Plant “growth” (total frequency of 7%) was the most frequently used keyword ( Figure 2 ). This is likely because drought and recovery represent different periods of plant water status and growth connects them and is also frequently used as an indicator of drought and recovery. Additionally, in contrast to the many shoot-only studies that are not part of this analysis, root-related studies usually take a whole-plant approach. When we ignore common words like “stress”, “drought”, “tolerance”, “response”, “plant” and “water-stress”, the high-frequency keywords in root plasticity during both drought and recovery research can be grouped into 3 types: 1) physiological factors: photosynthesis, gas-exchange, abscisic-acid, (stomatal) conductance, osmotic adjustment, accumulation, and transpiration; 2) different plant species: Arabidopsis, wheat, maize, and trees; 3) abiotic factors: temperature, salinity, nitrogen, and climate change ( Figure 2 ), which will be discussed below. No words related to biological interactions got into the top 50 list, and the first biological factor “fungi” was found in the top 150 list and appeared only 12 times and thereby had a frequency of 0.2%.

For physiological factors, higher frequency words “photosynthesis”, “gas-exchange”, “stomatal conductance”, and “transpiration” could reflect that aboveground traits received more attention than belowground traits. This shoot-dominated focus continued despite the vital role of roots in determining plant ecology, terrestrial ecosystem functioning, and their designation as the target for the second green revolution (Gašparíková et al., 2002; Lynch, 2007). This is partly because plant water uptake and transport are generally thought to be regulated by the stomata (Li et al., 2020), but probably also due to technological limitations in monitoring root growth dynamics and studying the hydraulic pathways in the root system. Using non-destructive technologies like MRI (Pflugfelder et al., 2017) and SWaP (Dusschoten et al., 2020), a recent study focussing on faba bean and maize has proved that stomatal sensitivity is partly explained by the sensitivity of root hydraulic conductance to soil drying (Müllers et al., 2022b). A better understanding of the role of root conductance in soil drying and rewatering is vital to complete the picture from soil to root to leaf. Associated with the focus on stomatal conductance and morphological leaf traits, “abscisic acid” was another important physiological keyword. For example, Correia et al. (2014) showed that both abscisic acid (ABA) and ABA-glucose ester (ABA-GE) are up-regulated during drought and down-regulated during recovery in the Eucalyptus globulus. Besides stomatal closure, ABA controls physiological processes like osmotic regulation, growth inhibition, and transcriptional regulation of stress-responsive gene expression (Zhang et al., 2006; Li et al., 2020).

Of the studied species, the model plant Arabidopsis (total frequency of 81, with a relative frequency of 2%) was the most popular ( Figure 2 ). Its popularity is associated with its small size, relatively short lifecycle, ease of growing under low-light lab conditions, and small genome which is instrumental for studying processes at the molecular level. In addition, wheat and maize as important cereals and worldwide staple food, were other popular species. From the first three species, it is clear that herbaceous plants are much more often researched than trees. However, as a group, “trees” still had a relatively high frequency (total frequency of 30, with a relative frequency of 1%). Trees are perennial plants with longer life spans and thus are likely to experience temporal drought, some mortality happens when these plants suffer from hydraulic failure; 2) Tree species generally have thicker stems and large enough vessels than tiny plants which make it easier to monitor water transport non-invasively (Brodribb et al., 2017).

Drought stress strongly interacts with other abiotic stresses. Drought stress is often accompanied by heat stress and is aggravated by salt stress. Salinity also can cause similar problems with drought stress due to the high osmotic potential in the soil, which leads to similar response patterns in plants, consequently, some scientists compare salinity and drought stress effects in their research (Sánchez-Blanco et al., 2014; Koevoets et al., 2016; Ma et al., 2020). Concomitantly, nutrient availability and uptake are inhibited by dry soil (Parrondo et al., 1975; Hira and Singh, 1977; He and Dijkstra, 2014). Climate change may exacerbate these interactions. Although it is challenging to study such interactions, they received relatively much attention as indicated by the frequencies of the words “temperature”, “salinity”, “nitrogen”, and “climate change” ( Figure 2 ).

The co-occurrence network of the keywords revealed three clusters which we labeled: “plant growth”, “drought tolerance”, and “root hydraulics” ( Figure 3 ). “Growth” was the most popular keyword in the plant growth cluster (i.e., red cluster). Like the word cloud results, this cluster described the plant growth response in association with aboveground traits like “photosynthesis”, “leaf”, “gas exchange”, “chlorophyll fluorescence”, “stomatal conductance”, and “transpiration”. Even though the initial literature search included the keyword “root”, leaf traits had a higher frequency in this cluster, indicating that the study of plant response to drought and recovery is strongly focused on aboveground parameters. The plant growth cluster included words like “trees”, “forest”, and “yield”, hinting at a more agroecological context. For forest ecosystems, climate change has caused more frequent drought events, and scientists have focused on different tree species’ growth under drought and other environmental factors like soil nitrogen deficit and higher temperature. In agricultural ecosystems, wheat production has been the main focus of research. Wheat is known to be deep rooting and relatively tolerant to drought compared to other major grains (Fan et al., 2016).

The plant physiology cluster (blue cluster) contains words associated with (molecular) plant physiology such as “drought”, and “tolerance”, which were the most popular keywords in this cluster and were strongly associated with words like “gene expression”, “osmotic adjustment”, “oxidative stress”, and “abscisic acid” (ABA). In contrast to the “red” cluster, which mainly focuses on plant growth response, the blue cluster mainly focuses on gene and hormone regulation. The keywords in the blue cluster were associated with the model plant Arabidopsis, indicating that molecular physiology is commonly studied in this plant. “Abscisic Acid” bridges back to the “red plant growth” cluster through keywords like “response” and “photosynthesis” ( Figure 3 ). Mechanistically, Abscisic Acid regulates the stomatal response and thereby directly influences photosynthesis and growth.

The root hydraulics cluster (green cluster) included “root” as the most popular keyword and its relation to “conductance”, “hydraulic conductivity”, “transport”, and “maize” ( Figure 3 ). This cluster comprises research on water transport from the soil through the roots into the shoot. Hydraulic conductance of soil, rhizosphere, roots, and xylem is of great importance to understanding water “transport”. The word “root” was only weakly connected to leaf and physiological traits, possibly indicating a discrepancy in the research. This is surprising, given that plant hydraulic conductance, transpiration, and CO₂ uptake, are regulated via stomatal opening and closure. The ABA-controlled regulation of stomata is, however, sensitive to soil hydraulic properties (Carminati and Javaux, 2020; Abdalla et al., 2021), interacting with the root length and morphology (Müllers et al., 2022a). The mechanisms are still strongly debated (Li et al., 2020). In addition, it seems that root and water-transport researchers have chosen “maize” as their favorite model species. Possibly, because maize has a rather sturdy root system which is more easily studied than the fine roots of Arabidopsis or wheat.

The frequently used keywords have changed over the past five decades. From 1975 to 2005, the topics ranged from whole plant physiology to molecular response, but roots were not in focus. Only in recent years, from 2006 to 2022, keywords related to root plasticity to drought and post-drought recovery became more frequent. During this period, the focus shifted to applied aspects such as “yield” and “climate change” as climate change made the need for resilient crop yield more imminent. The keyword “growth” was frequently used in all periods, as it is fundamental to our definitions of drought stress and recovery ( Figure 4 ).

During 1975-1995, the main focus was on plant physiology as indicated by the frequent use of words such as “osmotic adjustment”, “growth”, (cellular) “injury”, “abscisic acid” and “conductivity”. Some scientists also investigated the impact of microbes on drought response like endophytic fungi and plant growth-promoting rhizobacteria (PGPR), which explain the appearance of keywords like “infection” and “bacteria” (Ruizlozano et al., 1995). With the advancement of molecular technologies, “expression”, and “oxidative stress” increased from 1996-2005, indicating that more scientists devoted themselves to identifying genes involved in drought tolerance. During 2006-2015, research emphasized mechanisms using model plants and an increasing interest in roots, with frequent keywords: “root”, “gas exchange”, “mechanisms”, “Arabidopsis”, and “superoxide-dismutase”. After 2015, the number of studies increased, and more researchers demonstrated that the recovery period matters to the overall plant performance and that is especially true in the field station under climate change (Hagedorn et al., 2016; Guo et al., 2021; Santos-Medellin et al., 2021). During this period, keywords such as “recovery” and “climate change” were used frequently.

The temporal evolution of keywords’ frequency shows keywords that were gradually used less, like “abscisic acid”, ones that remained stable, like “growth”, and those that gained more attention in recent years, like “climate change” and “yield” ( Figure 4 ). Therefore, we added a thematic analysis to further understand the temporal evolution of keywords. The thematic analysis distinguishes “motor” from “basic” themes. Motor themes are both important, well-developed, and highly cited in recent years, whereas basic themes are the main and driving keywords for the research topic but receive a few citations (Cobo et al., 2011; Liu et al., 2021). To the motor themes belonged words like “water stress”, “stomatal conductance”, “abscisic-acid”, “chlorophyll fluorescence”, “biomass”, “proline”, and “hydraulic conductivity” whereas “drought”, “recovery”, “photosynthesis”, “root”, “climate change”, “growth” and “resilience” were the basic themes ( Table 2 ). Although “water stress”, “abscisic acid”, “proline” (osmotic adjustment), and “conductivity” were motor themes and important, they were studied intensively in the early stage ( Figure 4 ). The root supports growth through water and nutrient uptake, transport, perception, and signaling. Thematic evolution identified oxidative stress and nitrogen as basic themes, well developed in 1975-2005 ( Figure 4 , Table 2 ). In contrast, “recovery” and “root” (note these were part of our initial search terms), and “climate change” had a stronger development during recent years. This is likely to continue in the near future as climate change demands agro-ecological adjustment to the increasing risk of temporal drought.

We expect basic themes, like root, recovery, climate change, and their relationship to yield and leaf traits, will drive future studies. The challenges of root and rhizosphere dynamic measurement will require the deployment of innovative technologies to accelerate root science. New technologies and methods, like noninvasive root and rhizosphere phenotyping, will be key to understanding root dynamics during drought and rewatering (Wasson et al., 2020). Mathematical modeling will also be important to simulate mechanisms of water transport as well as the discovery of plant traits for greater crop resilience and faster recovery (Hall, 1982; Carminati and Javaux, 2020; Maurel and Nacry, 2020; Javaux and Carminati, 2021; Joshi et al., 2022). Additionally, understanding the interactions with other abiotic stresses will be crucial in the context of “climate change” and “yield”.

Over the past 120 years, the number of publications on “root plasticity during drought and recovery” increased strongly, but still, the topic seems to be underdeveloped compared to shoot-related research ( Figure 5 ). We distinguish three periods: 1900 to 1990, 1991 to 2004, and 2005 to 2022. For the first period, only 4 publications were found (in 1975, 1983, 1986, and 1988). During this period, few publications were uploaded to WOS and most researchers focused on drought but not recovery. The number of publications during that period was low in all sciences, but especially so in root research as measuring roots was challenging and few technologies were available. Thus, 99% of the analyzed publications were published in the last three decades with an annual increase of 7.20%/a, greater than the annual growth rate in Life Sciences of 5% (Bornmann et al., 2021). The results also revealed that the publications number improved very slightly from 1991-2006 with around 15 papers each year, while they dramatically increased during 2015-2022 ( Figure 5 ) with 80% of publications found after 2005. The trend of root plasticity during both drought and recovery research was consistent with the trend in the related emerging topic of “carbon exchange in global drylands” (Liu et al., 2021).

USA and China were the most productive and cited countries. Among the 10 countries with the highest number of publications, the USA (relative frequency=71%) and China (69%) had a similar frequency and were 3 times higher than the third most productive country. Spain, Brazil, and Australia, at position 3-5, had a similar frequency of around 200 ( Table 3 ). Although the most productive are also the most cited countries, USA-based publications were cited 2.2 times more often than those from China, with 14 vs 6 citations per publication. Spain, France, Australia, Italy, India, and Germany were positioned at 3-8 with 12-13 citations per article ( Table 3 ). Three reasons may explain these ranking patterns: all these countries 1) have arid regions with severe drought stress; 2) have advanced technology for root research; 3) have many researchers and a higher GDP that ensures enough human and material resources for related research.

The collaboration map shows a similar ranking: USA (164), China (136), Germany (105), Spain (87), and Australia (74) had the highest collaboration frequency of all countries. Even though Germany’s ranking in the most productive and cited countries is not high, it still ranked in the third position according to the collaboration frequency, which indicates that Germany had many collaborations with other countries but less direct ownership in the topic. Germany is well known for its development of technologies, root phenotyping and soil-plant hydrology and as such is a looked-for partner, even though its agriculture is less threatened by drought compared to the other listed countries ( Figure S1 ).

Our results showed that 1429 institutions all over the world have participated in root plasticity during both drought and recovery. The top ten most productive institutions contributed 27% (294 publications) of the total publications ( Tables 1 , 3 ). From 1975 to 2022, the Northwest A&F University in China ranked first with the most publications (48), followed by the University of California, Davis (USA), China Agriculture University (China), Khon Kaen University (Thailand), and University of Western Australia (Australia). Consistent with the most productive and cited countries, 9 of the top 10 institutions belong to China, USA, or Australia. Khon Kaen University in Thailand stood out, as Thailand did not rank high in the country ranking.

Some highly productive authors affiliated with Khon Kaen University are A. Patanothai and S. Jogloy. For example, they concluded that during drought and recovery, peanut pod yield was associated with increased root surface area deeper in the soil (Jongrungklang et al., 2012; Jongrungklang et al., 2014). This is the only study we found that proposed a root ideotype for both drought and recovery, although there are other ideotype studies that focused on drought only.

Studies on root plasticity during both drought and recovery were published in 334 journals, most related to botany or agronomy ( Table 1 ). The top ten most productive and cited journals focus on plant research across scales from molecules to ecology. Plant and Soil, Tree Physiology, Frontiers in Plant Science, Agricultural Water Management, and Journal of Plant Physiology were found to be the five most prominent journals, with a note that Frontiers in Plant science is a relatively new journal. Tree Physiology, Plant Cell and Environment, Journal of Experimental Botany, New Phytologist, and Plant and Soil were the five highest topic-h-index journals ( Table 4 ). In general, the h-index seems low, indicating that the topic is not receiving much attention despite its societal relevance. Plant Physiology, Tree Physiology, Journal of Experimental Botany, Plant Cell and Environment, and New Phytologist were the five journals that scored highest in total number citations, indicating the greater interest of publications in these journals as a source of bibliographic consultations.

Eight of the top ten journals have a relatively broad scope in plant sciences, except for Tree Physiology and Agricultural Water Management, which have a strong focus on trees and crop water management, respectively ( Table 4 ). Because drought stress is closely related to soil water content, drought events are easy to appear in a natural ecosystem like a forest ecosystem, so it’s not hard to understand why “Plant and Soil” and “Tree Physiology” were the most productive journals.

In the SCI-E database, 1086 publications were found when searching across the last 120 years. The oldest publication was published in 1975 by Parrondo et al. entitled “Rubidium absorption by corn root tissue after a brief period of water stress and during recovery” published in Physiologia Plantarum, showing that reductions in rubidium uptake during a short period of water deficit only partially recovered during the post-recovery period (Parrondo et al., 1975). The effect of drought and rewetting on nutrient availability remains an important topic today. In the past, the focus was strongly on soil physical effects, such as reduced effective diffusion in dry soil (Hira and Singh, 1977). Recently a more complicated picture emerged involving the microbiota which influences phosphorus concentrations and sorption rates (Chen et al., 2021). Thereby, it still remains a question of how nutrient availability and uptake influences plant growth during drought and recovery and few papers deal with the question after Parrondo’s initial work. The study conducted by Xu et al. “Expression of a late embryogenesis abundant protein gene, HVA1, from barley confers tolerance to water deficit and salt stress in transgenic rice” (Xu et al., 1996), published in Plant Physiology in 1996, was the most cited article with 655 citations or 24 citations per year ( Table 5 ). This study proved the important role of plant LEA (late embryogenesis abundant) proteins under drought and salt stress and its potentia1 for genetic crop improvement toward abiotic stress tolerance. Due to their versatility, LEA-related genes and their function have received much attention as potential drought and salt tolerance genes. Twenty six years later, LEA has been widely studied in the context of drought, still, its functioning is debated and real-world agronomic application is still a promise (Hernández-Sánchez et al., 2022). Among the top ten highly cited papers, two focused on genes (Xu et al., 1996; Swindell et al., 2007), one on proteins (Salekdeh et al., 2002), one on rhizosphere bacteria (Mayak et al., 2004), one on roots and branches hydraulic failure (Anderegg et al., 2012) and one on root-water transport modeling (Hsiao and Xu, 2000) ( Table 5 ). Surprisingly, despite our root search term, there wasn’t any highly cited paper (>295) on root morphology or architecture. Root research on this topic of drought and recovery should get more attention in the future.