Salt tolerance of seven genotypes of zoysiagrass (<i>Zoysia</i> spp.)

Triston Hooks; Joseph Masabni; Girisha Ganjegunte; Ling Sun; Ambika Chandra; Genhua Niu; Triston Hooks; Joseph Masabni; Girisha Ganjegunte; Ling Sun; Ambika Chandra; Genhua Niu

doi:10.48130/TIH-2022-0008

2022 Volume 2

Article Contents

Abstract

References

Next Previous

ARTICLE Open Access

Salt tolerance of seven genotypes of zoysiagrass (Zoysia spp.)

1.
Texas A&M AgriLife Research and Extension Center, 17360 Coit Rd, Dallas, TX 75252, USA
2.
Biosystems Engineering Department, University of Arizona, 1951 E. Roger Road, Tucson, AZ 85719, USA
3.
Texas A&M AgriLife Research, 1380 A&M Circle, El Paso, TX 79927, USA
4.
Department of Agriculture, Collin College, 391 Country Club Road, Wylie, TX 75098, USA

More Information

Corresponding author: genhua.niu@ag.tamu.edu

Received: 06 October 2022
Accepted: 24 November 2022
Published online: 13 December 2022
Technology in Horticulture 2, Article number: 8 (2022) | Cite this article

Abstract

Seven zoysiagrass genotypes were evaluated for salt tolerance in a greenhouse study. The plant materials included Zoysia matrella 'Diamond', Z. japonica 'Palisades', three Z. matrella × Z. japonica hybrids DALZ 1701, DALZ 1713, and 'Innovation', and two Z. minima × Z. matrella hybrids (DALZ 1309 and 'Lazer'). Treatments included a control (nutrient solution) and two saline treatments representing moderate and high salt levels. The electrical conductivity (EC) was 1.3 dS m⁻¹ for control and moderate (EC5) and high salinity (EC10) were 5.0 and 10.0 dS m⁻¹, respectively. At the end of eight-weeks of treatment, the relative (percent control) shoot dry weight (DW) was greatest in 'Diamond' in EC10, and the relative root DW was greatest in DALZ 1309 in EC5. A cluster analysis based on the relative tissue dry weight identified 'Diamond', DALZ 1309, and DALZ 1713 as the most salt tolerant genotypes. Additionally, the green leaf area (GLA) index of 'Diamond' and DALZ 1713 were 98.8% and 100%, respectively, indicating excellent visual appearance under high salt levels. Bi-weekly clipping DW showed that 'Diamond' continued to produce biomass throughout the duration of the study under the EC10 treatment. Sodium (Na⁺) and chloride (Cl⁻) content in the shoot tissue of the seven turfgrass genotypes indicated that lower concentrations corresponded to greater salt tolerance indicating exclusion of Na⁺ and Cl⁻ from the shoot tissue. Taken together, the genotypes 'Diamond' and DALZ 1713 were determined to be the most salt tolerant and recommended for use in areas with high soil or water salinity.
- Zoysiagrass,
- Turfgrass,
- Salinity,
- Salt exclusion

Introduction

In a recent report on Latin America's next petroleum boom, The Economist refers to the current and future situation in oil producing countries in the region. In the case of Argentina, the increase in oil and gas output 'have led to an increase in production in Vaca Muerta, a mammoth field in Argentina's far west. It holds the world's second-largest shale gas deposits and its fourth-largest shale oil reserves… Rystad Energy expects shell-oil production in Argentina will more than double by the end of the decade, to over a million barrels per day'^[1].

Oil production in Argentina is currently dominated by three Patagonian areas: Neuquén, San Jorge, and Austral. Based on 2021 information, 49% of oil reserves of Argentina are located in Neuquén, whereas San Jorge has 46%. Neuquén is also the largest source of oil (57%) and gas (37%) in the country. According to 2018 data, conventional oil produced in Argentina amounts to 87%, whereas non-conventional, shale production represents 13%; however, non-conventional oil is increasing due to Vaca Muerta shale oil exploitation.

This increase of production in the Patagonian fields requires the use of a large fluid storage capacity by means of vertical oil storage tanks having different sizes and configurations. Tanks are required to store not just oil but also water. The exploitation of nonconventional reservoirs, such as Vaca Muerta, involves massive water storage to carry out hydraulic stimulation in low-permeability fields, and for managing the return fluid and production water at different stages of the process (storage, treatment, and final disposal).

Storage tanks in the oil industry are large steel structures; they may have different sizes, and also different shell configurations, such as vertical cylinders with a fixed roof or with a floating roof and opened at the top^[2]. It is now clear that such oil infrastructure is vulnerable to accidents caused by extreme weather events^[3−5].

Data from emergencies occurring in oil fields shows that accidents due to regional winds, with wind speed between 150 and 240 km/h, may cause severe tank damage. Seismic activity in the region, on the other hand, is of less concern to tank designers in Patagonia.

Damage and failure mechanisms of these tanks largely depend on tank size and configuration, and their structural response should be considered from the perspective of shell mechanics and their consequences. In a report on damage observed in tanks following hurricanes Katrina and Rita in 2005^[6,7], several types of damage were identified. The most common damage initiation process is due to shell buckling^[8−11], which may progress into plasticity at higher wind speeds. In open-top tanks, a floating roof does not properly slide on a buckled cylindrical shell, and this situation may lead to different failure mechanisms. Further, damage and loss of integrity have the potential to induce oil spills, with direct consequences of soil contamination and also of fire initiation.

Concern about an emergency caused by such wind-induced hazards involves several stakeholders, because the consequences may affect the operation of oil plants, the local and regional economies, the safety of the population living in the area of a refinery or storage farm, and the environment^[6]. In view of the importance of preserving the shell integrity and avoiding tank damage, there is a need to evaluate risk of existing tanks at a regional level, such as in the Neuquén and San Jorge areas. This information may help decision makers in adopting strategies (such as structural reinforcement of tanks to withstand expected wind loads) or post-event actions (like damaged infrastructure repair or replacement).

The studies leading to the evaluation of risk in the oil infrastructure are known as vulnerability studies, and the most common techniques currently used are fragility curves^[12]. These curves evaluate the probability of reaching or exceeding a given damage level as a function of a load parameter (such as wind speed in this case).

Early studies in the field of fragility of tanks were published^[13] from post-event earthquake damage observations. Studies based on computational simulation of tank behavior under seismic loads were reported^[14]. The Federal Emergency Management Administration in the US developed fragility curves for tanks under seismic loads for regions in the United States, and more recently, this has been extended to hurricane and flood events in coastal areas^[15]. Seismic fragility in Europe has been reviewed by Pitilakis et al.^[16], in which general concepts of fragility are discussed. Bernier & Padgett^[17] evaluated the failure of tanks due to hurricane Harvey using data from aerial images and government databases. Fragility curves were developed based on finite element analyses and damage of the tank population was identified in the Houston Ship Channel. Flood and wind due to hurricane Harvey were also considered^[18] to develop fragility curves.

Because fragility curves for tanks under wind depend on the wind source (either hurricane or regional winds), and the type and size of tanks identified in a region, fragility curves developed for one area are not possible to be directly used in other areas under very different inventory and wind conditions.

This paper addresses problems of shell buckling and loss of integrity of open top tanks, with wind-girders and floating roof and it focuses on the development of fragility curves as a way to estimate damage states under a given wind pressure. The region of interest in this work covers the oil producing areas in Patagonia, Argentina. Damage of tanks under several wind pressures are evaluated by finite element analyses together with methodologies to evaluate the structural stability.

Tank data considered in this research

The construction of fragility curves requires information from the following areas: First, an inventory of tanks to be included in the analysis; second, data about the loads in the region considered; third, data about structural damage, either observed or computed via modeling; and fourth, a statistical model that links damage and load/structure data. This section describes the main features of the tank population considered in the study.

Strategies to establish a population in order to construct fragility curves

The construction of an inventory at a regional level is a very complex task, which is largely due to a lack of cooperation from oil companies to share information about their infrastructure. Thus, to understand the type of tanks in an oil producing region, one is left collecting a limited number of structural drawings and aerial photography. A detailed inventory of the Houston Ship Channel was carried out by Bernier et al.^[19], who identified 390 floating roof tanks. An inventory for Puerto Rico^[20] identified 82 floating roof tanks. Although both inventories used different methodologies and addressed very different tank populations, some common features were found in both cases.

An alternative strategy to carry out fragility studies is to develop a database using a small number of tanks, for which a detailed structural behavior is investigated using finite element analysis. This is a time-consuming task, but it allows identification of buckling pressures, buckling modes, and shell plasticity. This information serves to build approximate fragility curves, and it can also be used to develop what are known as meta-models, which predict structural damage based on tank/load characteristics. Such meta-models take the form of equations that include the tank geometry and wind speed to estimate damage. Meta-models were used, for example, in the work of Kameshwar & Padgett^[18].

This work employs a simplified strategy, and addresses the first part of the procedure described above. The use of a limited number of tanks in a database, for which a finite element structural analysis is carried out. This leads to fragility curves based on a simplified tank population (reported in this work) and the development of a meta-model together with enhanced fragility results will be reported and compared in a future work.

Tanks considered for the Patagonian region

Partial information of tanks in the Patagonian region was obtained from government sources, and this was supplemented by aerial photography showing details of tank farms in the region. As a result of that, it was possible to establish ranges of tank dimensions from which an artificial database was constructed.

The present study is restricted to open-top tanks with a wind girder at the top. They are assumed to have floating roofs, which are designed and fabricated to allow the normal operation of the roof without the need of human intervention. The main characteristics of tanks investigated in this paper, are illustrated in Fig. 1.

Figure 1. Geometric characteristics of open-topped oil storage considered in this paper.

DownLoad: Full-Size Img PowerPoint

The range of interest in terms of tank diameter D was established between 35 m < D < 60 m. Based on observation of tanks in the region, the ratios D/H were found to be in the range 0.20 < D/H < 0.60, leading to cylinder height H in the range 12 m < H < 20 m. These tanks were next designed using API 650^[21] regulations to compute their shell thickness and wind girder dimensions. A variable thickness was adopted in elevation, assuming 3 m height shell courses. The geometries considered are listed in Table 1, with a total of 30 tanks having combinations of five values of H and six values of D. The volume of these tanks range between 55,640 and 272,520 m³.

Table 1. Geometry and course thickness of 30 tanks considered in this work.

H (m)	Courses	Thickness t (m)
H (m)	Courses	D = 35 m	D = 40 m	D = 45 m	D = 50 m	D = 55 m	D = 60 m
12	V1	0.014	0.016	0.018	0.018	0.020	0.022
	V2	0.012	0.012	0.014	0.016	0.016	0.018
	V3	0.008	0.010	0.010	0.010	0.012	0.012
	V4	0.006	0.008	0.008	0.008	0.008	0.008
14	V1	0.016	0.018	0.020	0.022	0.025	0.025
	V2	0.014	0.014	0.016	0.018	0.020	0.020
	V3	0.010	0.012	0.012	0.014	0.014	0.016
	V4	0.008	0.008	0.008	0.010	0.010	0.010
	V5	0.006	0.008	0.008	0.008	0.008	0.008
16	V1	0.018	0.020	0.022	0.025	0.028	0.028
	V2	0.016	0.018	0.018	0.020	0.022	0.025
	V3	0.012	0.014	0.016	0.016	0.018	0.020
	V4	0.010	0.010	0.012	0.012	0.014	0.014
	V5	0.006	0.008	0.008	0.008	0.008	0.010
	V6	0.006	0.008	0.008	0.008	0.008	0.010
18	V1	0.020	0.022	0.025	0.028	0.030	0.032
	V2	0.018	0.020	0.022	0.025	0.025	0.028
	V3	0.014	0.016	0.018	0.020	0.020	0.022
	V4	0.012	0.012	0.014	0.016	0.016	0.018
	V5	0.008	0.010	0.010	0.010	0.012	0.012
	V6	0.008	0.010	0.010	0.010	0.012	0.012
20	V1	0.022	0.025	0.028	0.030	0.032	0.035
	V2	0.020	0.022	0.025	0.028	0.028	0.032
	V3	0.016	0.018	0.020	0.022	0.025	0.028
	V4	0.014	0.014	0.016	0.018	0.020	0.020
	V5	0.010	0.012	0.012	0.014	0.014	0.016
	V6	0.010	0.012	0.012	0.014	0.014	0.016
	V7	0.010	0.012	0.012	0.014	0.014	0.016

| Show Table

DownLoad: CSV

The material assumed in the computations was A36 steel, with modulus of elasticity E = 201 GPa and Poisson's ratio ν = 0.3.

For each tank, a ring stiffener was designed as established by API 650^[21], in order to prevent buckling modes at the top of the tank. The minimum modulus Z to avoid ovalization at the top of the tank is given by

$Z=\dfrac{{\mathrm{D}}^{2}H}{17}{\left(\dfrac{V}{190}\right)}^{2}$

(1)

where V is the wind speed, in this case taken as V = 172.8 km/h for the Patagonian region. Intermediate ring stiffeners were not observed in oil tanks in Patagonia, so they were not included in the present inventory.

Because a large number of tanks need to be investigated in fragility studies, it is customary to accept some simplifications in modeling the structure to reduce the computational effort. The geometry of a typical ring stiffener at the top is shown in Fig. 2a, as designed by API 650. A simplified version was included in this research in the finite element model, in which the ring stiffener is replaced by an equivalent thickness at the top, as suggested in API Standard 650^[21]. This approach has been followed by most researchers in the field. The equivalent model is shown in Fig. 2b.

Figure 2. Ring stiffener, (a) design according to API 650, (b) equivalent section^[22].

DownLoad: Full-Size Img PowerPoint

Wind pressures adopted for tanks located in the Patagonian region

The pressure distribution due to wind around a short cylindrical shell has been investigated in the past using wind tunnels and computational fluid dynamics, and a summary of results has been included in design regulations.

There is a vast number of investigations on the pressures in storage tanks due to wind, even if one is limited to isolated tanks, as in the present paper. For a summary of results, see, for example, Godoy^[¹¹^], and only a couple of studies are mentioned here to illustrate the type of research carried out in various countries. Wind tunnel tests were performed in Australia^[23], which have been the basis of most subsequent studies. Recent tests in Japan on small scale open top tanks were reported^[24,25]. In China, Lin & Zhao^[26] reported tests on fixed roof tanks. CFD models, on the other hand, were computed^[27] for open top tanks with an internal floating roof under wind flow. Although there are differences between pressures obtained in different wind tunnels, the results show an overall agreement.

The largest positive pressures occur in the windward meridian covering an angle between 30° and 45° from windward. Negative pressures (suction), on the other hand, reach a maximum at meridians located between 80° and 90° from windward. An evaluation of US and European design recommendations has been reported^[28,29], who also considered the influence of fuel stored in the tank.

The circumferential variation of pressures is usually written in terms of a cosine Fourier series. The present authors adopted the series coefficients proposed by ASCE regulations^[30], following the analytical expression:

$\mathrm{q}=\mathrm{\lambda }{\sum }_{\mathrm{n}}^{\mathrm{i}}{\mathrm{C}}_{\mathrm{i}}\mathrm{c}\mathrm{o}\mathrm{s}\left(\mathrm{i}\mathrm{\varphi }\right)$

(2)

in which λ is the amplification factor; the angle φ is measured from the windward meridian; and coefficients C_i represent the contribution of each term in the series. The following coefficients were adopted in this work (ASCE): C₀ = −0.2765, C₁ = 0.3419, C₂ = 0.5418, C₃ = 0.3872, C₄ = 0.0525, C₅ = 0.0771, C₆ = −0.0039 and C₇ = 0.0341. For short tanks, such as those considered in this paper, previous research reported^[31] that for D/H = 0.5 the variation of the pressure coefficients in elevation is small and may be neglected to simplify computations. Thus, the present work assumes a uniform pressure distribution in elevation at each shell meridian.

In fragility studies, wind speed, rather than wind pressures, are considered, so that the following relation from ASCE is adopted in this work:

${\mathrm{q}}_{\mathrm{z}}=0.613{\mathrm{K}}_{\mathrm{z}\mathrm{t}}{\mathrm{K}}_{\mathrm{d}}{\mathrm{V}}^{2}\mathrm{I}\mathrm{ }\mathrm{ }\mathrm{ }\mathrm{ }\mathrm{ }\mathrm{ }\mathrm{ }\mathrm{ }\mathrm{ }\Rightarrow \mathrm{V}=\sqrt{\dfrac{{\mathrm{q}}_{\mathrm{z}}}{0.613{\mathrm{K}}_{\mathrm{z}\mathrm{t}}{\mathrm{K}}_{\mathrm{d}}\mathrm{I}}}$

(3)

in which I is the importance factor; K_d is the directionality factor; and K_zt is the topographic factor. Values of I = 1.15, K_d = 0.95 and K_zt = 1, were adopted for the computations reported in this paper.

Because shell buckling was primarily investigated in this work using a bifurcation analysis, the scalar λ was increased in the analysis until the finite element analysis detected a singularity.

Evaluation of fragility curves

Fragility curves are functions that describe the probability of failure of a structural system (oil tanks in the present case) for a range of loads (wind pressures) to which the system could be exposed. In cases with low uncertainty in the structural capacity and acting loads, fragility curves take the form of a step-function showing a sudden jump (see Fig. 3a). Zero probability occurs before the jump and probability equals to one is assumed after the jump. But in most cases, in which there is uncertainty about the structural capacity to withstand the load, fragility curves form an 'S' shape, as shown in Fig. 3a and b probabilistic study is required to evaluate fragility.

Figure 3. Examples of fragility curves, (a) step-function, (b) 'S' shape function.

DownLoad: Full-Size Img PowerPoint

The construction of fragility curves is often achieved by use of a log-normal distribution. In this case, the probability of reaching a certain damage level is obtained by use of an exponential function applied to a variable having a normal distribution with mean value μ and standard deviation σ. If a variable x follows a log-normal distribution, then the variable log(x) has a normal distribution, with the following properties:

• For x < 0, a probability equal to 0 is assigned. Thus, the probability of failure for this range is zero.

• It can be used for variables that are computed by means of a number of random variables.

• The expected value in a log-normal distribution is higher than its mean value, thus assigning more importance to large values of failure rates than would be obtained in a normal distribution.

The probability density function for a log-normal distribution may be written in the form^[32]:

${\mathrm{f}}_{\left({\mathrm{x}}^{\mathrm{i}}\right)}=\dfrac{1}{\sqrt{2\mathrm{\pi }{\mathrm{\sigma }}^{2}}}\dfrac{1}{\mathrm{x}}\mathrm{e}\mathrm{x}\mathrm{p}\left[-{\left(\mathrm{l}\mathrm{n}\mathrm{x}-{\text{µ}}^{*}\right)}^{2}/\left(2{\mathrm{\sigma }}^{2}\right)\right]$

(4)

in which f(xⁱ) depends on the load level considered, and is evaluated for a range of interest of variable x; and μ* is the mean value of the logarithm of variable x associated with each damage level. Damage levels xⁱ are given by Eqn (5).

${\text{µ}}^{{*}}\left({\mathrm{x}}^{\mathrm{i}}\right)=\dfrac{1}{\mathrm{N}}{\sum }_{\mathrm{n}=1}^{\mathrm{N}}\mathrm{l}\mathrm{n}\left({\mathrm{x}}_{\mathrm{n}}^{\mathrm{i}}\right)$

(5)

where the mean value is computed for a damage level xⁱ, corresponding to I = DSi; summation in n extends to the number of tanks considered in the computation of the mean value. Damage levels in this work are evaluated using computational modeling and are defined in the next section. Variance is the discrete variable xⁱ (σ²), computed from:

${\mathrm{\sigma }}^{2}\left({\mathrm{x}}^{\mathrm{i}},{\text{µ}}^{{*}}\right)=\dfrac{1}{\mathrm{N}}{\sum }_{\mathrm{n}=1}^{\mathrm{N}}{\left(\mathrm{l}\mathrm{n}\left({\mathrm{x}}_{\mathrm{n}}^{\mathrm{i}}\right)-{\text{µ}}^{{*}}\right)}^{2}=\dfrac{1}{\mathrm{N}}{\sum }_{\mathrm{n}=1}^{\mathrm{N}}{\mathrm{l}\mathrm{n}\left({\mathrm{x}}_{\mathrm{n}}^{\mathrm{i}}\right)}^{2}-{{\text{µ}}^{{*}}}^{2}$

(6)

The probability of reaching or exceeding a damage level DSi is computed by the integral of the density function using Eqn (7), for a load level considered (the wind speed in this case):

$\mathrm{P}\left[\mathrm{D}\mathrm{S}/\mathrm{x}\right]={\int }_{\mathrm{x}=0}^{\mathrm{x}={\mathrm{V}}_{0}}\mathrm{f}\left(\mathrm{x}\right)\mathrm{d}\mathrm{x}$

(7)

where V₀ is the wind speed at which computations are carried out, and x is represented by wind speed V.

Damage in wind loaded oil storage tanks

Damage states considered in this work

Various forms of structural damage may occur as a consequence of wind loads, including elastic or plastic deflections, causing deviations from the initial perfect geometry; crack initiation or crack extension; localized or extended plastic material behavior; and structural collapse under extreme conditions. For the tanks considered in this work, there are also operational consequences of structural damage, such as blocking of a floating roof due to buckling under wind loads that are much lower than the collapse load. For this reason, a damage study is interested in several structural consequences but also in questions of normal operation of the infrastructure. Several authors pointed out that there is no direct relation between structural damage and economic losses caused by an interruption of normal operation of the infrastructure.

Types of damage are usually identified through reconnaissance post-event missions, for example following Hurricanes Katrina and Rita^[6,7]. Damage states reported in Godoy^[⁷^] include shell buckling, roof buckling, loss of thermal insulation, tank displacement as a rigid body, and failure of tank/pipe connections. These are qualitative studies, in which damage states previously reported in other events are identified and new damage mechanisms are of great interest in order to understand damage and failure modes not taken into account by current design codes.

In this work, in which interest is restricted to open top tanks having a wind girder at the top, four damage states were explored, as shown in Table 2. Regarding the loss of functionality of a tank, several conditions may occur: (1) No consequences for the normal operation of a tank; (2) Partial loss of operation capacity; (3) Complete loss of operation.

Table 2. Damage states under wind for open-top tanks with a wind girder.

Damage states (DS)	Description
DS0	No damage
DS1	Large deflections on the cylindrical shell
DS2	Buckling of the cylindrical shell
DS3	Large deflections on the stiffening ring

| Show Table

DownLoad: CSV

DS1 involves displacements in some area of the cylindrical body of the tank, and this may block the free vertical displacement of the floating roof. Notice that this part of the tank operation is vital to prevent the accumulation of inflammable gases on top of the fluid stored. Blocking of the floating roof may cause a separation between the fuel and the floating roof, which in turn may be the initial cause of fire or explosion.

DS2 is associated with large shell deflections, which may cause failure of pipe/tank connections. High local stresses may also arise in the support of helicoidal ladders or inspection doors, with the possibility of having oil spills.

DS3 is identified for a loss of circularity of the wind girder. The consequences include new deflections being transferred to the cylindrical shell in the form of geometrical imperfections.

In summary, DS1 and DS3 may affect the normal operation of a floating roof due to large shell or wind-girder deflections caused by buckling.

Finite element evaluation of damage states

Tank modeling was carried out in this work using a finite element discretization within the ABAQUS environment^[33] using rectangular elements with quadratic interpolation functions and reduced integration (S8R5 in the ABAQUS nomenclature). Two types of shell analysis were performed: Linear Bifurcation Analysis (LBA), and Geometrically Nonlinear Analysis with Imperfections (GNIA). The tank perimeter was divided into equal 0.35 m segments, leading to between 315 and 550 elements around the circumference, depending on tank size. Convergence studies were performed and errors in LBA eigenvalues were found to be lower than 0.1%.

The aim of an LBA study is to identify a first critical buckling state and buckling mode by means of an eigenvalue problem. The following expression is employed:

$\left( {{{\mathrm{K}}_0} + {\lambda ^{_C}}{{\mathrm{K}}_G}} \right)\,{{\mathrm{\Phi }}^C} = 0$

(8)

where K₀ is the linear stiffness matrix of the system; K_G is the load-geometry matrix, which includes the non-linear terms of the kinematic relations; λ^C is the eigenvalue (buckling load); and Φ^C is the critical mode (eigenvector). For a reference wind state, λ is a scalar parameter. One of the consequences of shell buckling is that geometric deviations from a perfect geometry are introduced in the shell, so that, due to imperfection sensitivity, there is a reduced shell capacity for any future events.

The aim of the GNIA study is to follow a static (non-linear) equilibrium path for increasing load levels. The GNIA study is implemented in this work using the Riks method^[34,35], which can follow paths in which the load or the displacement decrease. The geometric imperfection was assumed with the shape of the first eigenvector at the critical state in the LBA study, and the amplitude of the imperfection was defined by means of a scalar ξ ^[10]. To illustrate this amplitude, for a tank with D = 45 m and H = 12 m, the amplitude of imperfection is equal to half the minimum shell thickness (ξ = 4 mm in this case).

It was assumed that a damage level DS1 is reached when the displacement amplitudes do not allow the free vertical displacement of the floating roof. Based on information from tanks in the Patagonian region, the limit displacement was taken as 10 mm. This state was detected by GNIA, and the associated load level is identified as λ = λ^DS1.

The load at which damage state DS2 occurs was obtained by LBA, leading to a critical load factor λ^C and a buckling mode. An example of damage levels is shown in Fig. 4.

Figure 4. Damage computed for a tank with D = 45 m and H = 12 m. (a) Deflected shape for damage DS1; (b) Equilibrium path for node A (DS1); (c) Deflected shape for damage DS2 (critical mode).

DownLoad: Full-Size Img PowerPoint

An LBA study does not account for geometric imperfections. It is well known that the elastic buckling of shells is sensitive to imperfections, so that a reduction in the order of 20% should be expected for cylindrical shells under lateral pressure. This consideration allows to estimate DS0 (a state without damage) as a lower bound of the LBA study. An approach to establish lower bounds for steel storage tanks is the Reduced Stiffness Method (RSM)^[36−40]. Results for tanks using the RSM to estimate safe loads show that λ^DS0 = 0.5λ^DS2 provides a conservative estimate for present purposes.

DS3 was computed using a linear elastic analysis to evaluate the wind pressure at which a 10 mm displacement of the wind girder is obtained.

In a similar study for tanks with a fixed conical roof, Muñoz et al.^[41] estimated a collapse load based on plastic behavior. However, in the present case the top ring has a significant stiffness, and this leads to extremely high wind speeds before reaching collapse (higher than 500 km/h). For this reason, the most severe damage level considered here was that of excessive out-of-plane displacements of the wind girder, and not shell collapse.

Fragility curves for damage states DS0, DS1, DS2, and DS3

Methodology

The methodology to construct fragility curves has been presented by several authors^[42,43]. The following procedure was adapted here^[44]: (1) Establish qualitative damage categories (Table 2). (2) Compute a data base for different tanks, using LBA and GNIA. In each case, the damage category based on step (1) was identified (Table 3). (3) Approximate data obtained from step (2) using a log-normal distribution. (4) Plot the probabilities computed in step (3) with respect to wind speed x.

Table 3. Wind speed for each tank considered reaching a damage level.

H	D	ID	DS0	DS1	DS2	DS3
H12	D35	1	137.76	162.06	194.82	336.02
	D40	2	160.62	181.31	227.16	360.73
	D45	3	153.32	174.19	216.82	374.04
	D50	4	145.23	165.27	205.39	373.76
	D55	5	152.76	180.83	216.03	374.75
	D60	6	145.11	170.75	205.22	370.98
H14	D35	7	145.57	162.05	205.87	295.03
	D40	8	148.55	166.20	210.08	311.24
	D45	9	136.42	153.72	192.92	334.54
	D50	10	155.36	177.51	219.71	339.86
	D55	11	145.24	165.34	205.39	343,17
	D60	12	141.89	167.26	200.67	338.77
H16	D35	13	131.32	161.94	185.71	262.20
	D40	14	146.95	163.99	207.82	277.08
	D45	15	150.58	170.90	212.95	293.37
	D50	16	138.97	161.05	196.54	303.62
	D55	17	138.51	174.17	195.88	313.97
	D60	18	156.34	182.78	221.10	326.83
H18	D35	19	146.80	160.79	207.60	223.18
	D40	20	159.01	177.71	224.87	243.63
	D45	21	157.10	179.51	222.17	265.32
	D50	22	152.54	172.17	215.72	293.32
	D55	23	164.93	188.10	233.25	305.94
	D60	24	163.69	180.32	231.49	315.63
H20	D35	25	163.64	199.59	231.42	195.03
	D40	26	171.24	195.14	242.18	216.47
	D45	27	171.58	203.68	242.64	293.32
	D50	28	182.46	209.43	258.03	259.41
	D55	29	178.95	208.23	253.07	272.48
	D60	30	174.47	196.11	246.74	290.86

| Show Table

DownLoad: CSV

Results

Wind speeds for each tank, obtained via Eqn (3), are shown in Table 3 for the pressure level associated with each damage level DSi. A scalar ID was included in the table to identify each tank of the population in the random selection process. Wind speed was also taken as a random variable, so that wind speed in the range between 130 and 350 km/h have been considered at 5 km/h increase, with intervals of −2.5 and +2.5 km/h.

Out of the 30-tank population considered, a sample of 15 tanks were chosen at random and were subjected to random wind forces. The random algorithm allowed for the same tank geometry to be chosen more than once as part of the sample.

The type of damage obtained in each case for wind speed lower or equal to the upper bound of the interval were identified. Table 4 shows a random selection of tanks, together with the wind speed required to reach each damage level. For example, for a wind speed of 165 km/h, the wind interval is [162.5 km/h, 167.5 km/h]. This allows computation of a damage matrix (shown in Table 5). A value 1 indicates that a damage level was reached, whereas a value 0 shows that a damage level was not reached. In this example, 13 tanks reached DS0; six tanks reached DS1; and there were no tanks reaching DS2 or DS3. The ratio between the number of tanks with a given damage DSi and the total number of tanks selected is h, the relative accumulated frequency. The process was repeated for each wind speed and tank selection considered.

Table 4. Random tank selection for V = 165 km/h, assuming wind interval [162.5 km/h, 167.5 km/h].

ID	DS0	DS1	DS2	DS3
11	145.2	165.3	205.4	343.2
6	145.1	170.7	205.2	371.0
3	153.3	174.2	216.8	374.0
9	136.4	153.7	192.9	334.5
28	182.5	209.4	258.0	259.4
22	152.5	172.2	215.7	293.3
13	131.3	161.9	185.7	262.2
19	146.8	160.8	207.6	223.2
3	153.3	174.2	216.8	374.0
12	141.9	167.3	200.7	338.8
30	174.5	196.1	246.7	290.9
23	164.9	188.1	233.2	305.9
2	160.6	181.3	227.2	360.7
17	138.5	174.2	195.9	314.0
11	145.2	165.3	205.4	343.2

| Show Table

DownLoad: CSV

Table 5. Damage matrix for random tank selection (V = 165 km/h), assuming wind interval [162.5 km/h, 167.5 km/h].

	DS0	DS1	DS2	DS3
	1	1	0	0
	1	0	0	0
	1	0	0	0
	1	1	0	0
	0	0	0	0
	1	0	0	0
	1	1	0	0
	1	1	0	0
	1	0	0	0
	1	1	0	0
	0	0	0	0
	1	0	0	0
	1	0	0	0
	1	0	0	0
	1	1	0	0
Total	13	6	0	0
h_i	0.87	0.4	0	0

| Show Table

DownLoad: CSV

Table 6 shows the evaluation of the fragility curve for damage level DS0. This requires obtaining the number of tanks for each wind speed (f_i), the cumulative number as wind speed is increased (F_i), and the frequency with respect to the total number of the sample of 15 tanks is written on the right-hand side of Table 6, for relative frequency (h_i) and accumulated frequency (H_i).

Table 6. Damage DS0: Wind speed intervals [km/h] shown on the left; logarithm of wind speed; and relative and absolute frequencies (shown on the right).

V inf (km/h)	V m (km/h)	V sup (km/h)	Ln (Vm)	f_i	F_i	h_i	H_i
127.5	130	132.5	4.87	0	0	0.000	0
132.5	135	137.5	4.91	2	2	0.133	0.133
137.5	140	142.5	4.94	1	3	0.067	0.200
142.5	145	147.5	4.98	3	6	0.200	0.400
147.5	150	152.5	5.01	1	7	0.067	0.467
152.5	155	157.5	5.04	4	11	0.267	0.733
157.5	160	162.5	5.08	0	11	0.000	0.733
162.5	165	167.5	5.11	2	13	0.133	0.867
167.5	170	172.5	5.14	0	13	0.000	0.867
172.5	175	177.5	5.16	0	13	0.000	0.867
177.5	180	182.5	5.19	2	15	0.133	1.000

| Show Table

DownLoad: CSV

With the values of mean and deviation computed with Eqns (5) & (6), it is possible to establish the log normal distribution of variable V for damage level DS0, usually denoted as P[DS0/V]. Values obtained in discrete form and the log-normal distribution are shown in Fig. 5a for DS0. For the selection shown in Table 6, the media is μ* = 5.03 and the deviation is σ = 0.09.

Figure 5. Probability of reaching a damage level P[DSi/V], (a) DS0, (b) DS0, DS1, DS2 and DS3.

DownLoad: Full-Size Img PowerPoint

The process is repeated for each damage level to obtain fragility curves for DS1, DS2, and DS3 (Fig. 5b). Notice that the wind speeds required to reach DS3 are much higher than those obtained for the other damage levels. Such values should be compared with the regional wind speeds in Patagonia, and this is done in the next section.

Discussion

The oil producing regions in Argentina having the largest oil reserves are the Neuquén and the San Jorge regions, both located in Patagonia. This needs to be placed side by side with wind loads to understand the risk associated with such oil production.

Figure 6 shows the geographical location of these regions. The Neuquén region includes large areas of four provinces in Argentina (Neuquén, south of Mendoza, west of La Pampa, and Río Negro). The San Jorge region is in the central Patagonia area, including two provinces (south of Chubut, north of Santa Cruz). Another area is the Austral region covering part of a Patagonian province (Santa Cruz).

Figure 6. Oil producing regions in Argentina. (Adapted from IAPG^[47]).

DownLoad: Full-Size Img PowerPoint

A map of basic wind speed for Argentina is available in the Argentinian code CIRSOC 102^[45], which is shown in Fig. 7. Notice that the highest wind speeds are found in Patagonia, and affect the oil-producing regions mentioned in this work. For the Neuquén region, wind speeds range from 42 to 48 m/s (151.2 to 172.8 km/h), whereas for San Jorge Gulf region they range between 52 and 66 m/s (187.2 and 237.6 km/h).

Figure 7. Wind speed map of Argentina. (Adapted from CIRSOC 102^[45]).

DownLoad: Full-Size Img PowerPoint

The wind values provided by CIRSOC 102^[45] were next used to estimate potential shell damage due to wind. Considering the fragility curves presented in Fig. 4, for damage levels DS0, DS1, DS2 and DS3 based on a log-normal distribution, it may be seen that it would be possible to have some form of damage in tanks located in almost any region of Argentina because CIRSOC specifies wind speeds higher than 36 m/s (129.6 km/h). The fragility curve DS0 represents the onset of damage for wind speeds higher than 130 km/h, so that only winds lower than that would not cause tank damage.

Based on the fragility curves shown in Fig. 8, it is possible to estimate probable damage levels for the wind speed defined by CIRSOC. Because design winds in Patagonia are higher than 165.6 km/h (46 m/s), it is possible to conclude that there is 81% probability to reach DS0 and 25% to reach DS1.

Figure 8. Probability P[DSi/V] to reach damage levels DS1, DS2 and DS3 in tanks located in the Patagonia region of Argentina.

DownLoad: Full-Size Img PowerPoint

For the geographical area of the Neuquén region in Fig. 6, together with the wind map of Fig. 7, the expected winds range from 150 to 172.8 km/h (42 to 48 m/s). Such wind range is associated with a DS0 probability between 41% and 92%, whereas the DS1 probability is in the order of 48%.

A similar analysis was carried out for the San Jorge region, in which winds between 187.2 and 237 km/h (52 and 66 m/s). The probability of reaching DS1 is 87%, and the probability of DS2 is 88%. Wind girder damage DS3 could only occur in this region, with a lower probability of 18%.

Conclusions

This work focuses on open top tanks having a floating roof, and explores the probability of reaching damage levels for wind loads, using the methodology of fragility curves. A population of 30 tanks was defined with H/D ratios between 0.2 and 0.6; such aspect ratios were found to be the most common in the oil producing regions of Patagonia. The data employed assumed diameters D between 35 and 60 m, together with height between 12 and 20 m. The tanks were designed using current API 650 regulations which are used in the region, in order to define the shell thickness and wind girder. All tanks were assumed to be empty, which is the worst condition for shell stability because a fluid stored in a tank has a stabilizing effect and causes the buckling load to be higher.

Both structural damage (shell buckling) and operational damage (blocking of the floating roof due to deflections of the cylindrical shell) were considered in the analysis. The qualitative definition of damage levels in this work was as follows: The condition of no damage was obtained from a lower bound of buckling loads. This accounts for geometric imperfections and mode coupling of the shell. Shell buckling was evaluated using linear bifurcation analysis to identify damage level DS2. A geometrically non-linear analysis with imperfections was used to identify deflection levels that would block a floating roof, a damage level identified as DS1. Finally, deflections in the wind girder were investigated using a linear elastic analysis to define damage DS3.

The present results were compared with the wind conditions of Patagonia, to show that several damage levels may occur as a consequence of wind speeds higher than 130 km/h, which is the expected base value identified for the region. The most frequent expected damage is due to the loss of vertical displacements of the floating roof due to large displacements in the cylindrical shell of the tank, and this may occur for wind speed up to 200 km/h. Damage caused by shell buckling may occur for wind speeds higher than 190 km/h, and for that wind speed, further damage due to displacements in the wind girder may also occur, but with a lower probability. This latter damage form requires much higher wind speed to reach a probability of 20%, and would be more representative of regions subjected to hurricanes.

The number of tanks considered in the present analysis was relatively low, mainly because the aim of this work was to collect data to build a meta-model, i.e. a simple model that may estimate damage based on shell and load characteristics^[46]. In future work, the authors expect to develop and apply such meta-models to a larger number of tank/wind configurations, in order to obtain more reliable fragility curves.

Fragility studies for an oil producing region, like those reported in this work, may be important to several stakeholders in this problem. The fragility information links wind speed levels to expected infrastructure damage, and may be of great use to government agencies, engineering companies, and society at large, regarding the risk associated with regional oil facilities. At a government level, this helps decision makers in allocating funding to address potential oil-related emergencies cause by wind. This can also serve as a guide to develop further modifications of design codes relevant to the oil infrastructure. The engineering consequences may emphasize the need to strengthen the present regional infrastructure to reduce risk of structural damage and its consequences. The impact of damage in the oil infrastructure on society was illustrated in the case of Hurricane Katrina in 2005, in which a large number of residents had to be relocated due to the conditions created by the consequences of infrastructure failure.

Author contributions

The authors confirm contribution to the paper as follows: study conception and design: Jaca RC, Godoy LA; data collection: Grill J, Pareti N; analysis and interpretation of results: Jaca RC, Bramardi S, Godoy LA; draft manuscript preparation: Jaca RC, Godoy LA. All authors reviewed the results and approved the final version of the manuscript.

Data availability

All data generated or analyzed during this study are included in this published article.

Acknowledgments

The authors are thankful for the support of a grant received from the National Agency for the Promotion of Research, Technological Development and Innovation of Argentina and the YPF Foundation. Luis A. Godoy thanks Prof. Ali Saffar (University of Puerto Rico at Mayaguez) for introducing him to the field of fragility studies.

Conflict of interest

The authors declare that they have no conflict of interest.

Supplementary information

Supplemental Fig. S1 Photos of the seven genotypes that were treated for 8 weeks: control (left), EC5 (middle), and EC10 (right). The genotype arrangement in each treatment was randomized. Photos were taken right after trimming.

Rights and permissions
Copyright: © 2022 by the author(s). Published by Maximum Academic Press, Fayetteville, GA. This article is an open access article distributed under Creative Commons Attribution License (CC BY 4.0), visit https://creativecommons.org/licenses/by/4.0/.

References

[1]	Chavarria M, Wherley B, Thomas J, Chandra A, Raymer P. 2019. Salinity tolerance and recovery attributes of warm-season turfgrass cultivars. HortScience 54:1625−31 doi: 10.21273/hortsci13963-19 CrossRef Google Scholar
[2]	Chaudhary J. 2019. Application of reclaimed water for irrigation: A review. (Accessed 9 Nov. 2022) https://soils.ifas.ufl.edu/media/soilsifasufledu/sws-main-site/pdf/technical-papers/Chaudhary_Juhi_Immediate_Release.pdf
[3]	United States Environmental Protection Agency (EPA). 2012. Guidelines for water reuse. EPA/600/R-12/618. (Accessed 10 Oct. 2022) https://www.epa.gov/sites/default/files/2019-08/documents/2012-guidelines-water-reuse.pdf
[4]	Chen W, Lu S, Jiao W, Wang M, Chang AC. 2013. Reclaimed water: A safe irrigation water source. Environmental Development 8:74−83 doi: 10.1016/j.envdev.2013.04.003 CrossRef Google Scholar
[5]	Wu S, Sun Y, Niu G, Altland J, Cabrera R. 2016. Response of 10 aster species to saline water irrigation. HortScience 51:197−201 doi: 10.21273/HORTSCI.51.2.197 CrossRef Google Scholar
[6]	Qian YL, Fu JM, Wilhelm SJ, Christensen D, Koski AJ. 2007. Relative salinity tolerance of turf-type saltgrass selections. HortScience 42:205−9 doi: 10.21273/HORTSCI.42.2.205 CrossRef Google Scholar
[7]	Kronzucker HJ, Britto DT. 2011. Sodium transport in plants: a critical review. New Phytologist 189:54−81 doi: 10.1111/j.1469-8137.2010.03540.x CrossRef Google Scholar
[8]	Parida AK, Das AB. 2005. Salt tolerance and salinity effects on plants: a review. Ecotoxicology and Environmental Safety 60:324−49 doi: 10.1016/j.ecoenv.2004.06.010 CrossRef Google Scholar
[9]	Flowers TJ, Munns R, Colmer TD. 2015. Sodium chloride toxicity and the cellular basis of salt tolerance in halophytes. Annals of Botany 115:419−31 doi: 10.1093/aob/mcu217 CrossRef Google Scholar
[10]	Munns R, Tester M. 2008. Mechanisms of salinity tolerance. Annual Review of Plant Biology 59:651−81 doi: 10.1146/annurev.arplant.59.032607.092911 CrossRef Google Scholar
[11]	Bernstein L, François LE, Clark RA. 1972. Salt tolerance of ornamental shrubs and ground covers. Journal of the American Society for Horticultural Science 97:550−56 doi: 10.21273/JASHS.97.4.550 CrossRef Google Scholar
[12]	Niu G, Cabrera RI. 2010. Growth and physiological responses of landscape plants to saline water irrigation: A review. HortScience 45:1605−9 doi: 10.21273/HORTSCI.45.11.1605 CrossRef Google Scholar
[13]	Marcum KB, Pessarakli M, Kopec DM. 2005. Relative salinity tolerance of 21 turf-type desert saltgrasses compared to bermudagrass. HortScience 40:827−29 doi: 10.21273/HORTSCI.40.3.827 CrossRef Google Scholar
[14]	Wu Y, Chandra A, Genovesi AD, Meeks M. 2017. Freezing tolerance of field-and growth chamber-acclimated zoysiagrass hybrids. International Turfgrass Society Research Journal 13:421−26 doi: 10.2134/itsrj2016.05.0427 CrossRef Google Scholar
[15]	Patton AJ, Schwartz BM, Kenworthy KE. 2017. Zoysiagrass (Zoysia spp.) history, utilization, and improvement in the United States: A review. Crop Science 57:S37−S72 doi: 10.2135/cropsci2017.02.0074 CrossRef Google Scholar
[16]	Meeks M, Kong ST, Genovesi AD, Smith B, Chandra A. 2022. Low-input golf course putting green performance of fine-textured inter- and intra-specific zoysiagrass (Zoysia spp.). International Turfgrass Society Research Journal 14:610−21 doi: 10.1002/its2.63 CrossRef Google Scholar
[17]	Marcum KB, Anderson SJ, Engelke MC. 1998. Salt gland ion secretion: A salinity tolerance mechanism among five zoysiagrass species. Crop Science 38:806−10 doi: 10.2135/cropsci1998.0011183X003800030031x CrossRef Google Scholar
[18]	Maeda Y. 2019. Effects of calcium application on the salt tolerance and sodium excretion from salt glands in zoysiagrass (Zoysia japonica). Grassland Science 65:189−96 doi: 10.1111/grs.12234 CrossRef Google Scholar
[19]	Marcum KB, Murdoch CL. 1994. Salinity tolerance mechanisms of six C₄ turfgrasses. Journal of the American Society for Horticultural Science 119:779−84 doi: 10.21273/JASHS.119.4.779 CrossRef Google Scholar
[20]	Qian YL, Engelke MC, Foster MJV. 2000. Salinity effects on zoysiagrass cultivars and experimental lines. Crop Science 40:488−92 doi: 10.2135/cropsci2000.402488x CrossRef Google Scholar
[21]	Cavins TJ, Whipker BE, Fonteno WC. 2008. PourThru: A method for monitoring nutrition in the greenhouse. Acta Horticulturae 779:289−98 doi: 10.17660/actahortic.2008.779.35 CrossRef Google Scholar
[22]	Havlin JL, Soltanpour PN. 1980. A nitric acid plant tissue digest method for use with inductively coupled plasma spectrometry. Communications in Soil Science and Plant Analysis 11:969−980 doi: 10.1080/00103628009367096 CrossRef Google Scholar
[23]	Isaac RA, Johnson WC. 1975. Collaborative study of wet and dry ashing techniques for the elemental analysis of plant tissue by atomic absorption spectrophotometry. Journal of Association of Official Analytical Chemists 58:436−40 Google Scholar
[24]	Gavlak RG, Horneck DA, Miller RO. 1994. Plant, Soil, and Water Reference Methods for the Western Region. Corvallis, OR, USA: Western Regional Extension Publication.
[25]	Grismer ME. 1990. Leaching fraction, soil salinity, and drainage efficiency. California Agriculture 44:24−26 doi: 10.3733/ca.v044n06p24 CrossRef Google Scholar
[26]	Franco JA, Martínez-Sánchez, JJ, Fernández JA, Bañón S. 2006. Selection and nursery production of ornamental plants for landscaping and xerogardening in semi-arid environments. The Journal of Horticultural Science and Biotechnology 81:3−17 doi: 10.1080/14620316.2006.11512022 CrossRef Google Scholar
[27]	Sainju UM, Allen BL, Lenssen AW, Ghimire RP. 2017. Root biomass, root/shoot ratio, and soil water content under perennial grasses with different nitrogen rates. Field Crops Research 210:183−191 doi: 10.1016/j.fcr.2017.05.029 CrossRef Google Scholar
[28]	Beard JB, Green RL. 1994. The role of turfgrasses in environmental protection and their benefits to humans. Journal of Environmental Quality 23:452−60 doi: 10.2134/jeq1994.00472425002300030007x CrossRef Google Scholar
[29]	Marcum KB, Pessarakli M. 2006. Salinity tolerance and salt gland excretion efficiency of bermudagrass turf cultivars. Crop Science 46:2571−74 doi: 10.2135/cropsci2006.01.0027 CrossRef Google Scholar
[30]	Fu JM, Koski AJ, Qian YL. 2005. Responses of creeping Bentgrass to salinity and mowing management: growth and turf quality. HortScience 40:463−67 doi: 10.21273/HORTSCI.40.2.463 CrossRef Google Scholar

About this article

Cite this article

Hooks T, Masabni J, Ganjegunte G, Sun L, Chandra A, et al. 2022. Salt tolerance of seven genotypes of zoysiagrass (Zoysia spp.). Technology in Horticulture 2:8 doi: 10.48130/TIH-2022-0008

Hooks T, Masabni J, Ganjegunte G, Sun L, Chandra A, et al. 2022. Salt tolerance of seven genotypes of zoysiagrass (Zoysia spp.). Technology in Horticulture 2:8 doi: 10.48130/TIH-2022-0008

Figures(4) / Tables(3)

Download PDF

Article Metrics

Article views(5025) PDF downloads(753)

Other Articles By Authors

on this site
on Google Scholar

HTML

INTRODUCTION

Turfgrass is an important landscape groundcover that is widely used, not only for its aesthetics, but also for its function, such as in lawns, parks, athletic fields, and golf courses. However, in arid and semi-arid regions of the United States where fresh water sources are limited, reclaimed water sources that typically have elevated salt levels are increasingly being used for landscape irrigation^[1,2]. Reclaimed water, also known as recycled or reused water, is non-potable wastewater from a variety of sources, including residential, industrial, or stormwater runoff^[2]. Reclaimed water can be treated to varying degrees to be suitable for a wide range of uses, including agricultural and landscape irrigation, that effectually relieves the demand from freshwater sources^[3]. Because of these benefits, reclaimed water is increasingly being used for irrigation in arid and semi-arid regions, particularly for turfgrass areas^[4]. However, reclaimed water sources are typically saline, with high concentrations of sodium and chloride, which can be detrimental to the growth and aesthetic quality of salt-sensitive plants^[5]. Therefore, there is a continued need for salt tolerant turfgrasses for sustainable landscaping in arid or semi-arid regions in order to utilize saline, reclaimed water sources and conserve fresh water sources, especially in the Southwestern U.S.^[6].

Sodium and chloride are the two major soluble salts that can be detrimental to glycophytes at high concentrations^[7]. When these salts accumulate in the rhizosphere, they can impose osmotic stress on the plant, which leads to the inhibition of water uptake and can rapidly reduce plant growth and even lead to mortality^[8]. If salts are taken up by the roots and translocated to the shoots, then ionic stress can occur which can result in metabolic disruption in the cytosol of cells, as well as damage to chloroplasts by reactive oxygen species (ROS)^[9]. Ionic stress can lead to leaf burn, which appears as brown and necrotic tissue^[10] and can significantly degrade the quality of landscape and ornamental plants^[11]. For turfgrasses, it is important that salt tolerant cultivars are not susceptible to ionic stress and leaf burn and can maintain an appearance that is aesthetically appealing under saline conditions^[12].

Traditionally, bermudagrasses (Cynodon spp.) have been used as warm season turfgrasses for landscaping in arid and semi-arid regions, although they are considered high water consumption plants^[13]. Alternatively, zoysiagrasses (Zoysia spp.) are warm season turfgrasses that are moderately tolerant to salinity and have good potential for the selection and development of new salt tolerant cultivars that can be irrigated with saline, reclaimed water sources^[14,15]. Both bermudagrasses and zoysiagrasses are being used as breeding material for new cultivars with the aim of having warm season, salt tolerant varieties that can grow well and maintain high visual quality, or greenness, when irrigated with saline water.

Zoysiagrass was originally introduced to the U.S. in 1892, from East Asia, and have since been very influential in the turfgrass industry with more than 50 cultivars that have been developed, particularly for stress and pest tolerances. Two major species, Z. japonica and Z. matrella, readily hybridize with each other and are known for high quality turfgrasses primarily used in residential and commercial lawns, and golf courses^[15]. Another species, Z. minima, is native to New Zealand and has a diminutive growth habit that has potential for use in golf course putting greens^[16]. The recognized salt tolerance of zoysiagrasses is largely based on their ability to excrete salt out of their leaves through specialized salt glands^[17,18]. However, differences in salt tolerance have been noted among species and cultivars. Marcuum & Murdoch^[19] reported greater salt tolerance in Z. matrella compared to Z. japonica under solution culture up to 400 mM (approximately 30 dS m⁻¹) NaCl. Qian et al.^[20] reported differences in relative salt tolerance among 29 zoysiagrass experimental lines and cultivars under solution culture up to 42.5 dS m⁻¹.

In the present study, seven genotypes of zoysiagrass (Zoysia spp.), including several hybrids, were selected for evaluation for salt tolerance in a greenhouse study. The seven genotypes were developed by the turfgrass breeding program at Texas A&M AgriLife (Texas, USA). The parent species for these seven genotypes represent different leaf morphologies and traits: Z. japonica, wide leaf blade with drought and cold tolerance; Z. matrella, fine leaf blade with salt and shade tolerance; and Z. minima, very fine leaf blade with shade tolerance and good visual quality. The objectives of the study were to identify salt tolerant genotypes for the continued improvement of salt tolerance in turfgrass breeding programs, and for the potential use of these genotypes for landscaping under saline conditions in arid and semi-arid regions of the US.

MATERIALS AND METHODS

Plant material and culture

Seven turfgrass genotypes were acquired from the Texas A&M AgriLife Turfgrass Breeding Program and used in this study, including a Z. matrella cultivar 'Diamond', a Z. japonica 'Palisades', three Z. matrella × Z. japonica hybrids DALZ 1701 ([(Z. matrella × Z. matrella) × Z. japonica] × Z. japonica), DALZ 1713 ([Z. japonica × (Z. matrella × Z. matrella)], and 'Innovation' (Z. matrella × Z. japonica), and two Z. minima × Z. matrella hybrids (DALZ 1309 and 'Lazer'). Approximately 10 rhizomes were transplanted into 10-cm (top diameter) round plastic pots (volume: 450 mL, height: 8.5 cm) filled with potting mix (Sun Gro, Agawam, MA, USA) and fertigated through the surface of the pot with 20-10-20 (N-P₂O₅-K₂O) Peters Excel fertilizer (ICL, Sommerville, SC, USA) at a rate of 150 mg L⁻¹ N, on an as-needed basis (when the substrate surface became dry). The nutrient solution was made by mixing 1.0 g of the above fertilizer to 1 L of tap water. The final electrical conductivity (EC) and pH was 1.3 dS m⁻¹ and 6.5, respectively. Turfgrass cuttings were established and rooted for four weeks in a greenhouse at the Texas A&M AgriLife Research Center in Dallas, Texas (USA). A total of 24 pots with uniform growth of each genotype were selected. The genotypes were randomized on a greenhouse bench for the initiation of the saline treatments.

Saline treatments
Two saline treatments were used, in addition to a non-saline control, in this experiment. In the control group, plants were irrigated with the nutrient solution as mentioned above. The two saline treatments were prepared by the addition of NaCl to the nutrient solution to achieve EC levels of 5.0 dS m⁻¹ (EC5) and 10.0 dS m⁻¹ (EC10). These two salinity levels were chosen based on available information in the literature on salt tolerance of other turfgrasses. The treatment solution for EC5 or EC10 was prepared by adding 230 g or 550 g of NaCl to 100 L nutrient solution. The actual EC and pH were recorded each time. Treatments were applied to the plants overhead on an as-needed basis. Approximately 200 mL was applied to each plant/pot per treatment application which provided a leaching fraction of approximately 35% to reduce the accumulation of salts in the substrate throughout the experiment. The experiment was arranged in a split-plot design with treatments randomized in greenhouse benches and genotypes randomized within treatments. There were eight pots (replicates) per treatment. Weekly measurements of the leachate EC and pH were recorded to track the salinity level in the substrate and rhizosphere. The leachate was collected via the 'Pourthru' method as described by Cavins et al.^[21]. The treatments were initiated on 06 May 2020 and lasted eight weeks and was terminated on 02 July 2020.

Greenhouse environment
The experiment was conducted in a greenhouse at the Texas A&M AgriLife Research Center in Dallas, TX, USA (32°59'13.2" N 96°45'59.8" W; elevation 131 m). The greenhouse air temperature was controlled by an evaporative cooling wall and two exhaust fans. A 50% shade fabric was used throughout the experiment to reduce sunlight and heat in the greenhouse. Throughout the experiment, greenhouse air temperature and photosynthetic active radiation (PAR) were recorded by a datalogger (Campbell Scientific, Logan, UT, USA). The air temperature and quantum sensor (for PAR measurement) were installed right above the bench to capture the actual air temperature and light intensity near the plant canopy. The daily average air temperature during the experiment was 26.0 ± 3.56 °C (mean ± standard deviation) and the average daily light integral (DLI) was 12.0 ± 2.98 mol m⁻² d⁻¹.

Data collection
Throughout the experiment, the plants were clipped on a biweekly schedule and the clippings were collected and dried in a drying oven at 70 °C for dry weight determination. Clipping was accomplished by hand with scissors and a ruler, following a treatment application. The turfgrass was clipped to a 2-cm height and the perimeter of the pots were also trimmed. Additionally, after each biweekly trimming, the percent canopy green leaf area (GLA) was determined visually by two persons to assess the quality of the plants under the saline treatments. At harvest, shoot and root tissue were separated and dried in a drying oven at 70 °C for biomass determination. Roots were washed of substrate and rinsed briefly in reverse osmosis water before being placed in paper bags and dried in the drying oven. Following dry weight determination, three shoot samples from each treatment were ground in a Wiley mill (Thomas Scientific, Swedesboro, NH, USA) to pass a 40-mesh screen. Shoot tissue mineral contents were analyzed using inductively coupled plasma mass spectrometry (ICP-MS) using the methods described by Havlin & Soltanpour^[22] and Isaac & Johnson^[23]. Shoot tissue chloride content was determined by extraction with 2% acetic acid and analyzed using an M926 Chloride Analyzer (Cole Parmer Instrument Company, Vernon Hills, IL, USA) according to the methods described by Gavlak et al.^[24]

Statistical analysis
There was a total of three treatments and seven genotypes with eight replications each (N = 168). Data were analyzed as a two-way analysis of variance (ANOVA) with an alpha of 0.05 using JMP 15 (SAS, Cary, NC, USA). Means were separated using Tukey's Honest Significant Difference (HSD) test with an alpha of 0.05. Relative shoot dry weight (DW) was calculated as the shoot DW in the saline treatment/average shoot DW in control × 100%. Relative root DW and relative total DW in percentage were calculated in a similar fashion compared to control. Student's t-test was used for comparing the relative growth parameters between the two saline treatments.

DISCUSSION

Throughout the study, although the salt treatments remained fixed, salt accumulation occurred in the substrate as indicated by the increase in leachate EC of the salt treatments. Salt accumulation in the substrate depends on many factors, including salinity of the irrigation water, irrigation frequency, leaching fraction, and substrate type. The leaching fraction, simplified, is the percent of irrigation that drains out of the substrate^[25]. Higher leaching fractions can flush ions, including Na⁺ and Cl⁻, away from the root zone and out of the substrate. In this study, a leaching fraction of approximately 35% was applied to slow down the accumulation of salts in the substrate without wasting too much irrigation. Despite this, leachate EC increased throughout the study, most notably in the EC10 treatment. This imposed additional osmotic and/or ionic stress on the grasses beyond the fixed treatment salinities of 5.0 and 10 dS m⁻¹. Nevertheless, this is representative of irrigation regiments in arid landscaping, where low irrigation volumes and leaching fractions are commonly applied^[26].

Biomass was reduced by the high salt treatment more notably in the root tissue compared to the shoot tissue. In fact, shoot tissue increased marginally relative to the control in most genotypes when treated with salt, which is indicative of salt tolerance and halophytes^[24]. However, root tissue sensitivity to salt stress is rather unique, since typically shoot tissue is more sensitive^[10]. Chavarria et al.^[1] observed both increases and reductions in root mass among eight turfgrass genotypes when treated with 15 and 30 dS m⁻¹ salinity, when compared to the control. Additionally, in the present study the cluster analysis based on shoot and root tissue biomass identified the genotypes DALZ 1309, DALZ 1713, and Diamond as the most salt tolerant, which corresponds with the relative root tissue DW in the EC5 treatment. Therefore, our results indicate that root tissue biomass in turfgrasses might be a greater indication of salinity tolerance than shoot tissue biomass. This could be because grasses have relatively small leaf surface area and large root/shoot ratios compared to other plants^[27].

Salt tolerance for ornamental crops not only depends on growth under saline conditions, but also visual appearance, as salinity can impose ionic stress to plants which can lead to leaf burn^[12]. For turfgrasses, this is especially true due to its primary use for aesthetics and environmental benefits in residential, recreational, or commercial landscaping^[28]. For turfgrass managers, greenness can be a more important trait than shoot yield. Marcumm & Pessarakli^[29] reported GLA ranges of 7% to 84% in eight Distichlis spicata turfgrass genotypes treated with up to 1.0 mol L⁻¹ (58.5 g L⁻¹) NaCl for one week. In the present study, our results indicate relatively high GLA and hence, excellent visual quality in most genotypes even when treated with high (EC10) salinity for eight weeks. In contrast, the significant reductions in GLA in the genotype DALZ 1309 in the EC10 treatment indicates less salt tolerance and susceptibility to ionic stress.

Continued growth under saline conditions is another desirable trait in turfgrasses, indicating long-term establishment. However, mowing is a necessary management practice for turfgrass and has been shown to affect the salinity tolerance of turfgrass varieties. For example, clipping yield of creeping bentgrass (Agrostis palustris) was reduced the most under low mowing height (6.4 mm) compared to high mowing height (25.4 mm), when treated with salinity ranging from 5 to 15 dS m⁻¹^[30]. Similarly, our results showed that high salinity (EC10) reduced clipping DW compared to the control at a mowing height of 2.0 cm. However, at the end of the study, clipping DW tended to decline in the EC10 treatment in the genotypes DALZ 1701, Innovation, and Palisades, indicating less tolerance to salinity at this mowing height, whereas the remaining genotypes showed marginal gains in clipping DW, most notably Diamond, indicating greater tolerance to salinity at the respective mowing height.

A key mechanism of salt tolerance in plants is the ability to exclude Na⁺ and Cl⁻ from the leaf tissue by various means, such as sequestration in the cell vacuole or excretion through specific glands in the leaf^[6]. Therefore, concentration of salts in the shoot tissue can be an indication of salt tolerance and/or mechanisms of salinity tolerance in a specific plant. In the present study, Na⁺ and Cl⁻ concentrations in the shoot tissue increased in all genotypes when treated with high salt (EC10), indicating salt accumulation in the shoot tissue. However, our results also showed genotypic variation in salt accumulation, indicating different mechanisms for dealing with the salts. For example, genotype DALZ 1701 had the lowest concentration of Na⁺ which correlated with its excellent visual and growth parameter, and a potential explanation for this is it could more affectively exclude Na⁺ or excrete it from the leaves, which could contribute to its excellent visual quality (high GLA) and overall good salt tolerance in the present study. Chavarria et al.^[1] reported Na⁺ concentrations in the shoot tissue ranging from 7.2 to 22.4 mg g⁻¹ of eight warm-season turfgrasses when treated with 15 dS m⁻¹, which were comparable values to those reported here considering the higher salt treatment. However, they also reported that salt excretion correlated with salt tolerant genotypes. Nevertheless, the ability to maintain high Na⁺ and Cl⁻ concentrations in the shoot tissue while maintaining good visual quality and growth, indicates tolerance to osmotic and ionic salinity stress, which our results demonstrated.

Source	Shoot DW	Root DW	Total DW	R. Shoot DW	R. Root DW	R. Total DW	GLA	Clipping DW	Shoot Na⁺	Shoot Cl⁻
Model	0.0004	< 0.0001	< 0.0001	0.0003	0.0001	<.0001	< 0.0001	< 0.0001	< 0.0001	< 0.0001
Treatment (T)	0.008	0.0001	0.001	NS	0.0002	0.0013	< 0.0001	< 0.0001	< 0.0001	< 0.0001
Genotype (G)	0.0006	< 0.0001	< 0.0001	< 0.0001	0.0014	< 0.0001	< 0.0001	< 0.0001	< 0.0001	< 0.0001
T × G	NS	0.0424	0.0109	NS	NS	NS	< 0.0001	< 0.0001	< 0.0001	< 0.0001

Genotype	Control	EC5	EC10
Lazer	100.0 ± 0.0Aa	99.4 ± 0.6Aab	95 ± 2.1Ab
DALZ 1309	95.6 ± 2.0Ba	88.8 ± 3.0Ca	69.4 ± 6.4Bb
DALZ 1701	98.1 ± 0.9ABa	99.4 ± 0.6Aa	97.5 ± 0.9Aa
DALZ 1713	100.0 ± 0.0Aa	99.4 ± 0.6Aa	100.0 ± 0.0Aa
Diamond	99.4 ± 0.6ABa	100.0 ± 0.0Aa	98.8 ± 1.3Aa
Innovation	98.1 ± 1.3ABa	91.9 ± 2.5BCab	88.1 ± 3.3Ab
Palisades	100.0 ± 0.0Aa	98.1 ± 0.9ABa	98.1 ± 0.9Aa
Different letters indicate significant differences Tukey's HSD test; uppercase among genotypes and lowercase among treatments.

Genotype	Control	EC5	EC10
Na⁺
Lazer	1.70 ± 0.04BCb	8.76 ± 0.90Aa	9.90 ± 0.42CDa
DALZ 1309	2.07 ± 0.24ABCc	8.89 ± 0.13Ab	15.68 ± 1.43Aa
DALZ 1701	1.39 ± 0.06Cc	6.58 ± 0.41Ab	7.92 ± 0.07Da
DALZ 1713	2.79 ± 0.19Ac	9.33 ± 0.28Ab	12.86 ± 0.72ABCa
Diamond	1.81 ± 0.02BCc	7.58 ± 0.33Ab	10.78 ± 0.36CDa
Innovation	1.77 ± 0.15BCc	9.46 ± 1.02Ab	14.92 ± 0.56ABa
Palisades	2.45 ± 0.37ABc	8.84 ± 1.12Ab	12.19 ± 0.30BCa
Cl⁻
Lazer	7.55 ± 0.30Ac	9.72 ± 0.56Db	14.00 ± 0.28Ba
DALZ 1309	5.65 ± 0.23Ab	12.20 ± 0.02BCDb	23.97 ± 2.84Aa
DALZ 1701	5.67 ± 0.22Ac	9.68 ± 0.68Db	13.25 ± 0.77Ba
DALZ 1713	6.55 ± 0.10Ac	13.75 ± 0.31ABCb	19.15 ± 0.40ABa
Diamond	6.43 ± 0.10Ac	11.52 ± 0.10CDb	15.38 ± 0.15Ba
Innovation	5.72 ± 0.41Ac	14.88 ± 0.75ABb	25.18 ± 1.45Aa
Palisades	7.27 ± 0.99Ab	15.97 ± 1.36Aa	18.87 ± 0.96ABa
Means and standard errors are presented (n = 8). Different letters indicate significant differences Tukey's HSD test; uppercase among genotypes and lowercase among treatments.

Contents

Policies

Services

Partnerships

Salt tolerance of seven genotypes of zoysiagrass (Zoysia spp.)

Abstract

Introduction

Tank data considered in this research

Strategies to establish a population in order to construct fragility curves

Tanks considered for the Patagonian region

Wind pressures adopted for tanks located in the Patagonian region

Evaluation of fragility curves

Damage in wind loaded oil storage tanks

Damage states considered in this work

Finite element evaluation of damage states

Fragility curves for damage states DS0, DS1, DS2, and DS3

Methodology

Results

Discussion

Conclusions

Author contributions

Data availability

Acknowledgments

Conflict of interest

Supplementary information

Rights and permissions

References

About this article

Cite this article

Article Metrics

Access History

Other Articles By Authors

Salt tolerance of seven genotypes of zoysiagrass (Zoysia spp.)

HTML

Plant material and culture

Saline treatments

Greenhouse environment

Data collection

Statistical analysis

Leachate Electrical conductivity (EC)

Relative tissue dry weight (DW)

Hierarchal cluster analysis

Green Leaf Area (GLA) index

Clipping Dry Weight (DW)

Tissue sodium and chloride content

Catalog

Html

Plant material and culture

Saline treatments

Greenhouse environment

Data collection

Statistical analysis

Leachate electrical conductivity (ec)

Relative tissue dry weight (dw)

Hierarchal cluster analysis

Green leaf area (gla) index

Clipping dry weight (dw)

Tissue sodium and chloride content

Contents

Policies

Services

Partnerships

Export File

Citation

Format

Content