Guidelines for designing and interpreting drought experiments in controlled conditions

Abstract

As Journal of Experimental Botany (JXB) editors, we often receive manuscripts on drought tolerance and plant responses to water deficit. We have observed that the quality of research in this field frequently suffers from flawed experimental designs, inconsistent terminology, overinterpretation of data, or unrealistic lab-to-field extrapolations. To tackle these challenges, the JXB Editorial Board established a working group to guide better experiment design, data interpretation, and reporting of results, focusing on experiments performed in supposedly ‘controlled conditions’. Our recommendations include the following. Utilizing precise, consensus-driven terminology to clearly communicate objectives and hypotheses. Designing experiments that account for the complexities of genotype–environment (G×E) interactions, by including sufficient biological replicates, conducting multiple experiments, and measuring soil and plant water status as well as microclimate variables. Considering that whole-plant transpiration interacts with pot size and soil substrate to alter soil moisture and stress levels, and acknowledging that plant responses to drought depend on, and also affect, their growth dynamics. These guidelines aim to enhance research quality, contributing vital knowledge to combat the growing threat of drought to agriculture. Achieving agricultural sustainability is essential to meet the nutritional needs of a growing population (Scherer et al., 2018; Hinz et al., 2020). Among the threats to which agriculture is exposed, drought consistently claims first place. Despite this prominence, clear standards for conducting drought experiments are conspicuously absent. As a consequence, researchers often operate according to their own perspectives and principles, and this practice hinders reproducibility and reliable interpretation. As the name implies, the Journal of Experimental Botany has always had a focus on experimental botany. Therefore, as scientific editors of the journal, we believe it is our responsibility to raise, discuss, and advocate best practices in drought experimental design. This should help both researchers and decision-makers with scientifically grounded results. As JXB editors, we regularly receive manuscripts that address drought tolerance issues and/or plant responses to water deficit. However, we share a concern that a significant number of manuscripts show serious flaws. Two main reasons are that: (i) the precautions necessary for unbiased experimental design and conclusions are not taken; and (ii) conclusions are overstated, particularly in relation to the significance of results obtained in controlled conditions as they relate to field performance of crops. Here, our objective is to provide best-practice guidelines that address the research challenges specifically associated with drought experiments when they are performed in ‘controlled conditions’—in greenhouses or growth chambers which represent a large proportion of submitted papers on drought. These experiments are most often conducted in pots, where biases can occur due to improper management or characterization of drought, hampering comparisons of treatments or genotypes. A second problem is the gap between the ambition and achievements of many drought studies. A common claim is to have identified mechanisms and processes involved in securing or enhancing crop yields under drought stress, when in reality the study was based on reductionist experimental designs with limited data collection and/or analyses of limited general relevance, potentially leading to misinterpretation of the results. A third problem with the drought-related literature is a lack of consistency and coherence of terminology, despite well-defined terms that have been established for >50 years to characterize the different strategies plants adopt when faced with water deficit (e.g. escape, avoidance, or tolerance, see below). Moreover, the overall, and maybe overarching, issue is the lack of recognition of the fact that none of these strategies defines a drought resistance outcome, which depends on the objective and the climatic scenario. For instance, a ‘drought-avoidance’ phenotype related to early stomatal closure can be either beneficial when soil water is limited and water needs to be saved to complete the growth cycle, or detrimental if water is only lacking during limited periods (Hu et al., 2022). Similarly, other strategies, such as escape (e.g. early flowering) or tolerance (e.g. maintaining growth under low water potential), may have positive or negative impacts on plant survival and yield. Unfortunately, ‘drought resistance’ and ‘drought tolerance’ are often used indiscriminately. Proper use of generally accepted definitions in drought-related papers would be a major step in the right direction (Volaire, 2018). We hope that these guidelines will not only help authors to design their studies, but also help reviewers and editors to assess the robustness and transferability of results. Investigations of plant responses to water deficit have followed a common path, starting from the theoretical and basic research phase, through the pre-field phase, and culminating in field trials. Field experiments test whether laboratory findings apply in real-world scenarios. Like human clinical trials in medicine, field trials represent the ultimate test for our mechanistic understanding of drought responses, but their scope and capacity are inherently limited. While acknowledging the pivotal role of field experiments in validating strategies for crop yield improvement, this Editorial addresses the basic and pre-field experimental approaches, with particular focus on drought and high-temperature stress that is often associated with drought. Experimental botany is important during the pre-field phases, specifically in pot-based studies conducted in growth chambers and greenhouses with varying levels of environmental control. These studies facilitate systematic hypothesis testing and the dissection of complex mechanisms. By enhancing the experimental methodology during these foundational stages and accurately describing the growth conditions, we aim to foster more effective, efficient, and well-informed field studies in the future. This Editorial advocates classical methodical investigation, with particular emphasis on plant stress biology studies, in a strategy outlined here in four ‘D’s: ‘Define, Design, Develop, and Device’ (Fig. 1, and explained thereafter). Such a structured approach might assist researchers in accounting for the complex experimental dynamics. Sequential framework for designing and conducting plant drought experiments: the four ‘D’s checklist approach. To ensure scientific rigour and reproducibility of stress experiments (drought in particular), we propose the following four sequential steps. Firstly, defining the overarching research question or problem is crucial in any research endeavour. A classical question reflecting the issue discussed in this paper might be: ‘Can we improve yield production under drought stress?’ Following this, when defining the research objective, we recommend avoiding overly general terms (e.g. ‘improving plant drought tolerance’), and instead aim for more specific and quantitative objectives (e.g. ‘minimizing the yield penalty caused by drought’, or ‘understanding the molecular mechanisms regulating root length for reaching deep water’). This precise and, ideally quantitative, definition streamlines the subsequent stages, particularly hypothesis development. Furthermore, it also aids in conclusively determining, during the discussion phase, whether the study has achieved its research goals. The second stage involves developing specific, testable hypotheses or predictions. For instance, the hypothesis could be: ‘the cultivars with long roots will reach a deeper water table enabling the crop to maintain water uptake, thereby yielding better under the drought period compared with cultivars with short roots’. Another example would be. ‘Gene X expressed in the guard cells will increase stomatal sensitivity to ABA, thereby inducing faster closure at less negative soil water potential, thus enhancing plant survival’. In the latter example, special attention is needed for pleiotropic effects on leaf area or flowering time, both of which directly impact the targeted process (transpiration) yet via totally different mechanisms. In the third stage, the experiment should be designed to measure all key environmental variables as well as the biological variables of interest throughout the experiment period. The final stage selects appropriate monitoring devices/probes and their arrangement. Importantly, considering the dynamic nature of the G×E interactions, the variables of interest should be monitored as often as needed to allow unbiased interpretation of the results over the whole duration of the experiment. Precisely defining a research objective should be the first step in any study. For drought stress experiments, the performance or fitness of a crop or wild-type plant in its agronomic or natural environment should be evaluated with measurable criteria. For several decades, definitions describing a plant’s response to stress have been relatively broad. Levitt, a pioneer in the standardization of plant stress responses, suggested quantitative definitions that characterize plant reactions to stress in a more general way (Levitt, 1980). Nevertheless, many studies use the term ‘stress’ very broadly, which necessitates a more precise and quantitative definition (see Gaspar et al., 2002; Kranner et al., 2010), especially if the research objective is to minimize yield penalties under conditions of drought stress or to enhance survival under more severe conditions. In studies of stress responses in annual crop plants, vague or oversimplified research objectives such as: improving the plant’s ‘tolerance’, ‘resistance’, or ‘resilience’ to stress are all too common and generic (Box 1). Poorly defined research objectives impede the researcher by producing unfocused research, and confuse the reader with ambiguous information. For instance, desert plants known for their drought survival and high water-use efficiency (WUE; the ratio of carbon fixation, or biomass gain, to water loss, or transpiration) are often hypothesized to have advantageous traits to manage various environmental stressors (Nobel, 1988). However, these traits, while enhancing survival, are often associated with slow growth that causes low productivity and might not align with agronomic objectives aimed at optimizing yield in annual crops under milder stress. Thus, clearly defining research objectives is essential, as it helps distinguish between strategies focused on survival in harsh conditions and those aimed at maintaining yield under milder stresses. Such differentiation is key to tailoring research and interventions that meet specific agronomic or ecological goals. By establishing clear objectives, researchers gain insights into the intricate interactions between the plant and its environment, identifying genetic, epigenetic, and environmental factors contributing to observed responses. Defining the objective helps us to understand the complex interactions and adaptations that occur within ecosystems or field conditions. This analysis, however, is only as reliable as the shared language and concepts we use to frame our research questions. As we push the boundaries of our knowledge, it is essential that we ensure consistent and uniform terminology, similar to standards in physics. Deciphering plant responses to environmental stressors is not straightforward. The response network depends on evolutionary adaptations, survival strategies, and productivity trade-offs. To advance our understanding of these mechanisms, developing testable research hypotheses and predictions, in accordance with the research goal (see above), is critical. Hypotheses or predictions are not only essential for developing the proper experimental design, but they should also be explicitly outlined when the study. of precise definitions and terminology in stress biology research is essential throughout all stages of the research this from the very of a study helps the researcher to hypotheses that are not only and but also of and For example, terms such as and are often used to However, in plants that are more to environmental to more thereby enhancing their of ecological in the the term would more accurately plants that are by stress. The following the use of consensus-driven plants that are to stress will more response (e.g. stomatal closure with soil water thereby improving the plant’s of survival (e.g. of its size or a response (e.g. and stomatal closure with soil water may the and for example but plants to of detrimental their survival Therefore, stress sensitivity is a ecological resistance when the objective is to while stress may enhance agronomic tolerance (e.g. maintaining production despite the of the While stress may enhance yield under to conditions, it will in plant to the with stress Thus, this might be beneficial under to stress, but the plant of or to more severe stress (see more in et al., 2018). to stress can stress mechanisms, plants stress directly with These mechanisms include stomatal developing and deeper roots to early flowering to escape the stress or tolerance as in However, these mechanisms often to in potentially growth and and productivity and yield This their agronomic To accurately assess the and capacity of several stress mechanisms, it is to measure the plant’s or its stress plant’s to from stress et al., This example, in transpiration or biomass the and of the plant’s stress mechanisms throughout the stress the duration and of soil water deficit is essential to the of different to A stress mechanisms that maintain essential (e.g. or root and stress. slow less more or less of and a period. the intricate between plants and their environment a experimental due to the and of environmental factors deficit and soil water and the biological responses they and 2022). This a to the researcher when designing experiments and the The issue can be more complex as environmental conditions can depend on other traits, such as soil water status that depends on Moreover, with different whole-plant transpiration levels under the environmental conditions will to in soil moisture While researchers aim to various to stress, in whole-plant transpiration of soil it to different to similar of stress (see for in the experimental design and stress studies. The common experimental In pot experiments, the most common biases from a lack of the pot water the between plant transpiration and the water to it in the these experiments aim to apply stress levels several genotypes. However, in transpiration in of plants to varying soil water Here, we common drought this experiment was designed to assess the drought responses of genotypes. establishing plants in conditions, the soil in the was to at by the own transpiration and soil any control. However, this experimental design not the to similar stress it and biological between As a cultivars with transpiration which are also to more and thus will faster and more severe stress compared with those with transpiration this the However, it is a of transpiration not a drought Moreover, it to the under the stress the plants with transpiration not to drought conditions to those with Therefore, plants are to a faster of soil water This stress can alter important These include whole-plant transpiration and where is as the final Another example of stress experiment involves stress, which is by to high of in under a experiment. this all plants to a and However, plants with transpiration a those with a transpiration during the of the experiment. Thus, better growth and transpiration of these plants could to faster soil water deficit. In this drought causes in that from the experimental conditions in whole-plant transpiration and could that the plants with transpiration are they not to stress conditions drought and similar to the is important to that transpiration is with Thus, these experimental and may plants with transpiration and thus production as explained major potentially from the experimental design to the of is a of causes and Nevertheless, when working with the and can to in stress due to in soil While plants are often in greenhouses to provide environmental a For instance, in causes and which in the that plants et al., 2018). This environmental not only by the but also by and conditions, a for the Furthermore, the of a plant within the can its due to in and other plants and within the a approach and/or that account for these et al., these interactions a due to plant responses to their However, the is often limited in terms of plant and Therefore, we recommend that (i) and the in their experimental (ii) to assess the impact of complex environmental interactions (e.g. and account of G×E interactions when their Considering plant mechanisms in response to environmental variables is crucial for such as drought or occur and Furthermore, of experimental conditions, where the or of a process the or of the be For instance, can the conditions of their environment or plants can alter their soil et al., 2022). Thus, the experimental design should allow monitoring and soil water status and other related climatic conditions throughout the which may and interactions are be to designing the experiment and the the of in these dynamic experiments is due to different effects in response to various environmental conditions. To ensure when plant responses, it is crucial to to minimize and effects and account for in such as flowering time, by both and during Another important when designing a stress experiment the pot by the duration of the stress. all plants only at a in to many that plants within a group the of stress compared with the However, biological in stomatal and/or and in conditions in transpiration and thus different stress levels of soil water As the pot is any water the plants to different levels of stress. This of results may a biological (Fig. As a the of different transpiration This is a very common and a major of of drought stress manuscripts by Another in experimental design is stress This when within a group different from despite all to the plants can through transpiration and reporting only but not can in data that plant responses to stress. despite all plants to the their biological responses could due to in or other should be and (Fig. Moreover, the plant’s the thereby it is a major to and in growth chambers the which can often stress the very should be monitored and and its impact on transpiration should be such as to root et al., and root and a soil and between within the help to and more facilitate better to root may occur at the to to a pot may in soil moisture within the which can plant responses to soil water et al., While of these issues are best by pot size according to the experimental reporting soil such as soil moisture and water status will help the experimental results. but and precise reporting of all variables is as only the of these data the and reproducibility of the experiment. a standardization in stress experiments of several questions. is the experimental structured and is the biological are should these be to accurately different and these is essential to authors to and ensure of plant stress other important biological will such as the stage during and the of the compared other factors specific to the research While these biological they not alter the in the Therefore, while acknowledging their we will on these specific biological issues and in the following. the experimental conditions a significant particularly if for consistent of stress levels and to all plants throughout the experimental despite in and within the To address this can either or more These can be used to plant’s conditions to meet the stress For instance, a for cultivars under similar water stress conditions in could in the This that they are to stress levels and However, such a may and interactions, and other processes that are the focus of the experiment. are also These include measuring plant’s transpiration and soil water with of water in relation to the plant and et al., Such are in and can a large number of plants, thereby experiments et al., et al., and et al., 2020). For a limited number of pots, such and soil water can also be achieved In any plant stress the variables should be according to the specific of the experiment. These variables can and these variables should by the researcher (e.g. plants or may be to the impact of the stress on the of the but on their own can to and Therefore, quantitative data, such as those related to plant and soil water are needed as a for analysis, with a interpretation of the However, it is essential to understand that plant responses to stress are in nature Thus, such as leaf et al., and stomatal or specific molecular and (e.g. or are and to in stressors et al., and such as plant biomass or leaf may to is also important to that can be as stress For example, a less leaf might but it could of maintaining transpiration and instead of to which could a plant et al., As should not be the of stress Moreover, the of plant stress responses can be varying based on factors such as plant stage, and environmental conditions. In such as and water potential, is to stress levels et al., These are in the response and are the first to is important to that the are and which traits to measure and the researcher should their with of conditions. allow more monitoring of plant and environmental in the and the and of this goal within reach for many researchers et al., the appropriate and specific for particularly when stress in plant stress To understand plant stress responses, it may be to a more study that stress, and response a group not to soil water deficit or This design that any observed can be specifically to the and not to natural or other also understanding of the plant and potentially from stress over As due to the dynamic nature and of experimental conditions, based on the a is very to to for example soil water whole-plant in when stress should be based on plant For example, the when transpiration to should the of drought stress, as it that soil water has to the of The to this is to traits and their considering as is approach for plant stress responses. variables to measure and results be to the experimental and Nevertheless, to the suggested will enhance the experimental design, analysis, and of the study. The aim of this is to drought related stress such as effects on plants to of the in this and thereby help to design the experiment for the Our main recommendations are as (i) to the objective in precise (ii) to a and testable to design experiments that will allow and testing of the and to to the monitoring and of environmental conditions, thereby enhancing the and reproducibility of stress This approach not only the reproducibility of experiments but also the of the We that and can in the experimental design, such that not all recommendations can be This not of the however, in such it is essential that the data interpretation and discussion the in the experimental design. is threat to natural ecosystems and agricultural and its impact on both is to increase with The plant is at the in this but in the impact of drought on plant will depend on a proper of plant responses to drought. We hope that our recommendations for the design, interpretation, and reporting of experiments will help plant researchers to this We the Journal of Experimental Botany for us to these at and the for a at that discussed these The authors of

Topics & Concepts