yield? (2019 study) How does N only application impact yield? How does N&P impact yield? How does different tillage methods impact yield?

VIII

Background Literature Biomass Yield testing in Weyburn Saskatchewan determined that ACS mean dry matter (DM) biomass from 2007-2010 was 3000 kg/ha. The study had a drought year in 2009, which likely impacted yields. Comparatively, the study found ACS yields to be higher than Smooth Brome grass, which averaged 2500 kg/ha (Kayter, 2017). The literature shows that the average limitation zones for salinity:

•

no limitations 95 % 100% yield (0-3 dS m-1 )

•

slight limitations 77%-95% yield (3-5 dS m-1 ),

•

moderate limitations 55%-77% yield (5-7 dS m-1 ),

•

severe limitations 22%-50% yield (7-11 dS m-1 ),

•

extreme limitations 0-20 % yield (>12 dS m-1 )

Figure 1 yield salinity curve (Kayter, 2017)

IX

Weed Suppression Two Research sites in Alberta have done preliminary testing on ACS weed suppression. Initial results of the study found when ACS was planted with row spacing, with slender wheatgrass seeded in interspersed rows a 99% foxtail Barley control was achieved. Additionally, ACS suppresses

the

winter

annual

grass

Downy

Brome

(Bromus

tectorum

L.).

Figure 2 comparison of fields with and without ACS

Forage Quality The literature describes the main variables used for determining forage quality as Crude protein, Acid Detergent Fiber (ADF), Neutral Detergent Fiber (NDF), and Total Digestible Nutrients (TDN). Having high protein feed that can meet the needs of livestock is important because it impacts the amount of growth, lactation, and reproduction (Parish, 2008). ADF is the measure of the forage that is least digestible by livestock (cellulose and lignin). Subsequently, ADF has a high percent ADF results in less forge that can be digested. Additionally, NDF is the measure of the

X plants slowly digestible structural components, thus higher NDF results in less consumption. Generally, the smaller the NDF percentage, the better the feed, however, too low of NDF can cause digestion issues. TDN is derived from the Western Hay Equation using ADF to determine the percent of forages that can be digested by livestock (Peter Robinson, 1998). The variables above allow for prediction on how much livestock can eat before satiation, resulting in better management of energy maintenance and animal value (Saun, 2013). Lastly, if the cut too late or too early forage quality can suffer. The literature recommends cutting shortly after heading for best quality (Manitoba gov, 2017).

In 2001 and 2002 experiments on forage quality were conducted in Swift Current Research Centre. The results showed that forage quality in ACS had lower average protein content than Carlton Smooth Bromegrass. Additionally, the experiment did find ACS had a higher NDF and ADF reading (Steppuhn, 2006).

TABLE 1 MEAN FORAGE QUALITY SWIFTCURRENT

XI

Methods Site Description The Morden ACS study has an area of 8.5 acres, located on the North East side of the MRDC. Five different forage treatments, with 3 repetitions were seeded on June 26 th, 2017. Seeding in mid-May was missed due to equipment access. The Equipment used to seed was a 12ft Great Plains drill, at ½ inch seeding depth. The 2017 spring was extremely dry, the first precipitation event was 14mm on June 29th.

Local Geology The MRDC’s local geology consists of Mesozoic rock, which is mainly shale with trace amounts of sandstones, carbonates and evaporates. Parent material of the MRDC soil consists of glacial till, alluvial and lacustrine deposits. The ACS trial is located on the Horndean series, the soil is classified as an imperfectly drained gleyed black chernozem, ranging in texture from fine sand to silty clay loam. The ACS site topography is 963 meters above sea level, with the land sloping towards the east at a 0.3% gradient. Local hydrology consists of groundwater that is likely saline, as is majority of groundwater in aquifers running through southwestern and south-central Manitoba (Stantec, 2014).

Biomass In preliminary planning of the ACS trial. use of the Veris 3100 conductivity reading system was discussed as an alternative method. The two methods correlate with salinity; however, some key differences made the EM38 more suitable. The EM38 is non-contact sensor based on electromagnetic induction technology. The EM38 applies a current from a transmitter coil and is measured as apparent EC by a receiving coil. The Veris uses rotating metal discs 6cm in the soil.

XII An electric current is applied while the device is pulled through the soil. Consequently, the veris is better suited for large area use, while the EM38 is better suited for showing salinity of small areas. Consequently, the Em38 was selected for use in this study. With both devices numerous factors can influence the reading, soil moisture content, texture, bulk density and soil temperature can skew readings (Agronomics, 2004). The EM38 salinity data was collected by attaching a sled to a UTV and dragging the device across the plots. The device was connected to a Trimble GPS unit, and set to collect 1m/05m readings every 5 meters. The device applies an induction current through soil to measure ECa in mS/m. The data was then uploaded to ArcGIS and interpolated using the inverse distance weighting (IDW) method. Using the salinity map to determine high and low salinty, vegetation samples were collected. Three low and three high points in each plot were selected for vegetation samples. To confirm the validity of the EM38 data, 10 soil samples were sent to AGVISE labs. At AGVISE a saturated paste was completed to get pH, Ca, Mg, Na, SAR, and Electrical conductivity (Sager, 2018). Regression analysis was used to compare the EM38 to the soil sample results. The 1-meter reading shows a R2 = 0.9466 and at 05 meter depth R2 = 0.8675. Accordingly, the data collected by the EM38 shows a high enough correlation to use for vegetation sampling

XIII . Vegetation samples were collected on June. 19th & 20th for the first cut and September. 6th & 7th for the second cut. Additionally, to track the growth throughout the season game cameras were installed and set to take pictures every hour. Using a 0.25m² square, vegetation was trimmed 2 inches above the surface and placed into mesh bags. 3 samples were collected from high salinity and 3 from low salinity throughout the 15 treatments. Samples were then dried in the oven at 60 degree C for 48 hours and then weighed to determine the biomass. Lastly, a detailed assessment was conducted during collection of the samples to determine percent cover, percent of each grass species, and weed cover (Sager, 2018).

Figure 4 regression of em38 data and soil sample data

XIV

Figure 5 ec maps

XV

Weed Suppression The 2017 season had large quantities of weeds including: Foxtail Barley, Lambs Quarters, Japanese Brome and Stunted Redroot pigweed. The first herbicide application of roundup took place 2 days after seeding. Followed by the herbicide Pardner on July 21 st, the application was effective against lambs quarters; however, Redroot pigweed was only stunted but not eliminated. The last herbicide application of Barsagran had mixed results against common Pursane and Redroot. Japanese Brome was discovered in mid September and was rouged from the plots by hand, an estimated 1200 plants were pulled from the plots. The 2018 growing season saw a dramatic reduction of weeds, mainly due to a better growth from the forages outcompeting weed populations (Sager, 2018).

Forage Quality First, using the EM38 data, three low and three high points in each plot were selected for vegetation sample locations. Vegetation samples were taken in June and September simulating a first and second cut biomass. ACS and Slender Wheat Grass vegetation sample were sent to a lab in Saskatchewan to be analyzed for protein content, Acid Detergent Fiber (ADF), and Neutral Detergent Fibre (NDF). Using ADF and the western hay equation TDN was derived (Peter Robinson, 1998). The USDA guidelines for forage quality is used to determine the quality classification (USDA, n.d.).

TABLE 2 USDA QUALITY GUIDELINES

XVI Management Current plans for management methods include nitrogen, nitrogen and phosphorous, and a tillage treatment in the 2019 growing season. The management methods will run north/south across a high salinity zone, medium salinity zone, and low salinity zone. Soil and vegetation samples will be taken in each salinity zones to determine the efficacy of each management method.

Results Biomass Changes in biomass from first cut to second cut? At a 95 % confidence it was determined that there is a statistically significant difference from the first cut to the second cut in terms of biomass (p-value=8.1029E-10) (appendix 1). The average yield from all treatments collected in first cut (June 19th) is 6,690 Kg/ha. June biomass samples have less deviation in weight between highly saline and low saline samples compared to the September biomass samples. The second cut (September 6th) biomass had an average total biomass of 4965 kg/ha. In the low salinity condition, AC Saltlander at 10 lbs/acre had the highest biomass (6189 Kg/ha), and ACS Alf had the highest biomass in the high salinity condition (6872 Kg/ha). High Salinity

kg/ha

Average Biomass 9000 8000 7000 6000 5000 4000 3000 2000 1000 0

ACS 5 lb

ACS 10 lb

ACS & SWG ACS & ALF

High Salinity

6516

6832

6635

Low Salinity

7926

7700

6744

Low Salinity

Mix 5

ACS 10 lb

ACS & SWG

6872

5984

3121

3797

4830

5246

5635

5866

5832

6180

5418

5341

5068

5012

First cut-June 2018

ACS 5 lb

Mix 5

ACS & ALF

Second cut-September 2018

XVII Figure 6 average biomass

The impact of salinity on yield showed similar trends to what was observed in the literature. With little to no impact from 0-4 dS/m, moderate yield reduction from 4-7 dS/m, and severe reductions greater than 8 dS/m. However, there is weak R2 value (0.3825) and a hypothesis test have determined there is not a statistically significant relationship between salinity and yield (pvalue=0.07).

Figure 7 relationship between salinity and yield

Weed Suppression From the detailed assessment, a qualitative rating for severity of weeds was determined. Before seeding, the assessment rated the field as high in weed cover. In the first year of seeding the assessment of the field showed a slight reduction in weed cover. The 2018 study saw the weed severity reduced the rating to a low rate of weed cover.

XVIII

Forage Quality Is there a difference in forage quality between the first and second cut for ACS? The protein for ACS was higher in the second cut (14.76%) and lower in the first cut (12.44%). The ADF is higher in the first cut (37.63%) and lower in the second cut (28.92%). NDF was higher in the first cut (66.44%) and lower in the second cut (54.68%). The TDN in the first cut was slightly higher in the first cut (56.67%) and lower in the second cut (59.24%). There is a statistically significant difference between the first cut and the second cut TDN (p=0.013). Is there a difference in forage quality between ACS and SWG? The protein content was found to be higher in ACS (13.6%) than SWG (13.32%). ADF was similar in ACS (33.28%) and SWG (33.65%). NDF was also found to be similar in ACS (60.56%) and SWG (55.99%). Lastly, TDN was found to be 59.46% in ACS and 56.34 % was found in SWG. There is a statistically significant difference between the TDN of ACS and SWG with a p-value of 0.02. (fig.8). Is there a difference in forage quality between high and low salinity? The mean forage quality with standard error between high and low salinity is shown in table 3. There is a statistically significant difference in TDN in high salinity regions vs low salinity regions with a p-value of 0.03.

XIX

Figure 8 forage quality results

TABLE 3 FORAGE QUALITY MEAN WITH STD ERROR (MORDEN)

XX

Management The management methods are planned for the 2019 growing season, accordingly there will be no results from this year

Discussion Biomass The finding from the 2018 biomass samples show that ACS had exceptional biomass for the 2019 growing season. The biomass showed to be substantially higher in yield in the first cut, compared to the second cut. However, the first cut had lower overall forage quality demonstrating that the crop was too mature during the first cut. Additionally, when yield data was compared to biomass data, little to no impact was found in 0-4 (dS/m), moderate yield reduction from 4-7 (dS/m), and severe reductions greater than 8 (dS/m). ACS ability to produce high yields in saline conditions will prove useful as climate change worsens salinity and increase demand for quality livestock feed continues to rise.

Weed Suppression From the 2017 growing season to the 2018 growing season qualitative analysis showed a substantial reduction in Foxtail Barely and Downy Brome. By using ACS to control weeds longterm application of herbicides glyphosate, Imazapic, and Indaziflam can be avoided. This benefits producers by reducing the costs of herbicide application, and by allowing producers to make profits off forages in an area that would traditionally offer low yields (Harold Steppuhn, 2017).

Forage Quality From the first cut to the second cut forage quality performed better in the second cut in most quality

XXI parameters. Average forage quality between both cuts shows that ACS out performed SWG in the high salinity for protein content, and SWG was slightly higher in low salinity. According to the USDA protein quality guidelines, protein for both forage species was ranked as supreme quality (Table 2). ADF in ACS and SWG vary by less than a percent in high salinity and is higher in SWG in low salinity. Therefore, ACS performed better than SWG in terms of the low salinity ADF. The guidelines for ADF ranked ACS as fair in both salinity classes, and SWG as fair in high salinity and utility in low salinity. The ACS NDF was higher in both salinity classes than SWG levels. Hence, ACS performed worse than SWG in terms of NDF content. Lastly, ACS TDN was found to be higher in both salinity classes than SWG. The TDN in ACS received a ranking of premium in high salinity and a good rating in low salinity. Lastly, the first cut saw a higher biomass, but a lower forage quality. Therefore, the first cut was cut at too late of stage.

Management The use of Nitrogen, phosphorus and a tillage treatment will be used to evaluate the best management strategies for ACS. The analysis will be compared to the data we have collected in 2018 to determine if the different management methods impact biomass, weed suppression, and forage quality.

Summary The most successful treatment for biomass was in low salinity was ACS 5lbs/acre and with lower salinity content ACS mixed with alfalfa is the most productive. Salinity tolerant weeds saw a dramatic reduction from 2017 to 2018. Additionally, ACS protein content was ranked as supreme, premium TDN in high salinity / good TDN in low salinity, fair ADF, and utility NDF. The high NDF rates have potential to slow digestion process of nutrients. However, the high

XXII protein content and TDN content will give livestock nutritional value needed for growth. Lastly, the management methods in 2019 plan to increase the overall quality and yield for ACS.

Appendix 1.0- Hypothesis testing Biomass Is there a significant difference between first cut yields and second cut yields?

Ho = first cut yield =second cut yield Ha ≠ first cut yield ≠ second cut yield P=<0.05 reject null hypothesis Therefore, there is a significant difference between the first cut and the second cut biomass. Is there a significant difference between salinity and biomass?

Ho = high salinity yield = low salinity yield Ha= high salinity yield ≠ low salinity yield P(0.070)>0.05 fail to reject the null hypothesis Therefore, there is not significant difference between high salinity biomass and the low salinity biomass

XXIII

Forage quality Is there a significant difference between the first cut forage quality (TDN) and second cut total digestible nutrients?

Test Statisticsa second cut first cut Z Asymp. Sig. (2-tailed)

-2.490b .013

a. Wilcoxon Signed Ranks Test b. Based on positive ranks.

Ho= first cut forage quality= second cut forage quality Ha= first cut forage quality ≠ second cut forage quality P=.013<0.05 reject the null hypothesis Therefore, there is a significant difference between the forage quality from second cut to first cut at the Morden site. Is there a difference between ACS and SWG in terms of forage quality (TDN)?

Ho= ACS TDN different than SWG TDN HA= ACS TDN no difference from SWG TDN P=(.02)<0.05 reject the null hypothesis Therefore, there is a significant difference between the forage quality in ACS to SWG.

XXIV

Is there a difference in forage quality between high and low salinity?

Ho= High salinity TDN is different than low salinity TDN HA= High salinity TDN no difference from low salinity TDN P(.003)<0.05 reject the null hypothesis Therefore, there is a significant difference between the forage quality for low to high salinity.

XXV

References AAFC. (n.d.). Retrieved from https://www.cattlemen.bc.ca/docs/ac_saltlander_factsheet.pdf Agronomics, A. (2004). Amacher. (2000). Retrieved from https://www.researchgate.net/publication/237537841_SALINITY_AND_PLANT_TOLERANCE Harold Steppuhn, a. K. (2017). Retrieved from http://www.nrcresearchpress.com/doi/full/10.1139/cjps-20170057#.W1ndh0xFxaQhttp://millerseeds.com/pdf/Foxtail_Results.pdf Irrigation and Soil Salinization. (2015). Retrieved from World History: https://www.worldhistory.biz/ancienthistory/62007-irrigation-and-soil-salinization.html Kayter. (2017). AC-Saltlander:A tool for dealing with those saline areas. Weyburn: AAFC. Manitoba gov. (2017). Retrieved from https://www.gov.mb.ca/agriculture/crops/production/forages/gettingstarted-with-intensive-grazing.html Parish, D. J. (2008). Protein in Beef Cattle Diets. Pearson, K. E. (2003). Retrieved from http://waterquality.montana.edu/energy/cbm/background/soilprop.html Peter Robinson, D. P. (1998). California Forage Review. Retrieved from file:///I:/Soil_and_Water/Sagers/Hildebrand2018/AC%20saltlander/ac%20saltlander%20report/InterpretingFQReport.pdf Sager, S. (2018, August ). (R. H, Interviewer) Saun, R. (2013). Retrieved from https://extension.psu.edu/effects-of-forage-quality-on-a-camelid-feedingprogram Saun, R. (2013). Retrieved from https://extension.psu.edu/determining-forage-quality-understanding-feedanalysis Stantec. (2014). Detailed Soil Survey of the Morden Research Centre. Morden. Steppuhn, H. (2006). AC Saltlander green wheatgrass. Retrieved from http://www.nrcresearchpress.com/doi/pdfplus/10.4141/P05-160 USDA. (n.d.). Retrieved from https://www.ams.usda.gov/sites/default/files/media/HayQualityGuidelines.pdf USGS. (2017). Retrieved from https://water.usgs.gov/edu/wateruse-trends.html

XXVI Wiebe. (2007). Retrieved from http://www.usask.ca/soilsncrops/conferenceproceedings/2015%20pdf/2015%20posters/019-poster-gatzke.pdf Yeo. (1998). Retrieved from https://www.researchgate.net/publication/223784201_Predicting_the_interaction_between_the_ef fect_of_salinity_and_climate_change_on_crop_plants