Dual-virus co-infection reduces photosynthetic rate, yield, and sensitivity of photosynthetic rate to leaf-air VPD in <i>Pseudostellaria heterophylla</i>

Boqin Zheng; Zhenghua Wang; Lingfeng Zeng; Ding Wang; Zuyun Ye; Boqin Zheng; Zhenghua Wang; Lingfeng Zeng; Ding Wang; Zuyun Ye

doi:10.48130/mpb-0025-0002

2025 Volume 4

Article Contents

Abstract

References

Next Previous

ARTICLE Open Access

Dual-virus co-infection reduces photosynthetic rate, yield, and sensitivity of photosynthetic rate to leaf-air VPD in Pseudostellaria heterophylla

1.
The Engineering Technology Research Center of Characteristic Medicinal Plants of Fujian, School of Biological Science and Engineering, Ningde Normal University, Ningde 352100, China
2.
CAS Key Laboratory of Tropical Plant Resource and Sustainable Use, Xishuangbanna Tropical Botanical Garden, Chinese Academy of Sciences, Mengla, Xishuangbanna 666303, China
3.
College of Life Sciences, University of Chinese Academy of Sciences, Beijing 101408, China
^# Authors contributed equally: Boqin Zheng, Zhenghua Wang

More Information

Corresponding authors: zhenhua.wang@ndnu.edu.cn; zyye0593@163.com

Received: 18 October 2024
Revised: 02 January 2025
Accepted: 13 January 2025
Published online: 04 March 2025
Medicinal Plant Biology 4, Article number: e007 (2025) | Cite this article

Abstract

Viral infections exert a complex influence on plant growth, modifying tolerance to abiotic stresses, with effects varying depending on the specific virus. Pseudostellaria heterophylla, a medicinal herb, is often infected by Turnip mosaic virus and Broad bean wilt virus 2, leading to mosaic disease. This study comprehensively investigated the effects of diverse viral infections on plant growth and response to environmental factors, evaluating specific leaf weight, chlorophyll content, stomatal conductance, net photosynthetic rate, transpiration rate, yield, aqueous extract, and polysaccharide content. Results indicate that Turnip mosaic virus and Broad bean wilt virus 2 co-infection result in decreased chlorophyll content, stomatal conductance, net photosynthetic rate, transpiration rate, yield, and polysaccharide content in Pseudostellaria heterophylla, compared with Broad bean wilt virus 2 infection alone. Broad bean wilt virus 2 alone only reduces chlorophyll and polysaccharide content. Plants infected with both viruses show a reduced response to leaf-air vapor pressure deficit in stomatal conductance, net photosynthetic rate, and transpiration rate compared to singly infected plants. Thus, the eradication of Turnip mosaic virus should be prioritized for Pseudostellaria heterophylla cultivation.
- Pseudostellaria heterophylla,
- Yield,
- Turnip mosaic virus,
- Broad bean wilt virus,
- Photosynthesis,
- Vapor pressure deficit

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

Supplementary Table S1 PCR primers used in this study.
Supplementary Table S2 Effects of different viral infections on specific leaf weight, chlorophyll content, photosynthetic gas exchange parameters, yield and extract content of Pseudostellaria heterophylla.
Supplementary Table S3 The multiple linear regression model for net photosynthesis rate, dried tuberous root yield, aqueous extract content of tuberous roots and polysaccharide content of tuberous roots.
Supplementary Fig. S1 Detection of virus types infecting Pseudostellaria heterophylla seedlings after micro-stem tip tissue culture using duplex RT-PCR.
Supplementary Fig. S2 Correlation analysis of leaf functional traits with yield and extracts. The red circle indicates a negative Pearson correlation coefficient and the blue one a positive Pearson correlation coefficient.
Supplementary Fig. S3 Response of net photosynthesis rate (a, d), stomatal conductance (b, e) and transpiration rate (c, f) to VPDL among different viral infections.

Rights and permissions
Copyright: © 2025 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]	Lu DQ, Richard KR. 2001. Flora of China, ed. Wu ZY. Vol. 16. Beijing: Science Press. pp. 7–10
[2]	Xiao Q, Zhao L, Jiang C, Zhu Y, Zhang J, et al. 2022. Polysaccharides from Pseudostellaria heterophylla modulate gut microbiota and alleviate syndrome of spleen deficiency in rats. Scientific Reports 12:20217 doi: 10.1038/s41598-022-24329-9 CrossRef Google Scholar
[3]	Kan Y, Liu Y, Huang Y, Zhao L, Jiang C, et al. 2022. The regulatory effects of Pseudostellaria heterophylla polysaccharide on immune function and gut flora in immunosuppressed mice. Food Science & Nutrition 10:3828−41 doi: 10.1002/fsn3.2979 CrossRef Google Scholar
[4]	Li Y, Liu S, Guo K, Ding W, Wang R. 2022. Virome of Pseudostellaria heterophylla: Identification and characterization of three novel carlaviruses and one novel amalgavirus associated with viral diseases of Pseudostellaria heterophylla. Frontiers in Microbiology 13:955089 doi: 10.3389/fmicb.2022.955089 CrossRef Google Scholar
[5]	Nellist CF, Ohshima K, Ponz F, Walsh JA. 2022. Turnip mosaic virus, a virus for all seasons. Annals of Applied Biology 180:312−27 doi: 10.1111/aab.12755 CrossRef Google Scholar
[6]	Chen J, Zhou Y, Gu T, Guo X, Zhuang X, et al. 2022. First report of broad bean wilt virus 2 on Mirabilis jalapa in China. Plant Disease 107(6):149−58 doi: 10.1094/PDIS-06-22-1310-PDN CrossRef Google Scholar
[7]	Fuji SI, Mochizuki N, Fujinaga M, Ikeda M, Shinoda K, et al. 2007. Incidence of viruses in Alstroemeria plants cultivated in Japan and characterization of Broad bean wilt virus-2, Cucumber mosaic virus and Youcai mosaic virus. Journal of General Plant Pathology 73:216−21 doi: 10.1007/s10327-007-0009-9 CrossRef Google Scholar
[8]	Wang S, Cui W, Wu X, Yuan Q, Zhao J, et al. 2018. Suppression of nbe-miR166h-p5 attenuates leaf yellowing symptoms of potato virus X on Nicotiana benthamiana and reduces virus accumulation. Molecular Plant Pathology 19:2384−96 doi: 10.1111/mpp.12717 CrossRef Google Scholar
[9]	Nicaise V. 2014. Crop immunity against viruses: outcomes and future challenges. Frontiers in Plant Science 5:660 doi: 10.3389/fpls.2014.00660 CrossRef Google Scholar
[10]	Yang X, Huang J, Liu C, Chen B, Zhang T, et al. 2017. Rice stripe mosaic virus, a novel cytorhabdovirus infecting rice via leafhopper transmission. Frontiers in Microbiology 7:2140 doi: 10.3389/fmicb.2016.02140 CrossRef Google Scholar
[11]	Noman A, Aqeel M, Qasim M, Haider I, Lou Y. 2020. Plant-insect-microbe interaction: a love triangle between enemies in ecosystem. Science of the Total Environment 699:134181 doi: 10.1016/j.scitotenv.2019.134181 CrossRef Google Scholar
[12]	Wang LL, Wei XM, Ye XD, Xu HX, Zhou XP, et al. 2014. Expression and functional characterisation of a soluble form of Tomato yellow leaf curl virus coat protein. Pest Management Science 70:1624−31 doi: 10.1002/ps.3750 CrossRef Google Scholar
[13]	Llave C. 2016. Dynamic cross-talk between host primary metabolism and viruses during infections in plants. Current Opinion in Virology 19:50−55 doi: 10.1016/j.coviro.2016.06.013 CrossRef Google Scholar
[14]	Untiveros M, Fuentes S, Kreuze J. 2008. Molecular variability of sweet potato feathery mottle virus and other potyviruses infecting sweet potato in Peru. Archives of Virology 153:473−83 doi: 10.1007/s00705-007-0019-0 CrossRef Google Scholar
[15]	Zhu JK. 2016. Abiotic stress signaling and responses in plants. Cell 167:313−24 doi: 10.1016/j.cell.2016.08.029 CrossRef Google Scholar
[16]	Ma L, Zhang ZH, Yao BQ, Ma Z, Huang XT, et al. 2021. Effects of drought and heat on the productivity and photosynthetic characteristics of alpine meadow plants on the Qinghai-Tibetan Plateau. Journal of Mountain Science 18:2079−93 doi: 10.1007/s11629-020-6561-x CrossRef Google Scholar
[17]	Mott KA, Peak D. 2013. Testing a vapour-phase model of stomatal responses to humidity. Plant, Cell & Environment 36:936−44 doi: 10.1111/pce.12026 CrossRef Google Scholar
[18]	Wang Z, Wang C, Wang B, Wang X, Li J, et al. 2020. Interactive effects of air pollutants and atmospheric moisture stress on aspen growth and photosynthesis along an urban-rural gradient. Environmental Pollution 260:114076 doi: 10.1016/j.envpol.2020.114076 CrossRef Google Scholar
[19]	Atkinson NJ, Urwin PE. 2012. The interaction of plant biotic and abiotic stresses: from genes to the field. Journal of Experimental Botany 63:3523−43 doi: 10.1093/jxb/ers100 CrossRef Google Scholar
[20]	Mittler R. 2006. Abiotic stress, the field environment and stress combination. Trends in Plant Science 11:15−19 doi: 10.1016/j.tplants.2005.11.002 CrossRef Google Scholar
[21]	Mishra R, Shteinberg M, Shkolnik D, Anfoka G, Czosnek H, et al. 2022. Interplay between abiotic (drought) and biotic (virus) stresses in tomato plants. Molecular Plant Pathology 23:475−88 doi: 10.1111/mpp.13172 CrossRef Google Scholar
[22]	Bergès SE, Vasseur F, Bediée A, Rolland G, Masclef D, et al. 2020. Natural variation of Arabidopsis thaliana responses to Cauliflower mosaic virus infection upon water deficit. PLoS Pathogens 16:e1008557 doi: 10.1371/journal.ppat.1008557 CrossRef Google Scholar
[23]	Pandey P, Patil M, Priya P, Senthil-Kumar M. 2024. When two negatives make a positive: the favorable impact of the combination of abiotic stress and pathogen infection on plants. Journal of Experimental Botany 75:674−88 doi: 10.1093/jxb/erad413 CrossRef Google Scholar
[24]	Gorovits R, Sobol I, Altaleb M, Czosnek H, Anfoka G. 2019. Taking advantage of a pathogen: understanding how a virus alleviates plant stress response. Phytopathology Research 1:20 doi: 10.1186/s42483-019-0028-4 CrossRef Google Scholar
[25]	Shteinberg M, Mishra R, Anfoka G, Altaleb M, Brotman Y, et al. 2021. Tomato yellow leaf curl virus (TYLCV) promotes plant tolerance to drought. Cells 10:2875 doi: 10.3390/cells10112875 CrossRef Google Scholar
[26]	Kuang YB, Cheng MZ, Lu YR, Cheng J, Ye ZY. 2017. Detection of Turnip mosaic virus and Broad bean wilt virus in Pseudostellaria heterophylla by duplex RT-PCR. Acta Phytophylacica Sinica 44:184−791(in Chinese) doi: 10.16420/j.issn.0513-353x.2016-0940 CrossRef Google Scholar
[27]	Porra RJ, Thompson WA, Kriedemann PE. 1989. Determination of accurate extinction coefficients and simultaneous equations for assaying chlorophylls a and b extracted with four different solvents: verification of the concentration of chlorophyll standards by atomic absorption spectroscopy. Biochimica et Biophysica Acta (BBA) - Bioenergetics 975:384−94 doi: 10.1016/S0005-2728(89)80347-0 CrossRef Google Scholar
[28]	Chinese Pharmacopoeia Commission. 2020. Pharmacopoeia of the People's Republic of China. 2020 Edition. China Medical Science Press: Beijing, China
[29]	Liang TT, Zhou Y, Lin B, Feng TT, Liu XL, et al. 2013. Polysaccharide extraction from Radix pseudostellariae with water extracting-ethanol precipitation method. Journal of Mountain Agriculture and Biology 32:79−82(in Chinese) doi: 10.15958/j.cnki.sdnyswxb.2013.01.003 CrossRef Google Scholar
[30]	Benedetti CE, Costa CL, Turcinelli SR, Arruda P. 1998. Differential expression of a novel gene in response to coronatine, methyl jasmonate, and wounding in the Coi1 mutant of Arabidopsis. Plant Physiology 116:1037−42 doi: 10.1104/pp.116.3.1037 CrossRef Google Scholar
[31]	Su J, Yang L, Zhu Q, Wu H, He Y, et al. 2018. Active photosynthetic inhibition mediated by MPK3/MPK6 is critical to effector-triggered immunity. PLoS Biology 16:e2004122 doi: 10.1371/journal.pbio.2004122 CrossRef Google Scholar
[32]	Zhai Y, Yuan Q, Qiu S, Li S, Li M, et al. 2021. Turnip mosaic virus impairs perinuclear chloroplast clustering to facilitate viral infection. Plant, Cell & Environment 44:3681−99 doi: 10.1111/pce.14157 CrossRef Google Scholar
[33]	Tahmasebi AA, Dizadji A, Koohi Habibi M. 2013. Interaction of Cucumber mosaic virus and Bean yellow mosaic virus in co-infected plants of bean and broad bean. Archives of Phytopathology and Plant Protection 46:1081−92 doi: 10.1080/03235408.2012.757865 CrossRef Google Scholar
[34]	Abrahamian P, Sobh H, Seblani R, Abou-Jawdah Y. 2015. Co-infection of two criniviruses and a begomovirus enhances the disease severity in cucumber. European Journal of Plant Pathology 142:521−30 doi: 10.1007/s10658-015-0630-y CrossRef Google Scholar
[35]	Kiptui LJ, Toroitich FJ, Kilalo DC, Obonyo M. 2020. Interaction between cowpea aphid-borne mosaic virus isolates and its effect on passion fruit woodiness disease on Passiflora edulis Sims and Passiflora ligularis juss. Advances in Agriculture 2020:8876498 doi: 10.1155/2020/8876498 CrossRef Google Scholar
[36]	He W, Li J, Pu M, Xu ZG, Gan L. 2020. Response of photosynthate distribution in potato plants to different LED spectra. Functional Plant Biology 47:1128−37 doi: 10.1071/FP20131 CrossRef Google Scholar
[37]	Sugimoto H, Fujita T, Koesmaryono Y, Sato T. 1997. Canopy light distribution, photosynthesis and Tuber yield of eddoe plant characterized by clipping and non-clipping of daughter Tuber leaves. Journal of Agricultural Meteorology 52:889−92 doi: 10.2480/agrmet.52.889 CrossRef Google Scholar
[38]	Wilkinson S, Weston AK, Marks DJ. 2020. Stabilising urea amine nitrogen increases potato Tuber yield by increasing chlorophyll content, reducing shoot growth rate and increasing biomass partitioning to roots and tubers. Potato Research 63:217−39 doi: 10.1007/s11540-019-09436-x CrossRef Google Scholar
[39]	Qu M, Zheng G, Hamdani S, Essemine J, Song Q, et al. 2017. Leaf photosynthetic parameters related to biomass accumulation in a global rice diversity survey. Plant Physiology 175:248−58 doi: 10.1104/pp.17.00332 CrossRef Google Scholar
[40]	Takai T, Ohsumi A, Arai Y, Iwasawa N, Yano M, et al. 2013. QTL analysis of leaf photosynthesis in rice. Japan Agricultural Research Quarterly: JARQ 47:227−35 doi: 10.6090/jarq.47.227 CrossRef Google Scholar
[41]	Chini A, Grant JJ, Seki M, Shinozaki K, Loake GJ. 2004. Drought tolerance established by enhanced expression of the CC-NBS-LRR gene, ADR1 requires salicylic acid, EDS1 and ABI1. The Plant Journal: for Cell and Molecular Biology 38:810−22 doi: 10.1111/j.1365-313X.2004.02086.x CrossRef Google Scholar
[42]	Kasuga M, Liu Q, Miura S, Yamaguchi-Shinozaki K, Shinozaki K. 1999. Improving plant drought, salt, and freezing tolerance by gene transfer of a single stress-inducible transcription factor. Nature Biotechnology 17:287−91 doi: 10.1038/7036 CrossRef Google Scholar
[43]	Augustine SM, Tzigos S, Snowdon R. 2022. Heat-killed tobacco mosaic virus mitigates plant abiotic stress symptoms. Microorganisms 11:87 doi: 10.3390/microorganisms11010087 CrossRef Google Scholar
[44]	Aguilar E, Cutrona C, Del Toro FJ, Vallarino JG, Osorio S, et al. 2017. Virulence determines beneficial trade-offs in the response of virus-infected plants to drought via induction of salicylic acid. Plant, Cell & Environment 40:2909−30 doi: 10.1111/pce.13028 CrossRef Google Scholar
[45]	Koo YM, Heo AY, Choi HW. 2020. Salicylic acid as a safe plant protector and growth regulator. The Plant Pathology Journal 36:1−10 doi: 10.5423/PPJ.RW.12.2019.0295 CrossRef Google Scholar
[46]	Zhao M, Ren Y, Wei W, Yang J, Zhong Q, et al. 2021. Metabolite analysis of Jerusalem artichoke (Helianthus tuberosus L. ) seedlings in response to polyethylene glycol-simulated drought stress. International Journal of Molecular Sciences 22:3294 doi: 10.3390/ijms22073294 CrossRef Google Scholar
[47]	Piasecka A, Sawikowska A, Kuczyńska A, Ogrodowicz P, Mikołajczak K, et al. 2017. Drought-related secondary metabolites of barley (Hordeum vulgare L.) leaves and their metabolomic quantitative trait loci. The Plant Journal 89:898−913 doi: 10.1111/tpj.13430 CrossRef Google Scholar
[48]	Huang X, Rong W, Zhang X, Gao Y, Zhou Y, et al. 2024. Transcriptome and metabolome analysis reveal the dynamic changes and biosynthesis pathways of alkaloids in Sophora alopecuroides L. under drought stress. Industrial Crops and Products 212:118365 doi: 10.1016/j.indcrop.2024.118365 CrossRef Google Scholar

About this article

Cite this article

Zheng B, Wang Z, Zeng L, Wang D, Ye Z. 2025. Dual-virus co-infection reduces photosynthetic rate, yield, and sensitivity of photosynthetic rate to leaf-air VPD in Pseudostellaria heterophylla. Medicinal Plant Biology 4: e007 doi: 10.48130/mpb-0025-0002

Zheng B, Wang Z, Zeng L, Wang D, Ye Z. 2025. Dual-virus co-infection reduces photosynthetic rate, yield, and sensitivity of photosynthetic rate to leaf-air VPD in Pseudostellaria heterophylla. Medicinal Plant Biology 4: e007 doi: 10.48130/mpb-0025-0002

Figures(7)

Download PDF

Article Metrics

Article views(401) PDF downloads(97)

Other Articles By Authors

on this site
on Google Scholar

HTML

Introduction

Pseudostellaria heterophylla (Miq.) Pax. (P. heterophylla) is a medicinal herb belonging to the Caryophyllaceae family^[1] and previous experiments have demonstrated that extracts from its tuberous roots possess biological activity, including anti-inflammatory attributes, and immune enhancement^[2,3]. P. heterophylla undergoes a new generation of asexual reproduction through the formation of tuberous roots. It is worthy of note that within cultivation areas, P. heterophylla is frequently infected by viruses, including Turnip mosaic virus (TuMV) and Broad bean wilt virus 2 (BBWV2) in particular^[4]. Infection of plants with either of these two viruses results in the characteristic symptoms of chlorosis, yellowing, and mosaic patterns of leaves^[5−7]. Additionally, the chlorophyll content is typically significantly reduced in infected plants compared to uninfected controls^[8]. This also results in a reduction in the rate of photosynthesis, which ultimately affects the growth and development of the plant, potentially leading to stunted growth or even death. Although symptoms of TuMV or BBWV2 infection have been identified, it is still difficult to determine the specific virus species based on symptoms, which makes it difficult to determine the relative importance of different viruses in causing disease as well.

Viruses constitute a significant portion of plant pathogens and exert a considerable economic impact on agriculture^[9]. The detrimental effects of viruses on plant health manifest through a variety of symptoms, including leaf yellowing, crumpling, stunting, and in severe cases, plant death^[10−12]. These symptoms are collectively referred to as mosaic diseases, which represent a category of plant disorders caused by viral infections. In addition to the aforementioned effects on plant growth, it is crucial to recognize that viral infections can disrupt photosynthesis. This disruption occurs by damaging the structure of chloroplasts and decreasing chlorophyll content, ultimately leading to a significant reduction in crop yields^[13,14]. Understanding the wide-ranging impacts of viral infections on plant health is critical to developing effective agricultural management strategies.

In delving into the impact of viruses on plant growth, it becomes imperative to contemplate the role of environmental factors and their potential interplay with viral effects. Environmental conditions such as water deficit and elevated temperatures significantly curtail above-ground biomass, whereas high vapor pressure deficit triggers water loss and subsequent stomatal closure^[15−18]. Viral infections intricately modulate plant responses to environmental stimuli^[19,20]. For instance, the infection of Tomato yellow leaf curl virus has been observed to delay the decline in transpiration rate and biomass of tomatoes under drought stress^[21,22]. Prior research suggests that viral infections possess the capacity to bolster plant resilience against environmental stress like heat and drought^[23,24]. However, a comprehensive understanding of the distinct effects of various viral infections on plant growth and their responses to diverse environmental factors (such as temperature, relative air humidity, and vapor pressure deficit) necessitates further exploration^[25].

To explore the effects of different viral infections on plant growth and the response of plants infected with different viruses to changing environmental conditions, we measured the specific leaf weight, chlorophyll content, photosynthesis, yield, and extract content of P. heterophylla infected with different viruses to compare the effects of different viral infections on these traits. This study endeavors to address several key inquiries: (1) How does virus infection affect the net photosynthetic rate, yield, and extract content of P. heterophylla? (2) Which virus, TuMV or BBWV2, exerts a more substantial impact on plant growth? (3) How does virus infection regulate physiological indicators of P. heterophylla in response to environmental factors? We assume that either TuMV or BBWV2 infection would decrease photosynthetic rate, yield, and polysaccharide content. To elucidate these queries, we obtained P. heterophylla plants infected with diverse virus species and subsequently compared specific leaf weight, chlorophyll content, stomatal conductance, net photosynthetic rate, transpiration rate, yield, extract content, and the responsiveness of physiological indicators, such as net photosynthetic rates, to distinct environmental factors across various virus infections.

Materials and methods

Site description

The experiment was carried out at the Taizishen Agricultural Ecosystem Research Station (27°15'30'' N, 119°50'15'' E) in Zherong County, Ningde City, Fujian Province, China. The station, situated at an elevation of 986.3 m, experiences a subtropical monsoon climate with an annual mean temperature of 15.5 °C and annual rainfall of 2,061.9 mm. The soil in the area is classified as yellow earth with a pH range of 4.5 to 5.1.

Experimental design
For this study, P. heterophylla variety, Zhe-shen 2, was chosen as the plant material, known to carry TuMV and BBWV2. Zhe-shen 2 is extensively cultivated in Fujian Province (China), the primary planting area for P. heterophylla. In this experiment, 85 micro-stem tip tissue of Zhe-shen 2 were cultured and six of them eventually developed into seedlings. Subsequently, we expanded the number of these plants by tissue culture techniques and identified the type of viral infection of these six groups (namely, TS-1 to TS-6). Duplex reverse-transcription polymerase chain reaction (RT-PCR) with specific primer pairs (Supplementary Table S1) was used to identify the type of virus infecting P. heterophylla as described in Kuang et al.^[26]. TuMV was identified by amplified primer pairs TuMV-F (AGGTGAAAYGCTTGATGCAGGTY) and TuMV-R (GTTHCCATCARKCCGAACAAAT). BBWV2 was identified by amplified primer pairs BBWV-F (TTGGGHTCWAGYYTGGGACGYTTRT), and BBWV-R (TTRTARAACTTCTTGCTCCCACGM). The results of virus detection are illustrated in Supplementary Fig. S1. TS-1 was detected to carry BBWV2, TS-4 was detected to carry both TuMV and BBWV2 and the rest were not detected as carrying the virus (Supplementary Fig. S1). In the field experiment, tuberous roots of TS-4, TS-1, and TS-2 were planted and the corresponding plants were used as plants infected with two viruses (VI2), plants infected with one virus (VI1) and plants not infected with virus (VI0). These groups were isolated and propagated separately through tissue culture to obtain tuberous roots for the field experiments. In December 2020, the tuberous roots of VI0, VI1, and VI2 were planted in six blocks, each containing three plots where VI0, VI1, and VI2 were randomly planted. Each plot measured 0.5 m long and 0.65 m wide and the plants were cultivated under uniform management.

Measurements of specific leaf weight and chlorophyll content
During April 2021, at the mid-growth stage, one mature and undamaged leaves of one P. heterophylla from each of the six replicate plots for each treatment was randomly selected from each plot for determination of specific leaf weight (SLW), chlorophyll a content (Chl a), and chlorophyll b content (Chl b). The leaves were stored in ice boxes using zip-lock bags and transported to the laboratory. Leaf tissue samples were taken from each side of the leaf veins using a circular punch with an inner diameter of 1.06 cm. The leaf tissues were then dried in an oven at 60 °C until a constant weight was achieved. The dried tissues were then weighed using an electronic analytical balance. SLW was calculated as the ratio of leaf dry weight to leaf area). Chlorophyll was extracted using an acetone/ethanol mixture (1/1, V/V), and absorbance values at 647 and 664 nm were measured separately using a T6-New Century UV-visible spectrophotometer (PERSEE, China). The chlorophyll concentration formula was used to determine chlorophyll concentrations^[27].

Measurement of leaf photosynthesis
In July 2021, at the mid-growth stage of P. heterophylla, leaf photosynthesis was measured three times using the LI-6400XT portable photosynthesis measurement system (LI-COR, USA). A representative mature leaf from each of the six plots for each treatment was selected from each plot to measure stomatal conductance (Gs), net photosynthetic rate (Pn), and transpiration rate (Tr). Photosynthetic active radiation in the leaf chamber was set to 1,000 μmol m⁻² s⁻¹. The gas flow rate was set to 500 μmol s⁻¹ and the CO₂ concentration in the reference chamber to 400 ppm. The environmental parameters including leaf temperature (T_leaf), relative air humidity (RH), and leaf-air vapor pressure deficit (VPD_L) were also recorded by the photosynthesis measurement system.

Measurements of yield, aqueous extract, and polysaccharides
In July 2021, when P. heterophylla was harvested, tuberous roots were excavated from each plot to measure yield. The soil was washed off the fresh tuberous roots. The clean tuberous roots were sun-dried for 3 d to a constant weight. These dried tuberous roots were then used to determine the dried tuberous root yield (DY). To determine the aqueous extract (AE), and polysaccharide content (PS), the dried tuberous roots were ground to powder using a grinder and passed through a 24 mesh sieve. According to the Pharmacopoeia of the PRC^[28], AE was determined by the cold soak method. PS was determined by water extract-ethanol precipitation and modified phenol-sulfuric acid with D-glucose methods^[29].

Statistical analysis
All statistical analyses were performed using R (version 4.2.2). One-way ANOVA and least significant difference tests were performed using the 'agricolae' package with a significance level of 0.05. Box-and-line and violin plots were generated using the R package 'tidyverse' to show the distribution of the data. Pearson's correlation analysis was conducted using the 'corrplot' R package. Analysis of covariance was carried out using the R package 'HH'. Scatter plots were generated using the 'ggplot2' R package. Multiple linear regression was used to screen the optimal explanatory variables for Pn, DY, AE, and PS. Multiple covariance test (VIF < 5) was also performed for each explanatory variable using the 'car' R package and the relative importance of the optimal explanatory variables was analyzed and plotted using the 'relaimpo' R package.

Discussion

Leaf yellowing, a prevalent symptom in plants infected with TuMV or BBWV2, hints at a possible influence of these viruses on chlorophyll. Previous studies have documented that the detrimental effects of viruses on chloroplasts may contribute to this yellowing, primarily due to the accelerated degradation of chlorophyll, which can hinder photosynthetic processes^[30−32]. In our study, P. heterophylla infected with virus exhibit yellow spots and a significant decrease in chlorophyll content. It is noteworthy that the chlorophyll content of plants infected only with BBWV2 was significantly lower than that of plants without virus and the chlorophyll content of plants infected with both TuMV and BBWV2 was significantly lower than that of plants infected only with the BBWV2. Moreover, decreases in Gs, Pn, and Tr were more pronounced in plants infected with both TuMV and BBWV2 than in those infected with BBWV2 only, which is consistent with previous findings that dual or triple viral infections tend to result in more severe symptoms than those caused by single viral infections^[33−35].

The relationship between plant yield and carbohydrate content with photosynthetic activity is well established in research on potato and Colocasia esculenta^[36,37]. This relationship is closely associated with increased photosynthesis, which has been linked to enhanced carbohydrate transfer to subsurface organs^[38]. At harvest, the underground roots of the P. heterophylla have expanded into tubers, thereby completing the transfer and accumulation of polysaccharide. Our study observed a significant decline in DY and PS in P. heterophylla co-infected with TuMV and BBWV2 compared to those infected solely with BBWV2. Additionally, both DY and PS significantly increased with Pn, which is consistent with previous studies^[39,40].

Plants in nature and the field are subject to biotic and abiotic stresses. In response to these stresses, plants have developed signaling pathways that may share multiple metabolic pathways. The products of these metabolic pathways may overlap significantly, thereby reducing metabolic costs and enabling plants to survive in complex environments^[41]. In some cases, abiotic and biotic stresses can have synergistic effects. Previous studies have shown that plant immune responses triggered by viral infections increase plant tolerance to a variety of abiotic stresses such as drought and heat stress^[42,43]. Under drought stress, an increase in the abscisic acid induces stomatal closure. However, the accumulation of salicylic acid and the decrease in abscisic acid after infection with the virus can adjust the drought response strategy from stomatal regulation to one based on the synthesis of osmotic regulators (e.g., soluble polysaccharide)^[44,45]. In our study, co-infection diminished the sensitivity of stomatal conductance, net photosynthetic rate and transpiration rate to VPD_L considerable interaction between the effects of VPD_L and viral infection on Tr. The sensitivity of Pn, Gs, and Tr to VPD_L was reduced in TuMV and BBWV2 co-infected P. heterophylla compared to BBWV2 infected plants and the sensitivity of Pn to VPD_L was reduced in BBWV2 infected plants compared to health plants. These results suggest that viral infections do affect drought tolerance in P. heterophylla and are related to the species of virus and the synergistic or antagonistic relationship between different viruses under co-infection.

Under drought stress, plants have been observed to accumulate a variety of secondary metabolites, including soluble polysaccharide with osmoregulatory effects^[46], alkaloids, and phenols with antioxidant effects^[47]. These secondary metabolites are not only the active ingredients of medicinal plants but play a crucial role in the quality of herbs^[48]. In the future, it is crucial to analyze the changes in the content of secondary metabolites and the transcriptional regulation of these metabolites in P. heterophylla infected with different viruses. This analysis will provide a comprehensive and in-depth understanding of how the interactions among viruses, the abiotic environment, and P. heterophylla can affect the drought tolerance and quality of P. heterophylla.

In this study, a micro-stem tip tissue culture was taken to obtain TuMV-alone infected plants, BBWV2-alone infected plants, and virus-uninfected plants from TuMV and BBWV2 co-infected plants. However, all plants in the present study were TuMV and BBWV2 co-infected, BBWV2 infected or uninfected. The results suggest that co-infection of TuMV and BBWV2 may have a more deleterious impact on the growth of P. heterophylla than BBWV2 infection alone. Based on our findings, we suggest that: (1) Tuberous roots from plants without virus are preferred for the production of P. heterophylla; (2) Plants developed from tubers that have been used for cultivation for several years require tissue culture techniques to remove the virus; (3) Mark plants if they exhibit symptoms of virus disease, to prevent the use of tuberous roots from virus-infected plants for P. heterophylla production. Nevertheless, the absence of plants infected only with TuMV precludes the observation of the effects of TuMV infection alone or the confirmation of potential synergistic effects between the two viruses. Further investigation is required to determine whether the two viruses interact in affecting plant growth. In addition, the change of sensitivity of gas exchange parameters to leaf-air vapor pressure deficit is required for further investigation through transcriptome and secondary metabolite analysis to understand the interaction between P. heterophylla and viruses under drought stress.

Contents

Policies

Services

Partnerships

Dual-virus co-infection reduces photosynthetic rate, yield, and sensitivity of photosynthetic rate to leaf-air VPD in Pseudostellaria heterophylla

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

Dual-virus co-infection reduces photosynthetic rate, yield, and sensitivity of photosynthetic rate to leaf-air VPD in Pseudostellaria heterophylla

HTML

Site description

Experimental design

Measurements of specific leaf weight and chlorophyll content

Measurement of leaf photosynthesis

Measurements of yield, aqueous extract, and polysaccharides

Statistical analysis

Effect of different virus infections on morphological characteristics of leaves

Effect of different virus infections on leaf traits

Effect of different virus infections on yield and extract content of tuberous root

Relationships among leaf traits, tuberous root yield, and extract content

Effects of virus infection and leaf-air vapor pressure deficit on leaf physiological indicators

Major factors affecting net photosynthetic rate, yield, aqueous extract content, and polysaccharide content

Catalog

HTML

Site description

Experimental design

Measurements of specific leaf weight and chlorophyll content

Measurement of leaf photosynthesis

Measurements of yield, aqueous extract, and polysaccharides

Statistical analysis

Effect of different virus infections on morphological characteristics of leaves

Effect of different virus infections on leaf traits

Effect of different virus infections on yield and extract content of tuberous root

Relationships among leaf traits, tuberous root yield, and extract content

Effects of virus infection and leaf-air vapor pressure deficit on leaf physiological indicators

Major factors affecting net photosynthetic rate, yield, aqueous extract content, and polysaccharide content

Contents

Policies

Services

Partnerships

Export File

Citation

Format

Content