plant growth experiment controlled variables

Sciencing_Icons_Science SCIENCE

Sciencing_icons_biology biology, sciencing_icons_cells cells, sciencing_icons_molecular molecular, sciencing_icons_microorganisms microorganisms, sciencing_icons_genetics genetics, sciencing_icons_human body human body, sciencing_icons_ecology ecology, sciencing_icons_chemistry chemistry, sciencing_icons_atomic & molecular structure atomic & molecular structure, sciencing_icons_bonds bonds, sciencing_icons_reactions reactions, sciencing_icons_stoichiometry stoichiometry, sciencing_icons_solutions solutions, sciencing_icons_acids & bases acids & bases, sciencing_icons_thermodynamics thermodynamics, sciencing_icons_organic chemistry organic chemistry, sciencing_icons_physics physics, sciencing_icons_fundamentals-physics fundamentals, sciencing_icons_electronics electronics, sciencing_icons_waves waves, sciencing_icons_energy energy, sciencing_icons_fluid fluid, sciencing_icons_astronomy astronomy, sciencing_icons_geology geology, sciencing_icons_fundamentals-geology fundamentals, sciencing_icons_minerals & rocks minerals & rocks, sciencing_icons_earth scructure earth structure, sciencing_icons_fossils fossils, sciencing_icons_natural disasters natural disasters, sciencing_icons_nature nature, sciencing_icons_ecosystems ecosystems, sciencing_icons_environment environment, sciencing_icons_insects insects, sciencing_icons_plants & mushrooms plants & mushrooms, sciencing_icons_animals animals, sciencing_icons_math math, sciencing_icons_arithmetic arithmetic, sciencing_icons_addition & subtraction addition & subtraction, sciencing_icons_multiplication & division multiplication & division, sciencing_icons_decimals decimals, sciencing_icons_fractions fractions, sciencing_icons_conversions conversions, sciencing_icons_algebra algebra, sciencing_icons_working with units working with units, sciencing_icons_equations & expressions equations & expressions, sciencing_icons_ratios & proportions ratios & proportions, sciencing_icons_inequalities inequalities, sciencing_icons_exponents & logarithms exponents & logarithms, sciencing_icons_factorization factorization, sciencing_icons_functions functions, sciencing_icons_linear equations linear equations, sciencing_icons_graphs graphs, sciencing_icons_quadratics quadratics, sciencing_icons_polynomials polynomials, sciencing_icons_geometry geometry, sciencing_icons_fundamentals-geometry fundamentals, sciencing_icons_cartesian cartesian, sciencing_icons_circles circles, sciencing_icons_solids solids, sciencing_icons_trigonometry trigonometry, sciencing_icons_probability-statistics probability & statistics, sciencing_icons_mean-median-mode mean/median/mode, sciencing_icons_independent-dependent variables independent/dependent variables, sciencing_icons_deviation deviation, sciencing_icons_correlation correlation, sciencing_icons_sampling sampling, sciencing_icons_distributions distributions, sciencing_icons_probability probability, sciencing_icons_calculus calculus, sciencing_icons_differentiation-integration differentiation/integration, sciencing_icons_application application, sciencing_icons_projects projects, sciencing_icons_news news.

• Share Tweet Email Print
• Home ⋅
• Science ⋅
• Physics ⋅
• Fundamentals

What Are Dependent, Independent & Controlled Variables?

What Is a Responding Variable in Science Projects?

Say you're in lab, and your teacher asks you to design an experiment. The experiment must test how plants grow in response to different colored light. How would you begin? What are you changing? What are you keeping the same? What are you measuring?

These parameters of what you would change and what you would keep the same are called variables. Take a look at how all of these parameters in an experiment are defined, as independent, dependent and controlled variables.

What Is a Variable?

A variable is any quantity that you are able to measure in some way. This could be temperature, height, age, etc. Basically, a variable is anything that contributes to the outcome or result of your experiment in any way.

In an experiment there are multiple kinds of variables: independent, dependent and controlled variables.

What Is an Independent Variable?

An independent variable is the variable the experimenter controls. Basically, it is the component you choose to change in an experiment. This variable is not dependent on any other variables.

For example, in the plant growth experiment, the independent variable is the light color. The light color is not affected by anything. You will choose different light colors like green, red, yellow, etc. You are not measuring the light.

What Is a Dependent Variable?

A dependent variable is the measurement that changes in response to what you changed in the experiment. This variable is dependent on other variables; hence the name! For example, in the plant growth experiment, the dependent variable would be plant growth.

You could measure this by measuring how much the plant grows every two days. You could also measure it by measuring the rate of photosynthesis. Either of these measurements are dependent upon the kind of light you give the plant.

What Are Controlled Variables?

A control variable in science is any other parameter affecting your experiment that you try to keep the same across all conditions.

For example, one control variable in the plant growth experiment could be temperature. You would not want to have one plant growing in green light with a temperature of 20°C while another plant grows in red light with a temperature of 27°C.

You want to measure only the effect of light, not temperature. For this reason you would want to keep the temperature the same across all of your plants. In other words, you would want to control the temperature.

Another example is the amount of water you give the plant. If one plant receives twice the amount of water as another plant, there would be no way for you to know that the reason those plants grew the way they did is due only to the light color their received.

The observed effect could also be due in part to the amount of water they got. A control variable in science experiments is what allows you to compare other things that may be contributing to a result because you have kept other important things the same across all of your subjects.

Graphing Your Experiment

When graphing the results of your experiment, it is important to remember which variable goes on which axis.

The independent variable is graphed on the x-axis . The dependent variable , which changes in response to the independent variable, is graphed on the y-axis . Controlled variables are usually not graphed because they should not change. They could, however, be graphed as a verification that other conditions are not changing.

For example, after graphing the growth as compared to light, you could also look at how the temperature varied across different conditions. If you notice that it did vary quite a bit, you may need to go back and look at your experimental setup: How could you improve the experiment so that all plants are exposed to as similar an environment as possible (aside from the light color)?

How to Remember Which is Which

In order to try and remember which is the dependent variable and which is the independent variable, try putting them into a sentence which uses "causes a change in."

Here's an example. Saying, "light color causes a change in plant growth," is possible. This shows us that the independent variable affects the dependent variable. The inverse, however, is not true. "Plant growth causes a change in light color," is not possible. This way you know which is the independent variable and which is the dependent variable!

Related Articles

What are constants & controls of a science project..., what are independent & dependent variables in science..., difference between manipulative & responding variable, how to collect data from a science project, science fair projects on plants: do they grow faster..., how to write a testable hypothesis, how to grow a plant from a bean as a science project, what is the role of carotenoids in photosynthesis, proper way to label a graph, two week science projects, why should you only test for one variable at a time..., definitions of control, constant, independent and dependent..., what is a constant in a science fair project, science projects on which fertilizer makes a plant..., venus flytrap science projects, the effect of temperature on the rate of photosynthesis, phototropism experiments, cool science project ideas for k-4th grade, measuring wet bulb temperature.

• NCES Kids: What are Independent and Dependent Variables?
• Khan Academy: Dependent and independent variables review (article)

About the Author

Riti Gupta holds a Honors Bachelors degree in Biochemistry from the University of Oregon and a PhD in biology from Johns Hopkins University. She has an interest in astrobiology and manned spaceflight. She has over 10 years of biology research experience in academia. She currently teaches classes in biochemistry, biology, biophysics, astrobiology, as well as high school AP Biology and Chemistry test prep.

Find Your Next Great Science Fair Project! GO

• Science Notes Posts
• Contact Science Notes
• Todd Helmenstine Biography
• Anne Helmenstine Biography
• Free Printable Periodic Tables (PDF and PNG)
• Periodic Table Wallpapers
• Interactive Periodic Table
• Periodic Table Posters
• Science Experiments for Kids
• How to Grow Crystals
• Chemistry Projects
• Fire and Flames Projects
• Holiday Science
• Chemistry Problems With Answers
• Physics Problems
• Unit Conversion Example Problems
• Chemistry Worksheets
• Biology Worksheets
• Periodic Table Worksheets
• Physical Science Worksheets
• Science Lab Worksheets
• My Amazon Books

What Is a Control Variable? Definition and Examples

A control variable is any factor that is controlled or held constant during an experiment . For this reason, it’s also known as a controlled variable or a constant variable. A single experiment may contain many control variables . Unlike the independent and dependent variables , control variables aren’t a part of the experiment, but they are important because they could affect the outcome. Take a look at the difference between a control variable and control group and see examples of control variables.

Importance of Control Variables

Remember, the independent variable is the one you change, the dependent variable is the one you measure in response to this change, and the control variables are any other factors you control or hold constant so that they can’t influence the experiment. Control variables are important because:

• They make it easier to reproduce the experiment.
• The increase confidence in the outcome of the experiment.

For example, if you conducted an experiment examining the effect of the color of light on plant growth, but you didn’t control temperature, it might affect the outcome. One light source might be hotter than the other, affecting plant growth. This could lead you to incorrectly accept or reject your hypothesis. As another example, say you did control the temperature. If you did not report this temperature in your “methods” section, another researcher might have trouble reproducing your results. What if you conducted your experiment at 15 °C. Would you expect the same results at 5 °C or 35 5 °C? Sometimes the potential effect of a control variable can lead to a new experiment!

Sometimes you think you have controlled everything except the independent variable, but still get strange results. This could be due to what is called a “ confounding variable .” Examples of confounding variables could be humidity, magnetism, and vibration. Sometimes you can identify a confounding variable and turn it into a control variable. Other times, confounding variables cannot be detected or controlled.

Control Variable vs Control Group

A control group is different from a control variable. You expose a control group to all the same conditions as the experimental group, except you change the independent variable in the experimental group. Both the control group and experimental group should have the same control variables.

Control Variable Examples

Anything you can measure or control that is not the independent variable or dependent variable has potential to be a control variable. Examples of common control variables include:

• Duration of the experiment
• Size and composition of containers
• Temperature
• Sample volume
• Experimental technique
• Chemical purity or manufacturer
• Species (in biological experiments)

For example, consider an experiment testing whether a certain supplement affects cattle weight gain. The independent variable is the supplement, while the dependent variable is cattle weight. A typical control group would consist of cattle not given the supplement, while the cattle in the experimental group would receive the supplement. Examples of control variables in this experiment could include the age of the cattle, their breed, whether they are male or female, the amount of supplement, the way the supplement is administered, how often the supplement is administered, the type of feed given to the cattle, the temperature, the water supply, the time of year, and the method used to record weight. There may be other control variables, too. Sometimes you can’t actually control a control variable, but conditions should be the same for both the control and experimental groups. For example, if the cattle are free-range, weather might change from day to day, but both groups have the same experience. When you take data, be sure to record control variables along with the independent and dependent variable.

• Box, George E.P.; Hunter, William G.; Hunter, J. Stuart (1978). Statistics for Experimenters : An Introduction to Design, Data Analysis, and Model Building . New York: Wiley. ISBN 978-0-471-09315-2.
• Giri, Narayan C.; Das, M. N. (1979). Design and Analysis of Experiments . New York, N.Y: Wiley. ISBN 9780852269145.
• Stigler, Stephen M. (November 1992). “A Historical View of Statistical Concepts in Psychology and Educational Research”. American Journal of Education . 101 (1): 60–70. doi: 10.1086/444032

Related Posts

• Privacy Policy

Home » Control Variable – Definition, Types and Examples

Control Variable – Definition, Types and Examples

Table of Contents

Control Variable

Definition :

Control variable, also known as a “constant variable,” is a variable that is held constant or fixed during an experiment or study to prevent it from affecting the outcome. In other words, a control variable is a variable that is kept the same or held constant to isolate the effects of the independent variable on the dependent variable.

For example, if you were conducting an experiment to test how temperature affects plant growth, you might want to control variables such as the amount of water, the amount of sunlight, and the type of soil to ensure that these variables do not interfere with the results. By controlling these variables, you can isolate the effect of temperature on plant growth and draw more accurate conclusions from the experiment.

Types of Control Variables

Types of Control Variables are as follows:

Environmental Control Variables

These are variables related to the physical environment in which the experiment is conducted, such as temperature, humidity, light, and sound.

Participant Control Variables

These are variables related to the participants in the experiment, such as age, gender, prior knowledge, or experience.

Experimental Control Variables

These are variables that the researcher manipulates or controls to ensure that they do not affect the outcome of the experiment. For example, in a study on the effects of a new medication, the researcher might control the dosage, frequency, or duration of the treatment.

Procedural Control Variables

These are variables related to the procedures or methods used in the experiment, such as the order in which tasks are completed, the timing of measurements, or the instructions given to participants.

Equipment Control Variables

These are variables related to the equipment or instruments used in the experiment, such as calibration, maintenance, or proper functioning.

How to Control a Variable

To control a variable in a scientific experiment, you need to ensure that it is kept constant or unchanged throughout the experiment. Here are some steps to help you control a variable:

Identify the Variable

Start by identifying the variable that you want to control. This can be an environmental, subject, procedural, or instrumentation variable.

Determine the Level of Control Needed

Depending on the variable, you may need to exert varying levels of control. For example, environmental variables may require you to control the temperature, humidity, and lighting in your experiment, while subject variables may require you to select a specific group of participants that meet certain criteria.

Establish a Standard Level

Determine the standard level or value of the variable that you want to control. For example, if you are controlling the temperature, you may set the temperature to a specific degree and ensure that it is maintained at that level throughout the experiment.

Monitor the Variable

Throughout the experiment, monitor the variable to ensure that it remains constant. Use appropriate equipment or instruments to measure the variable and make adjustments as necessary to maintain the desired level.

Document the Process

Document the process of controlling the variable to ensure that the experiment is replicable. This includes documenting the standard level, monitoring procedures, and any adjustments made during the experiment.

Examples of Control Variables

Here are some examples of control variables in Scientific Experiments and Research:

• Environmental Control Variables Example: Suppose you are conducting an experiment to study the effect of light on plant growth. You would want to control environmental factors such as temperature, humidity, and soil nutrients. In this case, you might keep the temperature and humidity constant and use the same type and amount of soil for all the plants.
• Subject Control Variables Example : If you are conducting an experiment to study the effect of a new medication on blood pressure, you would want to control subject variables such as age, gender, and health status. In this case, you might select a group of participants with similar ages, genders, and health conditions to ensure that these variables do not affect the results.
• Procedural Control Variables Example : Suppose you are conducting an experiment to study the effect of distraction on reaction time. You would want to control procedural variables such as the time of day, the order of the tasks, and the instructions given to the participants. In this case, you might ensure that all participants perform the tasks in the same order, at the same time of day, and receive the same instructions.
• Instrumentation Control Variables Example : If you are conducting an experiment to study the effect of a new measurement device on the accuracy of readings, you would want to control instrumentation variables such as the type and calibration of the device. In this case, you might use the same type and model of the device and ensure that it is calibrated before each use.

Applications of Control Variable

Control variables are widely used in scientific research across various fields, including physics, biology, psychology, and engineering. Here are some applications of control variables:

• In medical research , control variables are used to ensure that any observed effects of a new treatment or medication are due to the treatment and not some other variable. By controlling subject variables such as age, gender, and health status, researchers can isolate the effects of the treatment and determine its effectiveness.
• In environmental research , control variables are used to study the effects of changes in the environment on various species or ecosystems. By controlling environmental variables such as temperature, humidity, and lighting, researchers can determine how different species adapt to changes in the environment.
• In psychology research, control variables are used to study the effects of different interventions or therapies on cognitive or behavioral outcomes. By controlling procedural variables such as the order of tasks, the length of time allotted for each task, and the instructions given to participants, researchers can isolate the effects of the intervention and determine its effectiveness.
• In engineering research, control variables are used to study the effects of different design parameters on the performance of a system or device. By controlling instrumentation variables such as the type of measurement device used and the calibration of instruments, researchers can ensure that the measurements are accurate and reliable.

Purpose of Control Variable

The purpose of a control variable in an experiment is to ensure that any observed changes or effects are a result of the manipulation of the independent variable and not some other variable. By keeping certain variables constant, researchers can isolate the effects of the independent variable and determine whether it has a significant effect on the dependent variable.

Control variables are important because they help to increase the reliability and validity of the experiment. Reliability refers to the consistency and reproducibility of the results, while validity refers to the accuracy and truthfulness of the results. By controlling variables, researchers can reduce the potential for extraneous or confounding variables that can affect the outcome of the experiment and increase the likelihood that the results accurately reflect the effect of the independent variable on the dependent variable.

Characteristics of Control Variable

Control variables have the following characteristics:

• Constant : Control variables are kept constant or unchanged throughout the experiment. This means that their values do not vary or change during the experiment. Keeping control variables constant helps to ensure that any observed effects or changes are due to the manipulation of the independent variable and not some other variable.
• Independent : Control variables are independent of the independent variable being studied. This means that they do not affect the relationship between the independent and dependent variables. By controlling for independent variables, researchers can isolate the effect of the independent variable and determine its impact on the dependent variable.
• Documented: Control variables are documented in the experiment. This means that their values and methods of control are recorded and reported in the results section of the research paper. By documenting control variables, researchers can demonstrate the rigor and transparency of their study and allow other researchers to replicate their methods.
• Relevant: Control variables are relevant to the research question. This means that they are chosen based on their potential to affect the outcome of the experiment. By selecting relevant control variables, researchers can reduce the potential for extraneous or confounding variables that can affect the outcome of the experiment and increase the reliability and validity of the results.
• Varied : Control variables can be varied across different conditions or groups. This means that different levels of control may be needed depending on the research question or hypothesis being tested. By varying control variables, researchers can test different hypotheses and determine the factors that affect the outcome of the experiment.

Advantages of Control Variable

The advantages of using control variables in an experiment are:

• Increased accuracy : Control variables help to increase the accuracy of the results by reducing the potential for extraneous or confounding variables that can affect the outcome of the experiment. By controlling for these variables, researchers can isolate the effect of the independent variable on the dependent variable and determine whether it has a significant impact.
• Increased reliability : Control variables help to increase the reliability of the results by reducing the variability in the experiment. By keeping certain variables constant, researchers can ensure that any observed changes or effects are due to the manipulation of the independent variable and not some other variable.
• Reproducibility: Control variables help to increase the reproducibility of the results by ensuring that the same results can be obtained when the experiment is repeated. By documenting and reporting control variables, researchers can demonstrate the rigor and transparency of their study and allow other researchers to replicate their methods.
• Generalizability : Control variables help to increase the generalizability of the results by reducing the potential for bias and increasing the external validity of the experiment. By controlling for relevant variables, researchers can ensure that their findings are applicable to a broader population or context.
• Causality : Control variables help to establish causality by ensuring that any observed changes or effects are due to the manipulation of the independent variable and not some other variable. By controlling for confounding variables, researchers can increase the internal validity of the experiment and establish a cause-and-effect relationship between the independent and dependent variables.

Disadvantages of Control Variable

There are some potential disadvantages or limitations of using control variables in an experiment:

• Complexity : Controlling for multiple variables can make an experiment more complex and time-consuming. This can increase the likelihood of errors and reduce the feasibility of the experiment, especially if the control variables require a lot of resources or are difficult to measure.
• Artificiality : Controlling for variables can make the experimental conditions artificial and not reflective of real-world situations. This can reduce the external validity of the experiment and limit the generalizability of the findings to real-world settings.
• Limited scope : Controlling for specific variables can limit the scope of the experiment and make it difficult to generalize the results to other situations or populations. This can reduce the external validity of the experiment and limit its practical applications.
• Assumptions: Controlling for variables requires making assumptions about which variables are relevant and how they should be controlled. These assumptions may not be valid or accurate, and the results of the experiment may be affected by uncontrolled variables that were not considered.
• Cost : Controlling for variables can be costly, especially if the control variables require additional resources or equipment. This can limit the feasibility of the experiment, especially for researchers with limited funding or resources.

Limitations of Control Variable

There are several limitations of using control variables in an experiment, including:

• Not all variables can be controlled : There may be some variables that cannot be controlled or manipulated in an experiment. For example, some variables may be too difficult or expensive to measure or control, or they may be affected by factors outside of the researcher’s control.
• Interaction effects : Control variables can interact with each other, which can lead to unexpected results. For example, controlling for one variable may have a different effect when another variable is also controlled, or when the two variables interact with each other. These interaction effects can be difficult to predict or control for.
• Over-reliance on statistical significance: Controlling for variables can increase the statistical significance of the results, but this may not always translate to practical significance or real-world significance. Researchers should interpret the results of an experiment in light of the practical significance, not just the statistical significance.
• Limited generalizability : Controlling for variables can limit the generalizability of the results to other populations or situations. If the control variables are not representative of other populations or situations, the results of the experiment may not be applicable to those contexts.
• May mask important effects : Controlling for variables can mask important effects that are related to the independent variable. By controlling for certain variables, researchers may miss important interactions between the independent variable and the controlled variable, which can limit the understanding of the causal relationship between the two.

About the author

Muhammad Hassan

Researcher, Academic Writer, Web developer

You may also like

Qualitative Variable – Types and Examples

Continuous Variable – Definition, Types and...

Polytomous Variable – Definition, Purpose and...

Confounding Variable – Definition, Method and...

Dependent Variable – Definition, Types and...

Discrete Variable – Definition, Types and...

• Skip to secondary menu
• Skip to main content
• Skip to primary sidebar

Statistics By Jim

Making statistics intuitive

Control Variables: Definition, Uses & Examples

By Jim Frost 4 Comments

What is a Control Variable?

Control variables, also known as controlled variables, are properties that researchers hold constant for all observations in an experiment. While these variables are not the primary focus of the research, keeping their values consistent helps the study establish the true relationships between the independent and dependent variables. The capacity to control variables directly is highest in experiments that researchers conduct under lab conditions. In observational studies, researchers can’t directly control the variables. Instead, they record the values of control variables and then use statistical procedures to account for them.

In science, researchers assess the effects that the independent variables have on the dependent variable. However, other variables can also affect the outcome. If the scientists do not control these other variables, they can distort the primary results of interest. In other words, left uncontrolled, those other factors become confounders that can bias the findings. The uncontrolled variables may be responsible for the changes in the outcomes rather than your treatment or experimental variables. Consequently, researchers control the values of these other variables.

Suppose you are performing an experiment involving different types of fertilizers and plant growth. Those are your primary variables of interest. However, you also know that soil moisture, sunlight, and temperature affect plant growth. If you don’t hold these variables constant for all observations, they might explain the plant growth differences you observe. Consequently, moisture, sunlight, and temperature are essential control variables for your study.

If you perform the study in a controlled lab setting, you can control these environmental conditions and keep their values constant for all observations in your experiment. That way, if you do see plant growth differences, you can be more confident that the fertilizers caused them.

When researchers use control variables, they should identify them, record their values, and include the details in their write-up. This process helps other researchers understand and replicate the results.

Related posts : Independent and Dependent Variables and Confounding Variables

Control Variables and Internal Validity

By controlling variables, you increase the internal validity of your research. Internal validity is the degree of confidence that a causal relationship exists between the treatment and the difference in outcomes. In other words, how likely is it that your treatment caused the differences you observe? Are the researcher’s conclusions correct? Or, can changes in the outcome be attributed to other causes?

If the relevant variables are not controlled, you might need to attribute the changes to confounders rather than the treatment. Control variables reduce the impact of confounding variables.

Controlled Variable Examples

How to Control Variables in Science

Scientists can control variables using several methods. In some cases, variables can be controlled directly. For example, researchers can control the growing conditions for the fertilizer experiment. Or use standardized procedures and processes for all subjects to reduce other sources of variation. These efforts attempt to eliminate all differences between the treatment and control groups other than the treatments themselves.

However, sometimes that’s not possible. Fortunately, there are other approaches.

Random assignment

In some experiments, there can be too many variables to control. Additionally, the researchers might not even know all the potential confounding variables. In these cases, they can randomly assign subjects to the experimental groups. This process controls variables by averaging out all traits across the experimental groups, making them roughly equivalent when the experiment begins. The randomness helps prevent any systematic differences between the experimental groups. Learn more in my post about Random Assignment in Experiments .

Statistical control

Directly controlled variables and random assignment are methods that equalize the experimental groups. However, they aren’t always feasible. In some cases, there are too many variables to control. In other situations, random assignment might not be possible. Try randomly assigning people to smoking and non-smoking groups!

Fortunately, statistical techniques, such as multiple regression analysis , don’t balance the groups but instead use a model that statistically controls the variables. The model accounts for confounding variables.

In multiple regression analysis, including a variable in the model holds it constant while the treatment variable fluctuates. This process allows you to isolate the role of the treatment while accounting for confounders. You can also use ANOVA and ANCOVA.

For more information, read my posts, When to Use Regression and ANOVA Overview .

Share this:

Reader Interactions

July 13, 2024 at 2:19 am

Sir you are doing a good job. much appreciated. Could you please tell us how to read the values of control variables like ranges and what do they mean. For instance how to read this (F=1.83; p= 0.07). Thank YOU

February 28, 2024 at 2:09 pm

In your explanation of control variables you use the example of a research study of plant fertilizers and their growth, wanting to control for moisture, sunshine and temperature. You state “Consequently, moisture, sunlight, and temperature are essential control variables for your study. These variables can be controlled by keeping their values constant for all observations in your experiment. You do not go further as to how you control for these values, particularly when such variables are continually changing. Al Wassler

February 28, 2024 at 2:13 pm

Presumably, this experiment would occur in a lab setting where you can control these variables. Plants would be raised with the same humidity, soil moisture, and light conditions.

I’ll add some text to the article to clarify that. Thanks!

January 26, 2023 at 7:00 pm

I have a question please about when a control variable is also itself part of the dependent variable. I see this referred to in the medical research literature as ‘mathematical coupling’, where – for example – the beats per minute (BPM) is the dependent variable and researchers want to put minutes also as a control variable. This seems to create a problem because ‘minutes’ appears on both sides of the equation, and the medical literature talks about spurious correlation, and the model needing to be redesigned. But do you have a simple text or reference – ideally just plain statistics/OLS rather than linked to medical research – where this could be explained in theory terms ? What goes wrong in the regression when a variable is both a control variable and part of the dependent variable (perhaps as part of a ratio or measurement of change)? I just haven’t found a textbook reference that says definitively ‘you can’t have the same variable in both sides of the regression simultaneously’, so I’m not sure whether this violates OLS and so is something to avoid entirely (with a new model design or different research question) or to live with.

Any help would be great, thank you for your work,

Comments and Questions Cancel reply

Plant Breeding and Genomics

Introduction to Experimental Design

Shawn C. Yarnes, The Ohio State University

Defining Variables and Experimental Units

Experimental design begins with the formulation of experimental questions , which help define the variables that will change in an experiment. Experimental treatments , or independent variables , are the controlled part of an experiment expected to affect the response , or dependent variables . The experimenter must identify which treatment and response variables will best answer experimental questions.

Consider the broad experimental question. How do plants respond to fertilizer application? This question must be made more specific to design an effective experiment. 

The dependent variable , plant response , can be defined and measured in numerous ways. If the experimenter is interested in plant growth and nitrogen content, the question can be made more specific by asking how does plant growth and nitrogen content change in response to fertilizer application? Determination of response variables is influenced by experimental objectives and practical considerations. For example, total dry weight is more accurate than height as a measurement of plant growth, but in the case of a tree experiment, height might be more practical.

The independent variable , fertilizer treatment , can also be defined in numerous ways that will help specify experimental questions. A single fertilizer treatment with different levels can be tested, or multiple fertilizers compared. Levels can be: qualitative, or categorical, as when denoting males and females in a population; or quantitative, such as different fertilizer concentrations. Levels can also be defined as fixed or random effects. Sex distribution in a population is generally a random effect ; while fertilizer application is an experimenter controlled, or  fixed effect . The decision to define a variable as fixed or random will affect future statistical analyses (See  Analysis of Variance (ANOVA): Experimental Design for Fixed and Random Effects ). 

Once response and experimental treatments are defined, proper control treatments must be determined. Controls are integral to the scientific method by providing baseline values against which other treatments are compared. Negative controls , such as non-fertilized plants in Example 1, are null treatments where no response is expected. The simplest experiment has one response variable, one negative control, and one treatment. If experimental results support a null hypothesis (H 0 ), no significant difference is observed between controls and other treatments. 

Positive controls are treatments where a known response is expected. Positive controls are often used to validate assays or equipment functioning. For example, many enzyme kits come with pre-digested substrates, so that experimental digestions can be deemed successful compared to the positive control. Positive controls can also be used to calibrate or standardize measurements. For example, a standard curve of known substrate concentrations can be used to calculate the amount of unknown substrate concentrations.  

Experimental units must be defined during experimental design. The experimental unit is an individual, object, or plot subjected to treatment independently of other units. The number of experimental units is the sum of all treatments, levels, and and replicates. When experimental units are sampled only once, the experimental unit and sampling unit  are the same. The experimental unit can also be comprised of multiple sampling units. When experimental units are heterogeneous for the response variable, the mean of multiple sample units can be more precise than a single measure.  For example, if leaf nitrogen content is variable between leaves, an experimenter may choose to measure the nitrogen content from multiple leaves, using the mean nitrogen content to represent the individual plant.  Increasing the sampling units does not increase replication.

Planning for Statistical Inference

The goal of an experiment is to detect differences between treatments. Statistical determination of these differences requires replication to compute experimental error and randomization to help ensure that the measure of experimental error is valid. Discussions of experimental error and replication become circular, because replications are needed to compute experimental error, and the number of replications needed is based on the magnitude of experimental error. Experimental design requires an a priori estimation of error. In some situations a preliminary study is used to estimate error. In other situations error is inferred using reasonable assumptions based on the current understanding of the study system.

Experimental Error

Experimental error is the variation among experimental units within the same treatment group. There are many possible reasons for error. Errors within an experiment are additive. Reducing the amount of error in an experiment increases your ability to detect significant differences between treatments. A well-designed experiment considers the error contributed by both natural variation and lack of experimental uniformity.

Natural variation is a large component of error in biological experiments. Genetic and developmental differences, as well as differences in species abundance and diversity, can vary between experimental units. In plant breeding, clones and inbreed lines are often utilized to reduce genetic variation between experimental units. 

Lack of experimental uniformity is the source of error over which an investigator has the most control. Although there is always an imperfect ability to provide identical environments for each experimental unit, identifying and controlling error is essential. Errors in technique and/or data recording can inflate estimated experimental error (decrease precision) and introduce bias into the results (decrease accuracy).

Relationship Between Error and Sample Size

The sample size needed to detect differences between treatments increases with error. This is the reason biological field experiments generally require larger sample sizes than more controlled laboratory experiments. Experimental effort and expense are directly proportional to sample size. For these reasons controlling error is the focus of every investigator.  

The graph below illustrates the realtionship between error (σ), sample size, and the ability to detect differences between two means. (See  Estimating Sample Size for Comparison of Two Means and  Equation to Estimate Sample Size Required for QTL Detection ).

Funding Statement

Development of this page was supported in part by the National Institute of Food and Agriculture (NIFA) Solanaceae Coordinated Agricultural Project, agreement 2009-85606-05673, administered by Michigan State University. Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the United States Department of Agriculture.  

PBGworks 1445

What Is The Controlled Variable In A Plant Growth Experiment 

What Is The Controlled Variable In A Plant Growth Experiment

In the realm of scientific investigations, experiments are often conducted to explore various hypotheses and gain a deeper understanding of certain phenomena. One such experiment involves assessing the growth of plants under different conditions. Fascinatingly, these experiments heavily rely on the concept of controlled variables to ensure accurate and reliable results. By manipulating certain factors while keeping others constant , scientists can effectively isolate and study the impact of specific variables on the growth of plants.

When conducting a plant growth experiment , the controlled variable, also known as the constant or standard variable, is the factor that remains consistent throughout the entire experiment. This means that it is not manipulated or changed in any way. The purpose of establishing a controlled variable is to provide a baseline against which the effects of other variables can be measured. By keeping certain factors unaltered, scientists can accurately determine the impact of the independent variable(s) on the plant’s growth.

In an experiment assessing plant growth, there are various factors that can be controlled, such as temperature, light exposure, humidity, soil type, and watering frequency. These variables can significantly influence a plant’s growth and must be kept constant to ensure that any changes observed can be attributed solely to the variable being investigated. For instance, if the study aims to examine the impact of light intensity on plant growth, all other factors, such as temperature and humidity, should remain unchanged to eliminate their potential influence.

Maintaining a controlled variable in a plant growth experiment is of utmost importance as it helps avoid confounding variables, which are factors that could unintentionally affect the results. Without a controlled variable, it would be difficult to determine the true cause-and-effect relationship between the manipulated variables and the observed changes in plant growth. Ultimately, by having a constant standard throughout the experiment, researchers can draw more accurate conclusions and make informed decisions based on their findings.

In conclusion, the controlled variable in a plant growth experiment plays a crucial role in ensuring the reliability and validity of the results obtained. By keeping specific factors constant while manipulating others, scientists can isolate the effects of independent variables on plant growth. This approach allows for a more thorough understanding of the factors that influence plant development and provides a solid foundation for further scientific exploration in this field.

key Takeaways

• The controlled variable in a plant growth experiment is the factor that remains constant throughout the entire experiment.
• It is crucial to identify and control the controlled variable to ensure accurate and reliable results.
• The controlled variable acts as a baseline for comparison and helps in determining the impacts of other variables on plant growth.
• In plant growth experiments, commonly controlled variables include light intensity, temperature, watering schedule, and soil type.
• Controlled variables should be carefully selected based on their potential impact on plant growth and their ability to be controlled.
• Failure to control the variables properly may lead to inaccurate conclusions and flawed experiment results.
• Controlling the variables allows researchers to observe the true effects of the independent variable on plant growth.
• Maintaining consistent and controlled conditions is essential to minimize any confounding factors that could influence the experiment results.

What is the controlled variable in a plant growth experiment?

In a plant growth experiment, the controlled variable refers to the aspect of the experiment that remains constant throughout. It is the factor or condition that is kept the same in order to ensure accurate and reliable results. By controlling this variable, researchers can isolate the effects of the independent variable, which is the factor being manipulated, and observe its impact on the dependent variable, which is the result or outcome of the experiment.

Controlled Variable Definition:

The controlled variable, also known as the constant variable, is the parameter that is intentionally kept consistent in order to prevent it from influencing the experimental outcomes. It serves as a baseline reference against which the effects of other variables can be measured. By maintaining it at a fixed level or value, researchers can accurately assess the impact of the independent variable on the dependent variable.

Importance of Controlled Variables:

Controlled variables are crucial in plant growth experiments as they ensure that any changes observed in the dependent variable are solely due to the manipulation of the independent variable. By keeping all other factors constant, researchers can minimize the influence of confounding variables and increase the validity and reliability of their findings. This allows for accurate comparisons and clear attribution of cause and effect relationships.

Examples of Controlled Variables in Plant Growth Experiments:

There are several parameters that can be controlled in a plant growth experiment, depending on the specific research question or hypothesis. Some common examples include:

• Temperature: Keeping the temperature constant ensures that any differences in plant growth can be attributed to the independent variable and not variations in environmental conditions.
• Light Intensity: Controlling the amount of light received by the plants helps eliminate the impact of light as a variable and allows for accurate measurement of the independent variable’s effects.
• Moisture: Maintaining consistent moisture levels ensures that differences in plant growth are not caused by variations in water availability.
• Soil Composition: Controlling the composition of the soil, including nutrients and pH levels, helps isolate the effects of the independent variable on plant growth.

Conclusion:

The controlled variable plays a crucial role in plant growth experiments. By maintaining certain factors constant, researchers can ensure that any changes observed in the dependent variable are solely due to the manipulation of the independent variable. This increases the reliability and validity of the results, allowing for accurate conclusions to be drawn.

Frequently Asked Questions

What is a controlled variable in a plant growth experiment.

A controlled variable, also known as a constant variable, is a factor that researchers deliberately keep the same or constant throughout the experiment. In the case of a plant growth experiment, the controlled variable is an aspect of the environment, such as light intensity or temperature, that is kept consistent for all plants being tested. By controlling this variable, scientists can isolate and observe the effects of other factors, such as the type of fertilizer used or the amount of water given, on plant growth.

Why is it important to have a controlled variable in a plant growth experiment?

Holding a variable constant in a plant growth experiment is crucial because it ensures that any observed changes in plant growth are the result of the manipulated or independent variable, and not due to variations in the environment. Without a controlled variable, it would be challenging to determine whether any observed differences in plant growth were influenced by the independent variable or factors outside of the experiment’s control. The controlled variable acts as a baseline or reference point, allowing researchers to compare the effects of different treatments or conditions on plant growth accurately.

How do you choose the controlled variable in a plant growth experiment?

Choosing the controlled variable in a plant growth experiment involves identifying the environmental factor you want to keep constant throughout the study. This depends on the specific question you are investigating and the variables you plan to manipulate. For example, if you are examining the effect of different types of soils on plant growth, you may choose to control the amount of water given to each plant by ensuring they all receive the same quantity. You could also keep light intensity constant by placing all the plants in a room with uniform lighting conditions.

Can there be more than one controlled variable in a plant growth experiment?

Yes, it is possible to have multiple controlled variables in a plant growth experiment. Researchers often aim to maintain a controlled environment by controlling several factors that may influence plant growth. However, it is crucial to keep in mind that the more variables you attempt to control, the more challenging it becomes to isolate the effects of the independent variable. It is essential to strike a balance between controlling variables and allowing for variation to achieve accurate and meaningful results.

What happens if the controlled variable is not properly controlled in a plant growth experiment?

If the controlled variable is not adequately controlled in a plant growth experiment, it could lead to inaccurate or misleading results. Any observed changes in plant growth may be falsely attributed to the independent variable being tested, even if they are actually due to uncontrolled environmental factors. Carefully controlling the variable ensures that any effects on plant growth are solely due to the manipulated factors, allowing for valid conclusions to be drawn from the experiment.

Types and Options for Achieving a Popular Subject Matter

Hydroponics: the future of sustainable crop production.

In recent years, hydroponics has gained significant attention as a highly efficient and sustainable method of crop production. This innovative technique involves growing plants in a carefully controlled environment without soil. By providing plants with a nutrient-rich water solution directly to their roots, hydroponics allows for maximum nutrient absorption, resulting in faster growth rates and higher crop yields. Water usage is also significantly reduced compared to traditional soil-based farming methods, making hydroponics an environmentally friendly and resource-efficient option for achieving sustainable crop production.

The Impact of Organic Farming on Agricultural Practices

Organic farming has emerged as a popular alternative to conventional agricultural practices in recent years. By eschewing synthetic fertilizers and pesticides, organic farming aims to minimize harm to the environment, promote biodiversity, and produce food that is free from chemicals. The use of organic fertilizers, such as compost and manure, not only provides plants with essential nutrients but also improves soil health and quality. Organic farming also emphasizes sustainable practices, such as crop rotation and the use of cover crops, which help reduce soil erosion, retain moisture, and prevent weeds.

Final Thoughts

In summary, a controlled variable in a plant growth experiment is a constant factor deliberately kept the same throughout the study. It is important to have a controlled variable to isolate the effects of the independent variable and ensure accurate conclusions. The choice of the controlled variable depends on the specific research question and variables being manipulated. It is possible to have multiple controlled variables, but balancing control and variation is crucial for meaningful results.

Additionally, exploring alternative methods like hydroponics and organic farming can offer sustainable solutions for crop production while minimizing environmental impact. Hydroponics allows for optimal nutrient absorption and water usage efficiency, while organic farming promotes biodiversity, soil health, and chemical-free food production. These options demonstrate the potential for innovation and sustainable practices in agriculture.

The Role of a Controlled Variable in an Experiment

• Scientific Method
• Chemical Laws
• Periodic Table
• Projects & Experiments
• Biochemistry
• Physical Chemistry
• Medical Chemistry
• Chemistry In Everyday Life
• Famous Chemists
• Activities for Kids
• Abbreviations & Acronyms
• Weather & Climate
• Ph.D., Biomedical Sciences, University of Tennessee at Knoxville
• B.A., Physics and Mathematics, Hastings College

A controlled variable is one which the researcher holds constant (controls) during an experiment. It is also known as a constant variable or simply as a "control." The control variable is not part of an experiment itself—it is neither the independent nor dependent variable —but it is important because it can have an effect on the results. It is not the same as a control group.

Any given experiment has numerous control variables, and it's important for a scientist to try to hold all variables constant except for the independent variable. If a control variable changes during an experiment, it may invalidate the correlation between the dependent and independent variables. When possible, control variables should be identified, measured, and recorded.

Examples of Controlled Variables

Temperature is a common type of  controlled variable . If a temperature is held constant during an experiment, it is controlled.

Other examples of controlled variables could be an amount of light, using the same type of glassware, constant humidity , or duration of an experiment.

Importance of Controlled Variables

Although control variables may not be measured (though they are often recorded), they can have a significant effect on the outcome of an experiment. Lack of awareness of control variables can lead to faulty results or what are called "confounding variables." Additionally, noting control variables makes it easier to reproduce an experiment and establish the relationship between the independent and dependent variables .

For example, say you are trying to determine whether a particular fertilizer has an effect on plant growth. The independent variable is the presence or absence of the fertilizer, while the dependent variable is the height of the plant or rate of growth. If you don't control the amount of light (e.g., you perform part of the experiment in the summer and part during the winter), you may skew your results.

• Null Hypothesis Examples
• What Is a Controlled Experiment?
• Understanding Simple vs Controlled Experiments
• Six Steps of the Scientific Method
• Random Error vs. Systematic Error
• What Is the Difference Between a Control Variable and Control Group?
• Scientific Method Vocabulary Terms
• Scientific Variable
• What Is a Hypothesis? (Science)
• What Are Examples of a Hypothesis?
• DRY MIX Experiment Variables Acronym
• What Is an Experimental Constant?
• What Are the Elements of a Good Hypothesis?
• Scientific Method Flow Chart
• What Is a Testable Hypothesis?
• Scientific Hypothesis Examples

Practical Biology

A collection of experiments that demonstrate biological concepts and processes.

Observing earthworm locomotion

Practical Work for Learning

Published experiments

Investigating the effect of minerals on plant growth, class practical.

All of these techniques involve a long-term project – prepared in one lesson, left for about a month (see Note 2 ), then with results gathered in one or more lessons after that time. There is scope for focus on the scientific methods involved in planning, controlling variables, collecting and analysing data, as well as on the biology of plant nutrient requirements. The methods include several different dependent variables – percentage cover, harvested mass, dry mass, turbidity, population count with haemocytometer. Each method produces a qualitative outcome as well.

Lesson organisation

In the first lesson, present the biological problem – how to investigate the effects of different minerals on plant growth. Give each group of students a different option for following plant growth. Ask each group to plan in detail how they would set up an investigation. Evaluate the methods in terms of controlled variables, reliability, and ease of data collection. Decide which method to work with. If you can manage the practicalities of two different investigations, choose two.

In the next lesson, set up an investigation (or two). Get all students involved – for example, if using the radish method, each student could set up one pot of seeds, for a particular culture medium, all seeds could be grown together and results pooled.

In the final lesson, collect the results and collate for the group – showing how to calculate means and discussing reliability of results and validity of any conclusions drawn.

Apparatus and Chemicals

Mineral nutrient mixes ( Note 1 )

For each group of students:

Plant material to investigate and associated materials. Choose from A , B , C or D .

A Germinating barley

Healthy barley seedlings, approximately 6, germinated a week in advance ( Note 3 ) test tubes (1 per culture solution) cotton wool aluminium foil or black card/ polythene to surround test tubes dropping pipette

B Radish Seeds – 2 per container Growing medium – peat/ vermiculite mix ( Note 4 ) Small container (for example a film canister) with hole cut in bottom, 1 per set of seeds Wicks – a piece of capillary matting/ cloth cut into narrow diamond shape, 1 per container Capillary matting and water reservoirs – one per culture medium ( Note 5 )

C Algal culture Algal suspension – in full mineral salts medium ( Note 6 ) Conical flask, 1 per culture solution Cotton wool Syringe to dispense 1 cm 3 of algal suspension Disinfectant for syringe Measuring cylinder, 100 cm 3 Microscope Microscope slide Cover slip

D Lemna (duckweed)

Healthy Lemna plants of similar size, 10 per culture solution Beakers or jam jars, 1 per culture solution Plastic film to cover the beakers or jars

Health & Safety and Tehnical notes

Read our standard health & safety guidance

1 Solid media to prepare Long Ashton water culture, or Sach’s water culture solutions, are available from Timstar or Philip Harris (see Suppliers). It can be cheaper, and is certainly much easier, to buy the ready-prepared nutrient solutions if not all the chemicals are available in-house. But you could make up your own solutions using the recipe from the CLEAPSS Recipe card.

Sach’s culture solution (complete recipe): Dissolve the following salts in 1 litre of distilled water.

• 0.25 g of calcium sulfate(VI)-2-water
• 0.25 g of calcium phosphate(V)-2-water CaH 4 (PO 4 ) 2 .2H 2 O
• 0.25 g of magnesium sulfate(VI)-7-water
• 0.08 g of sodium chloride
• 0.70 g of potassium nitrate(V) (see CLEAPSS Hazcard – OXIDISING and DANGEROUS with some metals and flammable substances)
• 0.005 g of iron(III) chloride-6-water (see CLEAPSS Hazcard – HARMFUL as a solid)

For Sach’s culture solution with mineral deficiencies , make the following changes.

• Deficient in calcium: 0.2 g of potassium sulfate(VI) replaces calcium sulfate(VI)-2-water and 0.71 g of sodium dihydrogenphosphate(V)-2-water replaces calcium phosphate(V).
• Deficient in iron: Omit iron(III) chloride-6-water.
• Deficient in nitrogen : 0.52 g of potassium chloride replaces potassium nitrate(V).
• Deficient in phosphorus : 0.16 g of calcium nitrate(V)-4-water (see CLEAPSS Hazcard – OXIDISING and IRRITANT) replaces calcium phosphate(V).
• Deficient in sulphur : 0.16 g of calcium chloride (see CLEAPSS Hazcard – IRRITANT as solid) replaces calcium sulfate(VI) and 0.21 g of magnesium chloride-6-water replaces magnesium sulfate(VI).
• Deficient in magnesium : 0.17 g of potassium sulfate(VI) (Hazcard 98B – low hazard) replaces magnesium sulfate(VI).
• Deficient in potassium : 0.59 g of sodium nitrate(V) (Hazcard 82 – oxidising and harmful as solid and dangerous with some metals and flammable materials) replaces potassium nitrate(V).

2 Each system requires a different lead-in time, a different length of time for results to develop and a different method for measuring the effects.

3 If you germinate barley seeds on cotton wool or blotting paper, the roots may stick in the damp medium. Using OASIS or vermiculite avoids this – although it costs a little more. Refresh the mineral solution every couple of days by tipping out and replacing. Aerating the solution before applying to the roots may improve the general uptake of solution, and reduce the risk of the barley seedlings rotting.

4 The peat/ vermiculite mix must be low in nutrients – for example a seed compost, rather than multipurpose (which has added nutrients).

5 Water reservoirs and wicks: Set up a series of ice-cream containers containing each culture medium to be tested. Cut slots in the lids of the containers. Cut pieces of capillary matting as shown in diagram. Insert the capillary matting and pour enough culture medium into the ice-cream container to ensure that the matting remains moist at all times.

Place the wicks in the bottom of the small containers before filling (to within 5 mm of the top) with growing medium. Add 2 seeds to each container. Add 2-3 mm more growing medium and firm gently. Place the container on the capillary matting so that the wick can draw liquid mineral salts medium from the container.

6 Inoculate 500 cm 3 of complete medium with Scenedesmus quadricaudus or Micrasterias thomasiana var. notata or Chlorella – about one week before required. Aerate continuously using a filter pump, or aquarium airstone and pump and keep in a light place or illuminate 24 hours a day. Sciento and Blades Biological provide suitable algae to culture. Do not use algae cultured on agar slopes.

SAFETY: Follow good hygiene practice after handling pond water or plants removed from ponds.

Preparation:

If using barley seedlings, germinate about one week before use. If using algal suspension, start culturing alga about 4 weeks before use.

Method A and B:

a Set up the plants (barley in liquid culture solution, or radish watered with culture solution) and allow to grow for about 3 weeks (for radish with 24 hour illumination or for barley).

b After 3 weeks, make qualitative observations of plant growth in each medium.

c Collect sample plant material, remove any adhering growth medium (radish) or blot off any liquid (barley). Measure the mass of the living material.

d Place the material in an oven at 80 – 90 °C to dry. Measure the mass every day until 3 readings are constant.

e Record the dry mass of plant material in each culture medium.

f Observe the algal suspension by eye and make qualitative observations of which has grown best.

g Measure the turbidity of each sample using a colorimeter.

h Estimate the population of algae using a microscope and a small grid square, or a haemocytometer.

i Make qualitative observations of the growth of each sample.

j Estimate the area of cover in each beaker/ jar by placing a grid underneath and counting the number of squares covered.

Teaching notes

In summary, any mineral deficiency will result in poor plant growth. It may be difficult for inexperienced botanists/ horticulturists to appreciate the subtle differences between one kind of poor growth and the next. Overall productivity is a simple measure of growth. You could also measure the total height (or length) of a plant leaf or stem (radish/ barley), and note the colour, and the pattern of loss of colour. Several deficiencies result in death of leaf tissue – so you may also notice different patterns of damage to the leaves. It is worth identifying veins and leaf margins and noting any changes in those areas.

Calcium deficiency shows in soft, dead, necrotic tissue at rapidly growing areas – such as on fruits, the tips of leaves and the heart of crops such as celery. If the margins of the leaves grow more slowly, the leaf tends to cup downwards. Calcium deficiency also leaves plants with a greater tendency to wilt than non-stressed plants.

Iron deficiency shows in strong chlorosis at the base of leaves – leading to completely bleached leaves. Bleached areas may develop necrotic spots.

Nitrogen deficiency results in generally poor growth – short, spindly plants – and general chlorosis (lack of chlorophyll). Plants show more tendency to wilt under water stress and to die more quickly. Young leaves at the growing point may still be green but will be small. Other leaves may lack colour entirely. In some plants, the underside of the leaves, and petioles and midribs may develop a purple colour.

Phosphorus deficiency produces dwarfed or stunted plants – perhaps with some necrotic spots on the leaves. They grow more slowly than similar plants not lacking phosphorus.

Sulfur deficiency shows in an overall chlorosis with veins and petioles gaining a reddish colouration. This includes young leaves. Leaves may be twisted and brittle.

Magnesium is an essential part of the chlorophyll molecule. Plants deficient in magnesium frequently show interveinal chlorosis (a lack of chlorophyll).

Potassium deficiency shows first in marginal chlorosis (loss of colour at the tips of the leaves). As this progresses, the leaves may curl and crinkle. Potassium is required for formation of healthy flowers and fruit– beyond the timescale of this investigation.

Related experiments

Identifying the conditions needed for photosynthesis

www.saps.org.uk/primary/teaching-resources/216-adding-mineral-salt-do-radishes-grow-better A link to the SAPS teacher notes on a related practical – investigating the effect of different amounts of mineral fertiliser on plant growth using the ‘radish in canister’ method.

(Website accessed October 2011)

Have a language expert improve your writing

Run a free plagiarism check in 10 minutes, generate accurate citations for free.

• Knowledge Base

Methodology

• Control Variables | What Are They & Why Do They Matter?

Control Variables | What Are They & Why Do They Matter?

Published on March 1, 2021 by Pritha Bhandari . Revised on June 22, 2023.

A control variable is anything that is held constant or limited in a research study. It’s a variable that is not of interest to the study’s objectives , but is controlled because it could influence the outcomes.

Variables may be controlled directly by holding them constant throughout a study (e.g., by controlling the room temperature in an experiment), or they may be controlled indirectly through methods like randomization or statistical control (e.g., to account for participant characteristics like age in statistical tests). Control variables can help prevent research biases like omitted variable bias from affecting your results.

Table of contents

Why do control variables matter, how do you control a variable, control variable vs. control group, other interesting articles, frequently asked questions about control variables.

Control variables enhance the internal validity of a study by limiting the influence of confounding and other extraneous variables . This helps you establish a correlational or causal relationship between your variables of interest and helps avoid research bias .

Aside from the independent and dependent variables , all variables that can impact the results should be controlled. If you don’t control relevant variables, you may not be able to demonstrate that they didn’t influence your results. Uncontrolled variables are alternative explanations for your results and affect the reliability of your arguments.

Control variables in experiments

In an experiment , a researcher is interested in understanding the effect of an independent variable on a dependent variable. Control variables help you ensure that your results are solely caused by your experimental manipulation.

The independent variable is whether the vitamin D supplement is added to a diet, and the dependent variable is the level of alertness.

To make sure any change in alertness is caused by the vitamin D supplement and not by other factors, you control these variables that might affect alertness:

• Timing of meals
• Caffeine intake
• Screen time

Control variables in non-experimental research

In an observational study or other types of non-experimental research, a researcher can’t manipulate the independent variable (often due to practical or ethical considerations ). Instead, control variables are measured and taken into account to infer relationships between the main variables of interest.

To account for other factors that are likely to influence the results, you also measure these control variables:

• Marital status

Prevent plagiarism. Run a free check.

There are several ways to control extraneous variables in experimental designs, and some of these can also be used in observational studies or quasi-experimental designs.

Random assignment

In experimental studies with multiple groups, participants should be randomly assigned to the different conditions. Random assignment helps you balance the characteristics of groups so that there are no systematic differences between them.

This method of assignment controls participant variables that might otherwise differ between groups and skew your results.

It’s possible that the participants who found the study through Facebook use more screen time during the day, and this might influence how alert they are in your study.

Standardized procedures

It’s important to use the same procedures across all groups in an experiment. The groups should only differ in the independent variable manipulation so that you can isolate its effect on the dependent variable (the results).

To control variables , you can hold them constant at a fixed level using a protocol that you design and use for all participant sessions. For example, the instructions and time spent on an experimental task should be the same for all participants in a laboratory setting.

• To control for diet, fresh and frozen meals are delivered to participants three times a day.
• To control meal timings, participants are instructed to eat breakfast at 9:30, lunch at 13:00, and dinner at 18:30.
• To control caffeine intake, participants are asked to consume a maximum of one cup of coffee a day.

Statistical controls

You can measure and control for extraneous variables statistically to remove their effects on other types of variables .

“Controlling for a variable” means modelling control variable data along with independent and dependent variable data in regression analyses and ANCOVAs . That way, you can isolate the control variable’s effects from the relationship between the variables of interest.

A control variable isn’t the same as a control group . Control variables are held constant or measured throughout a study for both control and experimental groups, while an independent variable varies between control and experimental groups.

A control group doesn’t undergo the experimental treatment of interest, and its outcomes are compared with those of the experimental group. A control group usually has either no treatment, a standard treatment that’s already widely used, or a placebo (a fake treatment).

Aside from the experimental treatment, everything else in an experimental procedure should be the same between an experimental and control group.

If you want to know more about statistics , methodology , or research bias , make sure to check out some of our other articles with explanations and examples.

• Normal distribution
• Degrees of freedom
• Null hypothesis
• Discourse analysis
• Control groups
• Mixed methods research
• Non-probability sampling
• Quantitative research
• Ecological validity

Research bias

• Rosenthal effect
• Implicit bias
• Cognitive bias
• Selection bias
• Negativity bias
• Status quo bias

A control variable is any variable that’s held constant in a research study. It’s not a variable of interest in the study, but it’s controlled because it could influence the outcomes.

Control variables help you establish a correlational or causal relationship between variables by enhancing internal validity .

If you don’t control relevant extraneous variables , they may influence the outcomes of your study, and you may not be able to demonstrate that your results are really an effect of your independent variable .

Internal validity is the extent to which you can be confident that a cause-and-effect relationship established in a study cannot be explained by other factors.

“Controlling for a variable” means measuring extraneous variables and accounting for them statistically to remove their effects on other variables.

Researchers often model control variable data along with independent and dependent variable data in regression analyses and ANCOVAs . That way, you can isolate the control variable’s effects from the relationship between the variables of interest.

Cite this Scribbr article

If you want to cite this source, you can copy and paste the citation or click the “Cite this Scribbr article” button to automatically add the citation to our free Citation Generator.

Bhandari, P. (2023, June 22). Control Variables | What Are They & Why Do They Matter?. Scribbr. Retrieved August 21, 2024, from https://www.scribbr.com/methodology/control-variable/

Is this article helpful?

Pritha Bhandari

Other students also liked, what is a controlled experiment | definitions & examples, independent vs. dependent variables | definition & examples, extraneous variables | examples, types & controls, what is your plagiarism score.

Control Variable: Simple Definition

Types of Variables > Control Variable

An experiment has several types of variables , including a control variable (sometimes called a controlled variable). Variables are just values that can change; a good experiment only has two changing variables: the independent variable and dependent variable . Let’s say you are testing to see how the amount of light received affects plant growth:

• The independent variable , in this case the amount of light, is changed by you, the researcher.
• As you change the independent variable, you watch what happens to the dependent variable . In this case you see how much the plants grow.
• A control variable is another factor in an experiment; it must be held constant. In the plant growth experiment, this may be factors like water and fertilizer levels.

The Control Variable and Experimental Design

Control Variables vs. Control Groups

In any experiment or research, it can be virtually impossible to account for all variables that may affect the outcome of your experiment. If it’s difficult to identify and control all potential confounding variables, it may be necessary to make a control group . A control group provides a baseline measurement for your experiment.

Information

• Author Services

Initiatives

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Original Submission Date Received: .

• Active Journals
• Find a Journal
• Proceedings Series
• For Authors
• For Reviewers
• For Editors
• For Librarians
• For Publishers
• For Societies
• For Conference Organizers
• Open Access Policy
• Institutional Open Access Program
• Special Issues Guidelines
• Editorial Process
• Research and Publication Ethics
• Article Processing Charges
• Testimonials
• Preprints.org
• SciProfiles
• Encyclopedia

Article Menu

• Subscribe SciFeed
• Recommended Articles
• Google Scholar
• on Google Scholar
• Table of Contents

Find support for a specific problem in the support section of our website.

Please let us know what you think of our products and services.

Visit our dedicated information section to learn more about MDPI.

JSmol Viewer

Engineering the rhizosphere microbiome with plant growth promoting bacteria for modulation of the plant metabolome.

Graphical Abstract

1. Introduction

2.1. plant biochemical profile—microcosm experiments, 2.2. plant biochemical profile—field experiments, 3. discussion, 3.1. metabolic shifts and plant growth, 3.2. metabolic shifts and stress tolerance, 3.3. biostimulation of salicornia europaea with pgpb: implications for economic value, 4. materials and methods, 4.1. reagents, 4.2. plant inoculation and growth, 4.3. sample preparation for phytochemical analyses, 4.4. silylation and gas-chromatography–mass-spectrometry (gc–ms) analysis, 4.5. extraction and ultra-high-performance liquid-chromatography–mass-spectrometry (uhplc–ms) analysis, 4.6. statistical analyses, 5. conclusions, supplementary materials, author contributions, data availability statement, acknowledgments, conflicts of interest.

• Vandenkoornhuyse, P.; Quaiser, A.; Duhamel, M.; Le Van, A.; Dufresne, A. The importance of the microbiome of the plant holobiont. New Phytol. 2015 , 206 , 1196–1206. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Rodríguez-Llorente, I.D.; Pajuelo, E.; Navarro-Torre, S.; Mesa-Marín, J.; Caviedes, M.A. Bacterial Endophytes from Halophytes: How Do They Help Plants to Alleviate Salt Stress? In Saline Soil-Based Agriculture by Halotolerant Microorganisms ; Springer: Berlin/Heidelberg, Germany, 2019; pp. 147–160. [ Google Scholar ] [ CrossRef ]
• Mesa-Marín, J.; Pérez-Romero, J.A.; Mateos-Naranjo, E.; Bernabeu-Meana, M.; Pajuelo, E.; Rodríguez-Llorente, I.D.; Redondo-Gómez, S. Effect of plant growth-promoting rhizobacteria on salicornia ramosissima seed germination under salinity, CO 2 and temperature stress. Agronomy 2019 , 9 , 655. [ Google Scholar ] [ CrossRef ]
• Stanković, M.; Stojanović-Radić, Z.; Jakovljević, D.; Zlatić, N.; Luković, M.; Dajić-Stevanović, Z. Coastal Halophytes: Potent Source of Bioactive Molecules from Saline Environment. Plants 2023 , 12 , 1857. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Hameed, A.; Hussain, S.; Rasheed, A.; Ahmed, M.Z.; Abbas, S. Exploring the Potentials of Halophytes in Addressing Climate Change-Related Issues: A Synthesis of Their Biological, Environmental, and Socioeconomic Aspects. World 2024 , 5 , 36–57. [ Google Scholar ] [ CrossRef ]
• Ruiu, L. Plant-Growth-Promoting Bacteria (PGPB) against Insects and Other Agricultural Pests. Agronomy 2020 , 10 , 861. [ Google Scholar ] [ CrossRef ]
• Ali, S.; Charles, T.C.; Glick, B.R. Amelioration of high salinity stress damage by plant growth-promoting bacterial endophytes that contain ACC deaminase. Plant Physiol. Biochem. 2014 , 80 , 160–167. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Ferreira, M.J.; Sierra-Garcia, I.N.; Cremades, J.; António, C.; Rodrigues, A.M.; Pinto, D.C.G.A.; Silva, H.; Cunha, Â. Biostimulation of Salicornia europaea L. crops with plant-growth-promoting bacteria in laboratory and field conditions: Effects on growth and metabolite profile. J. Appl. Microbiol. 2023 , 134 , lxad036. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Nephali, L.; Piater, L.A.; Dubery, I.A.; Patterson, V.; Huyser, J.; Burgess, K.; Tugizimana, F. Biostimulants for plant growth and mitigation of abiotic stresses: A metabolomics perspective. Metabolites 2020 , 10 , 505. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Díaz Herrera, S.; Grossi, C.; Zawoznik, M.; Groppa, M.D. Wheat seeds harbour bacterial endophytes with potential as plant growth promoters and biocontrol agents of Fusarium graminearum. Microbiol. Res. 2016 , 186–187 , 37–43. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Abdelshafy Mohamad, O.A.; Ma, J.-B.; Liu, Y.-H.; Zhang, D.; Hua, S.; Bhute, S.; Hedlund, B.P.; Li, W.-J.; Li, L. Beneficial Endophytic Bacterial Populations Associated with Medicinal Plant Thymus vulgaris Alleviate Salt Stress and Confer Resistance to Fusarium oxysporum. Front. Plant Sci. 2020 , 11 , 47. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Poria, V.; Dębiec-Andrzejewska, K.; Fiodor, A.; Lyzohub, M.; Ajijah, N.; Singh, S.; Pranaw, K. Plant Growth-Promoting Bacteria (PGPB) integrated phytotechnology: A sustainable approach for remediation of marginal lands. Front. Plant Sci. 2022 , 13 , 999866. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Mokrani, S.; Nabti, E.; Cruz, C. Recent Trends in Microbial Approaches for Soil Desalination. Appl. Sci. 2022 , 12 , 3586. [ Google Scholar ] [ CrossRef ]
• Alves, N.S.F.; Kaory Inoue, S.G.; Carneiro, A.R.; Albino, U.B.; Setzer, W.N.; Maia, J.G.; Andrade, E.H.; da Silva, J.K.R. Variation in Peperomia pellucida growth and secondary metabolism after rhizobacteria inoculation. PLoS ONE 2022 , 17 , e0262794. [ Google Scholar ] [ CrossRef ]
• Wang, J.; Zhao, H.; Chen, T.; Lin, W.; Lin, S. Effect of Burkholderia ambifaria LK-P4 inoculation on the plant growth characteristics, metabolism, and pharmacological activity of Anoectochilus roxburghii. Front. Plant Sci. 2022 , 13 , 1043042. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Mhlongo, M.I.; Piater, L.A.; Steenkamp, P.A.; Labuschagne, N.; Dubery, I.A. Metabolic Profiling of PGPR-Treated Tomato Plants Reveal Priming-Related Adaptations of Secondary Metabolites and Aromatic Amino Acids. Metabolites 2020 , 10 , 210. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Mashabela, M.D.; Tugizimana, F.; Steenkamp, P.A.; Piater, L.A.; Dubery, I.A.; Terefe, T.; Mhlongo, M.I. Metabolomic evaluation of PGPR defence priming in wheat ( Triticum aestivum L.) cultivars infected with Puccinia striiformis f. sp. tritici (stripe rust). Front. Plant Sci. 2023 , 14 , 1103413. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Alzate Zuluaga, M.Y.; Milani, K.M.L.; Miras-Moreno, M.B.; Lucini, L.; Valentinuzzi, F.; Mimmo, T.; Pii, Y.; Cesco, S.; Rodrigues, E.P.; de Oliveira, A.L.M. The adaptive metabolomic profile and functional activity of tomato rhizosphere are revealed upon PGPB inoculation under saline stress. Environ. Exp. Bot. 2021 , 189 , 104552. [ Google Scholar ] [ CrossRef ]
• Vaishnav, A.; Kumari, S.; Jain, S.; Varma, A.; Tuteja, N.; Choudhary, D.K. PGPR-mediated expression of salt tolerance gene in soybean through volatiles under sodium nitroprusside. J. Basic Microbiol. 2016 , 56 , 1274–1288. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Rahman, M.; Rahman, M.; Sabir, A.A.; Mukta, J.A.; Khan, M.M.A.; Mohi-Ud-Din, M.; Miah, M.G.; Islam, M.T. Plant probiotic bacteria Bacillus and Paraburkholderia improve growth, yield and content of antioxidants in strawberry fruit. Sci. Rep. 2018 , 8 , 2504. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Chamam, A.; Sanguin, H.; Bellvert, F.; Meiffren, G.; Comte, G.; Wisniewski-Dyé, F.; Bertrand, C.; Prigent-Combaret, C. Plant secondary metabolite profiling evidences strain-dependent effect in the Azospirillum-Oryza sativa association. Phytochemistry 2013 , 87 , 65–77. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Pang, Z.; Chen, J.; Wang, T.; Gao, C.; Li, Z.; Guo, L.; Xu, J.; Cheng, Y. Linking Plant Secondary Metabolites and Plant Microbiomes: A Review. Front. Plant Sci. 2021 , 12 , 621276. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Santhoshkumar, R.; Akash, P.; Viswam, P.; Soniya, E.V. Imprints of PGPB association on the metabolic dynamism of Piper nigrum. J. Plant Interact. 2022 , 17 , 967–979. [ Google Scholar ] [ CrossRef ]
• Putra, A.M.; Anastasya, N.A.; Rachmawati, S.W.; Yusnawan, E.; Syibli, M.A.; Trianti, I.; Setiawan, A.; Aini, L.Q. Growth performance and metabolic changes in lettuce inoculated with plant growth promoting bacteria in a hydroponic system. Sci. Hortic. 2024 , 327 , 112868. [ Google Scholar ] [ CrossRef ]
• Solomon, W.; Janda, T.; Molnár, Z. Unveiling the significance of rhizosphere: Implications for plant growth, stress response, and sustainable agriculture. Plant Physiol. Biochem. 2024 , 206 , 108290. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Abdelaal, K.; Alkahtani, M.; Attia, K.; Hafez, Y.; Király, L.; Künstler, A. The role of plant growth-promoting bacteria in alleviating the adverse effects of drought on plants. Biology 2021 , 10 , 520. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Gupta, S.; Schillaci, M.; Roessner, U. Metabolomics as an emerging tool to study plant-microbe interactions. Emerg. Top. Life Sci. 2022 , 6 , 175–183. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Jerónimo, D.; Lillebø, A.I.; Cremades, J.; Cartaxana, P.; Calado, R. Recovering wasted nutrients from shrimp farming through the combined culture of polychaetes and halophytes. Sci. Rep. 2021 , 11 , 6587. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Kim, S.; Lee, E.Y.; Hillman, P.F.; Ko, J.; Yang, I.; Nam, S.J. Chemical structure and biological activities of secondary metabolites from Salicornia europaea L. Molecules 2021 , 26 , 2252. [ Google Scholar ] [ CrossRef ]
• Hulkko, L.S.S.; Rocha, R.M.; Trentin, R.; Fredsgaard, M.; Chaturvedi, T.; Custódio, L.; Thomsen, M.H. Bioactive Extracts from Salicornia ramosissima J. Woods Biorefinery as a Source of Ingredients for High-Value Industries. Plants 2023 , 12 , 1251. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Giordano, R.; Aliotta, G.E.; Johannesen, A.S.; Voetmann-Jensen, D.; Laustsen, F.H.; Andersen, L.A.; Rezai, A.; Fredsgaard, M.; Lo Vecchio, S.; Arendt-Nielsen, L.; et al. Effects of Salicornia-Based Skin Cream Application on Healthy Humans’ Experimental Model of Pain and Itching. Pharmaceuticals 2022 , 15 , 150. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Giordano, R.; Saii, Z.; Fredsgaard, M.; Hulkko, L.S.S.; Poulsen, T.B.G.; Thomsen, M.E.; Henneberg, N.; Zucolotto, S.M.; Arendt-Nielsen, L.; Papenbrock, J.; et al. Pharmacological insights into halophyte bioactive extract action on anti-inflammatory, pain relief and antibiotics-type mechanisms. Molecules 2021 , 26 , 3140. [ Google Scholar ] [ CrossRef ]
• Sierra-García, I.N.; Ferreira, M.J.; Torres-Ballesteros, A.; Louvado, A.; Gomes, N.; Cunha, A. Brevibacterium EB3 inoculation enhances rhizobacterial community interactions leading to improved growth of Salicornia europaea. Appl. Soil Ecol. 2024 , 196 , 105306. [ Google Scholar ] [ CrossRef ]
• Balasubramaniam, T.; Shen, G.; Esmaeili, N.; Zhang, H. Plants’ Response Mechanisms to Salinity Stress. Plants 2023 , 12 , 2253. [ Google Scholar ] [ CrossRef ]
• Ferreira, M.J.; Cunha, A.; Figueiredo, S.; Faustino, P.; Patinha, C.; Silva, H.; Sierra-Garcia, I.N. The root microbiome of salicornia ramosissima as a seedbank for plant-growth promoting halotolerant bacteria. Appl. Sci. 2021 , 11 , 2233. [ Google Scholar ] [ CrossRef ]
• Morkunas, I.; Ratajczak, L. The role of sugar signaling in plant defense responses against fungal pathogens. Acta Physiol. Plant. 2014 , 36 , 1607–1619. [ Google Scholar ] [ CrossRef ]
• Gómez-Ariza, J.; Campo, S.; Rufat, M.; Estopà, M.; Messeguer, J.; San Segundo, B.; Coca, M. Sucrose-mediated priming of plant defense responses and broad-spectrum disease resistance by overexpression of the maize pathogenesis-related PRms protein in rice plants. Mol. Plant-Microbe Interact. 2007 , 20 , 832–842. [ Google Scholar ] [ CrossRef ]
• Lacrampe, N.; Lopez-Lauri, F.; Lugan, R.; Colombié, S.; Olivares, J.; Nicot, P.C.; Lecompte, F. Regulation of sugar metabolism genes in the nitrogen-dependent susceptibility of tomato stems to Botrytis cinerea. Ann. Bot. 2021 , 127 , 143–154. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Stein, O.; Granot, D. An overview of sucrose synthases in plants. Front. Plant Sci. 2019 , 10 , 95. [ Google Scholar ] [ CrossRef ]
• Jeandet, P.; Formela-Luboińska, M.; Labudda, M.; Morkunas, I. The Role of Sugars in Plant Responses to Stress and Their Regulatory Function during Development. Int. J. Mol. Sci. 2022 , 23 , 5161. [ Google Scholar ] [ CrossRef ]
• Ohto, M.A.; Onai, K.; Furukawa, Y.; Aoki, E.; Araki, T.; Nakamura, K. Effects of sugar on vegetative development and floral transition in Arabidopsis. Plant Physiol. 2001 , 127 , 252–261. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Gouffi, K.; Pica, N.; Pichereau, V.; Blanco, C. Disaccharides as a new class of nonaccumulated osmoprotectants for Sinorhizobium meliloti. Appl. Environ. Microbiol. 1999 , 65 , 1491–1500. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• O’Hara, L.E.; Paul, M.J.; Wingler, A. How Do Sugars Regulate Plant Growth and Development? New Insight into the Role of Trehalose-6-Phosphate. Mol. Plant 2013 , 6 , 261–274. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Zhu, G.-Y.; Geuns, J.M.C.; Dussert, S.; Swennen, R.; Panis, B. Change in sugar, sterol and fatty acid composition in banana meristems caused by sucrose-induced acclimation and its effects on cryopreservation. Physiol. Plant. 2006 , 128 , 80–94. [ Google Scholar ] [ CrossRef ]
• Morkunas, I.; Bednarski, W.; Kozłowska, M. Response of embryo axes of germinating seeds of yellow lupine to Fusarium oxysporum. Plant Physiol. Biochem. 2004 , 42 , 493–499. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Cho, Y.H.; Yoo, S.D. Signaling role of fructose mediated by FINS1/FBP in Arabidopsis thaliana. PLoS Genet. 2011 , 7 , e1001263. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Siddiqui, H.; Sami, F.; Hayat, S. Glucose: Sweet or bitter effects in plants-a review on current and future perspective. Carbohydr. Res. 2020 , 487 , 107884. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Hulkko, L.S.S.; Turcios, A.E.; Kohnen, S.; Chaturvedi, T.; Papenbrock, J.; Thomsen, M.H. Cultivation and characterisation of Salicornia europaea, Tripolium pannonicum and Crithmum maritimum biomass for green biorefinery applications. Sci. Rep. 2022 , 12 , 20507. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Walker, V.; Couillerot, O.; von Felten, A.; Bellvert, F.; Jansa, J.; Maurhofer, M.; Bally, R.; Moënne-Loccoz, Y.; Comte, G. Variation of secondary metabolite levels in maize seedling roots induced by inoculation with Azospirillum, Pseudomonas and Glomus consortium under field conditions. Plant Soil 2012 , 356 , 151–163. [ Google Scholar ] [ CrossRef ]
• Walker, V.; Bertrand, C.; Bellvert, F.; Moënne-Loccoz, Y.; Bally, R.; Comte, G. Host plant secondary metabolite profiling shows a complex, strain-dependent response of maize to plant growth-promoting rhizobacteria of the genus Azospirillum. New Phytol. 2011 , 189 , 494–506. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Han, Q.Q.; Lü, X.P.; Bai, J.P.; Qiao, Y.; Paré, P.W.; Wang, S.M.; Zhang, J.L.; Wu, Y.N.; Pang, X.P.; Xu, W.B.; et al. Beneficial soil bacterium Bacillus subtilis (GB03) augments salt tolerance of white clover. Front. Plant Sci. 2014 , 5 , 525. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Chen, L.; Liu, Y.; Wu, G.; Veronican Njeri, K.; Shen, Q.; Zhang, N.; Zhang, R. Induced maize salt tolerance by rhizosphere inoculation of Bacillus amyloliquefaciens SQR9. Physiol. Plant. 2016 , 158 , 34–44. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Kang, S.-M.; Shahzad, R.; Bilal, S.; Khan, A.L.; Park, Y.-G.; Lee, K.-E.; Asaf, S.; Khan, M.A.; Lee, I.-J. Indole-3-acetic-acid and ACC deaminase producing Leclercia adecarboxylata MO1 improves Solanum lycopersicum L. growth and salinity stress tolerance by endogenous secondary metabolites regulation. BMC Microbiol. 2019 , 19 , 80. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Sunita, K.; Mishra, I.; Mishra, J.; Prakash, J.; Arora, N.K. Secondary Metabolites from Halotolerant Plant Growth Promoting Rhizobacteria for Ameliorating Salinity Stress in Plants. Front. Microbiol. 2020 , 11 , 567768. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Aluko, O.O.; Li, C.; Wang, Q.; Liu, H. Sucrose utilization for improved crop yields: A review article. Int. J. Mol. Sci. 2021 , 22 , 4704. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• de Faria, R.T.; Rodrigues, F.N.; do V., R. Oliveira, L.; Müller, C. In vitro Dendrobium nobile plant growth and rooting in different sucrose concentrations. Hortic. Bras. 2004 , 22 , 780–783. [ Google Scholar ] [ CrossRef ]
• Bonaventure, G.; Salas, J.J.; Pollard, M.R.; Ohlrogge, J.B. Disruption of the FATB gene in Arabidopsis demonstrates an essential role of saturated fatty acids in plant growth. Plant Cell 2003 , 15 , 1020–1033. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Singh, P.; Arif, Y.; Bajguz, A.; Hayat, S. The role of quercetin in plants. Plant Physiol. Biochem. 2021 , 166 , 10–19. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Zuluaga, M.Y.A.; Milani, K.M.L.; Miras-Moreno, B.; Lucini, L.; Valentinuzzi, F.; Mimmo, T.; Pii, Y.; Cesco, S.; Rodrigues, E.P.; de Oliveira, A.L.M. Inoculation with plant growth-promoting bacteria alters the rhizosphere functioning of tomato plants. Appl. Soil Ecol. 2021 , 158 , 103784. [ Google Scholar ] [ CrossRef ]
• Peer, W.A.; Murphy, A.S. Flavonoids and auxin transport: Modulators or regulators? Trends Plant Sci. 2007 , 12 , 556–563. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Hann, E.C.; Overa, S.; Harland-Dunaway, M.; Narvaez, A.F.; Le, D.N.; Orozco-Cárdenas, M.L.; Jiao, F.; Jinkerson, R.E. A hybrid inorganic–biological artificial photosynthesis system for energy-efficient food production. Nat. Food 2022 , 3 , 461–471. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Chen, H.-C.; Zhang, S.-L.; Wu, K.-J.; Li, R.; He, X.-R.; He, D.-N.; Huang, C.; Wei, H. The effects of exogenous organic acids on the growth, photosynthesis and cellular ultrastructure of Salix variegata Franch. Under Cd stress. Ecotoxicol. Environ. Saf. 2020 , 187 , 109790. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Rekha, K.; Baskar, B.; Srinath, S.; Usha, B. Plant-growth-promoting rhizobacteria Bacillus subtilis RR4 isolated from rice rhizosphere induces malic acid biosynthesis in rice roots. Can. J. Microbiol. 2018 , 27 , 20–27. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• De Boeck, H.J.; Vicca, S.; Roy, J.; Nijs, I.; Milcu, A.; Kreyling, J.; Jentsch, A.; Chabbi, A.; Campioli, M.; Callaghan, T.; et al. Global Change Experiments: Challenges and Opportunities. Bioscience 2015 , 65 , 922–931. [ Google Scholar ] [ CrossRef ]
• He, M.; Ding, N.Z. Plant Unsaturated Fatty Acids: Multiple Roles in Stress Response. Front. Plant Sci. 2020 , 11 , 562785. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Ayala, A.; Muñoz, M.F.; Argüelles, S. Lipid Peroxidation: Production, Metabolism, and Signaling Mechanisms of Malondialdehyde and 4-Hydroxy-2-Nonenal. Oxidative Med. Cell. Longev. 2014 , 2014 , 360438. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Magni, N.N.; Veríssimo, A.C.S.; Silva, H.; Pinto, D.C.G.A. Metabolomic Profile of Salicornia perennis Plant’s Organs under Diverse In Situ Stress: The Ria de Aveiro Salt Marshes Case. Metabolites 2023 , 13 , 280. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Ali, M.S.; Baek, K.H. Jasmonic acid signaling pathway in response to abiotic stresses in plants. Int. J. Mol. Sci. 2020 , 21 , 621. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Rehman, M.; Saeed, M.S.; Fan, X.; Salam, A.; Munir, R.; Yasin, M.U.; Khan, A.R.; Muhammad, S.; Ali, B.; Ali, I.; et al. The Multifaceted Role of Jasmonic Acid in Plant Stress Mitigation: An Overview. Plants 2023 , 12 , 3982. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Mathesius, U. Flavonoid functions in plants and their interactions with other organisms. Plants 2018 , 7 , 30. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Laoué, J.; Fernandez, C.; Ormeño, E. Plant Flavonoids in Mediterranean Species: A Focus on Flavonols as Protective Metabolites under Climate Stress. Plants 2022 , 11 , 172. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Ameixa, O.M.C.C.; Rebelo, J.; Silva, H.; Pinto, D.C.G.A. Gall midge Baldratia salicorniae Kieffer (Diptera: Cecidomyiidae) infestation on Salicornia europaea L. induces the production of specialized metabolites with biotechnological potential. Phytochemistry 2022 , 200 , 113207. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Li, Y.; Beisson, F.; Ohlrogge, J.; Pollard, M. Monoacylglycerols are components of root waxes and can be produced in the aerial cuticle by ectopic expression of a suberin-associated acyltransferase. Plant Physiol. 2007 , 144 , 1267–1277. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Deng, Y.; Lu, S. Biosynthesis and Regulation of Phenylpropanoids in Plants. CRC Crit. Rev. Plant Sci. 2017 , 36 , 257–290. [ Google Scholar ] [ CrossRef ]
• Ferreira, M.J.; Sierra-Garcia, I.N.; Louvado, A.; Gomes, N.C.M.; Figueiredo, S.; Patinha, C.; Pinto, D.C.G.A.; Cremades, J.; Silva, H.; Cunha, Â. Domestication shapes the endophytic microbiome and metabolome of Salicornia europaea. J. Appl. Microbiol. 2023 , 134 , lxad178. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Lephatsi, M.; Nephali, L.; Meyer, V.; Piater, L.A.; Buthelezi, N.; Dubery, I.A.; Opperman, H.; Brand, M.; Huyser, J.; Tugizimana, F. Molecular mechanisms associated with microbial biostimulant-mediated growth enhancement, priming and drought stress tolerance in maize plants. Sci. Rep. 2022 , 12 , 10450. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Ferreira, M.J.; Pinto, D.C.G.A.; Cunha, Â.; Silva, H. Halophytes as Medicinal Plants against Human Infectious Diseases. Appl. Sci. 2022 , 12 , 7493. [ Google Scholar ] [ CrossRef ]
• Patel, S. Salicornia: Evaluating the halophytic extremophile as a food and a pharmaceutical candidate. 3 Biotech 2016 , 6 , 104. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Cárdenas-Pérez, S.; Rajabi Dehnavi, A.; Leszczyński, K.; Lubińska-Mielińska, S.; Ludwiczak, A.; Piernik, A. Salicornia europaea L. Functional Traits Indicate Its Optimum Growth. Plants 2022 , 11 , 1051. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Sun, W.; Shahrajabian, M.H. Therapeutic Potential of Phenolic Compounds in Medicinal Plants—Natural Health Products for Human Health. Molecules 2023 , 28 , 1845. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Kumar, N.; Goel, N. Phenolic acids: Natural versatile molecules with promising therapeutic applications. Biotechnol. Rep. 2019 , 24 , e00370. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Liu, W.; Li, J.; Zhang, X.; Zu, Y.; Yang, Y.; Liu, W.; Xu, Z.; Gao, H.; Sun, X.; Jiang, X.; et al. Current Advances in Naturally Occurring Caffeoylquinic Acids: Structure, Bioactivity, and Synthesis. J. Agric. Food Chem. 2020 , 68 , 10489–10516. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Hur, S.; Park, G.; Joo, S.-T. Biological activities of conjugated linoleic acid (CLA) and effects of CLA on animal products. Livest. Sci. 2007 , 110 , 221–229. [ Google Scholar ] [ CrossRef ]
• Natali, F.; Siculella, L.; Salvati, S.; Gnoni, G. V Oleic acid is a potent inhibitor of fatty acid and cholesterol synthesis in C6 glioma cells. J. Lipid Res. 2007 , 48 , 1966–1975. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Dilika, F.; Bremner, P.D.; Meyer, J.J.M. Antibacterial activity of linoleic and oleic acids isolated from Helichrysum pedunculatum : A plant used during circumcision rites. Fitoterapia 2000 , 71 , 450–452. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Kusumah, D.; Wakui, M.; Murakami, M.; Xie, X.; Yukihito, K.; Maeda, I. Linoleic acid, α-linolenic acid, and monolinolenins as antibacterial substances in the heat-processed soybean fermented with Rhizopus oligosporus. Biosci. Biotechnol. Biochem. 2020 , 84 , 1285–1290. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Zheng, C.J.; Yoo, J.-S.; Lee, T.-G.; Cho, H.-Y.; Kim, Y.-H.; Kim, W.-G. Fatty acid synthesis is a target for antibacterial activity of unsaturated fatty acids. FEBS Lett. 2005 , 579 , 5157–5162. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Kolar, M.J.; Konduri, S.; Chang, T.; Wang, H.; McNerlin, C.; Ohlsson, L.; Härröd, M.; Siegel, D.; Saghatelian, A. Linoleic acid esters of hydroxy linoleic acids are anti-inflammatory lipids found in plants and mammals. J. Biol. Chem. 2019 , 294 , 10698–10707. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Youssef, A.M.M.; Maaty, D.A.M.; Al-Saraireh, Y.M. Phytochemical Analysis and Profiling of Antitumor Compounds of Leaves and Stems of Calystegia silvatica (Kit.) Griseb. Molecules 2023 , 28 , 630. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Novotny, L.; Mahmoud, F.; Abdel-Hamid, M.; Hunakova, L. Anticancer potential of β-sitosterol. Int. J. Clin. Pharmacol. Pharmacother. 2017 , 2 , 2456–3501. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Shankar, E.; Goel, A.; Gupta, K.; Gupta, S. Plant Flavone Apigenin: An Emerging Anticancer Agent. Curr. Pharmacol. Reports 2017 , 3 , 423–446. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Tian, J.; Wang, X.-Q.; Tian, Z. Focusing on Formononetin: Recent Perspectives for its Neuroprotective Potentials. Front. Pharmacol. 2022 , 13 , 905898. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Di Rienzo, J.A.; Casanoves, F.; Balzarini, M.; Gonzalez, L.; Tablada, M.; Robledo, C.W. “INFOSTAT.” Grupo InfoStat, FCA, Universidad Nacional de Córdoba, Argentina. 2020.
• Hammer, Ø.; Harper, D.A.T.; Ryan, P.D. Past: Paleontological statistics software package for education and data analysis. Palaeontol. Electron. 2001 , 4 , 11. [ Google Scholar ]
• El-Sayed, M.A.; Abbas, F.A.; Refaat, S.; El-Shafae, A.M.; Fikry, E. UPLC-ESI-MS/MS Profile of the Ethyl Acetate Fraction of Aerial Parts of Bougainvillea “scarlett o’hara” Cultivated in Egypt. Egypt. J. Chem. 2021 , 64 , 793–806. [ Google Scholar ] [ CrossRef ]
• Ivan, I.; Jungfer, E.; Ritter, C.; Santiago-schübel, B.; Thiele, B.; Fett, R.; Galensa, R. Characterization of Fl Avan-3-Ols in Seeds of Grape Pomace by CE, HPLC-DAD-MS n and LC-ESI-FTICR-MS. FRIN 2012 , 48 , 848–855. [ Google Scholar ] [ CrossRef ]
• Llorent-Martínez, E.J.; Gouveia, S.; Castilho, P.C. Analysis of Phenolic Compounds in Leaves from Endemic Trees from Madeira Island. A Contribution to the Chemotaxonomy of Laurisilva Forest Species. Ind. Crops Prod. 2015 , 64 , 135–151. [ Google Scholar ] [ CrossRef ]
• Clifford, M.N.; Johnston, K.L.; Knight, S.; Kuhnert, N. Hierarchical Scheme for LC-MSn Identification of Chlorogenic Acids. J. Agric. Food Chem. 2003 , 51 , 2900–2911. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Llorent-Martinez, E.J.; Spinola, V.; Gouveia, S.; Castilho, P.C. HPLC-ESI-MSn Characterization of Phenolic Compounds, Terpenoid Saponins, and Other Minor Compounds in Bituminaria Bituminosa. Ind. Crops Prod. 2015 , 69 , 80–90. [ Google Scholar ] [ CrossRef ]
• Singh, A.; Kumar, S.; Kumar, B. LC-MS Identification of Proanthocyanidins in Bark and Fruit of Six Terminalia Species. Nat. Prod. Commun. 2018 , 13 , 555–560. [ Google Scholar ] [ CrossRef ]
• Di Lecce, G.; Arranz, S.; Jáuregui, O.; Tresserra-Rimbau, A.; Quifer-Rada, P.; Lamuela-Raventós, R.M. Phenolic Profiling of the Skin, Pulp and Seeds of Albariño Grapes Using Hybrid Quadrupole Time-of-Flight and Triple-Quadrupole Mass Spectrometry. Food Chem. 2014 , 145 , 874–882. [ Google Scholar ] [ CrossRef ]
• Clifford, M.N.; Knight, S.; Kuhnert, N. Discriminating between the Six Isomers of Dicaffeoylquinic Acid by LC-MSn. J. Agric. Food Chem. 2005 , 53 , 3821–3832. [ Google Scholar ] [ CrossRef ]
• Zhang, Y.; Xiong, H.; Xu, X.; Xue, X.; Liu, M.; Xu, S.; Liu, H.; Gao, Y.; Zhang, H.; Li, X. Compounds Identification in Semen Cuscutae by Ultra-High-Performance Liquid Chromatography (Uplcs) Coupled to Electrospray Ionization Mass Spectrometry. Molecules 2018 , 23 , 1199. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Gouveia, S.; Castilho, P.C. Characterisation of Phenolic Acid Derivatives and Flavonoids from Different Morphological Parts of Helichrysum obconicum by a RP-HPLC-DAD-(−)-ESI-MS n Method. Food Chem. 2011 , 129 , 333–344. [ Google Scholar ] [ CrossRef ] [ PubMed ]

Click here to enlarge figure

Share and Cite

Ferreira, M.J.; Veríssimo, A.C.S.; Pinto, D.C.G.A.; Sierra-Garcia, I.N.; Granada, C.E.; Cremades, J.; Silva, H.; Cunha, Â. Engineering the Rhizosphere Microbiome with Plant Growth Promoting Bacteria for Modulation of the Plant Metabolome. Plants 2024 , 13 , 2309. https://doi.org/10.3390/plants13162309

Ferreira MJ, Veríssimo ACS, Pinto DCGA, Sierra-Garcia IN, Granada CE, Cremades J, Silva H, Cunha Â. Engineering the Rhizosphere Microbiome with Plant Growth Promoting Bacteria for Modulation of the Plant Metabolome. Plants . 2024; 13(16):2309. https://doi.org/10.3390/plants13162309

Ferreira, Maria J., Ana C. S. Veríssimo, Diana C. G. A. Pinto, Isabel N. Sierra-Garcia, Camille E. Granada, Javier Cremades, Helena Silva, and Ângela Cunha. 2024. "Engineering the Rhizosphere Microbiome with Plant Growth Promoting Bacteria for Modulation of the Plant Metabolome" Plants 13, no. 16: 2309. https://doi.org/10.3390/plants13162309

Article Metrics

Article access statistics, supplementary material.

ZIP-Document (ZIP, 424 KiB)

Further Information

Mdpi initiatives, follow mdpi.

Subscribe to receive issue release notifications and newsletters from MDPI journals

• Open access
• Published: 24 August 2024

Exploring the impact of plant growth-promoting bacteria in alleviating stress on Aptenia cordifolia subjected to irrigation with recycled water in multifunctional external green walls

• Mansoure Jozay   ORCID: orcid.org/0000-0002-2513-6794 1 ,
• Hossein Zarei   ORCID: orcid.org/0000-0002-2792-9480 1 ,
• Sarah Khorasaninejad   ORCID: orcid.org/0000-0002-2786-4015 1 &
• Taghi Miri   ORCID: orcid.org/0000-0002-2428-1332 2  

BMC Plant Biology volume  24 , Article number:  802 ( 2024 ) Cite this article

Metrics details

Rapid urbanization and population growth exert a substantial impact on the accessibility of drinking water resources, underscoring the imperative for wastewater treatment and the reuse of non-potable water in agriculture. In this context, green walls emerge as a potential solution to augment the purification of unconventional waters, simultaneously contributing to the aesthetic appeal and enjoyment of urban areas. This study aims to optimize water management in green walls by investigating the impact of bacterial strains on the biochemical properties and performance of the ornamental accumulator plant, Aptenia cordifolia , grown with various unconventional water sources. The experiments were designed as split plots based on a completely randomized block design with three replications. The main factor was recycled water with three levels (gray water, wastewater from the Kashfroud region of Mashhad, and urban water (control)). The sub-factor included different bacterial strains at four levels, composed of various bacteria combinations, (B1: Psedoumonas flucrecens  +  Azosporillum liposferum  +  Thiobacillus thioparus  +  Aztobactor chorococcum , B2: Paenibacillus polymyxa  +  Pseudomonas fildensis  +  Bacillus subtilis  +  Achromobacter xylosoxidans  +  Bacillus licheniform , B3: Pseudomonas putida  +  Acidithiobacillus ferrooxidans  +  Bacillus velezensis  +  Bacillus subtilis  +  Bacillus methylotrophicus  +  Mcrobacterium testaceum , and the control level without bacterial application (B0).

The findings revealed significant differences at the 5% probability level across all morphophysiological traits, including plant height, the number and length of lateral branches, growth index, and plant coverage. Moreover, superior morphophysiological traits were observed in plants cultivated in substrates inoculated with wastewater irrigation. Substrates inoculated with bacteria exhibited the highest relative water content (RWC) and chlorophyll levels, coupled with the lowest relative saturation deficit (RSD), electrolyte leakage (EL), and carotenoid levels. Furthermore, plant growth-promoting bacteria (PGPB), from a biochemical perspective, were associated with increased carbohydrates, total protein, and anthocyanin. They also contributed to controlling oxidative stress caused by free radicals by enhancing the activity of antioxidant enzymes, such as guaiacol peroxidase (GPX), polyphenol oxidase (PPO), ascorbate peroxidase (APX), and peroxidase (POD), while reducing catalase enzyme (CAT) activity. This led to increased resistance to stress, as evidenced by a decrease in malondialdehyde and proline levels. The study concludes that the MIX B3, being both ecofriendly and economical, represents an effective strategy for mitigating the adverse effects of wastewater on plants.

This study showed that plant irrigation using wastewater increases the levels of proline, phenols and oxidative stress. However, the application of plant growth promoting bacteria (PGPB) reduced oxidative damage by increasing antioxidant activity and decreasing proline and phenol levels. These findings show the potential of bacterial treatments to improve plant growth and reduce adverse effects of recycled water irrigation.

Peer Review reports

Urban environments, with their current unsustainable developments, pose significant challenges that require fundamental environmental reforms. The primary reasons for these challenges include the purification of pollutants, urban heat islands, the expansion of impervious surfaces, climate change, loss of biodiversity, and the aesthetic damage associated with current urban transformations [ 1 ]. However, natural and non-built urban environments have the capacity for ecological restoration and improvement. Unfortunately, the unsustainable approaches of urban development have deprived us of these options [ 2 ]. To address this dilemma, Artmann and Sartison [ 3 ] proposed a nature-based solution. One of these solutions is sustainable green infrastructure in cities, including exterior green walls.

Considering that horizontal expansion of green spaces may not be feasible due to limited vacant land, vertical development of gardens remains the only viable solution in theory to fulfill this objective. The severe shortage of space in today's densely populated urban environments underscores the need for future cities to explore the expansion of rooftops and green walls [ 4 ]. Vertical green walls can be considered as sustainable environmental technologies that offer numerous economic, environmental, and social benefits and are also regarded as urban lungs [ 5 ]. However, the fundamental potential of these infrastructures, in terms of sustainable building elements, has received less attention in research with the aim of enhancing their environmental efficiency and multifunctionality [ 6 ].

The high water demand of green systems is a limiting factor for their development in water-scarce regions. To minimize water requirements, it is advisable to opt for drought-tolerant and heat-resistant plant species [ 7 ]. For example, Aptenia cordifolia is a CAM (Crassulacean Acid Metabolism) plant that is drought-resistant and also capable of phytoremediation. Aptenia cordifolia belonging to the Aizoaceae family, is a succulent and perennial species widely employed in green spaces in southern Iran. It is known as a hardy, flowering, perennial, trailing plant. In Iran it is also introduced as Ice Flower. A. cordifolia is resistant to heat and drought. It has light pink, white and red flowers that bloom throughout the year. The solitary to clustered flowers arise in leaf axils and generally open during the day. This plant needs full sun exposure.

Another solution can be the utilization of alternative water sources such as graywater and wastewater [ 5 ]. Since green walls are considered an environmental innovation, it is necessary to incorporate the possibility of rainwater harvesting and the reuse of recycled water [ 8 ].

Soil pollution with heavy metals, originating from fertilizers and non-conventional water sources used in urban green spaces, presents a significant issue due to the potential transfer of these contaminants to agricultural products in cities. Therefore, multiple challenges hinder the sustainability of this concept in urban environments, requiring effective planning to address them. Important factors such as agricultural and horticultural practices, including growth-promoting bacteria, design elements, and plant characteristics, play a crucial role in soil filtration [ 9 ] and the removal of pollutants. Recently, it has been reported that planting crops in Substrate containing certain growth-promoting bacteria can efficiently fulfill their nutrient requirements and reduce the need for chemical fertilizers [ 10 ]. Research on biofertilizers includes various types of microorganisms that can convert less accessible forms of essential nutrients into accessible forms through biological processes. This leads to the development of better root systems and seed germination [ 11 ]. Liu et al., [ 12 ] isolated forty-one bacterial strains from the rhizosphere soil and root tissues of five plant species ( Artemisia argyi Levl.، Gladiolus gandavensis Vaniot Houtt، Boehmeria nivea L.، Veronica didyma Tenore و Miscanthus floridulonizing Lab). Their results showed that among the bacteria, two strains, Klebsiella michiganensis TS8 and Lelliottia jeotgali MR2, exhibited higher tolerance to cadmium and were highly successful in cadmium phytoremediation of the soil. TS8 increased plant height and the dry weight of leaves from 9.39 to 1.99. It appears that strains of Pseudomonas, Mycobacterium, Staphylococcus, Micrococcus, Bacillus, Paenibacillus, and Klebsiella were widely used solely for phytoremediation, and the combined application of these strains were more successful than their individual application [ 13 ]. Some native plants and grasses also have the potential for phytoremediation in metal-contaminated water. Phytoremediation through biostimulation is a promising approach that can enhance the synergistic effects of microorganisms and plants [ 14 ].

On one hand, the most significant challenge is population growth and the increasing water demand for economic activities, particularly agriculture. Water scarcity is a major concern in densely populated urban areas. On the other hand, the disposal of graywater and urban wastewater can contaminate surface and groundwater sources. In this regard, green walls can improve the treatment of non-conventional waters, enhance the hydrological cycle, and increase the beauty and enjoyment of urban areas [ 15 ]. The benefits of using ornamental plant coverings with distinct aesthetic features are evident in satisfying the visual preferences of the community and providing pleasant landscape value [ 16 ]. Using alternative ornamental species in constructed wetlands (CWs), represents an effective system for pollutant removal [ 17 ]. Additionally, the use of these systems can significantly improve the visual quality of the landscape, which is often undervalued but has positive social and psychological impacts on people's daily lives [ 18 ].

One of the important aspects of non-conventional water is graywater and wastewater. Graywater refers to wastewater generated from laundry, toilets, showers, baths (also known as light graywater), and, in some cases, kitchen sinks and dishwashers (known as dark graywater). Light graywater is produced in significant amounts (45% to 60% of domestic wastewater) and contains a lower pollutant load compared to mixed domestic wastewater [ 19 ]. For this reason, considerable efforts have been made to reuse it on-site. Additionally, treated wastewater has good nutrient value that can enhance plant growth, reduce fertilizer consumption, and increase the productivity of nutrient-depleted soils [ 20 , 21 ].

Given the challenges posed by population growth and increased water demand for economic activities, particularly in agriculture, urban areas face significant hurdles. Freshwater scarcity emerges as a primary concern in these regions, where the disposal of graywater and urban wastewater poses a potential threat to surface and groundwater sources, exacerbating the issue. While green walls can effectively treat polluted water, enhance the hydrological cycle, and beautify urban areas, they may also introduce stress conditions. These stressors arise from microclimate factors and the use of unconventional water for plant growth, potentially increasing the transfer of heavy metals and other non-biological stresses Additionally, the use of wastewater in urban agriculture requires additional caution. Therefore, the primary objective of this study is to investigate the feasibility of utilizing organic biological fertilizers to improve soil organic matter content, providing a sustainable alternative to chemical fertilizers. The goal is to mitigate existing stresses and promote environmentally friendly soil management practices.

The study area and site

This research was conducted in Mashhad, located in northeastern Iran. Mashhad is the capital of Khorasan Razavi province and the second largest and most populous city in Iran. It has a semi-arid climate with cold winters and hot, dry summers (elevation of 995 m above sea level, geographical coordinates 36 degrees 18 min north, 59 degrees 36 min east). The average annual precipitation is approximately 255 mm. The average minimum and maximum temperatures annually are -4 and 22 degrees Celsius, respectively, and the relative humidity is reported to be about 40% [ 22 ]. The precise location of the experiment was on a 15 m wall in the outdoor area of the Armgan Greenhouse, located in the northern part of Mashhad (Fig.  1 ).

Location of the study site in Mashhad

The green wall systems

To conduct this research, vertical cultivation panels (mesh made of 5 mm steel wire) were installed at a suitable distance from shading factors, facing east–west. The vertical cultivation system used in this project was called the "Almich" system (initially developed in Malaysia). The experimental units consisted of Almich green wall pots, plastic pots made of new or recycled polypropylene, with dimensions of approximately 20 × 20 and a depth of about 21 cm. Leca was used as a drainage layer at the bottom of the pots, followed by a layer of geotextile as a soil filter.

In each plot, two plants of each tested species were cultivated. This study was conducted on an exterior green wall from March to December 2022. Each wall consisted of two panels measuring 44 × 106 cm, and each set of three walls constituted one replication of the experiment. Each wall had 4 vertical rows and 4 horizontal rows, resulting in a total of 16 plots (experimental units) per wall. Considering A. cordifolia and the mentioned treatments with three replications, the first experiment included 48 experimental units (Fig.  2 ).

A view of the experiment’s external green walls

The experiment

This experiment aimed to investigate the impact of irrigation water quality and different bacterial strains on the growth and performance of selected ornamental plants with phytoremediation characteristics, while also measuring the water consumption of these species under green wall conditions. This research was implemented as split-plot layout, based on a randomized complete block design with three replications from March to December 2022. The organic matter content in this experiment was 20%, which is very close to the maximum recommended 30% organic matter content (OC) by FLL (2018) [ 23 ].

The growth medium used in all experimental units was the same and consisted of the following components: 25% cocopeat [ 24 ], 5% vermicompost [ 25 ], 55% perlite [ 26 ], 10% vermiculite [ 27 ], and 5% zeolite [ 28 ].

Unconventional water treatments

The main-factor included graywater collected from rainwater, a twin sink designed for fruit and vegetable washing, wastewater from the Kashafrood region and urban water (control). The main factor was applied in three main tank reservoirs and irrigated to the plants in the form of drip irrigation, with irrigation levels at around 80% of the field capacity, adjusted based on the flow rate of the drip emitters.

Treatment of plant growth-promoting bacteria

In this research, growth-promoting bacteria were utilized. Apart from fulfilling the plant's requirements for chemical fertilizers, these bacteria also absorbed heavy metals from the soil [ 29 , 30 ]. Previous research suggests that the combined application of these bacterial strains has been more effective than their individual application [ 30 ]. The sub-factor of different biological strains of bacteria at four levels: Mix1 ( Psedoumonas flucrecens  +  Azosporillum Liposferum  +  Thiobacillus thioparus  +  Aztobactor chorococcum ), Mix2 ( Paenibacillus polymyxa  +  Pseudomonas fildensis  +  Bacillus subtilis  +  Achromobacter xylosoxidans  +  Bacillus licheniformis ), Mix3 ( Pseudomonas putida  +  Acidithiobacillus ferrooxidans  +  Bacillus velezensis  +  Bacillus subtilis  +  Bacillus methylotrophicus  +  Mcrobacterium testaceum ) and B0- control (without bacterial inoculation).

Each plant received 20 cc of the biofertilizer. The bacteria utilized in this experiment were procured from the Soil Biology and Biotechnology Laboratory at the Golestan Agricultural and Natural Resources Research Center LBSG, specifically identified by isolate number 041011 in the laboratory bank. The bacteria were extracted from the rhizosphere of agronomy plants like soybean and wheat using the method of Ju et al. [ 31 ]. The full scientific name and strains of the bacterial treatments in this study are given in Abbreviation. Except for the non-inoculated controls, the substrates around each plant specimen were watered and inoculated with 20 ml of bacterial suspension to obtain a bacterial concentration of 108 CFU/ml. The liquid was injected evenly around the plant root zone and the substrate surface around each plant using a syringe two weeks after planting. This method was selected to ensure an even distribution of the bacterial treatment in the substrate.

Selected accumulator ornamental plant

This plant, especially in phytoremediation, has effective applications for pollution extraction. They may act as a phytostabilizer, particularly in areas affected by metals [ 32 ] (Fig.  3 ). The selection of A. Cordifolia for the experiment was based on its ornamental qualities and accumulation characteristics. Additionally, this plant belongs to the CAM family, and in accordance with the traits of the Crassulaceae pathway, it exhibits rapid growth, high biomass production, robust root development, and high water use efficiency. These attributes enable effective pollutant removal, making it suitable for use in urban green walls. The Aptenia cordifolia seedlings utilized in this experiment were sourced from regions conducive to their growth within the country (Iran). Specifically, they were obtained from the Shandiz greenhouse in Mashhad, where such seedlings are abundantly produced.

Trends in Temperature, Relative Humidity, Precipitation, and Wind Speed in the Experimental Months of the Year 2022 in Mashhad City

The substrate was first passed through a 2 mm sieve after air drying. Physical and chemical characteristics were determined using standard laboratory methods as described in the subsequent sentences. Acidity and conductivity were measured using the extract and saturated mud [ 33 ], and the numbers were red using an Electrical Conductivity (E.C.) meter model JENWAY4510, and a pH meter model METROHM691. Particle density, bulk density, and total porosity were measured based on Chen et al. [ 34 ]. The field capacity and permanent wilting point of the substrate was measured according to the method of Salter and Haworth [ 35 ]. Organic carbon and organic matter measurements were based on Walkley and Black [ 36 ]. The digestion method determined macro and microelements. An instrument Inductively Coupled Plasma-Optical Emission Spectrometer (ICP-OES) Model 76,004,555 made in Germany, was used to take these measurements (Table  1 ). Throughout the study period, data on relative humidity, wind speed, precipitation, and air temperature at the experiment site were meticulously recorded (Fig.  4 ).

The appearance conditions of the four experimental plants studied in the under B3 and I2 treatments A , B , C and D during spring/summer/autumn and winter season!

Measurements

Morphological traits.

The measured morphological traits in this experiment included plant height, internode distance (measured using a ruler), stem diameter, leaf diameter, leaf length, leaf width (measured using a digital caliper), number of lateral branches, number of nodes, and number of leaves on lateral branches. These measurements were taken on a monthly basis [ 37 , 38 ]. The growth index (plant width × plant length × plant height) was also assessed on a monthly basis.

Plant coverage

The plant surface coverage, also known as the horizontal and vertical coverage of plants, was calculated using quadrat constructed to the size of 20 × 20 square centimeters for each experimental unit in each green wall. Each quadrat consisted of 100 chambers, with each chamber measuring 2 × 2 in width and length.

Physiological and Biochemical Traits

All physiological and biochemical traits were evaluated at the end of the experiment during the autumn season.

Total chlorophyll and carotenoid

The total chlorophyll concentration was assessed using the method described by Dere [ 39 ]. Fresh leaves weighing 0.2 g were thoroughly ground and homogenized in a mortar with 10 mL of 96% methanol. The grinding and homogenization process should be carried out in a cool and low-light environment. The resulting mixture was filtered through filter paper, and then subjected to centrifugation at 2500 rpm for 10 min. The supernatant was immediately collected, and the light absorption at wavelengths of 666 nm, 653 nm, and 470 nm was measured using a spectrophotometer for chlorophyll a, chlorophyll b, and carotenoids, respectively. Finally, the carotenoid and total chlorophyll concentrations were determined using the Eqs. ( 3 ) and ( 4 ):

Relative Water Content (RWC)

The calculation of leaf relative water content (RWC) was performed using the method described by Hossain et al., [ 40 ]. Initially, a leaf sample was weighed using a balance to obtain the initial weight. Then, to obtain the turgid weight, the samples were placed in closed containers containing distilled water for 12 h at a temperature of 21 °C (19 to 23 °C). After removing excess water from the leaf surface, the turgid weight of the samples was measured. To determine the dry weight, the samples were transferred to an oven at a temperature of 70 °C for 48 h (Eq.  5 ).

In which Fw represents the fresh weight, Dw represents the dry weight, and Tw represents the turgid weight of the leaf.

Relative Saturation Deficit (RSD)

The relative saturation deficit was calculated using the method described by Samar Raza et al., [ 41 ]. After collecting the leaves and weighing wet weight them (Fw), they are submerged in distilled water at room temperature for 5 h. Then, the leaves were removed from the water, re-weighed, and their turgid weight (Tw) is obtained. The relative saturation deficit is calculated using Eq.  6 .

Electrolyte Leakage (EL)

The stability of the cell membrane was measured using the method described by Sairam and Srivastava [ 42 ]. Leaf segments measuring 2 cm in size were prepared and washed. These segments were then placed in test tubes along with 10 ml of distilled water. At this stage, the electrical conductivity of the test samples (E1) was measured using a JENWAY conductivity meter. Then, the test tubes were transferred to an autoclave and subjected to a temperature of 121 degrees Celsius for 15 min to kill the leaf cells. After cooling down, the electrical conductivity was measured again (E2) in this stage. Finally, the electrolyte leakage values were calculated using Eq.  7 .

Soluble carbohydrates

To measure the soluble carbohydrates, 2.0 ml of methanol extract were mixed with 3 ml of anthrone reagent (0.15 g anthrone in 100 ml of 72% sulfuric acid). The mixture was then placed in a hot water bath at a temperature of 100 degrees Celsius for 20 min to allow the reaction to occur. The absorbance of the samples was measured at a wavelength of 620 nm using a spectrophotometer after cooling [ 43 ].

Total Protein Content

To measure the protein concentration, 5 ml of urine reference solution were added to a test tube, followed by the addition of 100 µl of protein extract, and it was quickly mixed. After 5 min, the absorption was read at a wavelength of 595 nm using a spectrophotometer. The protein concentration was calculated using the standard curve of albumin [ 44 ].

Anthocyanins

To measure anthocyanins in the leaves, the method described by Nadernejad et al., [ 45 ] was used. Fresh plant tissue weighing 0.1 g was ground in a mortar and pestle with 10 ml of methanol acid solution (pure methanol and pure hydrochloric acid in a volumetric ratio of 1:99). The extract was poured into a screw-capped test tube and kept in darkness at a temperature of 25 °C for 24 h. Then, it was centrifuged at 4000 rpm for 10 min, and the absorbance of the supernatant was measured at a wavelength of 550 nm using a spectrophotometer. The concentration was calculated using Eq.  8 , considering the extinction coeffici of 33,000 (ε) in cm/mol (Eq.  8 ).

where A represents absorbance, b is the cell width, and c is the concentration of the solution under investigation.

Total phenols

Determination of total phenols was performed using the Folin-Ciocalteu reagent at a wavelength of 765 nm, following the method described by Singleton and Rossi [ 46 ]. The measurement was carried out by calibrating the standard curve with gallic acid, and the amount of total phenolic compounds was expressed as milligrams of gallic acid equivalents per 100 g of dry weight.

Malondialdehyde (MDA)

To measure malondialdehyde (MDA) levels, approximately 0.2 g of fresh leaf tissue (the youngest leaves at the tip of the stem) were ground in a mortar containing 5 ml of 0.1% tri-chloroacetic acid (TCA). To the resulting centrifuged solution (1 ml), 5 ml of 20% TCA solution containing 5.0% thiobarbituric acid (TBA) were added. The concentration of malondialdehyde was measured at a wavelength of 532 nm. As other compounds besides malondialdehyde in the solution exhibit non-specific absorption, their absorption at a wavelength of 600 nm was also measured [ 47 ] (Eq.  9 ).

In Eq. ( 9 ), A represents the absorption of the sample of interest, £ represents the molar absorptivity coefficient, which is equal to 1.55 × 10 –5 Mcm −1 , and C represents the concentration of malondialdehyde.

Assays of antioxidant activities

To measure the antioxidant activity, the methanolic extract was first diluted at a ratio of 1:10. Then, to deactivate the free radicals, 4 ml of 2,2-Diphenyl-1-picrylhydrazyl (DPPH) solution was added to each sample [ 48 ]. The samples were kept in darkness for 30 min, and the absorbance of the resulting solutions and the absorbance of the control sample were measured at a wavelength of 517 nm using a spectrophotometer. The inhibition percentage of DPPH was obtained by comparing the absorbance of the extract sample with the absorbance of the control sample using Eq.  10 .

Assay GPX activity

To evaluate GPX activity, initially, 2 ml of 0.05 M sodium phosphate buffer with a pH of 5.6 were mixed with 2 ml of 3% hydrogen peroxide and 2 ml of 5-micromolar ascorbate in an ice bath. Immediately, 1.0 ml of enzyme extract from leaf tissue was added, and the absorbance changes at 256 nm were monitored by a spectrophotometer for 2 min with 10 s intervals. The enzyme activity is defined as the volume of enzyme required to hydrolyze one millimole of substrate per minute at 25 °C. Then, the enzyme activity was calculated in units per minute per milligram of protein [ 49 , 50 ].

Assay PPO activity

To measure the activity of the PPO enzyme, pyrogallol was used as the enzyme substrate. The reaction mixture consisted of 2.5 ml of 50 mM potassium phosphate buffer (pH 7), 200 µl of 0.02 M pyrogallol, and 100 µl of enzyme extract. The absorbance of the samples was read at a wavelength of 420 nm after three minutes using a spectrophotometer. The enzyme activity was calculated using the molar absorptivity coefficient of 6.2 Mm −1 Cm −1 [ 51 ].

Assay CAT activity

The CAT activity was measured by Kendal and Scandellius [ 52 ] method. Initially, 5.2 ml of 0.05 M potassium phosphate buffer with a pH of 7 and 0.3 ml of 3% hydrogen peroxide were mixed together in an ice bath. Immediately, 0.2 ml of enzyme extract were added to the mixture, and the absorbance changes at a wavelength of 240 nm were monitored for 4–3 min. Enzyme unit was defined per H 2 O 2 µmol ml- 1 decomposed per minute at 25 °C and then the enzyme activity was calculated in terms of unit changes per minute per mg of protein.

Assay APX activity

To evaluate APX activity, according to this method, initially, 2 ml of 0.05 M sodium phosphate buffer with a pH of 5.6 were mixed with 2 ml of 3% hydrogen peroxide and 2 ml of 5-micromolar ascorbate in an ice bath. Immediately, 1.0 ml of enzyme extract from leaf tissue was added, and the absorbance changes at 256 nm were monitored by a spectrophotometer for 2 min with 10 s intervals. The enzyme activity is defined as the volume of enzyme required to hydrolyze one millimole of substrate per minute at 25 °C. Then, the enzyme activity was calculated in units per minute per milligram of protein [ 50 ].

Assay POX activity

Measurement of POX activity was done according to Holley's method [ 53 ]. In this regard, initially, 2 ml of 0.2 M acetate buffer with a pH of 5, 0.2 ml of 3% hydrogen peroxide, and 0.1 ml of a 0.02 M benzidine solution in 50% methanol are mixed in an ice bath. Then, 0.1 ml of leaf enzyme extract is added to this mixture, and the absorbance curve of the samples is plotted using a spectrophotometer at a wavelength of 530 nm, at room temperature, every 30 s for 3 min. The specific enzyme activity is calculated by using the standard curve and determining the change in enzyme unit per minute per milligram of protein.

Statistical analysis

JMP 8 software was used for statistical analysis. Data analysis was performed using analysis of variance (ANOVA) and mean comparison with the Tukey test at a significance level of at least 5%. All graphs were plotted using Excel software.

According to the analysis of variance, the simple effects of recycled water treatments showed significant differences at the 5% probability level in the growth index characteristics of A. cordifolia . For plant height, only the simple effects of bacterial strains were statistically significant at the 5% level. When evaluating the number of lateral branches, significant differences were observed in both the interaction effects of the two treatments and the individual effects of bacterial strains (p ≤ 0.05). The simple effects of irrigation water type and bacterial strain showed significant differences at 5% probability in surface coverage and RSD. Additionally, in the case of total chlorophyll, RWC, carotenoids, and EL, not only the simple effects of irrigation and bacteria but also their interaction effects were significant at 5% level (p ≤ 0.05) (Table  2 ).

Type of irrigation water and bacteria on morphophysiological traits of Aptenia cordifolia

The impact of varied irrigation water and bacterial strains on the growth traits and surface coverage.

According to the analysis of variance, all morphological traits of the plant species were statistically significant at 5% probability level.

Plant height and growth index

As shown in Fig.  5 , the simple effects of bacterial strains on the height of the studied plant species in ice plant indicate that inoculation of the substrates with Mix B3 and Mix B2 resulted in an increase in plant height. Regarding the growth index in the ice plant, irrigation with wastewater and gray water resulted in a higher growth index compared to urban water.

Simple effects of recycled water and bacterial strains on growth traits, Plant height: (standard error: 2.95). Growth index: (standard error: 3575.99) and coverage level, the data are shown A and B (standard error: 1.06, 2.34). Data analysis was used JMP 8 software and analysis of variance (ANOVA) as split plots based on a completely randomized block design

As evident from the analysis of variance (Table  3 ), in relation to the biochemical data of A. cordifolia , the simple effects of bacterial strain treatments showed significant differences at 1% probability level in proline, GPX activity, and CAT. However, the simple effects of irrigation water type and bacterial strain showed significant differences at 1% level in POX activity. In the evaluation of carbohydrates and antioxidant activity, apart from the interactive effect of the two treatments, there were significant differences in the simple effects of bacterial strains (p ≤ 0.05). Additionally, in leaf phenolic, besides the simple effects of irrigation water type (p ≤ 0.01), the block effect and their interaction effects also showed significant differences (p ≤ 0.01). According to the analysis of variance (Table  3 ), regarding the total protein, anthocyanins, MDA, PPO enzyme, and APX, apart from the simple effects of irrigation water type and bacterial strain ( p  ≤ 0.01), their interaction effects also showed significant differences (p ≤ 0.05).

Surface coverage

As observed in Fig.  3 , the treatment involving the inoculation of the substrates with Mix B3 and irrigation with wastewater led to increased surface coverage (262.67 cm2) in the ice plant. It is noteworthy that the Mix B1 resulted in less surface coverage compared to the control (without bacterial application), suggesting that the B1 combination was not particularly successful in A. cordifolia .

Impact of varied irrigation water and bacterial strains on the number and length of lateral branches

Figure  6 demonstrates that the highest number of lateral branches in A. cordifolia . (11 branches) was observed in the presence of Mix B3 and irrigation with wastewater. The lowest number of lateral branches in the ice plant was observed in tap water in the presence of Mix B3 and B0, with an average of approximately 3 branches (Fig.  6 ). Additionally, the maximum lateral branch length in A. cordifolia . in all three water type treatments was related to the presence of the combined strains B3, B2, and B1, and the lowest was related to the control.

Interaction effects of recycled water and bacterial strains on n the number and length of lateral branches data are shown (standard error: 1.06, 2.34). Data analysis was used JMP 8 software and analysis of variance (ANOVA) as split plots based on a completely randomized block design

According to Fig.  7 , the highest total chlorophyll content was observed in Mix B1 in plants irrigated with wastewater (2.07 mg/g, FW), while the lowest total chlorophyll content was associated with the B0 of plants irrigated with urban water (0.62 mg/g, FW). Regarding carotenoid content in the A. cordifolia , the control treatment (without bacterial application) using urban water exhibited the highest carotenoid levels. Plants treated with the combined bacterial strains showed less leaf yellowing across various irrigation water types.

Interaction effects of recycled water and bacterial strains on photosynthetic pigments data are shown total chlorophyll: (standard error: 0.11). Carotenoid: (standard error: 0.54). Data analysis was used JMP 8 software and analysis of variance (ANOVA) as split plots based on a completely randomized block design

Impact of varied irrigation water and bacterial strains on RWC

According to Fig.  8 , in A. cordifolia , inoculation with Mix B3 in the substrate and irrigation with gray water resulted in the highest leaf water content (65.68%). The lowest leaf water content was observed in the control treatment (without bacterial application) and irrigation with urban water (26.69%).

Interaction effects of recycled water and bacterial strains on RWC, (standard error: 2.57). Data analysis was used JMP 8 software and analysis of variance (ANOVA) as split plots based on a completely randomized block design

Impact of varied irrigation water and bacterial strains on RSD

As evident in Fig.  9 , the highest relative water deficit in A. cordifolia was associated with the control treatment (without bacterial application). It is noteworthy that the highest RSD in the experiment was also observed in plants irrigated with urban water. It appears that recycled waters resulted in a lower RSD.

Interaction effects of recycled water and bacterial strains on RSD, data are shown A and B : (standard error: 1.41, 2). Data analysis was used JMP 8 software and analysis of variance (ANOVA) as split plots based on a completely randomized block design

Biochemical characteristics of ornamental plant

Different small Latin letters in each column of the table indicate a significant difference based on Tukey's test at a probability level of at least 5%. Data represent ± 1 standard error (SE). Data analysis was used JMP 8 software and analysis of variance (ANOVA) as split plots based on a completely randomized block design.

GPX: (guaiacol peroxidase), POD: (peroxidase), CAT: (catalase), B1:) Sedoumonas flucrecens  +  Azosporillum Liposferum  +  Thiobacillus thioparus  +  Aztobactor chorococcum ), B2: ( Paenibacillus polymyxa  +  Pseudomonas fildensis  +  Bacillus subtilis  +  Achromobacter xylosoxidans  +  Bacillus licheniformis ), B3: ( Pseudomonas putida  +  Acidithiobacillus ferrooxidans  +  Bacillus velezensis  +  Bacillus subtilis  +  Bacillus methylotrophicus  +  Mcrobacterium testaceum ) and B0: without the use of bacteria.

The impact of varied irrigation water and bacterial strains on soluble carbohydrates

According to Table  4 , in A. cordifolia , the highest soluble carbohydrates content was associated with Mix B3 and irrigation with wastewater (62.82 mg/g FW). On the other hand, urban water in the control treatment (without bacterial application) and Mix B1 in gray water and urban water exhibited similar behavior and resulted in the lowest soluble carbohydrates content (4.18 mg/g FW).

The impact of varied irrigation water and bacterial strains on total protein

In A. cordifolia , the highest total protein content was observed the substrate inoculated with Mix B3 and plants irrigated with wastewater (81.70 mg/g FW). The lowest total protein content was found in the control treatment (without bacterial application) and irrigation with gray water (0.16 mg/g FW). It appears that gray water was not very successful in enhancing cellular sap thickening in A. cordifolia.

The impact of varied irrigation water and bacterial strains on total phenol

The highest total phenolic content was observed in the substrate inoculated with Mix B2 and plants irrigated with wastewater in A. cordifolia (5.11 mg/g FW). However, the lowest content was found in the control treatment (without bacterial application) and irrigation with tap water (1.33 mg/g FW) (Table  4 ).

The impact of varied irrigation water and bacterial strains on anthocyanin

In A. cordifolia , the highest anthocyanin was observed in the substrate inoculated with Mix B3 and plants irrigated with gray water (7.93 μmol/g FW), while the lowest content was found in the control treatment (without bacterial application) and irrigation with urban water (0.52 μmol/g FW). It appears that the inoculation of the substrate with Mix B3 leads to an increase in anthocyanins in Plants of the CAM family.

The impact of varied irrigation water and bacterial strains on MDA

The MDA content was significantly affected by the main effects and interaction effects of different irrigation water types and bacterial strains (Table  3 ). The application of wastewater led to an increase in the MDA content in plants under irrigation (18.90 μmol/g FW), indicating a degree of tissue damage in the plant. In most cases, the MDA content decreased with bacterial treatments. The highest reduction in MDA content was observed, in order, with Mix B3, Mix B2, and irrigation with wastewater (0.82 μmol/g FW, 0.45 μmol/g FW).

The impact of varied irrigation water and bacterial strains on proline

In A. cordifolia , the control treatment (without bacterial application) and Mix B1 resulted in higher proline levels. The presence of Mix B3 and Mix B2 performed better in reducing proline. According to Table  5 , statistically, the three types of water sources had a similar effect and did not show any specific statistical differences. It is important to note that Mix B1 acted similarly to the control treatment in reducing proline stress indicators. It appears that Mix B1 may not be effective in reducing the stress caused by unconventional water sources in A. cordifolia (Table  5 ).

The impact of varied irrigation water and bacterial strains on antioxidant activity

The results obtained from ANOVA indicated that different bacterial strains had a significant effect (at a minimum level of 5%) on the activity of GPX, PPO, CAT, APX, and POX enzymes (Table  3 ). The application of different bacterial strains increased the activity of GPX, PPO, CAT, APX, and POX enzymes (Tables  4 and 5 ). Table 4 showed that in the ice plant, the highest antioxidant activity was observed in Mix B2 and Mix B3 and irrigation with wastewater (53.12% and 54.20%, respectively). In contrast, irrigation with municipal wastewater and tap water in the control treatment (without bacteria) exhibited the lowest antioxidant activity (22.19%).

Bacterial inoculation, in most cases, increased the activity of the four antioxidant enzymes in different irrigation water types (Tables  4 and 5 ). The highest increase in PPO and APX activity was observed, respectively, with Mix B3 and irrigation with wastewater (0.14 unit −1  mg protein) and (4.21 unit −1  mg protein). The lowest level of PPO activity was observed in the control treatment (without bacteria) and Mix B2 and Mix B1 in all three water types. Regarding APX, the lowest activity was associated with the control treatment (without bacteria) and irrigation with wastewater and urban water (Table  4 ).

Comparing the mean data shows that the substrate inoculated with Mix B2 and Mix B3 was associated with the highest GPX activity (0.012 and 0.01 unit −1  mg protein, respectively), while the lowest activity of this enzyme was observed in Mix B1 and the control medium (B0) (0.005 unit −1  mg protein). CAT activity exhibited an opposite trend compared to the other enzymes in this study, where Mix B1 and the control medium (B0) had the highest CAT activity (0.65 unit −1  mg protein) (Table  5 ). Additionally, the highest increase in POX activity was observed with Mix B3 (1.75 unit −1  mg protein), while other bacterial combinations did not significantly improve POX activity and were similar to the control. On the other hand, irrigation of plants with graywater resulted in increased POX activity (1.53 unit −1  mg protein) (Table  5 ).

Water scarcity has necessitated the use of unconventional water sources for agricultural. The issue of water crisis is a significant concern globally, particularly in Middle Eastern countries, where climate change and global warming are exacerbating the problem [ 54 ]. Overcoming this scarcity, farmers resort to irrigating their crops with treated or untreated wastewater due to its low cost and availability.

The impact of varied irrigation water and bacterial strains on morphological factors of Aptenia cordifolia in external green wall

The persistent application of recycled water has the potential to deteriorate soil quality and contribute to the accumulation of toxic heavy metals in the food chain [ 55 ]. This decline in soil quality caused by unconventional water sources results in reduced quality and quantity of crops grown in these soils and can be considered a form of stress for plants [ 56 ]. However, as observed in the present experiment, the presence of the mentioned microbial strains in wastewater not only did not reduce the quality and quantity of the cultivated plants compared to tap water but also improved certain traits (height, growth index, number and length of lateral branches, chlorophyll, leaf content, RWC, RSD and EL). These results are consistent with the findings of Urbano et al. (2017) [ 57 ] regarding increased plant growth due to higher nutrient content compared to freshwater, which can serve as a nutrient source for agriculture and reduce the demand for chemical fertilizers. Furthermore, studies have shown that plant growth-promoting bacteria enhance productivity in various environmental conditions, especially under stressful situations, to improve the detrimental effects of unconventional water on the quantitative and qualitative traits of plants [ 58 ]. In the present study, it was also demonstrated that the application of bacteria in the substrate resulted in improvement of the adverse effects of these unconventional waters, although the performance of all bacterial treatments was not identical. Mix B3 took the top position, followed by Mix B2 in the subsequent ranking. These bacteria appear to stimulate plant growth through nutrient provision, secretion of plant growth hormones, and various other mechanisms associated with PGPR [ 59 ]. In this study, an increase in chlorophyll content in all tested plants resulted in higher photosynthetic productivity, ultimately leading to better growth, increased plant organs, and improved surface coverage by the plant. The extent of surface coverage depends on the growth characteristics of plants, such as height, length, and width, which are influenced by nutrient conditions and substrate moisture [ 60 ]. Jozay et al., [ 61 ] also stated that the composition and content of substrate significantly affect the growth and coverage of plants on external green surfaces. Rapid coverage in vertical green structures, such as green walls, is a desirable trait. The utilization of bacterial combinations has proven effective in enhancing this characteristic. Kumar et al., [ 62 ] reported that the combined use of sewage sludge and plant growth-promoting rhizobacteria resulted in improved performance ( Luffa acutangula (L.) Roxb), with the highest fresh biomass (9.6 ± 0.3 g), growth rate (1.4 ± 0.1 g/day), plant length (15.5 ± 0.3 cm), root length (10.4 ± 0.3 cm) and total chlorophyll (3.2 ± 0.1 mg/g). The findings of this study regarding the increase in growth indices of the studied plants in the presence PGPB and irrigation align with the results of the mentioned researchers.

It can be said that A. cordifolia (Mesembryanthemum crystallinum) has been studied in favorable conditions in all four seasons and due to its greenness, trailing habit, and sufficient coverage to create a desirable visual landscape, it can be used for its visual attractiveness and appeal in combination substrates used in exterior green walls. Kazemi et al., [ 63 ] examined the growth and performance of four plant species in various substrate types in indoor green wall systems, and the results indicated that A. cordifolia is a desirable species for indoor green wall systems. One of the influential factors in the aesthetic performance of green wall systems is the percentage of plant coverage, which is dependent on plant growth indices, and this finding aligns with the study by Kazemi et al., [ 63 ] regarding the application of A. cordifolia and the current research results.

The leaf chlorophyll content has a close relationship with leaf nitrogen content in plants and can serve as an indicator of nitrogen availability in the soil [ 64 ]. Lubbe et al., [ 65 ] highlighted a decrease in chlorophyll content in the leaves of Amarantus dubius and Solanum nigrum under greywater irrigation treatments, suggesting that greywater irrigation may jeopardize nutrient accessibility compared to tap water irrigation. Vajpayee et al., [ 66 ] assessed the chl a/b ratio among plants exposed to different concentrations of tannery effluents and reported a greater reduction in chlorophyll-a compared to chlorophyll-b. The maximum reduction in carotenoid content in Spirodela polyrhiz a was observed after 7 days at a concentration of 75% tannery effluent. The present results support the degradation of carotenoids due to increased effluent in the soil and metal toxicity. In this study, an increase in pigments was observed in wastewater and greywater in the presence of PGPB, a trend that contradicted the significant reduction in photosynthetic pigments observed with effluent application. In present research, greywater and effluent irrigation resulted in increased plant greenness, which may be attributed to the presence of PGPB.

Yadav and Pandey [ 67 ] also observed a positive correlation between leaf relative water content and the concentration of chlorophyll, protein, and ribisco activity. Leaf water content in plants helps maintain physiological water balance under unfavorable environmental conditions. Jozay et al., [ 68 ] stated that there is a direct relationship between leaf water content and resistance to environmental stress conditions in PGPR (Plant Growth-Promoting Rhizobacteria). Plants inoculated with bacterial strains have the ability to modify lateral root system architecture and increase RWC [ 69 ]. Increased relative water content (RWC) in stressed leaves of plants inoculated with plant growth-promoting rhizobacteria (PGPR) has also been reported in other studies [ 70 ]. In general, an increase in the number of lateral roots and root hairs adds surface area for nutrient and water uptake. Enhanced water and nutrient uptake by inoculated roots improve the water status of plants, which in turn can be a primary factor in promoting plant growth [ 71 ].

The RSD impacts the water relations of plants, including the water content of plant tissues and gas exchange in leaves, leading to an increase in the relative water content (RWC) of leaves and a decrease in transpiration and leaf stomatal conductance [ 72 ]. Optimal growth conditions, along with efficient nutrient absorption and transport to plants, enhance the accumulation of ions and organic molecules in leaf vacuoles. This contributes to maintaining the water balance by reducing leaf osmotic potential [ 73 ]. Vishnupradeep et al., [ 74 ] found that MST-PGPB inoculation increased RWC under various stress conditions. PGPB can maintain water potential to prevent water loss by reducing root surface drying. Likewise, Woo et al., [ 75 ] suggested that PGPB inoculation improves RWC through various PGP metabolites, including siderophore production, ACC deaminase activity, phosphate solubilization, and IAA synthesis, resulting in enhanced plant tolerance to abiotic stress, biomass production, and protein content, particularly under non-biological stress conditions [ 76 ]. Substrate inoculated with Mix B3 and Mix B2, leading to an increase in pigments and RWC, as well as a reduction in RSD, confirming the findings of the mentioned researchers.

Impact of varied irrigation water and bacterial strains on Biochemical factors of Aptenia cordifolia in external green wall

Previous studies have reported an increase in growth associated with PGPB and an increase in secondary metabolites and anthocyanins [ 77 ]. The microbial population has a positive correlation with plant biomass, antioxidant enzyme activity, and anthocyanins, while it has a negative correlation with the production of free radicals [ 78 ], indicating the contribution of rhizosphere microbial population in reducing oxidative stress through increased anthocyanin production and enhanced antioxidant enzyme activity. In plants, the greatest decrease in proline and MDA content and the highest increase in antioxidant enzyme activity were observed substrate inoculated with Mix B3 and Mix B2. Singh and Malaviya [ 79 ] reported that due to acute toxicity of chromium present in wastewater, anthocyanins in the effluent were below detectable levels after 4 days. The increase in anthocyanins in this study, despite repeated use of wastewater and graywater, could be attributed to the presence of bacterial strains. The results of this study, which showed an increase in anthocyanins through substrate inoculated with Mix B3 in irrigated vegetable plants with recycled water, are consistent with the reports of Pagnani et al., [ 80 ] and Kumar et al., [ 62 ].

The result indicates that the increase in proline was in response to watering the plants with municipal wastewater. Proline may act as an osmolyte under stress conditions and increase proline activities thereby minimizing the side effects of stress [ 81 ]. In other findings, Islam et al., [ 82 ] reported that proline acts as a growth regulator and also protects cells against ROS accumulation. It is thought that proline is able to reduce the negative effects of cadmium toxicity on plant growth in plant tissues and reduce oxidative stress [ 83 ]. In this study, the use of growth-promoting bacteria reduced the adverse effects of wastewater irrigation, and it is worth noting Mix B3 and Mix B2 was more successful in reducing the amount of proline than the other two treatments. As a result, it can be said the use of growth-promoting bacteria has put the plant in a favorable condition by increasing the growth index and biomass and has reduced stress and proline content.

Proline and free phenols are non-enzymatic antioxidants that support organisms in unfavorable conditions by reducing the undesirable effects of reactive oxygen species (ROS) [ 84 ]. Environmental stresses, such as heavy metal exposure in unconventional water sources and urban gardening, have been documented to increase ROS production. Antioxidant molecules and enzymes play a crucial role in detoxifying ROS in plant cells. Antioxidant compounds, including proline and phenols, inhibit oxidation and play vital roles in stress responses [ 85 ]. After irrigating plants with wastewater in the control treatment without bacteria, the levels of these compounds significantly increased, confirming the findings of the aforementioned researchers [ 86 ]. It has been reported that the content of phenolic compounds tends to rise when plants are subjected to heavy metal stress, as phenolic compounds act as scavengers of reactive oxygen species and metal chelators. In contrast, the content of free proline and total phenols considerably decreased in response to our treatments compared to plants irrigated with water in the presence of growth-promoting bacteria. The highest reduction was observed in free proline when substrate inoculated with Mix B3 and Mix B2. These results indicate the role of these treatments in reducing the toxic effects of wastewater on plants. Our findings align with previous studies by Liu et al., [ 12 ], which reported that wastewater treatment with bacteria significantly reduces the stress caused by metal contaminants in wastewater. Regarding total phenols, the control treatment in urban water showed the highest reduction. The decrease in phenolic content may be attributed to oxidative stress, while a further increase in the content of phenolic compounds (as non-enzymatic antioxidants) was observed when substrate inoculated with Mix B3 and wastewater. These results are consistent with the reports of Khodamoradi et al., [ 87 ] regarding the reduction of phenols under stress conditions.

Malondialdehyde (MDA) is a byproduct of lipid oxidation and is responsible for cellular membrane damage, inducing alterations in membrane radical properties. These changes ultimately culminate in cell death [ 88 ]. In this study, the MDA content increased in response to wastewater irrigation in the control treatment B0. This may be attributed to the oxidative system's inability to reduce ROS levels, thereby failing to prevent damage to the cell membrane. This result is also supported by Yildirim et al., [ 89 ], who demonstrated that irrigation with contaminated water generally increased the MDA in plants. In contrast, the MDA content significantly decreased in response to treatments B3 and B2 compared to plants irrigated with wastewater in treatment B0. This may be due to the reduction in ROS production and the enhancement of the antioxidant system and repair mechanisms. These results are supported by a recent study by Malik et al., [ 90 ].

Some studies have found that PGPB strains can enhance plant growth and mitigate the negative effects of various stresses on plant growth. Enhanced growth associated with PGPB and increased secondary metabolites, carbohydrates, total protein, and antioxidant potential have been reported in Astragalus mongholicus [ 91 ]. Plant antioxidant enzymes GPX, PPO, CAT, APX, POX act as the first line of defense for tolerating unfavorable conditions. These enzymes stimulate the detoxification of ROS and reduce the detrimental effects of non-biological stress [ 91 ]. Another finding is that in most cases, bacterial treatments increased the activity of antioxidant enzymes. The greatest increase in biomass production and antioxidant enzyme activity was observed in substrate inoculated with Mix B3 and Mix B2 in wastewater, indicating a significant contribution of bacterial inoculation in activating antioxidant enzymes and promoting plant growth. The further increase in GPX, PPO, APX, and POX activities resulting from the most effective bacterial inoculation may indicate the major role of these enzymes in improving plant biomass through substrate inoculated with Mix B3.

Under wastewater stress, plants exposed to a stronger antioxidant system are less exposed to free radicals, leading to lower MDA production. High antioxidant activity in stress-tolerant plants may indicate their ability to neutralize harmful oxidants and maintain their growth and productivity at a normal level. Therefore, it can be concluded that these bacteria, in addition to promoting growth, provide new perspectives for the development of biofertilizers to alleviate environmental stresses. It can be inferred that greater protection was achieved under wastewater irrigation when bacterial strains were applied. Ultimately, Mix B3 and Mix B2 appear to be recommendable to farmers due to their economic and environmental compatibility, aiding in mitigating the adverse effects of recycled water use in urban gardening.

Phytoremediation is an environmentally friendly and cost-effective alternative to removing pollution from soil. Due to the non-degradable chemical nature of heavy metals in soil, we need to understand the function of antipollution facilitators such as plant growth promoting bacteria to improve or facilitate the removal of heavy metals by plants [ 92 , 93 ]. Heavy metal tolerant PGPR have also increased plant growth in heavy metal contaminated soil in recent years. These PGPR have successfully played an important role in promoting plant growth while reducing toxicity or damage to plants exposed to stress produced by various heavy metals in soil [ 94 ]. PGPR produce plant growth regulators, phytohormones, and various secondary metabolites that promote plant growth and reduce heavy metal toxicity. Various mechanisms are employed by PGPR to enhance plant growth under heavy metal stress. Many of them lead to the reduction of heavy metal toxicity [ 95 ].

Known mechanisms by which PGPR can benefit plants under a variety of stresses include: (1) bioremediation of heavy metal-contaminated soils by sequestering toxic heavy metal species and improving soil structure (by bacterial exopolysaccharides); (2) the synthesized enzyme ACC (1aminocyclopropane-1-carboxylate) deaminase, an enzyme that is involved in the reduction of stress-induced ethylene levels in the roots of growing plants; (3) supply of N2 to the plant through biological nitrogen fixation. (4) Production of siderophores. (5) production of phytohormones (such as ABA (abscisic acid), GA (gibberellic acid), auxin, for example, indole-3-acetic acid (IAA) and CK (cytokinin); (6) control of plant pathogens by various mechanisms such as the production of extracellular enzymes that hydrolyze the fungal cell wall, competition for nutrients in the rhizosphere, induction of systemic resistance (ISR), and production of antibiotics and siderophores; (7) dissolution and mineralization of nutrients, especially inorganic phosphate; and (8) improving resistance to abiotic stresses [ 92 , 96 ].

Therefore, growth-promoting bacteria, as living organisms, play an important role in plant nutrition balance. They can help improve their nutrition by reducing environmental stress and increasing the digestion and absorption of nutrients in plants. Using growth-promoting bacteria as an alternative to chemical fertilizers for soil repair and plant nutrition can be an effective method. These bacteria are commonly known as biologicals or biofertilizers. The use of microorganisms can reduce the stresses in various plants in stressed soils, thus opening a potential and promising strategy for sustainable agriculture [ 97 ].

In a recent report by Jozay [ 92 ], it was suggested that PGPR could be used as a bioremediation method for soils contaminated with toxic metals. PGPR contain bacteria that are rhizospheric and endophytic and facilitate bioremediation. Plants accumulate heavy metals in the roots and reduce their transfer to other parts of the plant. These microorganisms provide benefits to plants by providing nutrients and reducing the harmful effects of pollutants.

As pointed out by FAO [ 98 ], promoting safe and healthy agricultural practices in urban environments is essential to achieve sustainable urban development. At the same time, environmental pollutants in cities should be controlled while using city resources and inputs. The purpose of this study is to investigate the potential of green wall systems to produce horticultural materials that are aligned with the goals of sustainable urban development.

The results of this study indicated that plant irrigation using wastewater significantly increased the content of free proline, total phenol, GPX, PPO, CAT, APX, POX and MDA compared to the control. On the other hand, the PGPB reduced the oxidative damage effects of wastewater irrigation by increasing antioxidant activity and enhancing GPX, PPO, APX, POX. As a result, the content of free proline, total phenol, and MDA decreased significantly, while carbohydrates, proteins, and anthocyanins increased in the combined bacterial treatments. It can be concluded that better protection under wastewater irrigation was achieved using our treatments. Ultimately, Mix B3 and Mix B2 economically and environmentally friendly, can be recommended to farmers for mitigating the adverse effects of reclaimed water on urban gardening. Substrate inoculated with Mix B3, wastewater irrigation, and subsequent greywater irrigation, affects water availability, nutrient availability, and physiological traits of ornamental plants, including RWC, RSD, and plant pigment properties, leading to freshness and increased greening of the plant. It also has a considerable effect on morphological traits such as height, growth index, lateral branches, and surface coverage, and there are significant differences in plant growth improvement among the combined substrate inoculated with different bacteria types. This treatment can be used to improve the qualitative and morphological traits of ornamental plants used in external green walls under similar climatic conditions to Mashhad city. Water scarcity is expected to transform the reuse of reclaimed water for irrigation into a widespread and common practice globally. The incorporation of growth-promoting bacteria can serve as aneffective amendment for mitigating the toxicity associated with wastewater, facilitating improved plant growth in areas irrigated with wastewater and greywater. As a sustainable solution, green walls have the potential to mitigate water, soil, and air pollution, thereby enhancing environmental sustainability. By incorporating technology into multifunctional green walls, we can take a significant step towards sustainable urban development. Living green walls are not only introduced as effective tools for urban space management but also as instruments for enhancing the climate resilience of cities. The outcome of this research is the registration of a water recycling system utilizing nature-based methods, with Patent number 110287.

Future Prospects

Future research should focus on understanding the molecular mechanisms underlying these beneficial interactions and exploring the application of PGPB in other plant species and environmental conditions. Additionally, long-term field studies are needed to evaluate the sustainability and economic viability of using PGPB in urban green wall systems.

Availability of data and materials

The data that support the findings of this study are available from the corresponding author upon ‎reasonable request.‎

Abbreviations

Strain bacteria 0–3

Without the use of bacteria

Psedoumonas flucrecens  +  Azosporillum liposferum  +  Thiobacillus thioparus  +  Aztobactor chorococcum

Paenibacillus polymyxa  +  Pseudomonas fildensis  +  Bacillus subtilis  +  Achromobacter xylosoxidans  +  Bacillus licheniform

Pseudomonas putida  +  Acidithiobacillus ferrooxidans  +  Bacillus velezensis  +  Bacillus subtilis  +  Bacillus methylotrophicus  +  Mcrobacterium testaceum

Organic carbon

Organic matter

Field capacity

Permanent wilting poin

V´asquez A, Giannotti E, Gald´amez E, Vel´asquez P, Devoto, C. Green Infrastructure planning to tackle climate change in Latin American cities. In C. Henriquez, & H. Romero (Eds.), Urban Climate in Latin-American Cities. 2019;32–354. Springer.

Langemeyer J, Connolly JJT. Weaving notion of justice into urban ecosystem service research and practice. Environ Sci Policy. 2020;109:1–14.

Article   Google Scholar  

Artmann M, Sartison K. The role of urban agriculture as a nature-based solution: A review for developing a systemic assessment framework. Sustainability. 2018;10:1937.

Cojocariu M, Chelariu EL, Chiruţă C. Study on Behavior of Some Perennial Flowering Species Used in Vertical Systems for Green Facades in Eastern European Climate. Appl Sci. 2022;12(1):474.

Article   CAS   Google Scholar  

Mahmoudi Jabri K, Nolde E, Ciroth A, Bousselmi L. Life cycle assessment of a decentralized greywater treatment alternative for non-potable reuse application. Int J Environ Sci Tech. 2020;17(1):433–44.

Jozay M, Rabbani Khairkhah M, Kazemi F. The impact of humic acid solutions and types of growing media on some morphophysiological and biochemical features of Syngonium sp and Pothos sp plants in interior green wall conditions. Plant Arshive. 2021;21:2240–52.

Google Scholar  

Savi T, Casolo V, Luglio J, Bertuzzi S, Trifilo P, Lo Gullo M.A, ardini, A. Species-specific reversal of stem xylem embolism after a prolonged drought correlates to endpoint concentration of soluble sugars. Plant Physiology and Biochemistry. 2016; 106:198–207.

Jozay M. Rabbani Khairkhah SM. Vertical green systems (green wall). Iran Publishing House (Aristo Publications). 2019.

Li X, Chang Z, Lian X, Meng G, Ma J, Guo R, Wang Y. Phytoremediation of cadmium contaminated alkaline soil using the ornamental hyperaccumulator Mirabilis jalapa L. enhanced by double harvesting: a field study. Environmental Science and Pollution Research. 2022; 1–8.

Chouhan S, Agrawal L, Prakash A. Amelioration in traditional farming system by exploring the different plant growth-promoting attributes of endophytes for sustainable agriculture. Arch Microbiol. 2022;204:151.

Article   CAS   PubMed   Google Scholar  

Singh P, Singh RK, Zhou Y, Wang J, Jiang Y, Shen N, Wang Y, Yang L, Jiang M. Unlocking the strength of plant growth promoting Pseudomonas in improving crop productivity in normal and challenging environments: a review. Journal of Plant Interactions. 2022;17(1):220–38.

Liu Sh, Liu H, Chen R, Ma Y, Yang B, Chen Zh, Liang Y, Fang J, Xiao Y. Role of Two Plant Growth-Promoting Bacteria in Remediating Cadmium-Contaminated Soil Combined with Miscanthus floridulus (Lab). Plants. 2021;10:912.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Xiao Y, Liu H, Chen R, Liu S, Hao X, Fang J. Heteroauxin-producing bacteria enhance the plant growth and lead uptake of Miscanthus floridulus (Lab.). International Journal of Phytoremediation. 2022; 1–8.

Gavrilescu M. Enhancing phytoremediation of soils polluted with heavy metals. Curr Opin Biotechnol. 2022;74:21–31.

Zanin G, Bortolini L, Borin M. Assessing stormwater nutrient and heavy metal plant uptake in an experimental bioretention pond. Land. 2018;7:150.

Zhuang J, Qiao L, Zhang X, Su Y, Xia Y. Effects of visual attributes of flower borders in urban vegetation landscapes on aesthetic preference and emotional perception, Int. J. Environ. Res. Publ. Health. 2021; 18(17):9318. https://doi.org/10.3390/ijerph18179318 , 3.

Brix H, Arias CA. The use of vertical flow constructed wetlands for on-site treatment of domestic wastewater: new Danish guidelines. Ecol Eng. 2005;25:491–500. https://doi.org/10.1016/j.scitotenv.2015;09.154 .

Silvestri G, Aliata F. El paisaje como cifra de armonía, Ediciones Nueva Visi´on, Buenos Aires. 2001.

Noutsopoulos C, Andreadakis A, Kouris N, Charchousi D, Mendrinou P, Galani A, Mantziaras I, Koumaki E. Greywater characterization and loadings, physicochemical treatment to promote onsite reuse. J Environ Manage. 2018;216:337–46.

Soudi B. Appui a La Promotion De La Reutilisation Des Eaux Usees Par Le Renforcement Des Aspects Institutionnels, Reglementaires Et Financieres, Ainsi Que Des Démarches Participatives, Des Mesures Incitatives Et La Sensibilisation (Activité n_EFS-MO-2). 2018. Available online: https://www.swim-h2020.eu/wp-content/uploads/2018/09/SWIM-H2020-EFS-MO-2 .

Rosa PAL, Galindo FS, Oliveira CE dS, Jalal A, Mortinho ES, Fernandes GC, Marega EMR, Buzetti S, Teixeira Filho MCM. Inoculation with Plant Growth-Promoting Bacteria to Reduce Phosphate Fertilization Requirement and Enhance Technological Quality and Yield of Sugarcane. Microorganisms. 2022;10:192.

National Centers for Climatology. Climate information of the city of Mashhad, 2019. available on: https://www.ncdc.noaa.gov/climate-information , access date: 2019.07.25.

F.L.L. Guidelines for the planning, construction and maintenance of green roofing, Forschungsgesellschaft Landschaftsentwic. Bonnklung Landschaftsbau e.V.-FLL, Bonn. 2018.

Khorshid M, Oustan S, Najafi N, Khataee A. Kinetic characterization of hexavalent chromium stabilization in contaminated soils amended with cocopeat. Arab J Geosci. 2020;13(11):1–12.

Wang F, Miao L, Wang Y, Zhang M, Zhang H, Ding Y, Zhu W. Using cow dung and mineral vermireactors to produce vermicompost for use as a soil amendment to slow Pb2+ migration. Appl Soil Ecol. 2022;170: 104299.

Gong Y, Zhang G, Hao Y, Nie L. Enrichment Evaluation of Heavy Metals from Stormwater Runoff to Soil and Shrubs in Bioretention Facilities. Water. 2022;14(4):638.

Bai G, Luo F, Zou Y, Liu Y, Wang R, Yang H, Liu Z, Chang J, Wu Z, Zhang Y. Effects of vermiculite on the growth process of submerged macrophyte Vallisneria spiralis and sediment microecological environment. J Environ Sci. 2022;118:130–9.

Yu G, Wang G, Chi T, Du C, Wang J, Li P, Zhang Y, Wang S, Yang K, Long Y, Chen H. Enhanced removal of heavy metals and metalloids by constructed wetlands: A review of approaches and mechanisms. Sci Total Environ. 2022;821:153516.

Ekhlasi Nia A. Evaluation of arsenic and iron transport from sediments of a potable water treatment plant wastewater stabilization pond system. M.Sc. Thesis, In the Department of Civil, Geological and Environmental Engineering, Canada. 2021.

Huang Q, Qiu W, Yu M, Li S, Lu Z, Zhu Y, Kan X, Zhuo R. Genome-Wide Characterization of Sedum plumbizincicola HMA Gene Family Provides Functional Implications in Cadmium Response. Plants. 2022;11(2):215.

Ju W, Liu L, Fang L, Cui Y, Duan Ch, Wu H. Impact of co-inoculation with plant-growth-promoting rhizobacteria and rhizobium on the biochemical responses of alfalfa-soil system in copper contaminated soil. Ecotoxi & Environl Safety. 2019;167:218–26.

Esfandiari M, Hakimzadeh Ardakani MA, Sodaiezadeh H. Investigation of heavy metal Contamination in the ornamental plants of green spaces in the city of Yazd. Journal of Ornamental Plants. 2020;10(4):213–22.

Allison LE, Moodie CD. Carbonate. p. 1379–1400. In C.A. Black et al. (Ed.) Methods of Soil Analysis. Part 2. 2nd Ed. Agron. Monogr. 9. ASA, CSSA, and Soil Science Society of America, Madison, WI. 1965.

Chen Y, Inbar Y, Hadar Y. Composted agricultural wastes as potting media for ornamental plants. Soil Sci. 1988;145(4):298–303.

Salter PJ, Haworth F. The available water capacity of a sandy loam soil: I. A critical comparison of methods of determining the moisture content of soil at field capacity and at the permanent wilting percentage. Journal of Soil Science. 1961;12:326–34.

Walkley A, Black TA. An examination of Deglijareff method for determining soil organic matter and a proposed modification of the choromic acid titration method. Soil Sci. 1934;37:29–38.

Mozaffari S, Khorasaninejad S, Gorgini SH. The effects of irrigation regimes and humic acid on some of physiological and biochemical traits of Common Purslane in greenhouse. Journal of Crops Improvement (Journal of Agriculture). 2017;19(2):401–16.

Gorgini Shabankareh H, Khorasaninejad S, Sadeghi M, Tabasi AR. The effects of irrigation periods and humic acid on morpho- physiological and biochemical traits of Thyme ( Thymus vulgaris ). Journal of Plant Ecophysiological Research. 2018;13(51):67–82.

Dere S, Günes T, Sivaci R. Spectrophotometric determination of chlorophyll-a, b and total carotenoid contents of some algae species using different solvents. Turk J Bot. 1998;22:7–13.

Hossain MI, Khatun A, Talukder MSA, Dewan MMR, Uddin MS. Effect of drought on physiology and yield contributing characters of sunflower. Bangladesh J Agric Res. 2010;35(1):113–24.

Samar Raza MA, Saleem MF, Khan IH, Jamil M, Ijaz M, Khan MA. Evaluating the drought stress tolerance efficiency of wheat (Triticum aestivum L) Cultivars. Russ J  Agri Socio-Econ Sci. 2012;12(12):41–6.

Sairam RK, Srivastava GC. Changes in antioxidant activity in sub-cellular fractions of tolerant and susceptible wheat genotypes in response to long term salt stress. Plant Sci. 2002;162:897–904.

Sadasivam S, Manickam A. In: Biochemical Methods for Agricultural Sciences. New Delhi: Wiley Eastern Ltd.; 1992. p. 184–5.

Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem. 1976;72(1–2):248–54.

Nadernejad N, Ahmadimoghadam A, Hossyinifard J, Poorseyedi S. Study of the rootstock and cultivar effect in PAL activity, production of phenolic and flavonoid compounds on flower, leaf and fruit in Pistachio ( Pistacia vera L). Journal of Plant Biology. 2013;15:95–110.

Singleton VL, Rossi JA. Colorimetry of total phenolics with phosphomolybdicphosphotungstic acid reagents. Am J Enol Vitic. 1965;16:144–58.

Heath RL, Packer L. Photoperoxidation in isolated chloroplasts: I. Kinetics and stoichiometry of fatty acid peroxidation. Arch Biochem  Biophys. 1968;125(1):189–98.

Moon JH, Terao J. Antioxidant activity of caffeic acid and dihydrocaffeic acid in lard and human lowdensity lipoprotein. J Agric Food Chem. 1998;46:5062–5.

Jia L, Xu W, Li W, Ye N, Liu R, Shi L, Bin Rahman AR, Fan M, Zhang J. Class III peroxidases are activated in proanthocyanidin deficient Arabidopsis thaliana seeds. Ann Bot. 2013;111:839–47.

Asada K, Takahashi M. Production and scavenging of active oxygen in chloroplasts, In: D.J. Kyle, B. Osmond & C.J. Arntzen (Eds), photoinhibition, Elsevier Amsterdam. 1987;227–287.

Jebara S, Jebara M, Limam F, Aouani ME. Changes in ascorbate peroxidase, catalase, guaiacol peroxidase and superoxide dismutase activities in common bean (Phaseolus vulgaris) nodules under salt stress. J Plant Physiol. 2005;162(8):929–36.

Chandlee JM, Scandalios JG. Analysis of variants affecting the catalase developmental program in maize scutellum. Theor Appl Genet. 1984;69:71–7.

Holy MC. Indole acetic acid oxidase: a dual catalytic enzyme. Plant Phisiology. 1972;50:15–8.

Rezapour S, Kouhinezhad P, Samadi A. The potential ecological risk of soil trace metals following over five decades of agronomical practices in a semi-arid environment. Chem Ecol. 2017;34:68–78.

Zemanová V, Trakal L, Ochecova P, Száková J, Pavlíková D. A model experiment: competitive sorption of Cd, Cu, Pb and Zn by three different soils. Soil and Water Research. 2014;9:97–103.

Kanwal A, Farhan M, Sharif F, Hayyat MU, Shahzad L, Ghafoor GZ. Effect of industrial wastewater on wheat germination, growth, yield, nutrients and bioaccumulation of lead. Sci Rep. 2020; 10:11352–11361. https://doi.org/10.1038/s41598-020-68208-7 .

Urbano VR, Mendonça TG, Bastos RG, Souza CF. Effects of treated wastewater irrigation on soil properties and lettuce yield. Agric Water Manag. 2017;181:108–15.

Ramakrishna W, Yadav R, Li K. Plant growth promoting bacteria in agriculture: two sides of a coin. Appl Soil Ecol. 2019;138:10–8.

Chiappero J, Cappellari Ld. R Sosa Alderete LG, Palermo TB, Banchio E. Plant growth promoting rhizobacteria improve the antioxidant status in Mentha piperita grown under drought stress leading to an enhancement of plant growth and total phenolic content. Industrial Crops and Products. 2019;139:111553.

Rabbani Khairkhah SM, Kazemi F. The effect of the type of cultivation substrate and humic acid on some morphophysiological traits and water consumption of scorpion tail sedum ( Carpobrotus edulis L) on a green roof system. Journal of Plant Environmental Physiology. 2022;17(66):1–23.

Jozay M, Kazemi F, Fotovat A. The performance of cover plant ( Frankenia thymifolia ) and its effect on the physicochemical properties of the recycled substrates in four different seasons in external green wall systems. Journal of Plant Environmental Physiology. Ready to print publishing. 2022.

Kumar V, Eid EM, Al-Bakre DA, Abdallah SM, Širić I, Andabaka Ž, Kumar P, Goala M, Adelodun B, Singh J, Kumari S. Combined use of sewage sludge and plant growth-promoting rhizobia improves germination, biochemical response and yield of ridge gourd ( Luffa acutangula (L) Roxb) under field conditions. Agriculture. 2022;12(2):173.

Kazemi F, Rabbani M, Jozay M. Investigating plant and air-quality performances of an internal greenwall system under hydroponic conditions. Environmental Management. 2020;275:111230.

Marenco R, Antezana-vera S, Nascimento H. Relationship between specific leaf area, leaf thickness, leaf water content and SPAD-502 readings in six Amazonian tree spe¬cies. Photosynthetica. 2009;47:184–90.

Lubbe E, Rodda N, Naidoo S. Effects of greywater irrigation on germination, growth and photosynthetic characteristics in selected African leafy vegetables. Water SA. 2016;42(2):203–12.

Vajpayee P, Sharma SC, Tripathi RC, Rai UN, Yunus M. Bioaccumulation of chromium and toxicity to photosynthetic pigments, nitrate reductase activity and protein content of Nelumbo nucifera Gaertin. Chemosphere. 1999;39:2159–69.

Yadav R, Pandey P. Assessment of air pollution tolerance index (APTI) and anticipated performance index (API) of roadside plants for the development of greenbelt in urban area of bathinda city, punjab, India. Bull Environ Contam Toxicol. 2020;105:906–14.

Jozay M, Kazemi F, Fotovat A. Evaluating the environmental performance of the growing media in a green wall system in a dry climate region. Desert. 2019;20:217–30.

Bertrand A, Gatzke C, Bipfubusa M, Levesque V, Chalifour FP, Claessens A, Rocher S, Tremblay GF, Beauchamp CJ. Physiological and biochemical responses to salt stress of alfalfa populations selected for salinity tolerance and grown in symbiosis with salt-tolerant rhizobium. Agronomy. 2020;10(4):569.

Zilaie MN, Arani AM, Etesami H, Dinarvand M, Dolati A. Halotolerant plant growth-promoting rhizobacteria-mediated alleviation of salinity and dust stress and improvement of forage yield in the desert halophyte seidlitzia rosmarinus. Environ Exp Bot. 2022;201: 104952.

Chieb M, Gachomo EW. The role of plant growth promoting rhizobacteria in plant drought stress responses. BMC Plant Biol. 2023;23(1):407.

Article   PubMed   PubMed Central   Google Scholar  

Anjum SA, Xie XY, Wang L, Saleem MF, Man C, Lei W. Morphological, physiological and biochemical responses of plants to drought stress. Afr J Agric Res. 2011;6:2026–32.

Aeron A, Khare E, Jha CK, Meena VS, Aziz SMA, Islam MT, Kim K, Meena SK, Pattanayak A, Rajashekara H, Dubey RC. Revisiting the plant growth-promoting rhizobacteria: Lessons from the past and objectives for the future. Arch Microbiol. 2020;202(4):665–76.

Vishnupradeep R, Bruno LB, Taj Z, Karthik C, Challabathula D, Kumar A, Freitas H, Rajkumar M. Plant growth promoting bacteria improve growth and phytostabilization potential of Zea mays under chromium and drought stress by altering photosynthetic and antioxidant responses. Environ Technol Innov. 2022;25: 102154.

Woo OG, Kim H, Kim JS, Keum HL, Lee KC, Sul WJ, Lee JH. Bacillus subtilis strain GOT9 confers enhanced tolerance to drought and salt stresses in Arabidopsis thaliana and Brassica campestris. Plant Physiol Biochem. 2020;148:359–67.

Rajkumar M, Bruno LB, Banu R. Alleviation of environmental stress in plants: The role of beneficial Pseudomonas spp. Crit Rev Environ Sci Technol. 2017;47:372–407.

Lin Y, Zhang H, Li P, Jin J, Li Z. The bacterial consortia promote plant growth and secondary metabolite accumulation in Astragalus mongholicus under drought stress. BMC Plant Biol. 2022;22(1):1–14.

Mohsin SM, Hasanuzzaman M, Parvin K, Fujita M. Pretreatment of wheat (Triticum aestivum L) seedlings with 2,4-D improves tolerance to salinity-induced oxidative stress and methylglyoxal toxicity by modulating ion homeostasis, antioxidant defenses, and glyoxalase systems. Plant Physiol Biochem. 2020;152:221–31.

Singh A, Malaviya P. Chromium phytoaccumulation and its impact on growth and photosynthetic pigments of Spirodela polyrrhiza (L) Schleid on exposure to tannery effluent. Environmental Sustainability. 2019;2:157–66.

Pagnani G, Pellegrini M, Galieni A, D’Egidio S, Matteucci F, Ricci A, Stagnari F, Sergi M, Lo Sterzo C, Pisante M, Del Gallo M. Plant growth-promoting rhizobacteria (PGPR) in Cannabis sativa ‘Finola’ cultivation: an alternative fertilization strategy to improve plant growth and quality characteristics. Ind Crops Prod. 2018;123:75–83.

Khamsuk O, Sonjaroon W, Suwanwong S, Jutamanee K, Suksamrarn A. Effects of 24-epibrassinolide and the synthetic brassinosteroid mimic on chili pepper under drought. Acta Physiol Plant. 2018;40:1–12.

Islam MM, Hoque MDA, Okuma E, Banu MNA, Shimoishi Y, Nakamura Y, Murata Y. Exogenous proline and glycine betaine increase antioxidant enzyme activities and confer tolerance to cadmium stress in cultured tobacco cells. J Plant Physiol. 2009;166:1587–97.

Radic S, Babic M, Skobic D, Roje V, Pevalek-Kozlina B. Ecotoxicological effects of aluminum and zinc on growth and antioxidants in Lemna minor L. Ecotoxicol Environ Saf. 2010;73:336–42.

Zandi P, Schnug E. Reactive oxygen species, antioxidant responses and implications from a microbial modulation perspective. Biology (basel). 2022;11:1–30. https://doi.org/10.3390/biology11020155 .

Racchi M. 2013) Antioxidant defenses in plants with attention to prunus and citrus spp. Antioxidants 2:340–369. https://doi.org/10.3390/antiox2040340 .

Pratyusha S. Phenolic compounds in the plant development and defense: an overview. In: Plant stress physiology-perspectives in agriculture. Plant Stress Physiology-Perspectives in Agriculture. 2022;125–185.

Khodamoradi S, Sagharyan M, Samari E, Sharifi M. Changes in phenolic compounds production as a defensive mechanism against hydrogen sulfide pollution in Scrophularia striata. Plant Physiol Biochem. 2022; 177:23–31. https://doi.org/10.1016/j.plaphy.2022.02.013 .

Sharma P, Jha AB, Dubey RS, Pessarakli M. Reactive oxygen species, oxidative damage, and antioxidative defense mechanism in plants under stressful conditions. J Bot. 2012;1–26. https://doi.org/10.1155/2012/217037 .

Yildirim E, Ekinci M, Turan M, Agar G, Ors S, Dursun A, Kul R, Balcl T. Impact of cadmium and lead heavy metal stress on plant growth and physiology of rocket (Eruca sativa L). J Agric Nat. 2019; 22:843–850. https://doi.org/10.18016/ksutarimdoga.vi.548626 .

Malik Z, Afzal S, Dawood M, Abbasi GH, Khan MI, Kamran M, Zhran M, Hayat MT, Aslam MN, Rafay M. Exogenous melatonin mitigates chromium toxicity in maize seedlings by modulating antioxidant system and suppresses chromium uptake and oxidative stress. Environ Geochem Health. 2022; 44:1451–1469. https://doi.org/10.1007/s10653-021-00908-z .

Rai GK, Mushtaq M, Bhat BA, Kumar RR, Singh M, Rai PK. Reactive oxygen species: friend or foe. In: Thermotolerance in crop plants. Springer Nature Singapore, Singapore. 2022;129–162.

Jozay M. Combining Green Space and Urban Gardening in External Green Wall under the Influence of Growth-Promoting Bacteria and Types of Recycled Water. Gorgan, Iran: Gorgan University of Agricultural Sciences and Natural Resources; 2024.

Jozay M, Zarei H, Khorasaninejad S, Miri T. Maximising CO2 Sequestration in the City: The Role of Green Walls in Sustainable Urban Development. Pollutants. 2024;4(1):91–116.

Etesami H. Bacterial mediated alleviation of heavy metal stress and decreased accumulation of metals in plant tissues: mechanisms and future prospects. Ecotoxicol Environ Saf. 2018;147:175–91.

Etesami H, Maheshwari DK. Use of plant growth promoting rhizobacteria (PGPR) with multiple plant growth promoting traits in stress agriculture: Action mechanisms and future prospects. Ecotoxicol Environ Saf. 2018;156:225–46.

Glick BR. Bacteria with ACC deaminase can promote plant growth and help to feed the world. Microbiol Res. 2014;169:30–9.

Jozay M, Zarei H, Khorasaninejad S, Miri T. The environmental function of external green elements in facade architecture. 8th.International Conference on Researches in Science & Engineering & 5th.International Congress on Civil, Architecture and Urbanism in Asia 24 August. 2023, Kasem Bundit University, Bangkok, Thailand.

FAO. New Standards to Curb the Global Spread of Plant Pests and Diseases; UN Food and Agriculture Organization: Rome, Italy. 2019.

Download references

Acknowledgements

Not applicable.

This work was supported by Gorgan University of Agricultural sciences and Natural Resources] for their support and resources in conducting this research under a grant number 480. The facilities and expertise provided by the university have been instrumental in the successful completion of this study.

Author information

Authors and affiliations.

Horticultural Sciences Department, Faculty of Plant Production, Gorgan University of Agricultural Sciences and Natural Resources, Gorgan, Iran

Mansoure Jozay, Hossein Zarei & Sarah Khorasaninejad

School of Chemical Engineering, University of Birmingham, Birmingham, UK

You can also search for this author in PubMed   Google Scholar

Contributions

Mansoure Jozay: Contributed to the original writing and formatting of the article. Hossein Zarei: Assisted in manuscript discussion, review and corrections. Sarah Khorasaninejad: Provided input on article formatting. Taghi Miri: Reviewed the manuscript.

Corresponding authors

Correspondence to Hossein Zarei or Taghi Miri .

Ethics declarations

Ethics approval and consent to participate.

This research meets all the ethical guidelines, including adherence to the legal requirements of my country. The authors voluntarily agree to participate in this research study.

Consent for publication

Competing interests.

The authors declare no competing interests.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Jozay, M., Zarei, H., Khorasaninejad, S. et al. Exploring the impact of plant growth-promoting bacteria in alleviating stress on Aptenia cordifolia subjected to irrigation with recycled water in multifunctional external green walls. BMC Plant Biol 24 , 802 (2024). https://doi.org/10.1186/s12870-024-05511-9

Download citation

Received : 12 February 2024

Accepted : 12 August 2024

Published : 24 August 2024

DOI : https://doi.org/10.1186/s12870-024-05511-9

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Antioxidant activity
• External green wall
• Ornamental plant
• Oxidative stress

BMC Plant Biology

ISSN: 1471-2229

• General enquiries: [email protected]

• Methodology
• Open access
• Published: 20 August 2024

Prediction accuracy and repeatability of UAV based biomass estimation in wheat variety trials as affected by variable type, modelling strategy and sampling location

• Daniel T. L. Smith 1 ,
• Qiaomin Chen 1 ,
• Sean Reynolds Massey-Reed 2 ,
• Andries B. Potgieter 2 &
• Scott C. Chapman 1  

Plant Methods volume  20 , Article number:  129 ( 2024 ) Cite this article

129 Accesses

Metrics details

This study explores the use of Unmanned Aerial Vehicles (UAVs) for estimating wheat biomass, focusing on the impact of phenotyping and analytical protocols in the context of late-stage variety selection programs. It emphasizes the importance of variable selection, model specificity, and sampling location within the experimental plot in predicting biomass, aiming to refine UAV-based estimation techniques for enhanced selection accuracy and throughput in variety testing programs.

The research uncovered that integrating geometric and spectral traits led to an increase in prediction accuracy, whilst a recursive feature elimination (RFE) based variable selection workflowled to slight reductions in accuracy with the benefit of increased interpretability. Models, tailored to specific experiments were more accurate than those modelling all experiments together, while models trained for broad-growth stages did not significantly increase accuracy. The comparison between a permanent and a precise region of interest (ROI) within the plot showed negligible differences in biomass prediction accuracy, indicating the robustness of the approach across different sampling locations within the plot. Significant differences in the within-season repeatability (w 2 ) of biomass predictions across different experiments highlighted the need for further investigation into the optimal timing of measurement for prediction.

Conclusions

The study highlights the promising potential of UAV technology in biomass prediction for wheat at a small plot scale. It suggests that the accuracy of biomass predictions can be significantly improved through optimizing analytical and modelling protocols (i.e., variable selection, algorithm selection, stage-specific model development). Future work should focus on exploring the applicability of these findings under a wider variety of conditions and from a more diverse set of genotypes.

Introduction

Utilizing secondary physiological traits in grain crops to adapt to the environment is a crucial approach for securing food supply in an increasingly unpredictable world [ 52 ]. Biomass formed during the vegetative stage of crop development acts as a potential source of resources that can be translocated into harvestable yield [ 53 ]. Under non-limiting conditions, there is a positive relationship between biomass accumulation and grain yield [ 46 ], whereas excessive biomass formation under water or nutrient limitation may lead to asynchrony between resource use and availability, with potential to impede final yield. In wheat, biomass, along with fraction of intercepted radiation, is a key variable in defining radiation use efficiency (RUE), which relates to the ability of a plant to capture, and harness photosynthetic energy to produce the photosynthates necessary for growth [ 62 ]. The accurate monitoring of in-season biomass dynamics can thus provide a deeper understanding of the physiological dynamics of crop performance, especially when examined through the lens of genetic, environmental, and management-based interactions. This can lead to more appropriate selection of varieties more suitable for the target population of environments (TPE) for which they are destined [ 8 ].

Physical assessment of biomass formation at a plot level is laborious, time consuming and limited by the need for destructive sampling in the field [ 17 ]. The logistical difficulties and time constraints associated with measuring large numbers of biomass samples, mean that they are typically taken at a few key stages of crop growth, and from a limited number of plots. These limitations make it difficult to evaluate the differences between genotypes in large trials and where multiple environments are involved. As a result, approaches to estimating biomass using high throughput phenotyping (HTP) have become commonplace for a wide range of species, for example, in wheat [ 14 , 16 , 32 ], sorghum [ 36 ], maize [ 79 ] and soybean [ 75 ]. A large body of research has been conducted to develop indirect methods for the prediction of biomass using effective information extracted from RGB [ 39 ], multispectral [ 38 ], hyperspectral [ 75 ] and LIDAR sensors [ 13 , 14 ]. Indirect methods for biomass prediction provide opportunity for repeatable measurements that are less prone to human error, and are scalable, making them more amenable to use in large-scale variety selection programs [ 64 ].

Biomass prediction has been approached by modelling univariate relationships with HTP features, or more complex multivariate approaches. For example, [ 4 ] used UAV-based crop height as input for biomass prediction in the simple linear model. Such a univariate regression model presents an explicit biological relationship but cannot take advantage of the complimentary nature of data obtained from multiple sources [ 18 ]. Data fusion has been shown in many studies to improve the prediction accuracy for crop traits [ 47 , 50 , 76 ]. However, a major consideration in such datasets is they are often high dimensional and include both redundant and irrelevant variables, which reduce model interpretability and performance whilst increasing training time [ 26 , 35 ]. Thus, effective variable selection is vital in optimizing model development to improve model performance for predicting target crop traits (e.g., biomass) using high-dimensional HTP datasets.

In addition to variable selection, several factors regarding model selection have implications for biomass prediction accuracy. First, the algorithm used to develop prediction models play a key role, and machine learning models have been demonstrated to improve the accuracy of predictions over traditional statistical methods for a wide range of regression-related tasks [ 63 ]. Second, a major challenge in predictive modelling surrounds the trade-off between model specificity and generalizability. While a model that is trained on as many data points as possible, using a wide variety of growth stages and geographic locations, might be widely generalizable (scalability), the ability of such a model to identify important nuances in the data may be reduced. As such, multiple strategies exist for modelling experiment- level biomass: A single model could be trained across all available experiments and time-points with a focus on generalizability, or multiple models could be trained for individual circumstances (i.e., different growing stages, management practices, or years), with the aim of enhancing specificity. There is also the practical question of how to obtain high precision in the field, perhaps by establishing a ‘global’ prediction model, and enhancing it by a small set of biomass sampling in any given experiment, i.e., ‘real-time calibration’ by using a subset of the plots to build an enhanced model or check a ‘global’ model [ 29 ].

Another standing question in field-based phenotyping surrounds the region of interest (ROI) that image features are analysed [ 61 ]. While there are a range of options for automating the creation of ROIs within HTP workflows [ 70 ], the impact of the location of the ROI within the plot has not received sufficient attention. While biomass itself is only typically taken from a small area within a plot, provided the plot is homogeneous, it is often assumed that the ROI should be representative enough to correlate with the location where the sample was taken [ 28 , 59 ]. This approach has the advantage of allowing repeated measures over time, which may not be a major issue where within-plot variability is low, however, this may become an issue that reduces model accuracy where plot variability is high, for example [ 59 ]. Thus, how well a permanent ROI represents ground-truth measurements, and whether other approaches, such as sampling UAV indices directly from where the trait was measured would be more appropriate, remains an important factor to consider in HTP workflows.

In the context of variety selection, the goal of any phenotypic assessment is to determine genotypic differences to help understand the physiological basis of performance (i.e., yield) [ 17 , 52 ]. While in the earlier stages of a breeding program, genotypic differences in performance can be wide, given the diversity in early populations, at later stages, the variation can be reduced as selection may favour similar physiological adatptation pathways. While many studies have explored biomass prediction where variability is high, either at earlier stages of a breeding program [ 71 ], or where experimental treatments have been imposed to increase variability [ 4 ], we are not aware of studies that have attempted to estimate biomass between commercial varieties where variability is lower.

Given the spatial variability inherent in the field, there are a range of factors that can confound genetic effects, such as soils, moisture, shade, slope, and management [ 27 ]. To account for this variability, experimental design [ 9 ] and post-hoc analyses using linear mixed models is often a necessary step [ 56 ]. Based on this process, the genetic variance (V G ) and residual variance (V R ) of a trait can be estimated and used to predict within-season repeatability (w 2 ). While many studies have used w 2 to evaluate the efficacy of a prediction model [ 66 ], how these factors vary with regards to a model’s prediction accuracy has to our knowledge, not been explored [ 60 ].

In light of the issues surrounding variable selection, ROI determination and the estimation of w 2 for HTP derived predictions, in this paper we explored the sensitivity of biomass prediction models in wheat to a range of factors, by (a) exploring the effects of variable selection and various popular analytical algorithms to build prediction models (b) investigating the impact of within-plot position used to derive predictive variables; (c) making comparisons between a generic model and stage/experiment specific models to better understand the trade-offs between generalisability and specificity, and (d) Computed w 2 , V G and V R for each growth stage and experiment to better understand the optimal timing of measurements.

Field experiment

Wheat experiments were grown at Gatton Research Station, Queensland (27.56°S, 152.33°E) in 2020─2022 and at Boorowa Research Station, NSW (34.47°S, 148.69°E) in 2020. These sites contrast in the timing of rainfall and their temperature regimes throughout the season. Gatton receives most rainfall in the summer months leading up to the wheat growing season, and available crop water is normally dependent upon water stored in the soil profile during a summer fallow period. Alternatively, Boorowa typically receives rainfall during the growing season (Fig.  1 ). Gatton experiences higher in-season maximum and minimum temperatures than Boorowa, resulting in more rapid phenological development and a shorter growing season spanning May to October, whereas Boorowa, with its lower latitudes, experiences cooler mean temperatures that result in a longer growing season. While Gatton has a heavy, dar k Vertosol, Boorowa has a sandy loam with a sandy clay-loam subsoil described as a chromosol [ 42 ].

Minimum temperature (Min Temp) and maximum temperature (Max Temp), rainfall, and irrigation for each location and year combination. Dashed vertical lines represent planting dates and solid vertical lines represent harvest dates for individual experiments for a particular site and year combination. E1-E3 represent Gatton experiments in 2020, E4 represents Boorowa in 2020, E5-E7 represent Gatton experiments in 2021, E10 represents Gatton experiment in 2022

Table 1 describes the field experiments that were used for this study. At Gatton research station, three experiments (i.e., Early sowing date with high nitrogen, early sowing date with standard nitrogen, and a standard sowing date with standard nitrogen) were planted in 2020 and 2021, while a single experiment (i.e., late sown standard nitrogen) was planted in 2022. At Boorowa research station, one experiment (i.e., standard sowing standard nitrogen) was planted in 2020. Whereas Boorowa was grown under rainfed conditions, the Gatton site received supplementary irrigation to avoid the effects of drought stress. For each experiment, a variety trial (NVT) was planted, containing pre-commercial and commercial spring wheat varieties deemed suitable for the local environment by the GRDC NVT program ( https://nvt.grdc.com.au ). Compared to the original progeny of parental crosses and the several stages of selection that have been made by commercial breeders, these genotypes might be termed ‘elite’ given their history of selection.

The NVTs were designed with a row column configuration with 3 replicates using the R package DiGGer [ 9 ]. Additionally, a biomass calibration trial (BioCal) accompanied each NVT. In 2020 and 2021 for Gatton (E1, E2, E3, E5, E6 and E7), BioCal trials consisted of 6 genotypes × 3 densities (75, 150 and 225 plants per metre), in addition to a ‘check’ genotype planted again at 150 plants per meter, and 187 and 112 plants per meter. This resulted in a total of 21 plots for the BioCal trials. In 2022 (E10), the Gatton BioCal trial consisted of 8 genotypes × 3 densities (75, 150 and 300 plants per meter), with 7 out of 8 genotypes replicated once, and a single ‘check’ variety replicated twice (resulting in a total of 27 plots).

Biophysical measurements

For each experiment, dry weight of aboveground biomass ( DW AGB ) was measured by cutting plants at ground level within a quadrat area ( Quad area ) of a known size (Since row spacing and number varied, the quadrat area from which the sample was made varied for each experiment. The fresh weight of each sample was measured ( FW quad ) and the fresh weight of a subsample consisting of approximately 20 stems was taken ( FW sub ). This sample was oven dried at 750C until reaching a constant weight (DW sub ). DWAGB was thus calculated as:

At the time of each cut, biomass was measured from every plot within the BioCal trial, and additional biomass cuts were taken from a single replication of the NVT. In Gatton, growth stage was measured from only BioCal plots in 2020, and both BioCal and NVT plots in 2021 and 2022, using the Zadok’s growth scale on a weekly basis [ 77 ]. In Boorowa, Zadok’s growth scale was determined on two separate dates. To assign a particular DWAGB sample with a growth stage, the trial mean Zadok’s stage was calculated on each date with a Zadok’s score. Then, a generalized additive model (GAM) was built with using the following notation.

where \(Y\) represents the dependent Variable (Zadok’s stage), \({\beta }_{0}\) represents the intercept term, and \(f({TT}_{cumulative})\) represents the smoothing function of the predictor variable Cumulative thermal time ( \({TT}_{cumulative}\) ) using a spline made up of third-degree polynomial segments joined smoothly using 10 knots. TT cumulative and was calculated as per the method outlined by [ 80 ]. Using the relationship between TT cumulative and predicted Zadok’s stage (illustrated in Fig.  2 ), growth stage was determined to be ‘vegetative’ where Zadok’s was between 11 and 49 (1-leaf stage up to the end of booting), ‘flowering’ was determined where Zadok’s values were between 50 and 69 (head emergence to the end of anthesis), and ‘grain-fill’ where Zadok’s values fell between 70 and 99 (milk stage to the end of ripening). Solving for the equation of the GAM model resulted in categorizing the crop stages as follows: Vegetative stage when TT cumulative was less than or equal to 1189 °C day, Flowering stage when TT cumulative was greater than 1189 °C day but less than or equal to 1523 °C day, and Grain-Fill stage when TT cumulative exceeded 1523 °C day.

a Relationship between trial mean Zadok’s value and Cumulative thermal time (°C day) for measurements taken in each experiment (see coloured points). The solid line represents the line of best fit for a generalized additive model (GAM) and b the relationship between Cumulative thermal time (°C day) and aboveground biomass (DW AGB ) for each experiment in the study, with point colours representing the Growth-stage as classified

Image collection and processing

Unmanned Aerial Vehicle (UAV) flights were performed as close in time as logistically possible to the biomass cuts. In E4 (Boorowa), the mean difference between biomass cuts and flight dates was 4.75 days due to the remoteness of the site, whereas the mean difference for Gatton experiments was 1.15 days, with the final cut during grain-fill for E3 an outlier with 7 days difference (see Fig. S1 for overview of flight dates and biomass cut dates). Fixed ground control points were placed within each experiment and their GPS coordinates were measured using Propeller Aeropoints (Propeller, Australia). All flights were performed within 2 h of solar noon, under full sunlight and with a wind speed of less than 10 km/hr. (see Table  2 for an overview of flight parameters). In Gatton 2021 and 2022, a DJI M300 was used, which allows for the use of both an RGB and Multispectral sensor simultaneously in the same flight. In Gatton and Boorowa in 2020, two separate flights had to be made, given the fact that the RGB and multispectral sensors were carried on separate UAVs. For RGB flights, shutter speeds of ≤ 1/1600th of a second were used to reduce motion blur, and a front and side overlap of 80% was used for all flights to ensure sufficient matching of pixels between adjacent images. All multispectral imagery was radiometrically calibrated using a nadir image of a MicaSense calibration panel taken before and after each flight. A total of 36 RGB and 36 Multispectral datasets were processed for this study.

Raw imagery from each mission was processed using Agisoft Metashape (Agisoft, St Petersburg, Russia) which uses a structure from motion (SfM) algorithm to produce a 3d reconstruction of a scene from a set of images. First, multispectral imagery was calibrated through Metashape’s ‘calibrate reflectance’ option, which involves interpolation of the relationship between the known reflectance values of calibration panels and the timestamp of the before and after calibration panel images. This allows for in-flight reflectance values to be corrected based on estimated reflectance at a given time-point. Subsequently, Metashape was used to produce point clouds, which allow georectification and subsequent orthomosaic and digital elevation model (DEM) generation. A high degree of accuracy is enabled through SfM processing due to the presence of GCPs in the field. The average marker error across RGB orthomosaics was 1.62 cm and for MS orthomosaics was 1.37 cm. For each biomass cut taken within a plot, two ROIs were created for analysis: an ROI directly above the area where the biomass cut was taken (ROI precise ), and an ROI in a section of the plot that was not to be disturbed at any time (ROI permanent ) throughout the experiment until final harvest (see Fig.  3 for example trial design and illustration of the different ROIs used for the study). Each ROI precise was created by manually locating the location of biomass cut using the closest orthomosaic generated after the biomass sampling. This process was completed using ArcMap.

a Locations of Gatton and Boorowa sites in relation to the eastern and southern winter cropping region of Australia, b Gatton 2022 trial (E10), which includes a Biomass Calibration and NVT and c an example of paired plots with both an ROI permanent which was repeatedly measured for each biomass cut across the season and five ROI precise which were each analysed once for their respective biomass cuts

Variable extraction from proximal sensing

For both ROI types (ROI precise and ROI permanent ), spectral traits and geometric traits were calculated using a Python pipeline described by [ 12 ]. The spectral traits included 66 (33 × 2) variables, consisting of the median value for 33 vegetation indexes (Spectral traits) calculated from the entire or masked area within the ROI of the multispectral orthomosaic images. For the formulas of these Spectral traits refer to Table S4. The masked area represents the fraction of green plant matter within the ROI, which was distinguished based on a specific threshold value (OSAVI threshold ) of optimized soil-adjusted vegetation index (OSAVI) (i.e., OSAVI > OSAVI threshold for green plant matter, while OSAVI < OSAVI threshold for soil background). The threshold value was calculated by using Otsu’s method [ 49 ] and OSAVI was selected for this purpose due to its demonstrated ability to distinguish between green plant matter and soil across phenological stages [ 43 ]. For example cases of the thresholding results images please refer to figure Fig. S2.

For each ROI type, the geometric traits (9) included 4 height-related variables, 3 area-related variables, 1 volume-related variable, and 1 coverage-related variables (refer to Table S5 for details). The height-related variables consisted of the 50th, 75th, 95th and 98th percentile values as well as standard deviation of canopy height across the entire ROI. The canopy height was calculated by subtracting pixel values of the crop surface model (CSM) by corresponding pixel values of digital terrain models (DTM). A date specific CSM was derived from the 3d point cloud created using RGB images; while the DTM was produced from a fight made at the beginning of the season (before emergence). The area-related variables consisted of the area within the ROI where the height values were below the 25th, 50th, 75th percentiles. The volume-related variable indicated the sum of the pixel heights within a given ROI multiplied by the ground sample distance, divided by the area of the ROI. The coverage-related variable indicated the proportion of masked area of the total area within ROI, where the masked area was distinguished based on the Otsu threshold values of OSAVI (see Table S5 and Fig. S2 for further details).

Feature selection and biomass prediction models

The following section was performed using the Caret Package [ 34 ] in R Studio 3.0, and the R programming language version 4.2.3 and an overview of the workflow used in this study can be seen in Fig.  4 . The dataset was spilt into two parts: 80% as the training set and 20% as an independent test set. This was performed using stratified sampling so that each experiment was represented equally within both the training and test sets. The different combinations of broad growth-stages (i.e., vegetative, flowering, grain-filling, whole season) and predictive variables (i.e., spectral traits, geometric traits, spectral + geometric traits) resulted in 12 different combinations of broad growth-stages and features. Prior to the recursive feature selection using PLSR outlined below, for each of these combinations of features un-supervised filtering of variables based on pairwise correlations was performed to remove highly correlated variables using a correlation coefficient of 0.95 This final number of input parameters for each dataset can be seen in Table S1.

Overall workflow for this study. PLSR Partial least squares regression, GCP Ground Control Point, DEM Digital elevation model, ROI Region of Interest, XGBoost Extreme Gradient Boosting

Supervised feature selection was performed on each of the 12 Broad Growth-stage x variable subset groups by recursive feature elimination using a nested cross-validation approach, as per [ 1 ]. PLSR was used as the base-learner for this purpose with the number of components in the PLSR model tuned using tenfold cross-validation, and a grid containing 1:the number of features available for that dataset. Each dataset was resampled 30 times into training and test sets with an 80:20 split. For each resample, a PLSR model was initially trained using all features, then, from the set of original features, those with the lowest importance values were removed iteratively until only one feature remained in the model. Feature importance was calculated using the varImp function in caret, which is computed according to the weighted sums of absolute coefficients for each input variable [ 34 ]. After each iteration, feature importance was re-calculated, and performance was assessed on the corresponding validation set. The results of the 30 resamples were then aggregated to obtain a performance profile over the feature subset sizes and robust feature importance rankings. The optimal subset size was chosen by selecting the simplest model (with the lowest number of components) whose RMSE value was within 1 standard error of the absolute best model with the lowest RMSE value [ 5 ]. This process resulted in 12 sets of features corresponding with the different Broad Growth-stage x variable subsets.The variables selected using this approach served as predictive variables for the models explored below.

[ 34 ] PLSR, Random Forest, Support Vector Machine (SVM) and Extreme Gradient Boosting (XGBoost) models were trained for each variable and growth-stage combination using the selected variables based on the RFE feature selection outlined above For an overview of these models refer to the citations provided in Table  3 . Each of these models was trained on the training set using a tuning grid of model specific hyperparameters, and K-fold cross validation with 10 folds and 10 repeats (see Table  3 for an overview of model-specific hyperparameters). The average of each individual fold x repeat performance using k-fold validation was calculated as the mean performance of a specific model to reduce the prediction bias caused from random sampling in this small dataset. To decide upon optimal model performance, Root Mean Squared Error (RMSE) was used, while the coefficient of determination (R 2 ), and Relative Root Mean Squared Error (rRMSE) was calculated using the R Package Metrica [ 10 ].

where O i is the observed value for the i th observation, P i is the predicted value for the i th observation, n is the total number of observations, and O ¯ is the mean value of all observations.

Within-season repeatability

The R Package SpATS was used to fit a spatial model for the predictions of each prediction model [ 57 ]. SpATS uses restricted log-likelihood (REML) to estimate variance components in the model, and accounts for spatial trends using 2-dimensional p-splines with anisotropic penalties, implemented through the Separation of Anisotropic Penalties (SAP) algorithm [ 56 ]. This approach incorporates experimental rows and columns to model the spatial or environmental effect as a two-dimensional penalized tensor-product of marginal B-Spline basis functions.

Genotypic variance ( \(\sigma G\) ) in SpATS is estimated following [ 48 ], where genetic effects are treated as random effects. Within season repeatability is calculated using effective dimensions, defined as the trace of the corresponding hat matrix, reflecting the contribution of each model component to the phenotypic variation. This approach uses the following equation:

where \(E{D}_{g}\) is the effective dimension for the genetic component, \({m}_{g}\) is the total number of observations, and ζg​ is the shrinkage factor.

Predictive variable selection

The correlation coefficients of the 9 geometric variables as well as 66 median values of spectral variables with DW AGB across all growth stages are illustrated in Fig.  5 . These values were calculated on the entire dataset. For geometric traits, the weakest correlation was found for OSAVI canopy coverage. The different percentiles of height each had a positive correlation with DW AGB , whereas the area below a particular percentile of height had a negative correlation with DW AGB . For spectral traits, the median value of each trait provided the strongest correlation with DW AGB , so the sum and mean were disregarded for the analysis. Since masked and unmasked spectral traits provide information that relates to different aspects of the crop canopy and surrounding soil, we decided to use both the masked and unmasked median values in the prediction models. Overall, geometric variables displayed the strongest correlation with DW AGB , with 50th, 75th,98th percentile heights each having the greatest correlations (r 0.87–0.88). Meanwhile, spectral traits had a markedly lower correlation with DW AGB , with the strongest variable, GSAVI having an r value of 0.43.

Correlation coefficients (r) of a the 9 Geometric variables with dry biomass and b both the masked and unmasked median values of spectral traits

From the pairwise correlations seen in Fig. S3 (which utilized all geometric variables (9) and spectral variables (66)), a high degree of co-linearity was apparent amongst spectral traits and certain geometric traits across all growth stages (see Fig. S3). s.. For the four models where geometric and spectral traits were combined, the optimal set of variables always included both spectral traits and geometric traits, however overall, geometric traits were consistently ranked as being the most important, apart from at grain-fill, where red-edge reflectance was ranked as the most important variable (see Table S2). Canopy Volume, ranked as most important geometric trait for all growth-stages combined, vegetative and flowering stages, while at grain-fill, canopy coverage (%) was the most important variable. For spectral traits, rankings varied more considerably, and the overall number of variables chosen was also greater than for geometric variables.

The results of the RFE nested cross validation performance is illustrated in Fig.  6 . The optimal RMSE of the underlying PLSR models varied depending on the variable x broad growth-stage grouping investigated. A similar pattern for each of the broad growth-stage groupings was found, where spectral models had the highest RMSE, geometric models had slightly lower RMSE and combined geometric and spectral models had the lowest RMSE. When investigating all maturities combined, the optimal RMSE ranged from 202.8 g/m 2 using 13 variables when using spectral traits alone, to 208.27 g/m 2 when using 5 geometric variables alone, and was most accurate when using both geometric and spectral traits combined (147.77 g/m 2 using 16 variables). For vegetative models, RMSE ranged from 64.4 to 81.24 g/m 2 when using combined and geometric traits respectively, with the number of chosen variables ranging from 4 (geometric model) to 14 (combined model). For flowering time models, RMSE ranged from 112.89 to 149.44 g/m 2 when using Combined and Spectral traits respectively, with the number of chosen variables ranging from 1 (geometric model) to 3 (combined model). This stage resulted in the smallest number of chosen variables. For grain-fill models, RMSE ranged from 165.43 to 182.02 g/m 2 when using Combined and Spectral traits respectively, with the number of chosen variables ranging from 2 (geometric model) to 14 (spectral model).

Results of the Feature selection method using recursive feature elimination (RFE), on the x axis is the subset of features, and the y axis is the RMSE (g/m 2 ) achieved using nested cross validation. Error bars represent the standard deviation of the RMSE, while points represent the mean RMSE. The Vertical dashed line represents the optimal subset, chosen using the 1.se rule, with the final number of variables indicated in the label

Comparison of ML models for biomass prediction

Depending upon the broad growth-stage at which models were trained, models varied in their performance on the independent test set (see Fig.  7 ). In general, for any specific growth stage, the best model was always obtained using a combination of geometric and spectral variables. For the entire season, the XGBoost model exhibited the highest accuracy, with an RMSE of 31.2 g/m 2 on the test set and 29.47 g/m 2 on the training set. The RF model followed with an RMSE of 46.39 g/m 2 on the test set and 46.6 g/m 2 on the training set. In contrast, the PLSR_Refined model was the least accurate, with RMSE values of 164.45 g/m 2 and 159.36 g/m 2 for the test and train sets, respectively.

Performance metrics for the prediction of DW AGB (g/m 2 ) on the training and test set using the permanent ROI. Each horizontal facet represents the three different metrics used to evaluate model performance: R 2 , RMSE and rRMSE, while vertical facets represent the growth-stage stage at which the models were trained (vegetative, flowering, grain-fill, all (all maturities combined). The X axis includes the three different trait combinations, Geometric + Spectral (Combined), Geometric and Vi. The colour of each bar represents the respective model used to predict DWAGB, and the label represents the number of samples for a particular Growth-stage group

At the vegetative stage, the XGBoost model again demonstrated superior performance with RMSE values of 16.14 g/m 2 (test) and 14.05 g/m 2 (train). The RF model was slightly less accurate, with RMSEs of 22.11 g/m 2 (test) and 19.45 g/m 2 (train). The PLSR_Refined model had higher RMSE values of 68.76 g/m 2 (test) and 64.51 g/m 2 (train), indicating lower performance. During the flowering stage, the RF model showed robust accuracy, achieving RMSEs of 54.69 g/m 2 on the test set and 57.84 g/m 2 on the training set. XGBoost was comparable, with RMSE values of 67.74 g/m 2 (test) and 72.11 g/m 2 (train). Conversely, the PLSR_Refined model's performance was poorer, with RMSEs of 104.9 g/m 2 (test) and 111.33 g/m 2 (train). In the grain-fill stage, the XGBoost model maintained high accuracy, with RMSEs of 56.28 g/m 2 (test) and 51.47 g/m 2 (train). The RF model followed, with RMSEs of 81.31 g/m 2 (test) and 70.94 g/m 2 (train). The PLSR_Refined model, however, lagged with RMSE values of 180.22 g/m 2 (test) and 171.02 g/m 2 (train).

When comparing the use of spectral traits, geometric traits, and combined spectral and geometric traits, several general trends emerged. Models using combined spectral and geometric traits consistently outperformed those using only spectral or geometric traits across all stages. For instance, the combined trait models often achieved lower RMSE values, indicating higher predictive accuracy. Specifically, at the vegetative stage, the RF model with combined traits had an RMSE of 22.11 g/m 2 , compared to 48.99 g/m 2 for geometric traits and 25.79 g/m 2 for spectral traits.

Spectral traits alone generally provided better performance than geometric traits alone but were still inferior to the combined approach. For example, during the flowering stage, the RF model using spectral traits had an RMSE of 54.69 g/m 2 , while the geometric traits model had an RMSE of 63.04 g/m 2 . Similarly, at the grain-fill stage, spectral traits models showed an RMSE of 85.77 g/m 2 , compared to 110.72 g/m 2 for geometric traits.

Permanent ROI vs. Precise ROI

While XGBoost models provided the greatest accuracy when using a combination of geometric and spectral traits, we used RF models for the following section, given their similar performance, but faster training times in comparison to XGBoost. In addition to the existing RF models, we trained 12 additional models (on the same combination of 4 maturities and 3 trait combination) using the traits measured from the precise location where biomass cuts were taken (ROI precise ). Model performance when comparing the ROIs can be seen in Fig.  8 , which illustrates that overall, the differences in RMSE between ROI precise and ROI permanent were negligible, although we do see an increase in overall error on ROI precise when examining the test set performance. For example, when using training data from entire season with combined variables, the ROI permanent model had slightly higher accuracy (RMSE 46.6 g/m 2 ) than the ROI permanent model (RMSE 56.12 g/m 2 . The largest differences between the ROI types were found when using only geometric traits across the entire season where ROI precise had a RMSE 20.78 g/m 2 higher than ROI permanent . Similarly, across the entire season using spectral traits only, the difference in RMSE was 21.87 g/m 2 , with ROI permanent having the higher accuracy.

Observed versus predicted DW AGB on the permanent ROI , when looking at a Cross validation training set and b the independent test set and for the precise ROI when looking at c Cross-validation performance and d Test set performance, for random forest (RF) models trained on combinations of variable types (geometric, spectral, geometric + spectral (Combined)) and different growth-stage combinations (vegetative, flowering, grain-fill, all maturities combined)

Exploring model generalizability versus specificity

In the exploration of model generalizability, we compared the performance of models trained from data varying in growing stages or experiments (all models using RF with combined geometric and spectral variables). Figure  9 illustrates the prediction accuracy at specific growth stages when using a general model trained on data across all stages (stage-general model) and specific models trained only using data from a specific stage (stage-specific model). The analysis indicates that the optimal model type varies depending on the growth stage. At the vegetative stage, the stage-specific model outperformed the stage-general model with an RMSE of 22.111 g/m 2 compared to 25.924 g/m 2 . At the flowering stage, the stage-general model had a slightly lower RMSE of 53.341 g/m 2 compared to 58.782 g/m 2 for the stage-specific model. Similarly, at the grain-fill stage, the stage-general model achieved an RMSE of 76.207 g/m 2 , whereas the stage-specific model had an RMSE of 81.31 g/m 2 . These results suggest that while stage-specific models can provide more accurate predictions at certain stages, the stage-general model may offer better overall performance at other stages, highlighting the need for tailored modelling approaches depending on the specific growth stage being analysed.

A comparison of biomass prediction on the test set at different growth stages using RF regression using a a single model trained across all maturities, and b individual models trained only on biomass samples taken from that growth stage

In addition to growth stages, we trained different models using data based on experiments, i.e., a general model (experiment-general model) trained on data from all experiments, and individual models (experiment-specific models) trained only on data from specific experiments. The observed versus predicted biomass results are illustrated in Fig.  10 . Among the experiments, E1, E3, E4, E7, and E10 showed significant improvements in accuracy when using experiment-specific models over the experiment-general model. For example, in E4, the RMSE decreased from 56.59 g/m 2 in the experiment-general model to 31.86 g/m 2 in the experiment-specific model, showing the greatest benefit from having a specific model with a decrease in RMSE of 24.73 g/m 2 . In E2 and E5, the experiment-general model outperformed the experiment-specific models, with RMSEs decreasing from 51.2 to 42.46 g/m 2 in E2 and remaining nearly the same in E5 (44.13 g/m 2 vs. 44.04 g/m 2 ). For experiments E6 and E10, the differences in RMSE were smaller but still indicated improved performance with experiment-specific models. These results suggest that while the experiment-general model can perform well, experiment-specific models often provide better accuracy, likely due to the ability to capture subtle differences in experimental conditions.

Observed versus predicted DW AGB using a Random Forest (RF) model trained on a combination of Spectral traits and Geometric vars, trained on a ground truth data from all experiments, (‘experiment-general model’ shown in blue) and b ground truth data for a particular experiment (‘experiment-specific model’ shown in red)

We also examined the prediction accuracy of the RF model for each individual experiment x DW AGB cut. The model trained on combined geometric + spectral variables was able to predict DW AGB with relatively high accuracy across growth stages and experiments. A general increase in RMSE can be seen as cut numbers and crop-growth-stage increased (i.e. RMSE at cut 1 ranged from 4.4 to 9.16 g/m 2 and increased to 47.94–89.07 g/m 2 by cut 5. At the same time, the relative error remained stable, and saw a decrease as cut numbers increased (i.e. rRMSE ranged from 6 to 29% at cut 1, while at cut 5 it ranged from 3 to 6%).

Within season repeatability biomass prediction models

A significant difference in w 2 was identified across different experiments (p < 0.05), with no apparent relationship between w 2 and growth-stage. For V G , no relationship with the experiment was found, but a significant relationship with growth-stage was observed (p < 0.01). In the context of V R , an interaction effect was detected between growth-stage and experiment (p < 0.001). Specifically, experiments 4 and 7 showed higher V R for predicted biomass, while Experiments 3, 5, 6, and 7 reported increases in V R during the grain-fill stage. Conversely, E4 and E7 exhibited lower V R during the vegetative stage. The temporal dynamics of w 2 based on the predictions from the RF model trained on data points from all maturities and geometric + spectral traits are illustrated in Fig.  11 .

Relationship between biomass cut number and proportion of error variances of predicted biomass from the RF model trained on geometric + spectral indices across all dates and experiments. The x axis represents the biomass cut number and the y axis represents the variation for each respective variance component. White dots represent the within season repeatability (w 2 ) for predicted DW AGB at that timepoint

The significant increase in interest related to UAV based high throughput phenotyping is partly a result of the need for improved throughput and selection accuracy in breeding programs and NVTs [ 2 , 6 ]. While the accurate prediction of biomass using HTP platforms has been demonstrated for a range of agricultural crops [ 17 ], several practical details with potential to optimize biomass prediction results have yet to be discussed. HTP platforms generate large datasets and given the vast number of possible traits that can be calculated using multispectral and RGB sensors [ 2 ], the need for biological interpretability in predictive models is an essential task. Depending on the crop stage at which prediction is made, the ranking of various traits will vary, influencing the type of sensor that is used, and the traits that are necessary to be used.

Variable importance differs based on growth-stage and sensor type

The importance of using a combination of spectral and Geometric traits in predicting biomass at various growth stages was shown in this study by both the variable selection strategy, and the prediction results from different ML models. Predictions had lower accuracy when models used only geometric variables or spectral traits compared to when they were combined. For models across all maturities, geometric traits ranked as having the greatest importance after recursive feature elimination, however, spectral traits clearly added additional information that improved overall performance. These results reflect recent studies that combine canopy height with spectral traits to predict DW AGB . Canopy volume was consistently the first trait selected, however, more complex traits related to the area below a given percentile of height, the standard deviation of height and canopy coverage were also chosen. Canopy Volume has been identified in other literature [ 41 ] as an important trait for biomass prediction. Further studies that compare canopy surface provided by RGB cameras, to LIDAR derived volume could help to identify differences in actual biovolume to further improve model accuarcy [ 72 ].

When investigating models that used a combination of geometric and spectral traits, changes in the chosen variables may reflect the dynamic relationship between geometric and absorptive/ reflective properties of the canopy and the physiological state of the crop. At vegetative stage, vegetation indices related to canopy greenness (i.e. clg, ng) were highly favoured by the variable selection method. Corti et al. [ 11 ] found that spectral traits related to canopy greenness had the strongest correlations with DW AGB across multiple species. Canopy Volume the most important variable at this stage. In contrast, at flowering and grain-fill stages, vegetative indices contributed a smaller proportion to the final models. However, for the spectral indices that were chosen, Red Edge reflectance and traits related to canopy greenness (i.e., NG & masked VARIGREEN) were consistently chosen, which may be due to the relationship between the Red Edge band and the rate and onset of canopy senescence (and consequent photosynthetic function) between different varieties [ 1 ].

This study highlighted that combining geometric and spectral traits consistently led to improvements in prediction accuracy. This supports similar studies which found that sensor fusion can provide more comprehensive information surrounding canopy characteristics [ 11 , 14 , 50 ] than individual sensors alone. At the same time, our comparison of spectral and geometric traits highlighted that geometric traits were more closely correlated with canopy biomass at all growth stages. Given both the ubiquity and low cost of modern RGB UAVs, these results indicate RGB UAVs alone may provide a low-cost solution to biomass monitoring in the field through the calculation of geometric traits. It is considered more difficult to obtain consistent results with Multispectral imaging (MS) given the sensitivity to lighting changes and the need for radiometric calibration. While in this study spectral traits (OSAVI) were used to derive Coverage %, the use of classification methods on pixels of RGB imagery, would mean that ‘geometric traits’ could entirely come from an RGB sensor [ 24 , 40 ].

Phenology-based classification of growth stages

In this study, the objective of classifying broad growth stages into vegetative, flowering, and grain filling phases was to account for the variation in both structural characteristics and spectral response throughout the development of the canopy. This classification aimed to address the changes that occur as heads emerge (flowering phase) and as they mature and the canopy colour changes, marking the onset of rapid grain filling (grain filling phase). In our study, the vegetative phase was defined to end at Zadok's score of 50, which indicates that 50% of the plants in a plot have at least one awn appearing. The flowering phase was considered to end at Zadok's score of 70, when anthesis is completed on 50% of the plants, signifying that grain development has commenced in many spikelets. We considered further dividing the vegetative phase by introducing a threshold approximately halfway between stem elongation and booting.

However, this subdivision did not enhance the prediction accuracy for the resulting phases. We recognize that with a larger dataset, it might be possible to better optimize the structural and spectral parameters for biomass estimation. This could potentially involve adding an additional phase, such as late vegetative, or adjusting the existing phases to better reflect the relationships between remotely sensed proxies and biomass. A more refined approach could involve developing a generic biomass prediction model, which could then be adjusted using a phenology-derived parameter to optimize predictions continuously across the growth stages.

Comparison of permanent and precise ROI

While considerable work has been dedicated to the task of ROI generation, as highlighted by [ 67 ], the alignment of the ROI with the actual site of ground-truth measurement has not been extensively tested. Through analysis in this work, we found ROI precise (corresponding to the actual locations of biomass cuts) demonstrated accuracy comparable to ROI permanent despite utilization of set of variables across all growth stages. However, variations were noted at different growth stages, and when analysing geometric and spectral traits independently. The discrepancies in model performance between the ROIs are thought to be related to spatial variability within the plot and the distinct characteristics of the areas where ROIs were established. These observations suggest the need for additional research into alternative ROI selection methods and to understand the factors influencing the differences in model performance between ROIs. ROI precise may offer superior performance over ROI permanent particularly in scenarios where within-plot heterogeneity is pronounced.

Model generalizability

A key aim in a variety testing situations is the scaling up of predictions to encompass multiple experiments and time points. However, a significant portion of the literature in this field focuses on testing biomass prediction accuracy using a relatively small number of samples from single trials. Our strategy was to evaluate the accuracy of predictions with models trained using the same traits and approaches but distinguishing between specific experiments or growth stages and general models trained on a more comprehensive dataset. Our work indicated that models tailored to a particular growth stage or experiment exhibited comparable (in the case of broad growth-stage) or slightly higher (in the case of specific experiments) accuracy compared to those employing a general approach, presumably because specific models capture specific information that better represents those circumstances. However, the general models still provide adequate predictions across growth stages (Fig.  9 ), experiments (Fig.  10 ), and at individual time-points in experiments (Fig.  12 ) indicating the possibility of using a generic model to make predictions under a wide range of conditions.

Observed versus predicted DW AGB (g/m 2 ) using the Random Forest model trained using geometric and spectral variables on the cross-validated training dataset. Vertical facets represent the different experiments in the study and the horizontal facets represent the DW AGB cuts in order. Point colours represent the cumulative thermal time (TT cumulative ). Metrics for each cut x experiment are shown in each facet. For test set performance see \* MERGEFORMAT Fig. S4

Breeding and variety selection

The findings from the study, particularly those related to biomass prediction and the importance of various variables at different growth stages, can significantly influence breeding decisions and variety selection in wheat crops. w 2 , a key parameter in breeding for trait improvement, was explored in the study, revealing variations across different experiments and maturities. The observed w 2 can provide insights into the genetic control of the traits under investigation, thereby guiding breeders in selecting varieties that not only exhibit desirable traits but also have a higher probability of passing these traits to subsequent generations. It should be noted that this study focused on a sample of ‘elite’ germplasm, and future research should investigate whether w 2 is as high for earlier stage breeding material. This study was also limited by the fact that ground-truth DW AGB w 2 or broad-sense heritability (H 2 ) could not be calculated, since only a single replicate was physically sampled at one time. A comparison of the ground-truth w 2 to HTP derived proxy traits is a key step in determining the optimal timing of sampling, as it defines the ceiling by which the accuracy of remotely-sensed traits can be estimated. This is an important area for future research that requires extra attention.

Since w 2 is calculated based on the ratio of V G to the total variance (V G  + V R ), to achieve high w 2 , the ratio of V G to V R must also be high. The significant relationship between both V G and growth-stage and V R and growth-stage * experiment means that while w 2 can be highest at the latest growth stages, the risk of increased V R due to experiment level factors (i.e., lodging or pest damage) can increase. In our experiments, while not significantly different from other growth stages, the growth stage at which mean w 2 across experiments was highest, was during the flowering stage, which may be a result of this simultaneous increase of V G and V R . This may be due to the presence of larger differences among reflectance of genotypes as the differentially reach new phenological stages. This study was limited by the number of locations and years in which testing occurred, and as such, there is a need to investigate the relationship between V G and V R in a wider set of environments to examine whether the timing of measurement can be optimized to maximize w 2 .

Limitations and future research directions

While the study provides valuable insights into biomass prediction using UAV-based high-throughput phenotyping, several limitations warrant acknowledgment and consideration. One potential limitation is related to the geographical and environmental specificity of the study. The findings and models developed might be highly tailored to the specific environmental conditions and wheat varieties studied, potentially limiting their applicability to different geographical regions, environmental conditions, or wheat varieties. Furthermore, the varieties used in this study were limited to ‘elite’ breeding materials, which potentially lack the range of variation that might be present in earlier-stage breeding trials where selection may be warranted. Future research should also compare the w 2 and H 2 of ground-truth samples to the estimated values for DW AGB , so as to confirm the genetic-gain, or selection advantage for using HTP derived traits.

In future research, we also recommend exploring the use of parametric models to further analyse the relationships between biomass and selected predictors. This approach would involve selecting a subset of variables that maintains performance close to that of the full model and then fitting a parametric model to these predictors. Such a model would allow for clearer interpretation of the impact of each predictor on biomass at different stages of crop development, enhancing the overall interpretability and robustness of the results.

This methodology, whilst providing a proof of concept for scaling up an empirical calibration for biomass, still requires manual ground-truth measurements for model building. Even taking a small number of biomass samples can be a costly undertaking, especially where multiple experiments and locations are concerned. As such, there is a need to determine an economically viable number of biomass samples that can be taken, whilst still maintaining model accuracy and w 2 . One such approach, as demonstrated by Hu et al. [ 29 ] is to take a self-calibration approach, which takes advantage of empirical ground-truth measurements, but optimizes the number of samples necessary to maintain accuracy whilst maintaining low cost. In a situation with 1000 s of plots, you can use a generalized model to estimate biomass and immediately choose a diverse set of plots to measure the biomass (20 to 50 plots) to create an improved model calibration. In addition, the topic of sub-plot selection and analysis warrants further exploration. Underlying within-plot variability is a source of variability that may not be properly accounted for when building the ROI to analyse secondary traits from UAVs. Whilst in experiments with low variability this may not be a major issue, but where variability is high, alternative approaches such as the one proposed in this paper might lead to higher prediction accuracies and w 2 values.

This study explored the sensitivity of UAV-based biomass prediction in wheat, exploring the influence of variable type, modelling strategy and sampling location on model accuracy. We utilised robust feature selection, using recursive feature elimination, to identify key features associated with biomass at varying growth stages and using different sensor traits. A combination of RGB and multispectral traits was confirmed to provide the greatest accuracy across growth stages, with canopy height and volume having the greatest importance, but being supplemented by growth-stage specific spectral indices. The comparison of the specific and permanent ROI did not result in significant differences in model accuracy, however this may have been due to the relative homogeneity in experiments, and further investigation into this approach in heterogenous situations may provide greater accuracy. In this case a general model trained across all available data performed comparably with stage-specific and experiment-specific models, highlighting the ability for Machine Learning methods to capture complex relationships in the data. Overall, biomass prediction using UAV offers a non-destructive and scalable alternative to manual measurements, however, the need for careful modelling to demonstrate physiological relevance, and ability to be used for variety selection is still necessary.

Availability of data and materials

The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request. This ensures the accessibility of data while maintaining the necessary controls for their ethical and responsible use.

Code availability

The custom code developed for this study is available from the corresponding author upon reasonable request. This approach ensures that the code can be accessed for legitimate research purposes while maintaining appropriate oversight and adherence to ethical standards in its use and application.

Anderegg J, Yu K, Aasen H, Walter A, Liebisch F, Hund A. Spectral vegetation indices to track senescence dynamics in diverse wheat germplasm. Front Plant Sci. 2020. https://doi.org/10.3389/fpls.2019.01749 .

Article   PubMed   PubMed Central   Google Scholar  

Araus JL, Cairns JE. Field high-throughput phenotyping: the new crop breeding frontier. Trends Plant Sci. 2014;19(1):52–61. https://doi.org/10.1016/j.tplants.2013.09.008 .

Article   CAS   PubMed   Google Scholar  

Baret F, Jacquemoud S, Hanocq J. The soil line concept in remote sensing. Remote Sens Rev. 1993;7(1):65–82.

Article   Google Scholar  

Bendig J, Bolten A, Bennertz S, Broscheit J, Eichfuss S, Bareth G. Estimating biomass of barley using crop surface models (CSMs) derived from UAV-based RGB imaging. Remote Sens. 2014;6(11):10395–412. https://doi.org/10.3390/rs61110395 .

Breiman L, Friedman J, Olshen RA, Stone CJ (1984) Classification and Regression Trees (1st ed.). Chapman and Hall/CRC. https://doi.org/10.1201/9781315139470

Cabrera-Bosquet L, Crossa J, von Zitzewitz J, Serret MD, Araus JL. High-throughput phenotyping and genomic selection: the frontiers of crop breeding converge. J Integr Plant Biol. 2012;54(5):312–20. https://doi.org/10.1111/j.1744-7909.2012.01116.x .

Article   PubMed   Google Scholar  

Chen T, He T, Benesty M, Khotilovich V, Tang Y, Cho H, Chen K, Mitchell R, Cano I, Zhou T. Xgboost: extreme gradient boosting. R package version 0.4-2. 2015; 1(4): 1–4.

Comstock RE. Quantitative genetics and the design of breeding programs. In E. Pollak (Eds). Proc. Int. Conf. on Quant. Genet., Ames, Iowa. 16–21 Aug. Iowa State University Press. 1977; 705–718.

Coombes N.. DiGGer design search tool in R. 2009. http://nswdpibiom.org/austatgen/software . Accessed 20 Mar 2024

Correndo AA, Rosso LHM, Hernandez CH, Bastos LM, Nieto L, Holzworth D, Ciampitti IA. metrica: an R package to evaluate prediction performance of regression and classification point-forecast models. J Open Sour Softw. 2022;7(79):4655.

Corti M, Cavalli D, Cabassi G, Bechini L, Pricca N, Paolo D, Marinoni L, Vigoni A, Degano L, Marino Gallina P. Improved estimation of herbaceous crop aboveground biomass using UAV-derived crop height combined with vegetation indices. Precision Agric. 2023;24(2):587–606.

Das S, Massey-Reed SR, Mahuika J, Watson J, Cordova C, Otto L, Zhao Y, Chapman S, George-Jaeggli B, Jordan D. A high-throughput phenotyping pipeline for rapid evaluation of morphological and physiological crop traits across large fields. IGARSS 2022–2022 IEEE International Geoscience and Remote Sensing Symposium. 2022.

Deery D, Jimenez-Berni J, Jones H, Sirault X, Furbank R. Proximal remote sensing buggies and potential applications for field-based phenotyping. Agron Basel. 2014;4(3):349–79. https://doi.org/10.3390/agronomy4030349 .

Deery DM, Rebetzke GJ, Jimenez-Berni JA, Condon AG, Smith DJ, Bechaz KM, Bovill WD. Ground-based LiDAR improves phenotypic repeatability of above-ground biomass and crop growth rate in wheat. Plant Phenomics. 2020. https://doi.org/10.34133/2020/8329798 .

Fitzgerald G, Rodriguez D, Christensen L, Belford R, Sadras V, Clarke T. Spectral and thermal sensing for nitrogen and water status in rainfed and irrigated wheat environments. Precision Agric. 2006;7:233–48.

Fu Y, Yang G, Song X, Li Z, Xu X, Feng H, Zhao C. Improved estimation of winter wheat aboveground biomass using multiscale textures extracted from UAV-based digital images and hyperspectral feature analysis. Remote Sens. 2021;13(4):581. https://doi.org/10.3390/rs13040581 .

Furbank RT, Jimenez-Berni JA, George-Jaeggli B, Potgieter AB, Deery DM. Field crop phenomics: enabling breeding for radiation use efficiency and biomass in cereal crops. New Phytol. 2019;223(4):1714–27. https://doi.org/10.1111/nph.15817 .

Gill T, Gill SK, Saini DK, Chopra Y, de Koff JP, Sandhu KS. A comprehensive review of high throughput phenotyping and machine learning for plant stress phenotyping. Phenomics. 2022;2(3):156–83.

Gitelson AA, Kaufman YJ, Merzlyak MN. Use of a green channel in remote sensing of global vegetation from EOS-MODIS. Remote Sens Environ. 1996;58(3):289–98.

Gitelson AA, Kaufman YJ, Stark R, Rundquist D. Novel algorithms for remote estimation of vegetation fraction. Remote Sens Environ. 2002;80(1):76–87.

Gitelson AA, Merzlyak M, Zur Y, Stark R, Gritz U. Non-destructive and remote sensing techniques for estimation of vegetation status. Proceedings of the 3rd European Conference on Precision Agriculture, Montpelier, France. 2001.

Gitelson AA, Viña A, Ciganda V, Rundquist DC, Arkebauer TJ. Remote estimation of canopy chlorophyll content in crops. Geophys Res Lett. 2005. https://doi.org/10.1029/2005GL022688 .

Gobron N, Pinty B, Verstraete MM, Widlowski J-L. Advanced vegetation indices optimized for up-coming sensors: design, performance, and applications. IEEE Trans Geosci Remote Sens. 2000;38(6):2489–505.

Guo W, Zheng B, Duan T, Fukatsu T, Chapman S, Ninomiya S. EasyPCC: benchmark datasets and tools for high-throughput measurement of the plant canopy coverage ratio under field conditions. Sensors. 2017;17(4):798.

Haboudane D, Miller JR, Pattey E, Zarco-Tejada PJ, Strachan IB. Hyperspectral vegetation indices and novel algorithms for predicting green LAI of crop canopies: modeling and validation in the context of precision agriculture. Remote Sens Environ. 2004;90(3):337–52.

Hall MA. Correlation-based feature selection for machine learning The University of Waikato. 1999.

Hoefler R, González-Barrios P, Bhatta M, Nunes JA, Berro I, Nalin RS, Borges A, Covarrubias E, Diaz-Garcia L, Quincke M. Do spatial designs outperform classic experimental designs? J Agric Biol Environ Stat. 2020;25(4):523–52.

Hu P, Chapman SC, Jin H, Guo Y, Zheng B. Comparison of modelling strategies to estimate phenotypic values from an unmanned aerial vehicle with spectral and temporal vegetation indexes. Remote Sens. 2021;13(14):2827.

Hu P, Chapman SC, Wang X, Potgieter A, Duan T, Jordan D, Guo Y, Zheng B. Estimation of plant height using a high throughput phenotyping platform based on unmanned aerial vehicle and self-calibration: example for sorghum breeding. Eur J Agron. 2018;95:24–32.

Huete AR. A soil-adjusted vegetation index (SAVI). Remote Sens Environ. 1988;25(3):295–309.

Jiang Z, Huete AR, Didan K, Miura T. Development of a two-band enhanced vegetation index without a blue band. Remote Sens Environ. 2008;112(10):3833–45.

Jin X, Madec S, Dutartre D, de Solan B, Comar A, Baret F. High-throughput measurements of stem characteristics to estimate ear density and above-ground biomass. Plant Phenomics. 2019. https://doi.org/10.34133/2019/4820305 .

Jordan CF. Derivation of leaf-area index from quality of light on the forest floor. Ecology. 1969;50(4):663–6.

Kuhn M. Building predictive models in R using the caret package. J Stat Softw. 2008;28:1–26.

Kursa MB, Rudnicki WR. Feature selection with the Boruta package. J Stat Softw. 2010;36:1–13.

Li J, Shi Y, Veeranampalayam-Sivakumar A-N, Schachtman DP. Elucidating sorghum biomass, nitrogen and chlorophyll contents with spectral and morphological traits derived from unmanned aircraft system. Front Plant Sci. 2018. https://doi.org/10.3389/fpls.2018.01406 .

Liu HQ, Huete A. A feedback based modification of the NDVI to minimize canopy background and atmospheric noise. IEEE Trans Geosci Remote Sens. 1995;33(2):457–65.

Liu Y, Liu S, Li J, Guo X, Wang S, Lu J. Estimating biomass of winter oilseed rape using vegetation indices and texture metrics derived from UAV multispectral images. Comput Electron Agric. 2019;166:105026.

Lu N, Zhou J, Han Z, Li D, Cao Q, Yao X, Tian Y, Zhu Y, Cao W, Cheng T. Improved estimation of aboveground biomass in wheat from RGB imagery and point cloud data acquired with a low-cost unmanned aerial vehicle system. Plant Methods. 2019;15(1):1–16.

Article   CAS   Google Scholar  

Madec S, Irfan K, Velumani K, Baret F, David E, Daubige G, Samatan LB, Serouart M, Smith D, James C. VegAnn, vegetation annotation of multi-crop RGB images acquired under diverse conditions for segmentation. Sci Data. 2023;10(1):302.

Maimaitijiang M, Sagan V, Sidike P, Maimaitiyiming M, Hartling S, Peterson KT, Maw MJ, Shakoor N, Mockler T, Fritschi FB. Vegetation index weighted canopy volume model (CVMVI) for soybean biomass estimation from unmanned aerial system-based RGB imagery. ISPRS J Photogramm Remote Sens. 2019;151:27–41.

Malone B, Stockmann U, Glover M, McLachlan G, Engelhardt S, Tuomi S. Digital soil survey and mapping underpinning inherent and dynamic soil attribute condition assessments. Soil Secur. 2022;6:100048. https://doi.org/10.1016/j.soisec.2022.100048 .

Marcial-Pablo MDJ, Gonzalez-Sanchez A, Jimenez-Jimenez SI, Ontiveros-Capurata RE, Ojeda-Bustamante W. Estimation of vegetation fraction using RGB and multispectral images from UAV [Marcial-Pablo2019]. Int J Remote Sens. 2019;40(2):420–38.

MaxMax. Enhanced Normalized Difference Vegetation Index. 2015. https://www.maxmax.com/endvi.htm . Accessed 20 Mar 2024

Mevik B-H, Wehrens R. The pls package: principal component and partial least squares regression in R. J Stat Softw. 2007;18(2):1–23

Mondal S, Dutta S, Crespo-Herrera L, Huerta-Espino J, Braun HJ, Singh RP. Fifty years of semi-dwarf spring wheat breeding at CIMMYT: grain yield progress in optimum, drought and heat stress environments. Field Crop Res. 2020;250:107757.

Montes JM, Technow F, Dhillon BS, Mauch F, Melchinger AE. High-throughput non-destructive biomass determination during early plant development in maize under field conditions. Field Crop Res. 2011;121(2):268–73.

Oakey H, Verbyla A, Pitchford W, Cullis B, Kuchel H. Joint modeling of additive and non-additive genetic line effects in single field trials. Theor Appl Genet. 2006;113:809–19.

Otsu N. A threshold selection method from gray-level histograms. IEEE Trans Syst Man Cybern. 1979;9(1):62–6.

Pauli D, Chapman SC, Bart R, Topp CN, Lawrence-Dill CJ, Poland J, Gore MA. The quest for understanding phenotypic variation via integrated approaches in the field environment. Plant Physiol. 2016;172(2):622–34. https://doi.org/10.1104/pp.16.00592 .

Article   CAS   PubMed   PubMed Central   Google Scholar  

Qi J, Chehbouni A, Huete AR, Kerr YH, Sorooshian S. A modified soil adjusted vegetation index. Remote Sens Environ. 1994;48(2):119–26.

Reynolds M, Langridge P. Physiological breeding. Curr Opin Plant Biol. 2016;31:162–71. https://doi.org/10.1016/j.pbi.2016.04.005 .

Reynolds MP, Slafer GA, Foulkes JM, Griffiths S, Murchie EH, Carmo-Silva E, Asseng S, Chapman SC, Sawkins M, Gwyn J. A wiring diagram to integrate physiological traits of wheat yield potential. Nature Food. 2022;3(5):318–24.

Richards JA, Richards JA. Remote sensing digital image analysis. Berlin: Springer; 2022. https://doi.org/10.1007/978-3-030-82327-6

Book   Google Scholar  

Richardson AJ, Wiegand C. Distinguishing vegetation from soil background information. Photogramm Eng Remote Sens. 1977;43(12):1541–52.

Google Scholar  

Rodriguez-Alvarez MX, Boer M, Eilers P, van Eeuwijk F. Spatial analysis of field trials with splines. 2020. https://cran.r-project.org/web/packages/SpATS/SpATS.pdf

Rodríguez-Álvarez MX, Boer MP, van Eeuwijk FA, Eilers PH. Correcting for spatial heterogeneity in plant breeding experiments with P-splines. Sp Stat. 2018;23:52–71. https://doi.org/10.1016/j.spasta.2017.10.003 .

Rouse J, Haas RH, Schell JA, Deering DW. Monitoring vegetation systems in the Great Plains with ERTS. NASA Spec Publ. 1974;351:309.

Roy Choudhury M, Das S, Christopher J, Apan A, Chapman S, Menzies NW, Dang YP. Improving biomass and grain yield prediction of wheat genotypes on sodic soil using integrated high-resolution multispectral, hyperspectral, 3D point cloud, and machine learning techniques. Remote Sens. 2021;13(17):3482.

Schmidt P, Hartung J, Rath J, Piepho H-P. Estimating broad-sense heritability with unbalanced data from agricultural cultivar trials. Crop Sci. 2019;59(2):525–36.

Sharma P, Leigh L, Chang J, Maimaitijiang M, Caffé M. Above-ground biomass estimation in oats using UAV remote sensing and machine learning. Sensors. 2022;22(2):601.

Sinclair TR, Muchow RC. Radiation use efficiency. Adv Agron. 1999;65:215–65.

Singh A, Ganapathysubramanian B, Singh AK, Sarkar S. Machine learning for high-throughput stress phenotyping in plants. Trends Plant Sci. 2016;21(2):110–24.

Smith DT, Potgieter AB, Chapman SC. Scaling up high-throughput phenotyping for abiotic stress selection in the field. Theor Appl Genet. 2021. https://doi.org/10.1007/s00122-021-03864-5 .

Sripada RP. Determining in-season nitrogen requirements for corn using aerial color-infrared photography. North Carolina State University. 2005.

Toda Y, Kaga A, Kajiya-Kanegae H, Hattori T, Yamaoka S, Okamoto M, Tsujimoto H, Iwata H. Genomic prediction modeling of soybean biomass using UAV-based remote sensing and longitudinal model parameters. Plant Genome. 2021;14(3):e20157.

Tresch L, Mu Y, Itoh A, Kaga A, Taguchi K, Hirafuji M, Ninomiya S, Guo W. Easy MPE: extraction of quality microplot images for UAV-based high-throughput field phenotyping. Plant Phenomics. 2019. https://doi.org/10.34133/2019/2591849 .

Tucker CJ. Red and photographic infrared linear combinations for monitoring vegetation. Remote Sens Environ. 1979;8(2):127–50. https://doi.org/10.1016/0034-4257(79)90013-0 .

Wang D, Li R, Zhu B, Liu T, Sun C, Guo W. Estimation of wheat plant height and biomass by combining UAV imagery and elevation data. Agriculture. 2022;13(1):9.

Wang H, Duan Y, Shi Y, Kato Y, Ninomiya S, Guo W. EasyIDP: a python package for intermediate data processing in UAV-based plant phenotyping. Remote Sens. 2021;13(13):2622.

Wang T, Crawford MM, Tuinstra MR. A novel transfer learning framework for sorghum biomass prediction using UAV-based remote sensing data and genetic markers. Front Plant Sci. 2023;14:1138479.

Wiering NP, Ehlke NJ, Sheaffer CC. Lidar and RGB image analysis to predict hairy vetch biomass in breeding nurseries. Plant Phenome J. 2019;2(1):1–8.

Woebbecke DM, Meyer GE, Von Bargen K, Mortensen DA. Color indices for weed identification under various soil, residue, and lighting conditions. Trans ASAE. 1995;38(1):259–69.

Wright MN, Ziegler A. Ranger: a fast implementation of random forests for high dimensional data in C++ and R. J Stat Softw. 2017;77(1):1–17. https://doi.org/10.18637/jss.v077.i01 .

Yoosefzadeh-Najafabadi M, Tulpan D, Eskandari M. Using hybrid artificial intelligence and evolutionary optimization algorithms for estimating soybean yield and fresh biomass using hyperspectral vegetation indices. Remote Sens. 2021;13(13):2555.

Yue J, Yang G, Tian Q, Feng H, Xu K, Zhou C. Estimate of winter-wheat above-ground biomass based on UAV ultrahigh-ground-resolution image textures and vegetation indices. ISPRS J Photogramm Remote Sens. 2019;150:226–44.

Zadoks JC, Chang TT, Konzak CF. A decimal code for the growth stages of cereals. Weed Res. 1974;14(6):415–21.

Zarco-Tejada PJ, Miller JR, Noland TL, Mohammed GH, Sampson PH. Scaling-up and model inversion methods with narrowband optical indices for chlorophyll content estimation in closed forest canopies with hyperspectral data. IEEE Trans Geosci Remote Sens. 2001;39(7):1491–507.

Zhang Y, Xia C, Zhang X, Cheng X, Feng G, Wang Y, Gao Q. Estimating the maize biomass by crop height and narrowband vegetation indices derived from UAV-based hyperspectral images. Ecol Ind. 2021;129:107985.

Zheng B, Chenu K, Doherty A, Chapman S. The APSIM-wheat module (7.5 R3008). Agricultural Production Systems Simulator (APSIM) Initiative. 2014; 615. http://www.apsim.info/Portals/0/Documentation/Crops/WheatDocumentation.pdf

Download references

Acknowledgements

Carla Gho for her technical and administrative support, Alex Chamanmah, Lleyton Cave, Jonothan Powell and Bill Bovill for their support in the field.

This research has been supported by The University of Queensland and funds from the Grains Research and Development Corporation (GRDC) as a part of the INVITA project—A technology and analytics platform for improving variety selection (UOQ2003-011RTX). DS is a recipient of an Australian Government research training program (RTP) scholarship together with a PhD top-up scholarship from the GRDC (UOQ2004-013RSX).

Author information

Authors and affiliations.

School of Agriculture and Food Sustainability, The University of Queensland, St Lucia, QLD, 4072, Australia

Daniel T. L. Smith, Qiaomin Chen & Scott C. Chapman

Center for Crop Science, Queensland Alliance for Agriculture and Food Innovation, The University of Queensland, St Lucia, QLD, 4072, Australia

Sean Reynolds Massey-Reed & Andries B. Potgieter

You can also search for this author in PubMed   Google Scholar

Contributions

Daniel Smith: conceptualized the idea, collected the data, performed the analysis, and prepared the manuscript. Qiaomin Chen: assisted with the analysis, contributed to conceptualization, and participated in manuscript editing. Sean Reynolds Massey-Reed: assisted with analysis, developed code for trait extraction. Andries Potgieter: contributed to manuscript editing. Scott Chapman: conceptualized the idea and contributed to manuscript editing.

Corresponding authors

Correspondence to Daniel T. L. Smith or Scott C. Chapman .

Ethics declarations

Ethics approval and consent to participate.

Not applicable.

Consent for publication

Competing interests.

The authors declare that there are no competing interests regarding the publication of this paper. This research was conducted independently and impartially, with no financial or personal relationships that could have influenced the work reported in this document.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

13007_2024_1236_moesm1_esm.docx.

Supplementary material 1 Fig. S1 Relationship between the Date of biomass cut and the date of the corresponding UAV flight for each experiment. Point colours represent the trial mean Zadok’s stage at a given timepoint. Fig. S2 Examples of the thresholding methodology that was used in the study. The panels include a true colour image, the OSAVI vegetation index and binary masks after thresholding using Otsu’s method. Table S1 Overview of number of samples included for each Variable and Growth-stage combination assessed in this study, including number of observations in the train set and test set, along with the number of input features included after filtering for co-linearity. Table S2 Overview of selected variables after recursive feature elimination (RFE) for each of the 12 variable x growth stage combinations. Text colour indicates whether a variable is geometric (blue) or spectral (orange). The numbering of the columns indicates the order in which each variable was selected. Table S3 Train and Test performance metrics for each of the Growth-stage x variable set x Model types on the Permanent ROI. Table S4 Abbreviation, calculation, name, and key reference for the Spectral indices used as input to biomass prediction models in this study. Table S5 Calculation of Geometric Variables for use as input variables for biomass prediction models. Table includes the name of the trait, a description of how it was calculated, and a relevant reference. Fig. S3 Correlation Matrix illustrating the relationship between all variables and all timepoints, ordered by the angle of eigenvectors (AOE). Fig. S4 Observed versus predicted DW AGB (g/m 2 ) using the Random Forest model trained using geometric and spectral variables on the independent test set. Vertical facets represent the different experiments in the study and the horizontal facets represent the DW AGB cuts in numeric order. Point colours represent the cumulative thermal time (TT cumulative ). Metrics for each cut x experiment are shown in each facet.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Smith, D.T.L., Chen, Q., Massey-Reed, S.R. et al. Prediction accuracy and repeatability of UAV based biomass estimation in wheat variety trials as affected by variable type, modelling strategy and sampling location. Plant Methods 20 , 129 (2024). https://doi.org/10.1186/s13007-024-01236-w

Download citation

Received : 23 January 2024

Accepted : 11 July 2024

Published : 20 August 2024

DOI : https://doi.org/10.1186/s13007-024-01236-w

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Wheat biomass estimation
• UAV based monitoring
• High throughput phenotyping
• Field-based

Plant Methods

ISSN: 1746-4811

• Submission enquiries: [email protected]