Physiological, Morphological and Root System Architecture Acclimation Responses to Drought in the African Orphan Millet White Fonio (Digitaria exilis)

doi:10.21203/rs.3.rs-4850590/v1

Download PDF

Research Article

Physiological, Morphological and Root System Architecture Acclimation Responses to Drought in the African Orphan Millet White Fonio (Digitaria exilis)

https://doi.org/10.21203/rs.3.rs-4850590/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Aims

White fonio is an ancient West African orphan millet crop. As one of the world's fastest-maturing cereals, it helps smallholders mitigate mid-season hunger. There are no reported studies on drought acclimation responses in fonio to identify traits that can enable breeding for climate change adaptation.

Methods

Here, two white fonio accessions from wetter (Guinea) and dryer (Mali) environments were grown indoors under three moisture levels in replicated trials. Physiological, morphological and metabolomic phenotyping was undertaken, including root system architecture analysis, culminating in measuring ~ 27,000 root hairs.

Results

Fonio responded to drought by dramatically upregulating glycine- and alanine-betaine leaf concentrations. Both accessions exhibited increased root:shoot ratio and leaf angle, but reduced shoot branching, leaf width, transpiration, and stomatal conductance. Grain yield most positively correlated with SPAD chlorophyll. Lower grain fill was observed in the Guinea accession, while the Mali accession showed a surprising increase in the harvest index when exposed to extreme drought. In the Mali accession, drought resulted in fewer but longer crown roots, increased lateral root branching, and a greater density and lengthening of root hairs. In particular, leaf width, angle and SPAD were identified as cost- and time effective selection traits.

Conclusions

This paper has identified above- and below-ground drought acclimation traits in white fonio. These results build a foundation for future efforts to breed this crop to tolerate accelerating climate change, ultimately to assist vulnerable West African farmers.

Fonio

millet

drought

root system architecture

physiology

Africa

The impact of climate change on people’s livelihoods is greatest in Africa especially, Sub-Saharan Africa (Omotoso et al. 2023). The rate of temperature increase is 1.5 times greater in the African Sahel compared to the rest of the world (IPCC, 2018). Total rainfall in Western Africa has decreased over the past few decades resulting in a higher frequency of early summer drought, more intense and intermittent rains, and more frequent late-season rains associated with crop harvest, adding adversity to crop production (Biasutti, 2019).

Fonio is a small grain cereal grown by smallholder farmers in the semi-arid regions of West Africa for the past 4000 years (NRC, 1996). It is one of the fastest maturing cereals, as some landraces are harvested in less than two months (NRC, 1996; Blench, 2012). In the hot and arid subtropics of Mali, Guinea, Niger and Burkina Faso, fonio alleviates mid-season hunger prior to harvesting of primary crops like sorghum or pearl millet (NRC, 1996; Cruz and Beavogui, 2016). Fonio also has cultural significance for different ethnic groups, like the Dogon and Malinke peoples of Guinea and Mali, where it is used in different ceremonies (Cruz and Beavogui, 2016). It is a C4 grass that produces tiny grains (1000 grain weight: <0.5 g) (Fig. 1B, D) (Jideani and Akingbala, 1993). Its yield is low compared to other cereals (under optimal conditions), ranging from 0.7 to 1.3 t/ha (FAOSTAT, 2022; Gueye et al. 2015). Cultivation and post-harvest processing (threshing, dehulling and cleaning) (Fig. C-E) of fonio are highly labour-intensive, which are additional limitations of this crop that hinder its expansion (Cruz and Beavogui, 2016). On the other hand, demand for fonio as a nutritious and tasty whole grain cereal is increasing in local and international markets (in Europe and America), and modern recipes are evolving (Fig. 1F) (CBI, 2023; Maritz, 2022). There are two types of cultivated fonio: (i) black fonio (Digitaria iburua) which is mostly limited to parts of Nigeria, Benin and Togo (Vodouhe et al. 2007, Wang et al. 2021, NRC, 1996), and (ii) white fonio [Digitaria exilis (Kippist) Stapf], grown in the Sahel from Senegal to Nigeria (Fig. 1A). White fonio is a staple food for almost half a million people; it is rich in selected micronutrients (Fe, Zn), and essential amino acids (methionine, cysteine) (Zhu, 2020; Temple and Bassa, 1991).

Plants adapted to dry environments exhibit several acclimation responses which include biochemical, physiological and morphological changes (Berger et al. 2016). Osmotic adjustment is a known mechanism in plants to decrease osmotic potential in cells by increasing the concentrations of osmotic metabolites including organic acids, soluble sugars, trehalose, amino acids, proline and glycine betaine (Luo et al. 2010; Sanders and Arndt, 2012; Parthasarathy et al. 2019). These metabolites protect cellular proteins, enzymes, lipids and membranes from dehydration and damage during drought stress (Rajendrakumar et al. 1994; Morgan, 1984). Physiological acclimation responses include changes in the rate of stomatal conductance, photosynthesis, transpiration, and leaf chlorophyll (SPAD) content (Sallam et al. 2019; Khadka et al. 2020). Stomatal closure is one of the earliest and most common responses in plants, to prevent dehydration during drought (Medrano et al. 2002; Bertolino et al. 2019). Restricted opening of stomates prevents transpirational water loss during drought. Stomatal closure also limits CO₂ from entering leaves, resulting in a reduced rate of photosynthesis (Bertolino et al. 2019). Reducing the rate of photosynthesis is another strategic physiological adjustment to slow down metabolism when water is scarce (Reddy et al. 2004). On the other hand, ‘stay green’ is a genetic trait found in maize and other cereals that maintains chlorophyll to stabilize photosynthesis and delay drought-induced senescence (Thomas and Ougham, 2014).

Physiological changes are oftentimes complemented by aboveground and below-ground plant morphological changes in plants. These alterations under drought include reductions in leaf number and size/area as well as tillering/branching, and an increase in leaf angle in various cereal crops, including millets (Tiwari et al. 2022). Steeper leaf angles help to prevent damage to photosystems from intense sunlight during drought (Chaves et al. 2003). Similarly, reduced tillering/branching and smaller/narrower leaves contribute to limited shoot growth, leading to decreased aboveground demand for carbon metabolites during drought (Chaves et al. 2003; Pastenes et al. 2004). In general, plants alter their carbon partitioning to prioritize root growth rather than shoot growth to maintain water scavenging during drought (Hasibeder et al. 2015). Therefore, an increased root/shoot biomass ratio is another usual sign of drought adaptation in plants (Comas et al. 2013). Additionally, plants modify their root system architecture to further support efficient water scavenging. In cereal crops, the root system consists initially of a primary root and seminal roots, which are replaced by post-embryonic long crown roots that give rise to lateral root branches, all of which can be covered by root hairs, that are epidermal cell protusions that increase the surface area for nutrient and water uptake (Marin et al. 2021 and Lynch, 2022). Increased crown root length has been observed in cereals, which promotes water scavenging from a larger soil volume; this increase is is metabolically compensated by a reduction in the number of crown roots (Lynch, 2022). Finally, these macro- and meso-scale root traits are combined with increased root hair length and density under drought (Lynch, 2022).

Fonio is recognized for its capacity to thrive in marginal/nutrient-limited and dry growing conditions (Gigou et al. 2009; Cruz and Beavogui, 2016). Despite its value in terms of food security (especially in the context of climate change), nutrition and culture of West African peoples, it is considered an orphan crop in terms of scientific attention (Adoukonou Sagbadja et al. 2007). Therefore fonio farmers have been struggling with the same challenges for centuries. In particular, fonio is not as economically competitive as other high-yielding cereals like maize and sorghum, primarily due to the presence of ancestral wild alleles in fonio landraces that contribute to small grain size, seed shattering, and lodging (Barnaud et al. 2017; Vodouhe et al. 2007). However, as reported by various studies, there are hundreds of fonio landraces that potentially possess remarkable genetic variation related to stress tolerance and grain yield (Wang et al. 2021; Ibrahim Bio Yerima et al. 2020). This suggests potential for selection breeding in fonio. However, for targeted selection, specific traits that can contribute to drought tolerance and grain yield must first be identified, which has not been reported in fonio to the best of our knowledge.

Here, we hypothesized that white fonio possesses biochemical, physiological and morphological acclimation responses that support its survival during drought. To test this hypothesis, we selected two accessions from West Africa, and grew them under three moisture conditions (drought, moderate and optimum). Since the greatest diversity of white fonio has been shown to be in modern Guinea and Mali (Adoukonou-Sagbadja et al. 2007; Olodo et al. 2019; Abrouk et al. 2020), the accessions were from Mali (CM05836) and Guinea (CM07354) (Fig. 1A) (Abrouk et al. 2020). The Mali accession originated from the Mopti region (Fulani people), and it was chosen because this region has the lowest rainfall in a survey of 166 white fonio accessions (Abrouk et. al., 2020). Additionally, the Mali accession has now become the genetic model for this crop, after its genome sequence was reported (Abrouk et al. 2020). By contrast, the Guinea accession was chosen because it originates from a wet environment (top quartile, in the same accession survey), specifically from the Konsankoro region (Malinke ethnic group) (Abrouk et al. 2020); furthermore, the majority of white fonio farming occurs in Guinea (Adoukonou-Sagbadja et al. 2007; FAOSTAT, 2022). Population structure analysis revealed that the two accessions belonged to distinct genetic clusters (Abrouk et al. 2020).

The first experiment (Experiment A) evaluated above-ground traits in both accessions, followed by a separate experiment (Experiment B) that characterized root system architecture in detail in the Mali accession.

Plant Genetic Materials

Two accessions (landraces) of white fonio (Digitaria exilis) were generously provided by The French Research Institute for Development (IRD). Accession CM05836 originated from a dry region (mean precipitation 333 mm, mean temp 27.9℃) of Mali associated with the Fulani indigenous group (Mopti region, latitude 15.05; longitude − 3.88; altitude, 263 masl), whereas CM07354 originated from a high rainfall region (mean precipitation 998 mm, mean temp 24.1˚C) from Guinea associated with the Malinke indigenous group (Konsankoro region, latitude 9.03; longitude − 9.00; 537 masl) (Abrouk et al. 2020).

Experiment A: Fonio Shoot Phenotyping Experiments

A controlled indoor trial was conducted at the Crop Science Plant Growth Facility (University of Guelph) during July-October 2021 and repeated during June-September 2022 using both fonio accessions.

Experimental design

In a split-plot design, three moisture treatments (6%, 16% and 26% volumetric water content, VWC) were set as the main plot, and the two fonio accessions as subplots (randomized). The treatments were replicated 10 times. Plants were rotated twice within the main plot. For destructive samples (root and shoot biomass at 41 days after emergence, DAE), a separate set of 10 replicates for each treatment was planted.

Plant growth conditions: Fonio seeds were sown in 4 L pots (~ 10 seeds sown, subsequently reduced to one seedling per pot) containing a 2:1:2 mix, respectively, of sand, perlite and field soil obtained from the University of Guelph Elora Research Station, Elora, Canada. Plants were provided with mixed fluorescent lamps (LED18ET8/4/850, GE, USA) and incandescent bulbs (Ultra LED A21 15W, Sylvania, USA) in an indoor growth room (Crop Science Growth Facility, University of Guelph) with a light intensity of 300 µmol/m²/s at canopy level (12/12 h day/night photoperiod). The daytime temperature was 30˚C, and night temperature was 25℃, and the RH was 60%. Soil moisture levels were measured each day (9–11 am) as follows: the entire probe (12 cm) of a FieldScout TDR 150 Soil Moisture Meter (Spectrum Tech. Inc., UK) was inserted into the soil. The calibration mode was set to ‘standard soil’. Readings were taken 2–3 times per pot, and then averaged. Guided by the moisture readings, pots were watered by hand to maintain the desired VWC during the drought treatment period (21–41 DAE). However, for the drought treatment, starting at 14 DAE, watering was decreased to achieve the target moisture levels in the pots gradually by 21 DAE. With respect to mineral nutrients, 10 ml/pot of a fertilizer stock solution [50 g/L of NPK 24:10:20 (#10535, Plant-Prod, Canada) + 1.2 g/L MgSO₄] was added at planting time, whereas no fertilizer was added after planting until the termination of the drought treatment (i.e. until tissue harvest) to prevent adding additional water.

Measurements

Yield parameters

Accessions were harvested upon maturity, defined as 80% of the panicles having transitioning from a green to grey colour which corresponded to grain detaching from panicles upon threshing. The Mali and Guinea accessions were harvested on day 96 and day 110, respectively, since the two accessions had different flowering times and hence maturity dates. After harvest, grain was thrashed by hand, and cleaned, then the grain weight, panicle length, number of panicles per plant and grain weight per panicle were measured. Shoots were simultaneously collected at harvesting (Mali 96 DAE; Guinea 110 DAE), then oven dried for 72 hours at 60℃ and weighed. The harvest index (HI) was calculated by using the following formula

Harvest Index (%) = [grain weight / (grain weight + shoot dry weight)] × 100

Plant phenotyping: The following parameters were measured at 10 day intervals from the start to end of the drought treatment (21, 31 and 41 DAE): number of shoot branches, plant height (soil to tip of the top leaf), leaf width (2 middle leaves from 2 branches, measured at the midpoint of each blade) and leaf angle (using a protector from the stem to the leaf blade). An additional measurement was recorded at 61 DAE to evaluate recovery after the drought treatment. Similarly, leaf width was measured at five day intervals from the start to the end of the drought treatment (21–41 DAE), with an additional recovery measurement made 20 days after the end of the drought treatment (61 DAE). The root/shoot ratio was measured at the end of the drought treatment (41 DAE).

Leaf chlorophyll

Leaf chlorophyll was measured using a SPAD meter (SPAD-502Plus, Konica Minolta, Japan). The top 3 fully emerged leaves in a mature branch were measured at 21, 31 and 41 and 61 DAE, and averaged for each time point.

Photosynthesis, stomatal conductance and transpiration

Rates of CO₂ assimilation (µmol CO₂ m^− 2 s^− 1), stomatal conductance to water vapour (mol H₂O m^− 2 s^− 1) and transpiration (mmol H₂O m^− 2 s^− 1) were measured using a LI-6400 Measurement System (LI-COR, Inc., Lincoln, NE, USA) on the first fully opened leaf at 9 am at 5-day intervals from 21 to 41 DAE. The leaf chamber was set at 400 µmol reference CO₂, 300 µmol light intensity (PAR), 30 ºC temperature and 60% relative humidity. Every LI-COR measurement was standardized to the LI-COR leaf chamber area (3 cm x 2 cm) by using the following formula

Standardized measurements = (actual measurement/leaf area) x chamber area

Metabolomics: Twenty metabolites (Table S1A-B) with osmoprotectant potential were measured from fonio leaf samples harvested at 41 DAE (termination of the drought treatment). For each treatment, three sample pools, each consisting of 12 leaves from 3 plants, were collected and stored at -80˚ C Falcon tubes. The selected metabolites were quantified using ultra-performance liquid chromatography electrospray ionization multiple reaction monitoring/mass spectrometry (UPLC-MRM/MS) at The Metabolomics Innovation Centre, University of Victoria, Canada, and the values were averaged (n = 3). Briefly, to 50 mg of each ground leaf sample, prior to LC-MS analysis, internal standards were added. Quantification of sugars and succinic acid was carried out using a 3-nitrophenylhydrazine derivatization – LC-MRM/MS method with modifications as previously described (Han, et al. 2016; Han, et al. 2013). A PFP LC column was loaded onto an Agilent 1290 UHPLC system coupled to an Agilent 6495B triple-quadrupole (QQQ) mass spectrometer operated in the negative-ion ESI mode. The mobile phase employed (A) 0.1% formic acid in water and (B) 0.1% formic acid in methanol for binary solvent gradient elution. Quantitation of amino acids was undertaken using the same chemical derivatization - LC-MS method as previously described (Han et al. 2017) using a C18 LC column for LC-ESI-MRM/MS using the same Agilent machine but in positive-ion mode. The mobile phase was (A) 0.1% formic acid in water and (B) 0.1% formic acid in acetonitrile-isopropanol (1:1) for gradient elution. For quantitation of betaines and polyamines, 50 µL of the clear supernatant of each sample or calibration was dried under nitrogen gas flow. The dried residues were reconstituted in 100 µL of 80% aqueous acetonitrile. Thereafter, 5 µL-aliquots of the resulting solutions were injected into an amide-type HILIC column (2.1*100 mm, 1.7 µm) to run LC-ESI-MRM/MS using a Waters UPLC system coupled to a Sciex 6500 plus mass spectrometer which was operated in the positive-ion mode. The mobile phase used for LC separation was (A) ammonium formate buffer and (B) acetonitrile for binary-solvent gradient elution (80–5% B over 15 min) at 0.35 mL/min and 40°C. Calibration curves of individual compounds were constructed from the data acquired. For data presentation, the average values (n = 3) from 6% VWC-treated plants was divided by the average value from 26% VWC-treated plants.

Experiment B: Detailed Fonio Root Phenotyping Experiments

A follow up experiment was conducted using the fonio Mali accession to conduct detailed phenotyping of the root system architecture in response to drought. Plants were grown in two controlled indoor trials at the University of Guelph Crop Science Growth Facility from March to April 2022 and repeated from September to October 2022.

Treatments, experimental design and growth conditions: There were two moisture treatments with four replicates each, arranged in a completely randomized design (CRD). As the different time points required destructive sampling, multiple sets of plants required for serial harvesting were planted in parallel. The same moisture measurement procedures and growth conditions (potting media, temperature, light and light hours, fertilizer) were used as in the above shoot experiment. The Mali accession (CM05836) was exposed to two moisture treatments: 6% and 26% VWC. All the pots had 26% VWC at germination. Thereafter, watering was stopped in pots targeted to achieve the 6% VWC treatment after germination until they reached the desired water level (achieved at 6 DAE), whereas 26% VWC pots were watered every day. In the first trial, roots were harvested every 3rd day, starting from 3 DAE until 15 DAE (5 harvest time points). In the second trial, roots were again harvested every 3rd day, but from 3 DAE until 24 DAE (8 harvest time points). There were 4 replicates per harvest time point (1 replicate = 1 plant/pot) from each moisture treatment. There were two independent trials.

Root structure analysis

Roots were excavated from pots by placing them in buckets submerged in water, and then they were gently washed in clean water. Root length, root surface area and number of branches/tips were analyzed using a WinRHIZO scanner and software (Version PRO2009, Regent Instruments Inc., Quebec, Canada). Root diameter classes were defined in WinRHIZO for calculating root surface area and total root length for each diameter category ≤ 0.2 mm and > 0.2 mm separately. It was confirmed that the measurements of diameter for all the lateral roots were ≤ 0.2 mm and for all the crown roots > 0.2 mm by transferring the WinRHIZO pictures to ImageJ software (version 1.53K, National Institutes of Health, USA) followed by tracing and measuring.

Root hair length and density measurements

Light microscopy was used to assess the density and lengths of root hairs on fonio crown roots, adapted from a prior protocol (Goron et al. 2015). First, root systems were stained with toluidine blue (Soukup, 1992), then individual crown roots were separated. Each crown root was measured, and then 3–4 equally spaced segments (1 cm each) were sampled; the distal end of the crown root was excluded (Fig. 10A). Root segments were placed in 50% glycerol on a microscope slide and then viewed under a Leica MZ8 stereomicroscope at 5x magnification. Each slide was gradually moved, and then pictures taken in order to visualize root hairs along the entire root segment; in total 4 photos were taken for each root segment using AmScope Microscope Digital Camera (MU313-BI, AmScope, USA) and software (v4.11.18421). Thereafter, from each photo, 10 root hairs were traced using ImageJ software (version 1.53K, National Institutes of Health, USA) to calculate root hair length. In total, 40 root hairs were measured per crown root segment and averaged. For root hair density, root hairs were counted in a randomly selected 300 µm subsegment from each photo, and then the density across 4 photos per crown root segment was averaged. Root hair length and density values for the middle two root segments were averaged, to simplify the presentation to top-middle-base of the crown roots. This entire process was repeated for each crown root, for up to the 4 longest crown roots per plant. There were 4 replicates (root systems) per treatment per time point, × 5–8 time points, × 2 treatments × 2 trials. Measurements of root hairs from the longest crown root are presented separately from the average of the ≤ 4 crown roots per plant.

Statistical Analysis

Data were analyzed using a generalized linear mixed model (GLMM) using PROC GLIMMIX, then analyzed using Two-Way ANOVA to analyse the effect of treatments and their interactions. and means were compared using Tukey’s pairwise

comparison in SAS 9.4 (SAS Institute, Cary, North Carolina USA) with a significance level of P ≤ 0.05 and ≤ 0.10. Principal Component Analysis (PCA) was conducted in PRISM 10.1.1 software (Graphpad, USA). Also, all the graphical representations for the results were obtained from PRISM.

Fonio Mali and Guinea accessions were phenotyped for 18 metabolomic, physiological and morphometric traits once the soil volumetric water content (VWC) achieved the target (at 21 DAE), during the drought period (21–41 DAE), and/or after recovery (61 DAE):

Metabolomics

Analysis of 30 metabolites in fonio leaf tissue known to contribute to drought tolerance in other plants showed that fonio leaves accumulate betaines, namely, alanine betaine, glycine betaine and proline betaine under drought stress (Fig. 2A-B, Table S1A-B). There were remarkable 8- and 60-fold increases (P ≤ 0.05) in alanine betaine in the drought-exposed (6%-VWC) plants of the Mali accession in trials 1 and 2, respectively, whereas it was 9 and 6-fold in the case of the Guinea accession (Fig. 2A-B). Similarly, glycine betaine increased 7–12 times across accessions, which was significantly greater (P ≤ 0.05) than in optimally watered (26%-VWC) plants (Fig. 2A-B, Table S1A-B). Proline betaine concentrations also increased 2-fold (P ≤ 0.10) Proline betaine concentrations also increased 2-fold(P ≤ 0.10)under drought, except in the Guinea accession in Trial 1. No significant difference were observed in the concentrations of the above three metabolites when comparing the two accessions under drought (Table S1A-B). A 2-fold increase in proline was noted in trial 1 for both genotypes, but this was not consistent in Trial 2 (Fig. 2A).

Short term above-ground effects

Physiological traits

SPAD at 10 day intervals was measured to determine the impact of drought on fonio leaf chlorophyll content. The Mali accession showed a general trend of superior SPAD values over Guinea under water limitation (Fig. 3A, Table 1A). Compared to optimal water, drought caused a general trend of decreased SPAD chlorophyll in both accessions, but it was only significant for the Mali accession in Trial 2. As Mali plants aged, they showed a trend of increasing SPAD values, regardless of the water treatment, whereas the Guinea accession was more static (Table 1A).

With respect to rates of photosynthesis, no significant differences were observed between plants grown at 26% vs 16%-VWC, but significant declines were observed for 6%-VWC plants starting after 10 days of drought exposure (31 DAE), with a progressive decline until the end date of the treatment. In the Mali accession, 6%-VWC plants showed a 47–51% decline in photosynthesis compared to 26%-VWC plants, whereas this decline was 28–34% for the Guinea accession at 41 DAE (both significant at p ≤ 0.05) (Fig. 3B, Table 1B). When comparing the accessions, no consistent significant differences in photosynthesis were observed, though the Mali accession had a significantly lower rate at the end of the drought treatment (41 DAE).

Next, we asked whether the fonio accessions could acclimate to drought by reducing transpiration. Consistent with photosynthesis, whereas optimally watered plants showed increases in transpiration as plants aged (21–41 DAE), notable declines in transpiration were observed in both accessions (52–57% for Mali; 37–45% for Guinea) after 20 days of drought exposure (Fig. 3C, Table 1C). However, within each accession, significant declines in photosynthesis between 6% and 26%-VWC were inconsistently observed starting after 10 days of drought exposure (31 DAE), becoming consistent only at 41 DAE.

As a potential contributor to restricted transpiration, stomatal conductance was assessed at 5 day intervals during the drought treatment. Similar to transpiration, stomatal conductance in optimally watered plants showed increases in transpiration as plants developed (21–41 DAE), whereas stomatal conductance declined in both accessions exposed to drought as the drought interval progressed (Fig. 3D, Table 1D). By 41 DAE, stomatal conductance declined 63–64% for the Mali accession and 53–54% for Guinea when comparing 26–6%-VWC plants. No significant differences were observed in stomatal conductance when comparing the two accessions.

Shoot morphology

In addition to altered stomatal conductance, we hypothesized that the reduced transpiration observed in white fonio under drought was partly due to shoot acclimation responses. Drought caused significant declines in shoot biomass in both accessions (Fig. 4A Table S3). The Mali accession at 6%-VWC had a 48% and 45% reduction in shoot weight per

Table 1

**Experiment A: Effect of moisture treatments on physiological traits of white fonio.** A. Leaf Chlorophyll (SPAD)
	21 DAE		31DAE		41 DAE		61 DAE (recovery)
Mali	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2
6% VWC	39.61_a	37.78_ab	40.65_a	41.57_b	42.36_a	41.71_b	44.46_a	48.88_b
16% VWC	38.71_a	37.98_ab	42.20_a	46.96_a	44.96_a	47.54_a	45.83_a	54.06_a
26% VWC	38.97_a	39.36_a	41.63_a	47.81_a	45.67_a	46.12_a	46.38_a	51.89_ab
Guinea
6% VWC	36.63_a	36.28_ab	34.13_b	36.85_c	35.15_b	37.16_c	38.44_b	38.60_c
16% VWC	37.35_a	35.84_b	36.81_b	38.56_c	37.34_b	38.18_c	39.28_b	39.59_c
26% VWC	37.22_a	37.69_ab	36.00_b	39.19_bc	38.34_b	39.01_c	39.03_b	39.71_c
B. Photosynthesis (µmolCO₂m^− 2s^− 1)

	21 DAE		26 DAE		31DAE		36 DAE		41 DAE
Mali	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2
6% VWC	15.16_ab	12.67_a	15.94_ab	11.75_a	13.87_c	11.29_b	9.12_d	11.27_b	7.78_d	8.22_c
16% VWC	15.13_ab	11.49_a	16.88_a	11.68_a	15.62_ab	14.94_a	15.33_ab	15.31_a	14.95_ab	14.79_a
26% VWC	17.12_a	13.75_a	15.77_ab	13.10_a	16.69_a	15.40_a	16.12_a	15.34_a	15.77_a	15.36_a
Guinea
6% VWC	16.57_a	14.17_a	15.70_ab	11.86_a	13.78_c	12.56_b	11.24_c	10.61_b	10.26_c	10.25_b
16% VWC	15.38_ab	11.5_a	15.49_ab	13.06_a	14.6_bc	15.30_a	13.87_b	15.05_a	13.13_b	14.57_a
26% VWC	14.17_b	13.97_a	14.56_b	12.63_a	15.41_abc	15.58_a	14.09_b	15.82_a	14.16_ab	15.58_a
C. Transpiration (mmolH₂Om^− 2s^− 1)

	21 DAE
Mali	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2
6% VWC	1.71_a	1.53_a	1.61_ab	1.58_b	1.35_a	1.24_c	1.20_b	1.17_b	0.74_c	0.73_b
16% VWC	1.72_a	1.71_a	1.77a	2.52_a	1.53_a	2.11_a	1.60_a	2.03_a	1.69_a	2.24_a
26% VWC	1.79_a	1.72_a	1.70_ab	2.54_a	1.62_a	2.09_a	1.47_ab	2.18_a	1.70_a	2.29_a
Guinea
6% VWC	1.71_a	1.71_a	1.45_b	1.71_b	1.36_a	1.46_bc	1.30_b	1.27_b	1.07_b	0.94_b
16% VWC	1.57_a	1.59_a	1.56_ab	2.27_ab	1.48_a	1.99_ab	1.66_a	1.90_a	1.59_a	1.85_a
26% VWC	1.62_a	1.74_a	1.52_ab	2.25_ab	1.51_a	2.18_a	1.64_a	2.03_a	1.67_a	2.03_a
D. Stomatal Conductance (molH₂Om^− 2s^− 1)

Mali	21 DAE		26 DAE		31DAE		36 DAE		41 DAE
Mali	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2
6% VWC	0.083_a	0.075_a	0.088_c	0.140_a	0.077_c	0.102_c	0.065_c	0.081_d	0.051_c	0.052_b
16% VWC	0.089_a	0.087_a	0.109_ab	0.124_a	0.121_ab	0.145_a	0.135_a	0.144_a	0.137_ab	0.140_a
26% VWC	0.086_a	0.083_a	0.113_a	0.124_a	0.128_a	0.137_a	0.139_a	0.140_ab	0.139_a	0.146_a
Guinea
6% VWC	0.086_a	0.086_a	0.092_bc	0.092_a	0.087_c	0.112_bc	0.073_c	0.097_cd	0.063_c	0.066_b
16% VWC	0.085_a	0.078_a	0.097_abc	0.124_a	0.116_b	0.129_ab	0.118_b	0.119_bc	0.122_b	0.139_a
26% VWC	0.081_a	0.085_a	0.103_abc	0.115_a	0.117_ab	0.134_ab	0.119_b	0.128_ab	0.136_ab	0.142_a

Different letters symbolize statistical differences at P ≤ 0.05. Abbreviations: VWC, volumetric water content; DAE, days after emergence.

plant in trials 1 and 2, respectively, compared to 26%-VWC grown plants. The Guinea accession also had 51% and 32% lower shoot weight at 6%-VWC trial 1 and trial 2, respectively. The effect of drought on shoot growth was clearly visible by 41 DAE (Fig. 4B), though drought-treated plants showed substantial recovery after 20 days of restoration of optimal watering (Fig. 4C).

Morphologically, reduced shoot branching was visible after 10 days of drought initiation (31 DAE) for both accessions (P≤0.05) (Fig. 4D, Table 2A). For the Mali accession, in both trials, the number of branches was 60% lower at 6%-VWC compared to 26%-VWC, whereas for the Guinea accession, the reduction was 43%-53% (P≤0.05) by 41 DAE (Fig. 4D, Table 2A). There was no significant difference between the two accessions in the total number of branches across the moisture levels at 41 DAE. However, for the 6%-VWC treated plants, partial recovery in branching was observed after 20 days of optimal irrigation (61 DAE) in both accessions (Fig. 4D, Table 2A).

Following a similar trend as branching, plant height was also significantly reduced after 10 days of drought (31 DAE) for both accessions (P≤0.05) (Fig. 4E, Table 2B). By 41 DAE, for the Mali accession at 41 DAE, plants exposed to 6%-VWC were 29%-37% (16–19 cm) shorter than the plants exposed to 26%-VWC, and for the Guinea accessions, they were 27%-32% (16 cm) shorter (Fig. 4E, Table 2B). In both accessions, 20 days after restoration of normal watering, plant height of 6%-VWR treated plants remained significantly reduced compared to plants that had been optimally watered throughout the experiment (Fig. 4E, Table 2B). No significant difference in plant height was noted between the two accessions across different moisture levels.

The leaves of the Guinea accession had significantly greater width than the Mali accession across the moisture treatments for all time point measurements except for a few exceptions (Fig. 4F, Table 2C). Similar to previous traits, leaf width was significantly reduced (P≤0.05) after 10 days of drought exposure (at 31 DAE) for both accessions (6% versus 26% VWC). By 41 DAE, the Mali accession had 10%-12% (0.6–0.8 mm) reduced leaf width in plants exposed to 6%-VWC than the plants exposed to 26%-VWC, similar to the 9%-13% (0.6–1.1 mm) reductions observed for the Guinea accession. After normal watering (61 DAE), in both accessions, the width of new leaves from drought-exposed plants recovered, to match those from the other two moisture levels (16% and 26%-VWC) (P≤0.05) (Fig. 4F, Table 2C).

Whereas optimally watered plants (26%-VWC) showed progressively reduced leaf angle as the plants developed, leaf angle significantly increased (p ≤ 0.05) during drought (6%-VWC versus 26%-VWC) for both accessions (Fig. 4G, Table 2D), starting after 10 days of drought initiation. For the Mali accession at 41 DAE, a 50%-83% higher leaf angle was observed in 6%-VWC exposed plants, compared to 26%-VWC treated plants, whereas for the Guinea accession, it was 41%-62% higher. Numerically, the greatest leaf angles observed in the experiment belonged to the Mali accession exposed to drought during the drought exposure at 31–41 DAE (i.e. angles of 50–65˚). For both accessions, after 20 days of normal watering (61 DAE), plants previously exposed to drought had decreased leaf angles but remained the highest among the three moisture treatments (Fig. 4G, Table 2D).

Long term above-ground impacts

Next, we asked to what extent drought had long term impacts on fonio, approximately 8–10 weeks after the drought treatments had ended after optimal watering had resumed. At 96–110 DAE, the accessions were harvested, and shoot and grain traits were measured.

Table 2

**Experiment A: Effect of moisture treatments on shoot morphological traits in white fonio.**
Mali	21DAE		31DAE		41DAE		61DAE (Recovery)
Mali	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2
6% VWC	5.90_a	5.60_a	11.0_b	7.80_c	15.60_b	14.50_b	48.00_c	42.60_b
16% VWC	6.80_a	7.40_a	16.7_a	14.60_ab	37.50_a	32.80_a	67.70_ab	66.00_a
26% VWC	6.30_a	7.40_a	17.9_a	16.40_a	38.90_a	36.40_a	72.70_a	68.90_a
Guinea
6% VWC	6.80_a	5.50_a	9.7_b	7.50_c	20.10_b	16.90_b	51.10_c	42.60_b
16% VWC	6.10_a	6.80_a	17.8_a	13.30_ab	35.90_a	30.40_a	67.20_ab	61.80_a
26% VWC	6.50_a	6.40_a	17.0_a	12.00_b	35.20_a	35.00_a	64.70_b	62.10_a
A. Number of branches
B. Plant height (cm)

Mali	21 DAE		31 DAE		41DAE		61 DAE (Recovery)
Mali	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2
6% VWC	28.75_a	21.22_b	30.91_c	33.56_b	32.42_b	39.93_c	49.01_c	43.90_d
16% VWC	30.45_a	24.43_ab	38.13_a	41.82_a	47.36_a	52.42_b	52.83_bc	58.53_c
26% VWC	30.2_a	26.70a	40.10_a	42.85_a	51.72_a	55.81_ab	58.96_ab	65.86_bc
Guinea
6% VWC	30.7_a	24.20_ab	32.80_bc	33.93_b	33.0_b	43.72_c	50.80_c	58.51_c
16% VWC	29.95_a	27.23_a	37.12_ab	41.74_a	48.4_a	56.24_ab	53.55_bc	81.44_a
26% VWC	29.10_a	24.45_a	38.80_a	43.91_a	48.5_a	59.73_a	61.61_a	74.41_ab
C. Leaf width (cm)

Mali	21DAE		26DAE		31DAE						61DAE (Recovery)
Mali	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2
6% VWC	0.56_b	0.5_b	0.57_b	0.56_d	0.53_d	0.57_c	0.53_c	0.58_c	0.55_d	0.59_c	0.61_b	0.69_b
16% VWC	0.60_b	0.5_b	0.61_b	0.6_cd	0.65_abc	0.66_b	0.6_b	0.68_b	0.59_cd	0.68_b	0.59_b	0.7_b
26% VWC	0.57_b	0.5_b	0.61_b	0.63_bc	0.64_bc	0.66_b	0.64_ab	0.67_b	0.61_c	0.67_b	0.60_b	0.71_b
Guinea
6% VWC	0.57_b	0.58_a	0.63_b	0.64_abc	0.61_c	0.66_b	0.62_b	0.66_b	0.63b_c	0.68_b	0.69_a	0.74_ab
16% VWC	0.67_a	0.55_ab	0.70_a	0.66_ab	0.71_a	0.77_a	0.69_a	0.79_a	0.68_ab	0.80_a	0.68_a	0.79_a
26% VWC	0.67_a	0.59_a	0.71_a	0.70_a	0.70_ab	0.74_a	0.70_a	0.77_a	0.69_a	0.79_a	0.70_a	0.8_a
D. Leaf Angle (º)

Mali	21 DAE		31 DAE		41DAE		61 DAE (Recovery)
Mali	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2
6% VWC	42.0_a	47.5_a	50.0_a	54.5_a	49.5_a	65.0_a	39.0_a	41.5_a
16% VWC	43.0_a	45.0_a	39.0_b	41.5_bcd	35.5_b	42.5_bc	34.5_bc	32.0_bc
26% VWC	43.5_a	50.5_a	35.0_b	49.0_ab	33_b	35.5_cd	32.0_c	35.0_b
Guinea
6% VWC	42.0_a	47.5_a	46.5_a	48.5_bc	46.5_a	49.5_b	37.0_ab	35.5_b
16% VWC	39.5_a	44.5_a	36.0_b	37.0_d	34_b	33.0_d	32.5_c	29.0_c
26% VWC	44.0_a	43.5_a	34.0_b	40.5_cd	33_b	30.5_d	31.5_c	28.0_c

Table S1

A. Experiment A: Concentrations of selected metabolites in white fonio leaves from drought exposed (6%-VWC) and optimally watered (26%-VWC) plants at 41 DAE (end of drought treatment) in the Mali and Guinea accession (Trial 1).
Metabolites (nmol/per mg tissue)	Mali		Guinea		SEM (±)
Metabolites (nmol/per mg tissue)	VWC6%	VWC26%	VWC6%	VWC26%	SEM (±)
Glucose	21.9545_a	16.1543_a	20.6480_a	14.3926_a	1.9333
Fructose	9.2732_a	6.1355_a	8.9343_a	6.1801_a	0.9906
Succinic acid	0.1532_a	0.0897_a	0.1648_a	0.1251_a	0.0141
Alanine betaine *	0.0462_a	0.0056_ab	0.023_ab	0.0025_b	0.0064
Proline betaine	0.0002_a	0.0001_a	0.0003_a	0.0002_a	0.0001
Glycine betaine*	3.7862_a	0.4992_b	3.2952_a	0.4927_b	0.3018
Ectoine	0.0001_a	0.0002_a	0.0002_a	0.0002_a	0.0001
Spermidine	0.0027_a	0.0017_a	0.0019_a	0.0021_a	0.0004
Spermine	0.0013_a	0.0009_a	0.0010_a	0.0008_a	0.0001
4-OH Pro	0.0025_a	0.0022_a	0.0025_a	0.0021_a	0.0003
GABA	0.2356_a	0.2110_a	0.2840_a	0.2362_a	0.0221
Arg	0.0416_a	0.0446_a	0.0433_a	0.0548_a	0.0049
Asn	0.3177_a	0.695_a	0.4436_a	0.4656_a	0.0780
Asp	0.8906_a	0.8555_a	0.8742_a	0.9594_a	0.0842
Cit	0.0105_a	0.0107_a	0.0187_a	0.0093_a	0.0036
Gln	0.3994_a	0.2904_a	0.4634_a	0.3132_a	0.0407
Glu	1.3078_a	1.2060_a	1.4276_a	1.4155_a	0.1263
Gly	0.0656_a	0.0672_a	0.0854_a	0.0725_a	0.0107
Ile	0.0386_a	0.0498_a	0.0348_a	0.0525_a	0.0063
Leu	0.1261_a	0.1097_a	0.1208_a	0.1138_a	0.0094
Orn	0.0026_a	0.0021_a	0.0026_a	0.0019_a	0.0003
Phe	0.0897_a	0.0805_a	0.0733_a	0.0757_a	0.0118
Pipecolic acid	0.0659_a	0.0567_a	0.0506_a	0.0338_a	0.0063
Pro*	0.3132_a	0.1330_ab	0.2315_ab	0.1023_b	0.0287
Putrescine	0.0242_a	0.0217_a	0.0167_a	0.0224_a	0.0028
Ser	0.1990_a	0.1850_a	0.219_2a	0.1970_a	0.0183
Thr	0.2983_a	0.2522_a	0.3056_a	0.2958_a	0.0221
Trp	0.0145_a	0.0270_a	0.0091_a	0.0241_a	0.0038
Tyr	0.0551_a	0.0525_a	0.0460_a	0.0548_a	0.0051
Val	0.1080_a	0.1250_a	0.0873_a	0.1254_a	0.0138
Different letters within each row (*) symbolize statistical differences at P ≤ 0.10. Abbreviation: VWC, volumetric water content.

Table S2

**Experiment A: Effect of moisture treatments on shoot biomass, root biomass and root/shoot ratio of white fonio at the end of the moisture treatment (at 41 DAE)**.
VWC	Accession	Shoot dry weight/plant 41 DAE (g)		Root dry weight/plant 41 DAE (g)		Root/shoot ratio 41DAE
VWC	Accession	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2
6%	Mali	2.98_c	3.02_b	0.42_a	0.56_ab	0.144_a	0.187_a
6%	Guinea	3.42_c	3.61_b	0.33_b	0.48_b	0.095_b	0.134_b
16%	Mali	5.65_b	5.03_a	0.48_a	0.63_a	0.087_b	0.126_b
16%	Guinea	5.91_ab	5.40_a	0.48_a	0.64_a	0.084_b	0.121_b
26%	Mali	5.77_b	5.44_a	0.50_a	0.60_a	0.087_b	0.109_b
26%	Guinea	7.04_a	5.37_a	0.53_a	0.66_a	0.077_b	0.124_b
Different letters symbolize statistical differences at P ≤ 0.05. Abbreviations: VWC, volumetric water content; DAE, days after emergence.

Table S3

**Experiment A: Effect of moisture treatments on harvest index, shoot yield and grain yield traits of white fonio at final harvest (96–110 DAE).**
VWC	Accession	Harvest Index (%)		Shoot weight/plant (g)		Grain weight/plant (g)		No of panicles/plant		Panicle length (cm)		Grain weight/ panicle (mg)
		Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2	Trial 1	Trial 2
6%	Mali	14.51_a	14.76_a	64.96_c	58.09_d	11.01_b	10.06_bc	153.50_ab	132.50_bc	8.58_bc	7.44_b	75.00_a	77.59_a
6%	Guinea	6.33_c	7.78_c	76.07_c	72.26_c	5.15_c	6.13_d	122.50_b	127.90_c	10.44_a	11.33_a	43.75_b	48.17_b
16%	Mali	10.92_b	9.82_b	94.59_b	110.00_b	11.59_a	11.95_ab	146.30_b	149.80_ab	8.33_bc	8.33_b	80.74_a	80.06_a
16%	Guinea	6.05_c	7.55_c	105.45_ab	109.72_b	6.63_c	8.96_c	159.10_ab	160.20_a	10.2_a	10.93_a	42.12_b	56.15_b
26%	Mali	11.88_b	11.34_b	100.43_b	104.14_b	13.53_a	13.33_a	192.70_a	170.40_a	8.08_c	7.63_b	73.67_a	78.64_a
26%	Guinea	7.78_c	7.77_c	114.21_a	122.34_a	9.61_b	10.33_bc	159.40_ab	168.60_a	9.74_ab	11.12_a	60.49_ab	61.83_b
Different letters symbolize statistical differences at P ≤ 0.05. Abbreviations: VWC, volumetric water content; DAE, days after emergence.

Table S4

**Experiment B: Effect of moisture treatments on crown root architecture (count, diameter, length and surface area) in the Mali accession of white fonio at different time points.**
Trial 1	VWC	3 DAE	6 DAE	9 DAE	12 DAE	15 DAE
No of CR	6%VWC	1.00_a	1.00_a	1.00_b	1.25_b	1.25_b
No of CR	26%VWC	1.00_a	1.25_a	2.75_a	3.00_a	3.50_a
CR diameter	6%VWC	0.18_a	0.19_b	0.19_b+	0.19_b	0.19_b
CR diameter	26%VWC	0.19_a	0.25_a	0.28_a+	0.32_a	0.33_a
Length of the longest CR	6%VWC	5.00_a	8.355_a	11.54_a	15.98_a	16.90_a
Length of the longest CR	26%VWC	4.22_a	7.40_a	8.20_a	10.88_b	10.74_b
Average length of CRs	6%VWC	5.00_a	8.36_a	11.54_a	15.48_a	16.12_a
Average length of CRs	26%VWC	4.18_a	7.08_a	6.18_b	8.14_b	7.46_b
Root area-diameter > 0.2 mm (cm²)	6%VWC	0.09_a	0.15_a	0.18_a+	0.19_b	0.26_b
Root area-diameter > 0.2 mm (cm²)	26%VWC	0.12_a	0.24_a	0.48_a+	0.78_a	1.23_a

Trial 2	VWC	3DAE	6DAE	9DAE	12DAE	15DAE	18DAE	21DAE	24DAE
No of CR	6%	1.00_a	1.00_b	1.00_b	1.00_b	1.25_b	1.25_b	1.50_b	1.50_b
No of CR	26%	1.25_a	1.70_a	2.50_a	2.25_a	3.00_a	3.00_a	3.00_a	3.50_a
CR diameter	6%	0.16_a	0.15_a	0.15_a	0.16_a	0.14_a	0.15_b	0.15_b+	0.17_b
CR diameter	26%	0.18_a	0.18_a	0.22_a	0.25_a	0.25_a	0.29_a	0.28_a+	0.31_a
Length of the longest CR	6%	6.56_a	8.33_a	10.50_a	11.36_a	14.03_a	14.71_a	16.05_a	17.48_a
Length of the longest CR	26%	5.81_a	6.90_a	8.96_a	9.48_a	11.54_a	11.92_a	11.70_b	12.64_b
Average length of CRs	6%	4.59_a	5.45_a	9.62_a	11.50_a	12.85_a	13.03_a	13.85_a	14.75_a
Average length of CRs	26%	3.70_a	4.11_a	6.69_a	6.81_b	8.65_b	8.56_b	9.50_b	9.20_b
Root area-diameter > 0.2 mm (cm²⁾)	6%	0.08_a	0.11_a	0.19_b	0.23_b+	0.43_b	1.28_b	1.80_b	2.52_b
Root area-diameter > 0.2 mm (cm²⁾)	26%	0.20_a	0.39_a	0.57_a	0.78_a+	1.77_a	3.60_a	3.83_a	4.04_a
Different letters symbolize statistical differences at P ≤ 0.05. Abbreviations: VWC, volumetric water content; DAE, days after emergence.

Table S5

**Experiment B: Effect of moisture treatments on lateral root architecture (tips/branching, surface area) in the Mali accession of white fonio at different time points**
Trial 1	VWC	3 DAE	6 DAE	9 DAE	12 DAE	15 DAE
No of tips	6%	336.75_a	499.00_a	761.00_a+	957.50_a	1143.50_a
No of tips	26%	346.75_a	437.00_a	582.75_b+	669.50_b	723.75_b
Root area-diameter ≤ 0.2 mm (cm²)	6%	0.69_a	1.65_a	3.76_a	5.37_a	7.34_a
Root area-diameter ≤ 0.2 mm (cm²)	26%	0.86_a	1.86_a	3.03_a	3.64_b	4.08_b

Trial 2	VWC	3DAE	6DAE	9DAE	12DAE	15DAE	18DAE	21DAE	24DAE
No of tips	6%	454.25_a	565.00_a	827.00_a	999.00_a	1157.00_a	1249.00_a	1349.00_a	1503.25_a
No of tips	26%	263.75_b	413.25_a	474.50_b	509.50_b	682.00_b	777.50_b	843.75_b	920.25_b
Root area-diameter ≤ 0.2 mm (cm²)	6%	0.71_a	2.24_a	4.55_a	5.47_a	7.45_a	9.64_a	10.75_a	11.24_a
Root area-diameter ≤ 0.2 mm (cm²)	26%	1.10_a	1.56_a	2.74_b	2.34_b	3.50_b	5.43_b	6.68_b	7.87_b
Different letters symbolize statistical differences at P ≤ 0.05. Abbreviations: VWC, volumetric water content; DAE, days after emergence.

Table S6

Experiment B: Effect of moisture treatments on the root hair architecture (length and density) of white fonio at different time points measured on three crown root segments (S1-S3). Root hair measurements on the longest crown root:
Trial 1	VWC	3DAE	6 DAE	9DAE	12DAE	15DAE
Root hair length at S1 (x100 µm)	6%	3.64_a	3.82_a	4.40_a+	3.50_a	4.56_a+
Root hair length at S1 (x100 µm)	26%	3.80_a	4.05_a	3.05_b+	3.35_a	3.47_b+
Root hair length at S2 (x100 µm)	6%	4.18_a	0.41_a	4.23_a	4.18_a	5.90_a+
Root hair length at S2 (x100 µm)	26%	3.13_a	3.90_a	3.65_a	3.75_a	3.78_b+
Root hair length at S3 (x100 µm)	6%	2.73_a	2.22_a	3.76_a	4.00_a	4.78_a
Root hair length at S3 (x100 µm)	26%	1.50_b	2.03_a	1.69_b	2.71_a	2.12_b
Root hair density at S1 (count/300 µm)	6%	9.23_a	7.95_a	8.78_a	9.15_a+	9.13_a
Root hair density at S1 (count/300 µm)	26%	6.03_a	5.03_a	4.53_a	5.15_b+	4.28_b
Root hair density at S2 (count/300 µm)	6%	3.72_a	5.69_a	5.82_a	7.00_a	7.54_a+
Root hair density at S2 (count/300 µm)	26%	2.47_a	2.44_a	4.12_a	2.75_b	2.94_b+
Root hair density at S3 (count/300 µm)	6%	2.19_a	5.06_a	8.69_a	6.69_a	9.50_a
Root hair density at S3 (count/300 µm)	26%	0.63_a	1.81_a	2.33_b	2.58_b	2.94_b

Trial 2	VWC	3DAE	6DAE	9DAE	12DAE	15DAE	18DAE	21DAE	24DAE
Root hair length at S1 (x100 µm)	6%	3.69_a	3.39_a	1.72_a	3.11_a	2.12_a	3.21_a	1.96_a	6.10_a
Root hair length at S1 (x100 µm)	26%	2.74_a	3.10_a	1.87_a	4.16_a	2.18_a	3.36_a	1.59_a	3.84_a
Root hair length at S2 (x100 µm)	6%	2.75_a	2.87_a	1.90_a	2.46_a	2.17_a	3.15_a	2.29_a	4.17_a+
Root hair length at S2 (x100 µm)	26%	3.34_a	3.34_a	1.91_a	3.27_a	1.84_a	2.69_a	1.55_b	3.17_b+
Root hair length at S3 (x100 µm)	6%	2.62_a	2.82_a	2.12_a	2.29_b	1.85_a	2.23	1.85_a	3.04_a
Root hair length at S3 (x100 µm)	26%	2.56_a	3.37_a	1.69_a	3.02_a	1.70_a	NA	1.40_b	2.00_b
Root hair density at S1 (count/300 µm)	6%	11.23_a	8.69_a	7.75_a	6.38_a	7.15_a	6.19_a	9.69_a	11.25_a
Root hair density at S1 (count/300 µm)	26%	9.17_a	9.21_a	7.75_a	5.75_a	6.63_a	6.38_a	9.42_a	4.59_b
Root hair density at S2 (count/300 µm)	6%	8.58_a	3.23_a	7.86_a	2.87_a	4.88_a	7.39_a+	7.79_a	11.85_a
Root hair density at S2 (count/300 µm)	26%	6.75_a	6.01_a	4.85_a	3.67_a	2.38_a	3.72_b+	4.82_a	3.19_b
Root hair density at S3 (count/300 µm)	6%	5.81_a	4.71_a	7.00_a	6.63_a+	5.38_a	5.21_a	7.11_a	7.56_a
Root hair density at S3 (count/300 µm)	26%	0.42_b	3.13_a	1.25_b	2.94_b+	3.63_a	0.00_b	2.19_b	0.44_b
Different letters symbolize statistical differences at P ≤ 0.05. Abbreviations: VWC, volumetric water content; DAE, days after emergence.

Table S7

Experiment B: Effect of moisture treatments on the root hair architecture (length and density) of white fonio at different time points measured on three crown root segments (S1-S3). Root hair measurement; average of ≤ 4 crown roots/plant:
Trial 1	VWC	3DAE	6DAE	9DAE	12DAE	15DAE
Root hair length at S1 (x100 µm)	6%VWC	3.78_a	3.90_a	4.40_a	3.50_a	4.17_a
Root hair length at S1 (x100 µm)	26%VWC	3.65_a	3.75_a	3.45_a	3.03_a	3.43_a
Root hair length at S2 (x100 µm)	6%VWC	4.18_a	4.05_a	4.23_a	4.18_a	5.90_a
Root hair length at S2 (x100 µm)	26%VWC	3.13_a	3.90_a	3.45_a	3.58_a	3.73_b
Root hair length at S3 (x100 µm)	6%VWC	2.75_a	2.95_a	3.76_a	3.00_a	4.78_a
Root hair length at S3 (x100 µm)	26%VWC	1.50_b	2.23_a	2.40_b	2.33_a	2.10_b
Root hair density at S1 (count/300 µm)	6%VWC	9.19_a	7.94_a	8.75_a+	9.13_a	9.13_a
Root hair density at S1 (count/300 µm)	26%VWC	6.00_a	4.47_a	4.08_b+	3.87_b	4.34_b
Root hair density at S2 (count/300 µm)	6%VWC	3.72_a	5.69_a	5.82_a	7.00_a	8.29_a
Root hair density at S2 (count/300 µm)	26%VWC	2.47_a	2.19_a	3.74_a	2.70_b	3.39_b
Root hair density at S3 (count/300 µm)	6%VWC	2.19_a	5.06_a	8.69_a	6.69_a	9.50_a
Root hair density at S3 (count/300 µm)	26%VWC	0.63_a	1.81_a	2.15_b	1.74_b	3.06_b

Trial 2	VWC	3DAE	6DAE	9DAE	12DAE	15DAE	18DAE	21DAE	24DAE
Root hair length at S1 (x100 µm)	6%VWC	2.92_a	3.23_a	1.77_a	3.11_a	2.12_a	3.21_a	1.93_a	5.65_a
Root hair length at S1 (x100 µm)	26%VWC	2.68_a	3.28_a	2.05_a	3.92_a	2.07_a	3.13_a	1.62_a	3.23_b
Root hair length at S2 (x100 µm)	6%VWC	2.56_a	2.87_a	1.97_a	2.46_a	2.17_a	3.15_a	2.70_a	4.36_a
Root hair length at S2 (x100 µm)	26%VWC	3.07_a	3.34_a	2.01_a	3.38_a	1.87_a	2.84_a	1.59_b	2.78_b
Root hair length at S3 (x100 µm)	6%VWC	2.28_a	2.63_a	1.69_a	2.29_b	1.85_a	2.23_a	1.86_a	3.76_a+
Root hair length at S3 (x100 µm)	26%VWC	2.90_a	3.98_a	3.10_a	3.74_a	1.82_a	1.86_b	1.18_b	2.76_b+
Root hair density at S1 (count/300 µm)	6%VWC	9.74_a	7.52_a	8.33_a	6.38_a	7.15_a	6.19_a	10.34_a	11.75_a
Root hair density at S1 (count/300 µm)	26%VWC	5.69_a	7.71_a	7.17_a	3.56_a	5.71_a	5.92_a	8.26_a	5.68_a
Root hair density at S2 (count/300 µm)	6%VWC	8.32_a	4.42_a	7.18_a	3.82_a	4.88_a	7.39_a	8.50_a	10.19_a
Root hair density at S2 (count/300 µm)	26%VWC	5.63_a	6.01_a	4.11_a	3.33_a	2.55_a	3.15_b	4.743_a	2.97_b
Root hair density at S3 (count/300 µm)	6%VWC	5.05_a	4.70_a	4.58_a	6.39_a	7.17_a	4.92_a	6.36_a	6.88_a
Root hair density at S3 (count/300 µm)	26%VWC	0.67_b	3.33_a	1.87_a	2.02_b	2.95_b	0.14_b	0.68_b	1.37_b
Different letters symbolize statistical differences at P ≤ 0.05. Abbreviations: VWC, volumetric water content; DAE, days after emergence.
Supplemental Figures

Different letters symbolize statistical differences at P ≤ 0.05. Abbreviations: VWC, volumetric water content; DAE, days after emergence.

The harvest index (HI) for the Mali accession was consistently superior across the moisture treatments than the Guinea accession (P ≤ 0.05) in both trials (Fig. 5A, Table S3). Interestingly, for the Mali accession, plants exposed to 6%-VWC had the highest HI (14.5%-14.8%) among all the moisture treatments. By contrast, the Guinea accession had significantly lower HI (6.3%-7.8%) at 6%-VWC, which was not significantly different than other moisture levels. The higher HI for the Mali accession could be partially explained by it having a numerically lesser shoot biomass at harvest compared to the Guinea accession (though inconsistently significant), in combination with having significantly higher grain yield (p ≤ 0.05) (Fig. 5B-C, Table S3).

Drought exposed plants did not fully recover (Fig. 5B, Table S3). The Mali accession that had earlier been exposed to 6% VWC had a 35–44% reduction in shoot weight per plant, as compared to 26% VWC treated plants; for the Guinea accession, it was 33–41% (Fig. 5B, Table S3). Similarly, grain weight per plant for the Mali accession remained decreased by 19%-25% at harvest for plants previously exposed to drought (6%-VWC) compared to those that had been optimally watered (26% VWC) throughout; 46%-41% reductions were noted for the Guinea accession exposed to drought (Fig. 5C, Table S3).

Across all moisture levels, the average grain weight per plant for the Mali accession was significantly greater than the Guinea accession (Fig. 5C, Table S3); for plants that had been continuously grown under optimum moisture (26%-VWC), the Mali accession was 35% greater. To understand this difference, the components of grain yield were analyzed. The number of panicles per plant was not significantly different between the two genotypes across the moisture treatments (P ≤ 0.05) (Fig. 5D, Table S3). However, the earlier drought treatment showed a trend for reduced panicle number per plant. Upon visual examination, the Guinea panicles appeared to be longer than those from the Mali accession, but had lower grain fill (Fig. 5E). Subsequent measurements verified these observations: the panicle length was significantly greater for the Guinea accession than the Mali accession across the moisture treatments (P ≤ 0.05) (Fig. 5D-E, Table S3). Indeed, the Guinea accession had 21–46% longer panicles than the Mali accession under optimum moisture (26%-VWC) (Fig. 5F, Table S3). Soil moisture did not affect panicle length. Furthermore, as predicted, grain weight per panicle was consistently higher for the Mali accession than the Guinea accession across the moisture treatments with one exception (Fig. 5G, Table S3). In fact, grain weight per panicle for the Guinea accession was 18–21% lower than the Mali accession when continuously grown under optimum water (26%-VWC). Grain weight per panicle showed a strong positive correlation (r = 0.6–0.8, P < 0.0001) with grain weight per plant, whereas panicle length had a negative correlation (r = -0.5, P ≤ 0.0001) (Fig. 5H-I).

Principal Component Analysis (PCA)

Multivariate spatial PCA was undertaken to understand the relationships amongst physiological and morphological measurements at 41 DAE, and yield-related traits recorded at harvest (total = 17 traits). Principal components 1 and 2 explained a mean 36% and 30% variation of the data, respectively, and hence were presented (Fig. 7A-D). Several physio-morphological traits measured at 41 DAE clustered including: shoot weight, root weight, leaf width, plant height, photosynthesis, conductance, transpiration, and number of branches; these mostly decreased during drought stress. Along the PC1 axis, leaf angle and root/shoot ratio at 41 DAE at vector angles opposite to the above cluster, consistent with these trait values increasing during drought; along this axis, the harvest index clustering pattern differed between the Mali and Guinea accession (Fig. 7A-D). The traits that consistently clustered with grain weight included: root weight-41DAE, leaf width, shoot weight-41DAE, plant height and SPAD, suggestive of their potential positive correlation. Significant positive correlations (r = 0.7 and 0.8, P ≤ 0.001) between grain weight and SPAD chlorophyll were found in trials 1 and 2, respectively (Fig. 7E). In contrast, a negative correlation between grain weight and leaf width was revealed, although the correlation was not strong (Fig. 7F). Positive and significant but lower correlations with grain weight were found for plant height and root weight-41DAE (Fig. 7G-H).

Underground Effects

Root biomass

To evaluate the potential underground acclimation responses of fonio against drought, root biomass was calculated after the end of the drought treatment (41-DAE). The Guinea accession had a significant (P ≤ 0.05) reduction in root weight (27–37%) from plants exposed to 6%-VWC compared to 26% VWC. By contrast, the Mali accession did not significantly change root biomass after exposure to drought (Fig. 6A-B, Table S2). Furthermore, Mali accession plants exposed to drought (6%-VWC) showed dramatic increases in the root/shoot ratio, which were significantly greater (p ≤ 0.05) than all other treatments, and 40%-52% greater than the Guinea accession (Fig. 6C, Table S2).

Root system architecture

The above measurements suggested that the Mali accession root system was acclimating to help plants survive under drought by prioritizing root growth for water scavenging. When excavated Mali accession roots were visually examined, it appeared as if drought exposure caused roots to be more fibrous, suggestive of a change in root system architecture (RSA) (Fig. 6C). Therefore, a new experiment was designed to examine RSA in detail, focusing only on the Mali accession. Plants were exposed only to the extreme moisture treatments (6% and 26%-VWC) which were initiated at germination, and root system architecture was examined at 3-day intervals for finer temporal resolution, and at macro-, meso- and micro-scales for morphological resolution. In Trial 1, the experiment terminated at 15 DAE, which was extended to 24 DAE in Trial 2:

Macro scale

For the macro scale root system study, crown root number, diameter and length were evaluated (Fig. 8). The number of crown roots from plants exposed to 6%-VWC was significantly lower (2–3 fold) (p ≤ 0.05) than under 26%-VWC starting at 6–9 DAE and continuing until the final data point (Fig. 8A-B, S1, Table S4). A similar trend was observed for the crown root diameter in both trials, except it was significant only after 18 DAE in Trial 2; drought was associated with a 45% reduction in crown root diameter by 24 DAE in Trial 2 (Fig. 8A, C, Table S4). Contrastingly, the length of the longest crown root was greater for drought-exposed plants, which started to be significant at 12 DAE and 21 DAE in Trials 1 and 2, respectively; longest crown root length was 29% greater under drought by 24 DAE in Trial 2 (Fig. 8A, D, Table S4). Looking at the average crown root lengths, the difference between moisture treatments was more visible, starting at 9–12 DAE (Fig. 8A, E, Table S4). At 24 DAE, the average crown root length for drought-exposed plants was 60% greater than the plants watered optimally (Table S4). Total root surface area for roots with a diameter > 0.2 mm, corresponding to crown roots, was significantly greater (p ≤ 0.10) in plants grown at 26%-VWC compared to 6%-VWC starting at 9 DAE which was 38% lower at 24 DAE (Fig. 8F, Table S4).

Meso scale

At the meso scale, corresponding to lateral root branching, the earlier observation (Fig. 6C) that root systems under drought were more fibrous/highly branched than optimally-watered plants, was again observed (Fig. 9A). Upon counting, the number of root tips was significantly greater for drought-treated plants than optimally-watered plants starting at 9 DAE (P ≤ 0.10), becoming 63% greater by 24 DAE (Fig. 9B, Table S5). The total surface area of thin roots (diameter ≤ 0.2 mm) was significantly greater (P ≤ 0.05) in plants exposed to drought than optimal water, starting at 9–12 DAE; by 24 DAE, total thin root surface area was 43% greater under drought (Fig. 9C, Table S5).

Microscale

At the micro scale, light microscope images combined with manual tracing were used to measure root hair length and density on the longest crown roots and an average from the longest ≤ 4 crown roots per plant. For each crown root, root hairs from older (S1), intermediate (S2) and younger (S3) segments were phenotyped (Fig. 10A-C). In total, an estimated 27,000 root hairs were manually traced. Root hair length and density were generally significantly greater (p ≥ 0.05–0.10) under drought than optimal water at the latest time point in each trial, i.e. the longest exposure to drought (Fig. 10B-H). In general, in Trial 1, the trend of increased root hair length and density under drought was observed at earlier time points, than Trial 2 (Fig. 10D-H).

Numerically, at the end of trial 1 (15 DAE), root hair lengths on the longest crown root were, respectively, 31, 56 and 125% greater in S1, S2, and S3 for the plants exposed to drought, whereas at the end of trial 2 (24 DAE) they were 59, 32 and 52% greater (Fig. 10D-F, Table S6). Similar observations were made for root hairs averaged across crown roots (Fig. S2A-C, Table S7).

Similarly, in all crown root segments, there was a > 2-fold increase in root hair density by 15 DAE in trial 1 and 24 DAE in trial 2 for plants grown under drought (Fig. 10G-H); it became significantly greater starting at 9 DAE in S3 (Table S6). Root hair density, when averaged across crown roots, increased 2–5 fold in S2 and S3 by 24 DAE (Fig. S2D-F, Table S7).

Fonio is known to survive the dry growing conditions of the African Sahel (Cruz and Beavogui, 2016) but prior to this study, there was a lack of scientific detail to explain the traits that may contribute to drought acclimation in this crop. Here, several metabolic, physiological and morphological traits have been identified for the first time as potential drought acclimation strategies in white fonio (Fig. 11). Two independent trials showed consistent results for some traits, but the magnitude sometimes differed, with the responses suggesting that plants in Trial 2 were exposed to more severe stress (e.g. as shown by leaf angle, alanine betaine). The two tested accessions, representing different rainfall regimes in West Africa, exhibited similar trends in responding to drought; however, the degree of response for selected traits (e.g. betaine accumulation, leaf width, root/shoot ratio) was greater for the Mali accession (originating from a low rainfall environment) than the Guinea accession (higher rainfall origin), which was associated with the superior grain yield and harvest index observed for the Mali accession after drought exposure. Such differences in acclimation traits provide preliminary evidence that fonio landraces possess genetic diversity that offers opportunities for selection in order to develop more adapted and higher-yielding varieties for African smallholder farmers.

Grain yield and harvest index

The stable and superior harvest index of the Mali accession, despite the severe drought imposed, is a remarkable indicator of post-drought recovery compared to the Guinea accession. The reduction in grain yield (per plant) due to drought was < 25% for the Mali accession, whereas it was > 40% for the Guinea accession. However, the shoot yield reduction was similar (~ 38%) for both accessions. These findings suggest that the Mali accession is comparatively superior with respect to prioritizing grain production over shoot biomass during the recovery phase, and explains the improved harvest index observed in plants grown at 6%-volumetric water content (VWC). In a prior study in foxtail millet (Setaria italica), the phylogenetically closest mainstream crop to fonio (Saarela et al. 2018), increased carbon partitioning to grain during drought, resulting in improved harvest index, was also observed (Krishnamurthy et al. 2016). Similarly, a positive correlation between drought tolerance and harvest index has also been reported in foxtail millet (Qin et al. 2020; Zhang et al. 2022). The Guinea accession generally showed a trend of superior shoot weight compared to the Mali accession, in part due to its later flowering time (~ 2 weeks later for Guinea), which may have contributed to the overall lower harvest index of the Guinea accession (i.e. greater time for vegetative growth, but similar grain fill duration).

Here, panicle traits were examined in detail as it is another parameter associated with differences in the grain yields of a crop. The Guinea accession had > 34% longer panicles than the Mali accession under drought. As panicle length and grain weight/plant had a negative correlation, longer panicles in the Guinea accession contributed to lower grain weight per panicle. In addition, grain weight per panicle, which was positively correlated with the grain weight per plant, was lower in the Guinea accession than in the Mali accession mainly because of poor grain filling. Similar findings were also observed in foxtail millet, where genotypes with superior grain-filling traits had greater grain yield under drought (Zhang et al. 2020). Since our findings do not show a significant association between moisture treatments and grain filling, but rather only differences between accessions, variation in grain filling may be a fixed genetic trait.

This finding opens an opportunity for landrace screening and selection.

Osmoprotectants and physiological responses to drought

Accumulation of osmolytes in shoot tissue is a known acclimation strategy observed in many grasses under drought stress (Blum, 2017; Ghosh et al. 2021). Therefore, in white fonio exposed to drought, the observed remarkable increases (up to 60-fold) in the leaf concentrations of betaines (including alanine betaine, glycine betaine and proline betaine) may contribute to moisture stress acclimation in this crop. Prior studies have shown that osmolytes such as glycine betaine and proline protect membranes and cell organelles from dehydration and damage in different plant species exposed to drought stress (Sharma et al. 2019; Ghosh et al. 2021). Similarly, in maize and foxtail millet, accumulation of proline, glycine betaine and amino acids in leaves during drought has been reported (Ajithkumar and Panneerselvam 2013; Salika and Riffat, 2021), including induction of proline biosynthesis genes (Qin et al. 2020). Furthermore, betaine accumulation has been suggested to be a domestication trait in foxtail millet (Li et al. 2022). Numerically, the Mali accession showed greater induction of these osmoprotectants than the Guinea accession, particularly in the apparently more stressful second trial, but these differences were not significant. Therefore, evaluation of a wider genetic pool is necessary to understand the potential osmoprotectant variation observed, which if validated, could be another target of selection breeding.

Among all the physiological parameters studied, leaf chlorophyll (SPAD) showed a strong positive correlation with grain yield in white fonio. Similar observations have been reported in other crops including barley, maize and wheat (Sallam et al. 2019; Khadka et al. 2020; Pallavolu et al. 2023). SPAD declined during the drought period in both accessions as previously observed in foxtail millet (Ajithkumar and Panneerselvam, 2013; Abreha et al. 2022). In our study, the decline in chlorophyll content was associated with reduced photosynthesis during drought, consistent with reports from other crops (Mu et al. 2016; Sallam et al. 2019). However, across moisture treatments, the Mali accession exhibited greater SPAD chlorophyll compared to the Guinea accession, similar to grain yield and harvest index. A possible reason for this superiority may be altered genetic regulation for chlorophyll (e.g. Staygreen genes in maize) resulting in delayed foliar senescence and prolonged leaf greening during abiotic stresses (Thomas and Ougham, 2014; Khadka et al. 2020), resulting in improved grain yield during drought in the Mali accession. In sorghum, delayed leaf senescence was shown to increase late-season nitrogen supply to promote grain filling (Borrell et al. 2000). Therefore, in future breeding efforts in white fonio, SPAD improvement could be a selectable trait to target grain yield improvement, and has the advantage of low equipment cost compared to many other yield trait evaluations.

Contradicting the SPAD result, the rate of photosynthesis decline under drought was greater for the Mali accession than the Guinea accession. A reasonable explanation for this finding was that the Mali accession had a ~ 20% steeper leaf angle compared to the Guinea accession under drought, and hence reduced incidental sunlight capture (Falster and Westoby, 2003). Under high light and dry conditions, as would be expected in Mali, prior studies have demonstrated that a steeper leaf angle prevents damage to photosystems (Falster and Westoby, 2003; Pastenes et al. 2004).

We could not find information in the literature on the specific relationship between leaf width and grain yield in related millets, but a few studies note a positive correlation between total leaf area and grain yield in millets (Zhang et al. 2022; Tiwari et al. 2022). In this experiment, reduced leaf width was observed at the end of the drought treatment (41DAE), which negatively correlated with grain yield (weak but significant) in both accessions. Both accessions had a similar reduction in leaf width (10–13%) during drought and recovered after optimal watering was restored by 61 DAE; however, leaf width was consistently lower in the Mali accession than the Guinea accession across water treatments, contradictory to the finding that the Mali accession consistently had greater grain yield. However, the reduced leaf width of the Mali accession may have contributed to its overall greater drought tolerance in terms of final grain yield. Reduced leaf width contributes to reduced leaf surface area which is a well-known acclimation response to drought, limiting transpiration and carbon demand which would require open stomates (Chaves et al. 2003; Saleem et al. 2021). Dynamic changes in leaf width in response to drought should be explored in other fonio landraces as a potential selection trait, because it is visual, cost- and time-effective.

Finally, this study revealed that white fonio is capable of dramatically reducing stomatal conductance during drought. Reduced stomatal conductance is well known to be a major contributor to drought tolerance in plants (Bertolino et al. 2019), and this crop should be added to future studies that examine the underlying mechanisms of this trait. Additionally, the Mali accession showed a trend towards lower transpiration and stomatal conductance than the Guinea accession under drought, which may have substantially contributed to the greater yield stability of the Mali accession under water limitation. Previous studies have shown that plants reduce water loss by limiting incidental sunlight exposure via increased leaf angle (Chaves et al. 2003). Indeed, as already noted earlier, the Mali accession had a steeper leaf angle compared to the Guinea accession under drought, and identifies this trait as another potential visual, cost- and time-effective selection trait in fonio.

Underground traits

Total root biomass at the end of the drought treatment was reduced for both fonio accessions in Experiment A; however, the reduction was significantly lower for the Mali accession. Compared to optimally watered plants, this stability of root biomass, while shoot biomass diminished, led to a relatively greater (> 40%) root/shoot ratio in the Mali accession compared to the Guinea accession. An increase in the root/shoot ratio has been reported to be helpful in water scavenging during drought in various crops including wheat, sorghum and maize (Comas et al. 2013; Khadka et al. 2020; Ndlovu et al. 2021; Salika and Riffat, 2021). Given these results, a subsequent thorough investigation of the fonio root system was undertaken at the seedling stage, which included manual tracing of ~ 27,000 root hairs. These measurements revealed that drought caused a shift in the fonio root system architecture at three levels: macro scale (crown roots), meso scale (lateral roots) and micro-scale (root hair).

At the macro scale, in drought-exposed plants, growth of the juvenile primary root was promoted but initiation of nodal roots (later emerging crown roots) was restricted, whereas it was the opposite in optimally watered plants. The number, surface area and thickness of crown roots of Mali accession juvenile plants were significantly reduced under drought, but crown root length increased. In cereals crops, decreased crown root number is a strategic drought acclimation trait as reported in previous studies (Comas et al. 2013; Lynch, 2022). As observed in wheat, maize and sorghum, a reduction in the crown root number has been shown to help conserve and divert assimilates to root elongation to enable scavenging of water at greater depth (Gao and Lynch, 2016; Ndlovu et al. 2021; Lynch, 2022).

At the meso scale, the fonio root system was observed to be more fibrous in plants exposed to drought than optimally watered plants. Contrasting the crown root response, the surface area of lateral roots on the juvenile primary root increased significantly under drought, inconsistent with the previous literature (Lynch, 2022; Elakhdar et al. 2022). Potentially, this trait helps to increase total root surface area for water uptake as the root penetrates the water column (Comas et al. 2013) but perhaps also to scavenge subsurface water from sporadic rainfall at the start of the growing season in the dry Sahel, functionally similar to what has been observed in desert plants (Kirschner et al. 2021).

At the micro scale, root hair length and density were significantly increased in white fonio under drought, mainly around the distal part of the crown roots (segments 1 and 2). As a primary site of aquaporins, increased root hair length and density potentially supports scavenging of soil water during drought, as reported in different cereals, including barley and foxtail millet (Postma et al. 2014; Liu et al. 2018; Marin et al. 2021; Kou et al. 2022). However, since water can diffuse across large distances in soil, longer root hairs would provide limited benefit for water uptake during drought (Comas et al. 2013). It may be that fonio produces longer root hairs during drought as an acclimation response to improve uptake of mineral nutrients, especially immobile nutrients (e.g. phosphorus, zinc) which can diffuse at only a micro-scale in soil (Brown et al. 2012; Wissuwa and Kant, 2021).

In this study, among different acclimation traits observed in response to drought in fonio (Fig. 7, 11), a few traits were identified as being practically relevant for future selection breeding in white fonio. Amongst these traits, narrower leaves should be explored for its potential to improve grain weight and harvest index under drought. Similarly, greater leaf angle holds potential to reduce transpiration, though no significant positive correlation was observed with grain weight, perhaps because of its negative impact on light capture (photosynthesis). Advantages of these traits is that they can be selected at the juvenile stage, and are cost and time effective. Similarly, leaf chlorophyll (SPAD), which was highly and positivel correlated with grain yield, may be another trait that can be targeted in fonio breeding programmes as only a modest investment is required to purchase a SPAD meter. However, future studies are needed to understand the relationship between SPAD, photosynthesis and grain yield, as we could not fully explain the apparent contradiction of greater SPAD chlorophyll but a lower rate of photosynthesis in the Mali accession during drought. Similarly, the cost versus benefit of increased concentrations of betaine osmoprotectants in drought-exposed plants has to be quantified, before consideration as a selection trait, ideally simultaneously with the development/adoption of inexpensive colorimetric diagnostics to quantify these osmoprotectants. Finally, this study has identified poor grain filling as a major potential genetic limitation of the Guinea accession, hindering its total grain yield, regardless of soil moisture. This grain fill trait variation between the two genotypes provides preliminary evidence that germplasm screening and landrace selection for this trait may be beneficial to help small scale farmers who cultivate white fonio in West Africa.

ANOVA Analysis of Variance

DAE Days After Emergence

HI Harvest Index

PCA Principal Component Analysis

RSA Root System Architecture

VWC Volumetric Water Content

Funding

This work was supported by a grant to M.N.R. from the Natural Sciences and Engineering Research Council (NSERC) of Canada (Discovery Program 4000924).

Competing Interests

The authors have no relevant financial or non-financial interests to disclose

Author contributions

R.P. conducted experiments, statistically analyzed the data, created figures and tables and wrote the manuscript. J.A. Contributed to conducting root hair experiments and writing the root hair section. D.J.L.B assisted in conducting experiments. M.N.R. conceived and supervised the study and edited the manuscript.

Data Availability

The datasets generated during and/or analysed during the current study are available in the study itself including supplemental materials.

Acknowledgements

We thank Dr. Adeline Barnaud (IRD, France) for the gift of fonio seeds, and the farmers in West Africa who were the original source of the landraces used in this study. We thank the University of Victoria-Genome BC Proteomics Centre, Victoria, BC, Canada, for conducting metabolomic analysis. We thank Sue Couling (University of Guelph) for growth facility assistance, as well as Michael Gebre and Prof. Hugh Earl (University of Guelph) for technical assistance with the LI-COR instrument. R.P. was the recipient of the Arrell Graduate Scholarship from the Arrell Food Institute, University of Guelph, Canada. Additional funding was provided by grants to M.N.R. from the Natural Sciences and Engineering Research Council (NSERC) of Canada.

Abreha KB, Enyew M, Carlsson AS, Vetukuri RR, Feyissa T, Motlhaodi T, Ng’uni D, Geleta M (2022) Sorghum in dryland: morphological, physiological, and molecular responses of sorghum under drought stress. Planta 255:20. https://doi.org/10.1007/s00425-021-03799-7
Abrouk M, Ahmed HI, Cubry P, Šimoníková D, Cauet S, Pailles Y, Bettgenhaeuser J, Gapa L, Scarcelli N, Couderc M et al (2020) Fonio millet genome unlocks African orphan crop diversity for agriculture in a changing climate. Nat Commun 11:4488. https://doi.org/10.1038/s41467-020-18329-4
Adoukonou-Sagbadja H, Wagner C, Dansi A, Ahlemeyer J, Daïnou O, Akpagana K, Ordon F, Friedt W (2007) Genetic diversity and population differentiation of traditional fonio millet (Digitaria spp.) landraces from different agro-ecological zones of West Africa. Theor Appl Genet 115:917–931. https://doi.org/10.1007/s00122-007-0618-x
Ajithkumar IP, Panneerselvam R (2013) Osmolyte accumulation, photosynthetic pigment and growth of Setaria italica (L.) P. Beauv. under drought stress. Asian Pac J Reprod 2(3):220–224. https://doi.org/10.1016/S2305-0500(13)60151-7
Barnaud A, Vigouroux Y, Diallo MT, Saidou SI, Piquet M, Barry MB, Bakasso Y, Zekraoui L, Rivallan R, Kane NA, Billot C (2017) High selfing rate inferred for white fonio [Digitaria exilis (Kippist.) Stapf] reproductive system opens up opportunities for breeding programs. Genet Res Crop Evol 64:1485–1490. https://doi.org/10.1007/s10722-017-0515-3
Berger J, Palta J, Vadez V (2016) Review: An integrated framework for crop adaptation to dry environments: Responses to transient and terminal drought. Plant Sci 253:58–67. https://doi.org/10.1016/j.plantsci.2016.09.007
Bertolino LT, Caine RS, Gray JE (2019) Impact of stomatal density and morphology on water-use efficiency in a changing world. Front Plant Sci 10:225. https://doi.org/10.3389/fpls.2019.00225
Biasutti M (2019) Rainfall trends in the African Sahel: Characteristics, processes, and causes. WIRES Clim Change 10:e591. https://doi.org/10.1002/wcc.591
Blum A (2017) Osmotic adjustment is a prime drought stress adaptive engine in support of plant production. Plant Cell Environ 40:4–10. https://doi.org/10.1111/pce.12800
Borrell AK, Hammer GL, Henzell RG (2000) Does maintaining green leaf area in sorghum improve yield under drought? II. Dry matter production and yield’. Crop Sci 40:1037–1048. https://doi.org/10.2135/cropsci2000.4041037x
Brown LK, George TS, Thompson JA, Wright G, Lyon J, Dupuy L, Hubbard SF, White PJ (2012) What are the implications of variation in root hair length on tolerance to phosphorus deficiency in combination with water stress in barley (Hordeum vulgare)? Ann Bot. 110:319–328. https://doi.org/10.1093/aob/mcs085
CBI (2023) The European market potential for fonio. accessed: 20 December 2023. https://www.cbi.eu/market-information/grains-pulses-oilseeds/fonio/market-potential
Chaves MM, Maroco JP, Pereira JS (2003) Understanding plant responses to drought - from genes to the whole plant. Funct Plant Biol 30:239–264. https://doi.org/10.1071/FP02076
Comas LH, Becker SR, Cruz VMV, Byrne PF, Dierig DA (2013) Root traits contributing to plant productivity under drought. Front Plant Sci 4:442. https://doi.org/10.3389/fpls.2013.00442
Cruz J, Béavogui F (2016) Fonio, an African cereal. The French Agricultural Research Centre for International Development CIRAD, Montpellier. http://www.cirad.fr/en/
Elakhdar A, Solanki S, Kubo T, Abed A, Elakhdar I, Khedr R, Hamwieh A, Capo-chichi LJA, Abdelsattar M, Franckowiak JD, Qualset CO (2022) Barley with improved drought tolerance: Challenges and perspectives. Environ Exp Bot 201:104965. https://doi.org/10.1016/j.envexpbot.2022.104965
Falster DS, Westoby M (2003) Leaf size and angle vary widely across species: What consequences for light interception? New Phytol 158:509–525. https://doi.org/10.1046/j.1469-8137.2003.00765.x
FAOSTAT (2022) Crop and livestock products. Food and Agriculture Organization of the United Nations. Accessed on November 17, 2023. https://www.fao.org/faostat/en/#data
Gao Y, Lynch JP (2016) Reduced crown root number improves water acquisition under water deficit stress in maize (Zea mays L). J Exp Bot 67(15):4545–4557. https://doi.org/10.1093/jxb/erw243
Ghosh UK, Islam MN, Siddiqui MN, Khan MAR (2021) Understanding the roles of osmolytes for acclimatizing plants to changing environment: a review of potential mechanism. Plant Signal Behav 16(8):e1913306. https://doi.org/10.1080/15592324.2021.1913306
Gigou J, Stilmant D, Diallo TA, Cisse N, Sanogo MD, Vaksmann M, Dupuis B (2009) Fonio millet (Digitaria exilis) response to N, P and K fertilizers under varying climatic conditions in West Africa. Exp Agric 45:401–415. https://doi.org/10.1017/S0014479709990421
Goron TL, Bhosekar VK, Shearer CR, Watts S, Raizada MN (2015) Whole plant acclimation responses by finger millet to low nitrogen stress. Front Plant Sci 6:652. https://doi.org/10.3389/fpls.2015.00652
Gueye M, Kanfany G, Fofana A, Gueye M, Noba K, Grove J (2015) Effect of planting date on growth and grain yield of fonio millet (Digitaria exilis Stapf) in the Southeast of Senegal. Int J Biol Chem Sci 9(2):581–592. https://doi.org/10.4314/ijbcs.v9i2.1
Han J, Gagnon S, Eckle T, Borchers CH (2013) Metabolomic analysis of key central carbon metabolism carboxylic acids as their 3-nitrophenylhydrazones by UPLC/ESI-MS. Electrophoresis 34:2891–2900. https://doi.org/10.1002/elps.201200601
Han J, Higgins R, Lim MD, Lin K, Yang J, Borchers CH (2018) Short-term stabilities of 21 amino acids in dried blood spots. Clin Chem 64:400–402. https://doi.org/10.1373/clinchem.2017.278457
Han J, Lin K, Sequria C, Yang J, Borchers CH (2016) Quantitation of low molecular weight sugars by chemical derivatization-liquid chromatography/multiple reaction monitoring/mass spectrometry. Electrophoresis 37:1851–1860. https://doi.org/10.1002/elps.201600150
Hasibeder R, Fuchslueger L, Richter A, Bahn M (2015) Summer drought alters carbon allocation to roots and root respiration in mountain grassland. New Phytol 205:1117–1127. https://doi.org/10.1111/nph.13146
Ibrahim Bio Yerima AR, Achigan-Dako EG, Aissata M, Sekloka E, Billot C, Adje COA, Barnaud A, Bakasso Y (2020) Agromorphological characterization revealed three phenotypic groups in a region-wide germplasm of fonio (Digitaria exilis (Kippist) Stapf) from West Africa. Agronomy 1653:10. https://doi.org/10.3390/agronomy10111653
IPCC (2018) In: V Masson-Delmotte P, Zhai H-O, Portner D, Roberts J, Skea et al (eds) Summary for Policymakers. in Global Warming of 1.5°C. An IPCC Special Report on the impacts of global warming of 1.5°C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change, sustainable development, and efforts to eradicate poverty. Cambridge University Press, Cambridge, UK and New York, NY, USA, pp 3–24. https://doi.org/10.1017/9781009157940.001
Jideani AI, Akingbala JO (1993) Some physiochemical properties of Acha (Digitaria exilis Stapf) and Iburu (Digitaria iburua Stapf) grains. J Sci Food Agric 63:369–374
Khadka K, Earl HJ, Raizada MN, Navabi A (2020) A physio-morphological trait-based approach for breeding drought tolerant wheat. Front Plant Sci 11:715. https://doi.org/10.3389/fpls.2020.00715
Kirschner GK, Xiao TT, Blilou I (2021) Rooting in the desert: A developmental overview on desert plants. Genes 12:709. https://doi.org/10.3390/genes12050709
Kou X, Han W, Kang J (2022) Responses of root system architecture to water stress at multiple levels: A meta-analysis of trials under controlled conditions. Front Plant Sci 13:1085409. https://doi.org/10.3389/fpls.2022.1085409
Krishnamurthy L, Upadhyaya HD, Kashiwagi J, Purushothaman R, Dwivedi SL (2016) Variation in drought-tolerance components and their interrelationships in the core collection of foxtail millet (Setaria italica) germplasm. Crop Pasture Sci 67:834–846. https://doi.org/10.1071/CP15338
Li X, Gao J, Song J, Guo K, Hou S, Wang X, He Q, Zhang Y, Zhang Y, Yang Y et al (2022) Multi-omics analyses of 398 foxtail millet accessions reveal genomic regions associated with domestication, metabolite traits, and anti-inflammatory effects. Mol Plant 15:1367–1383. https://doi.org/10.1016/j.molp.2022.07.003
Liu TY, Ye N, Song T, CaoY, Gao B, Zhang D, Zhu F, Chen M, Zhang Y, Xu W, Zhang J (2019) Rhizosheath formation and involvement in foxtail millet (Setaria italica) root growth under drought stress. J Integr Plant Biol 61(4):449–462. https://doi.org/10.1111/jipb.12716
Luo Y, Li F, Wang GP, Yang XH, Wang W (2010) Exogenously-supplied trehalose protects thylakoid membranes of winter wheat from heat-induced damage. Biol Plant 54(3):495–501
Lynch JP (2022) Harnessing root architecture to address global challenges. Plant J 109:415–431. https://doi.org/10.1111/tpj.15560
Marin M, Feeney DS, Brown LK, Naveed M, Ruiz S, Koebernick N, Bengough AG, Hallett PD, Roose T, Puértolas J et al (2021) Significance of root hairs for plant performance under contrasting field conditions and water deficit. Ann Bot 128:1–16. https://doi.org/10.1093/aob/mcaa181
Maritz J (2022) Company bets on surging demand for West African ancient grain fonio in US market. in web portal: How we made it in Africa.com. accessed: 30 December 2023. for%20fonio%20is%20growing,want%20to%20be%20on%2Dtrend. https://www.howwemadeitinafrica.com/company-bets-on-surging-demand-for-western-african-ancient-grain-fonio-in-us-market/139319/#:~:text=Demand%20
Medrano H, Escalona JM, Bota J, Gulías J, Flexas J (2002) Regulation of photosynthesis of C₃ plants in response to progressive drought: Stomatal conductance as a reference parameter. Ann Bot 89:895–905. https://doi.org/10.1093/aob/mcf079
Morgan JM (1984) Osmoregulation and water stress in higher plants. Ann Rev Plant Physiol 35:299–319
Mu X, Chen Q, Chen F, Yuan L, Mi G (2016) Within-leaf nitrogen allocation in adaptation to low nitrogen supply in maize during grain-filling stage. Front Plant Sci 7:699. https://doi.org/10.3389/fpls.2016.00699
Ndlovu E, van Staden J, Maphosa M (2021) Morpho-physiological effects of moisture, heat and combined stresses on Sorghum bicolor [Moench (L.)] and its acclimation mechanisms. Plant Stress 2:100018. https://doi.org/10.1016/j.stress.2021.100018
NRC (1996) Lost Crops of Africa: Volume I: Grains. National Research Council. National Academies, Washington, DC. https://doi.org/10.17226/2305
Olodo KF, Gueye MC, Calatayud C, Diop BM, Kane NA, Ngom A, Ntui VO, Barreto MMS, Uyoh EA, Abraham S et al (2019) EST-SSR development for Digitaria exilis and its relatives D. iburua and D. longiflora from transcriptome sequences. Plant Genet Resour: Characteterisation Util 17(3):280–284. https://doi.org/10.1017/S1479262118000400
Omotoso AB, Letsoalo S, Olagunju KO, Tshwene CS, Omotayo AO (2023) Climate change and variability in sub-Saharan Africa: A systematic review of trends and impacts on agriculture. J Clean Prod 414:137487. https://doi.org/10.1016/j.jclepro.2023.137487
Pallavolu LA, Pasala R, Kulasekaran R, Pandey BB, Virupaksham U, Perika S (2023) Analysing the SPAD dynamics of water-stressed vs. well-watered sesame (Sesamum indicum L.) accessions and establishing their relationship with seed yield. PeerJ 11:14711. https://doi.org/10.7717/peerj.14711
Parthasarathy A, Savka MA, Hudson AO (2019) The synthesis and role of β-alanine in plants. Front Plant Sci 10:921. https://doi.org/10.3389/fpls.2019.00921
Pastenes C, Pimentel P, Lillo J (2005) Leaf movements and photoinhibition in relation to water stress in field-grown beans. J Exp Bot 56:425–433. https://doi.org/10.1093/jxb/eri061
Postma JA, Dathe A, Lynch JP (2014) The optimal lateral root branching density for maize depends on nitrogen and phosphorus availability. Plant Physiol 166:590–602. https://doi.org/10.1104/pp.113.233916
Qin L, Chen E, Li F, Yu X, Liu Z, Yang Y, Wang R, Zhang H, Wang H, Liu B et al (2020) Genome-wide gene expression profiles analysis reveal novel insights into drought stress in foxtail millet (Setaria italica L). Int J Mol Sci 21:8520. https://doi.org/10.3390/ijms21228520
Rajendrakumar CSV, Reddy BVB, Reddy AR (1994) Proline protein interactions protection of structural and functional integrity of m₄ lactate dehydrogenase. Biochem Biophys Res Commun 201(2):957–963
Reddy AR, Chaitanya KV, Vivekanandan M (2004) Drought-induced responses of photosynthesis and antioxidant metabolism in higher plants. J Plant Physiol 161:1189–1202. https://doi.org/10.1016/j.jplph.2004.01.013
Saarela JM, Burke SV, Wysocki WP, Barrett MD, Clark LG, Craine JM, Peterson PM, Soreng RJ, Vorontsova MS, Duvall MR (2018) A 250 plastome phylogeny of the grass family (Poaceae): Topological support under different data partitions. PeerJ 6:e4299. https://doi.org/10.7717/peerj.4299
Saleem S, UI Mushtaq N, Shah WH, Rasool A, Hakeem KR, Rehman Reiaz UI (2021) Morpho-physiological, biochemical and molecular adaptation of millets to abiotic stresses: A review. Phyton-Int J Exp Bot 90(5):1363–1385. https://doi.org/10.32604/PHYTON.2021.014826
Salika R, Riffat J (2021) Abiotic stress responses in maize: a review. Acta Physiol Plant 43:130. https://doi.org/10.1007/s11738-021-03296-0
Sallam A, Alqudah AM, Dawood MFA, Baenziger PS, Börner A (2019) Drought stress tolerance in wheat and barley: Advances in physiology, breeding and genetics research. Int J Mol Sci 20:3137. https://doi.org/10.3390/ijms20133137
Sanders G, Arndt SK (2012) Plant responses to drought stress: From morphological to molecular features. In: Aroca R (ed) Plant Responses to Drought Stress: From Morphological to Molecular Features. Springer, Heidelberg, pp 199–229. https://doi.org/10.1007/978-3-642-32653-0
Sharma A, Shahzad B, Kumar V, Kohli SK, Sidhu GPS, Bali AS, Handa N, Kapoor D, Bhardwaj R, Zheng B (2019) Phytohormones regulate accumulation of osmolytes under abiotic stress. Biomolecules 9:285. https://doi.org/10.3390/biom9070285
Tadele Z (2016) Drought Adaptation in Millets. in Abiotic and Biotic Stress in Plants - Recent Advances and Future Perspectives. eds. A. K. Shanker, C. Shanker. InTechOpen. 639–662. https://doi.org/10.5772/61929
Temple VJ, Bassa JD (1991) Proximate Chemical Composition of Acha (Digitaria exilis) Grain. J Sci Food Agric 56:561–563
Thomas H, Ougham H (2014) The stay-green trait. J Exp Bot 65:3889–3900. https://doi.org/10.1093/jxb/eru037
Tiwari A, Kesarwani K, Sharma A, Ghosh T, Bisht N, Punetha S (2022) Drought Stress in Millets and Its Response Mechanism. in Advances in Plant Defense Mechanisms. ed. JN Kimatu. IntechOpen. https://doi.org/10.5772/intechopen.105942
Vodouhe SR, Achigan-Dako GE, Dansi A, Adoukonou-Sagbadja H (2007) Fonio: A treasure for West Africa. in Plant genetic resources and food security in West and Central Africa, regional conference, 26–30 April, 2004. eds. RS Vodouhe, K Atta-Krah, GE Achigan-Dako, O Eyog-Matig, H Avohou. Bioversity International, Rome, Italy. pp. 219–223
Wang X, Chen S, Ma X, Yssel AEJ, Chaluvadi SR, Johnson MS, Gangashetty P, Hamidou F, Sanogo MD, Zwaenepoel A et al (2021) Genome sequence and genetic diversity analysis of an under-domesticated orphan crop, white fonio (Digitaria exilis). GigaScience 10: 1–12. https://doi.org/10.1093/gigascience/giab013
Wissuwa M, Kant J (2021) Does half a millimetre matter? Root hairs for yield stability. A commentary on ‘Significance of root hairs for plant performance under contrasting field conditions and water deficit’. Ann Bot 128(1):III–V. https://doi.org/10.1093/aob/mcab027
Zhang W, Wang B, Liu B, Chen Z, Lu G, Ge Y, Bai C (2022) Trait selection for yield improvement in foxtail millet (Setaria italica beauv.) under climate change in the north China plain. Agronomy 12:1500. https://doi.org/10.3390/agronomy12071500
Zhu F (2020) Fonio grains: Physicochemical properties, nutritional potential, and food applications. Compr Rev Food Sci Food Saf 19:3365–3389. https://doi.org/10.1111/1541-4337.12608

FigureS1.pdf
Figure S1. Experiment B: Effect of moisture treatments on crown root architecture (macro scale) in the Mali accession of white fonio at different time points: Representative images to show crown root development at different time points (3-15 DAE): each pair consists of the root system of an optimally watered plant (left), and a drought exposed plant (right) from Trial 1. Scale: 1 cm. Abbreviations: VWC, volumetric water content; DAE, days after emergence.
FigureS2.pdf
Figure S2. Experiment B: Effect of moisture treatments on the root hair architecture (micro scale) of the Mali accession of white fonio at different time points (3-24 DAE), with data averaged for ≤4 crown roots/plant. (A-F) Effect of moisture treatments on root hair length and density in different crown root segments (S1-S3)-Trial 1 (top) and Trial 2 (bottom): (A) root hair length on S1, (B) root hair length on S2, (C) root hair length on S3, (D) root hair density on S1, (E) root hair density on S2, (F) root hair density on S3. For statistical significance, see Table S7. (*) and (+) signs symbolize statistical difference at P≤0.05 and P≤0.10 respectively. Abbreviations: VWC, volumetric water content; DAE, days after emergence.
SupplementalTables.docx

Download PDF

Editorial decision: Major revisions
30 Sep, 2024
Reviewers agreed at journal
27 Aug, 2024
Reviewers invited by journal
27 Aug, 2024
Editor invited by journal
07 Aug, 2024
Editor assigned by journal
07 Aug, 2024
First submitted to journal
06 Aug, 2024

You are reading this latest preprint version

Physiological, Morphological and Root System Architecture Acclimation Responses to Drought in the African Orphan Millet White Fonio (Digitaria exilis)

Status:

Version 1

Abstract

Aims

Methods

Results

Conclusions

Figures

Introduction

Materials and Methods

Plant Genetic Materials

Experiment A: Fonio Shoot Phenotyping Experiments

Measurements

Experiment B: Detailed Fonio Root Phenotyping Experiments

Statistical Analysis

Results

Metabolomics

Short term above-ground effects

Physiological traits

Shoot morphology

Long term above-ground impacts

Principal Component Analysis (PCA)

Underground Effects

Root biomass

Root system architecture

Meso scale

Microscale

Discussion

Grain yield and harvest index

Osmoprotectants and physiological responses to drought

Underground traits

Conclusion

Abbreviations

Declarations

References

Supplementary Files

Status:

Version 1